blob: 7df5d0061aed987fc4a0116d6519c41e58b3786d [file] [log] [blame] [edit]
/*
* Copyright (c) 2021, Alliance for Open Media. All rights reserved
*
* This source code is subject to the terms of the BSD 3-Clause Clear License
* and the Alliance for Open Media Patent License 1.0. If the BSD 3-Clause Clear
* License was not distributed with this source code in the LICENSE file, you
* can obtain it at aomedia.org/license/software-license/bsd-3-c-c/. If the
* Alliance for Open Media Patent License 1.0 was not distributed with this
* source code in the PATENTS file, you can obtain it at
* aomedia.org/license/patent-license/.
*/
#include "av1/encoder/encodetxb.h"
#include "aom_ports/mem.h"
#include "av1/common/blockd.h"
#include "av1/common/idct.h"
#include "av1/common/pred_common.h"
#include "av1/common/scan.h"
#include "av1/common/reconintra.h"
#include "av1/encoder/bitstream.h"
#include "av1/encoder/cost.h"
#include "av1/encoder/encodeframe.h"
#include "av1/encoder/hash.h"
#include "av1/encoder/rdopt.h"
#include "av1/encoder/tokenize.h"
#if NEWHR
static INLINE int get_low_range(int abs_qc, int lf) {
int base_levels = lf ? 6 : 4;
int parity = abs_qc & 1;
int abs2 = abs_qc & ~1;
int low = AOMMIN(abs2, COEFF_BASE_RANGE + base_levels - 2) + parity;
low -= base_levels - 1;
return low;
}
static INLINE int get_high_range(int abs_qc, int lf) {
int base_levels = lf ? 6 : 4;
int low_range = get_low_range(abs_qc, lf);
int high_range = (abs_qc - low_range - (base_levels - 1)) >> 1;
return high_range;
}
#endif
typedef struct {
tran_low_t qc;
tran_low_t dqc; // dequantized qc
int64_t delta_cost; // cost change between between coding up and low level
int delta_rate; // rate change between coding up and low level
bool upround; // is quantized into up level
bool tunable; // tunable mark
} coeff_info;
// set rd related information for the coefficient at current position.
void set_coeff_info(tran_low_t qc_low, tran_low_t dqc_low, tran_low_t qc_up,
tran_low_t dqc_up, int64_t cost_low, int64_t cost_up,
int rate_low, int rate_up, bool upround,
coeff_info *coef_info, const int scan_idx) {
if (!scan_idx) {
return;
}
coef_info[scan_idx].upround = upround;
if (upround) {
coef_info[scan_idx].qc = qc_low;
coef_info[scan_idx].dqc = dqc_low;
} else {
coef_info[scan_idx].qc = qc_up;
coef_info[scan_idx].dqc = dqc_up;
}
coef_info[scan_idx].tunable =
(abs(coef_info[scan_idx].qc) < MAX_BASE_BR_RANGE) ||
((abs(qc_up) == MAX_BASE_BR_RANGE) && upround);
if (coef_info[scan_idx].tunable) {
if (upround) {
coef_info[scan_idx].delta_cost = (cost_low - cost_up);
coef_info[scan_idx].delta_rate = (rate_low - rate_up);
} else {
coef_info[scan_idx].delta_cost = (cost_up - cost_low);
coef_info[scan_idx].delta_rate = (rate_up - rate_low);
}
}
}
typedef struct LevelDownStats {
int update;
tran_low_t low_qc;
tran_low_t low_dqc;
int64_t dist0;
int rate;
int rate_low;
int64_t dist;
int64_t dist_low;
int64_t rd;
int64_t rd_low;
int64_t nz_rd;
int64_t rd_diff;
int cost_diff;
int64_t dist_diff;
int new_eob;
} LevelDownStats;
static INLINE int get_dqv(const int32_t *dequant, int coeff_idx,
const qm_val_t *iqmatrix) {
int dqv = dequant[!!coeff_idx];
if (iqmatrix != NULL)
dqv =
((iqmatrix[coeff_idx] * dqv) + (1 << (AOM_QM_BITS - 1))) >> AOM_QM_BITS;
return dqv;
}
void av1_alloc_txb_buf(AV1_COMP *cpi) {
AV1_COMMON *cm = &cpi->common;
// We use the frame level sb size here instead of the seq level sb size. This
// is because fr_sb_size <= seq_sb_size, and we want to avoid repeated
// allocations. So we prefer to to allocate a larger memory in one go here.
int size = ((cm->mi_params.mi_rows >> cm->mib_size_log2) + 1) *
((cm->mi_params.mi_cols >> cm->mib_size_log2) + 1);
av1_free_txb_buf(cpi);
// TODO(jingning): This should be further reduced.
CHECK_MEM_ERROR(cm, cpi->coeff_buffer_base,
aom_memalign(32, sizeof(*cpi->coeff_buffer_base) * size));
}
void av1_free_txb_buf(AV1_COMP *cpi) { aom_free(cpi->coeff_buffer_base); }
static void write_golomb(aom_writer *w, int level) {
int x = level + 1;
int length = 0;
#if CONFIG_BYPASS_IMPROVEMENT
length = get_msb(x) + 1;
assert(length > 0);
aom_write_literal(w, 0, length - 1);
aom_write_literal(w, x, length);
#else
int i = x;
while (i) {
i >>= 1;
++length;
}
assert(length > 0);
for (i = 0; i < length - 1; ++i) aom_write_bit(w, 0);
for (i = length - 1; i >= 0; --i) aom_write_bit(w, (x >> i) & 0x01);
#endif
}
static INLINE int64_t get_coeff_dist(tran_low_t tcoeff, tran_low_t dqcoeff,
int shift) {
const int64_t diff = (tcoeff - dqcoeff) * (1 << shift);
const int64_t error = diff * diff;
return error;
}
static const int8_t eob_to_pos_small[33] = {
0, 1, 2, // 0-2
3, 3, // 3-4
4, 4, 4, 4, // 5-8
5, 5, 5, 5, 5, 5, 5, 5, // 9-16
6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6 // 17-32
};
static const int8_t eob_to_pos_large[17] = {
6, // place holder
7, // 33-64
8, 8, // 65-128
9, 9, 9, 9, // 129-256
10, 10, 10, 10, 10, 10, 10, 10, // 257-512
11 // 513-
};
static INLINE int get_eob_pos_token(const int eob, int *const extra) {
int t;
if (eob < 33) {
t = eob_to_pos_small[eob];
} else {
const int e = AOMMIN((eob - 1) >> 5, 16);
t = eob_to_pos_large[e];
}
*extra = eob - av1_eob_group_start[t];
return t;
}
#if CONFIG_ENTROPY_STATS
void av1_update_eob_context(int cdf_idx, int eob, TX_SIZE tx_size,
#if CONFIG_EOB_POS_LUMA
int is_inter,
#endif // CONFIG_EOB_POS_LUMA
PLANE_TYPE plane, FRAME_CONTEXT *ec_ctx,
FRAME_COUNTS *counts, uint8_t allow_update_cdf) {
#else
void av1_update_eob_context(int eob, TX_SIZE tx_size,
#if CONFIG_EOB_POS_LUMA
int is_inter,
#endif // CONFIG_EOB_POS_LUMA
PLANE_TYPE plane, FRAME_CONTEXT *ec_ctx,
uint8_t allow_update_cdf) {
#endif
int eob_extra;
const int eob_pt = get_eob_pos_token(eob, &eob_extra);
TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
const int eob_multi_size = txsize_log2_minus4[tx_size];
#if CONFIG_EOB_POS_LUMA
const int pl_ctx = get_eob_plane_ctx(plane, is_inter);
#else
const int pl_ctx = plane;
#endif // CONFIG_EOB_POS_LUMA
switch (eob_multi_size) {
case 0:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi16[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf) {
update_cdf(ec_ctx->eob_flag_cdf16[pl_ctx], eob_pt - 1,
EOB_MAX_SYMS - 6);
}
break;
case 1:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi32[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf)
update_cdf(ec_ctx->eob_flag_cdf32[pl_ctx], eob_pt - 1,
EOB_MAX_SYMS - 5);
break;
case 2:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi64[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf)
update_cdf(ec_ctx->eob_flag_cdf64[pl_ctx], eob_pt - 1,
EOB_MAX_SYMS - 4);
break;
case 3:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi128[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf) {
update_cdf(ec_ctx->eob_flag_cdf128[pl_ctx], eob_pt - 1,
EOB_MAX_SYMS - 3);
}
break;
case 4:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi256[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf) {
update_cdf(ec_ctx->eob_flag_cdf256[pl_ctx], eob_pt - 1,
EOB_MAX_SYMS - 2);
}
break;
case 5:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi512[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf) {
update_cdf(ec_ctx->eob_flag_cdf512[pl_ctx], eob_pt - 1,
EOB_MAX_SYMS - 1);
}
break;
case 6:
default:
#if CONFIG_ENTROPY_STATS
++counts->eob_multi1024[cdf_idx][pl_ctx][eob_pt - 1];
#endif
if (allow_update_cdf) {
update_cdf(ec_ctx->eob_flag_cdf1024[pl_ctx], eob_pt - 1, EOB_MAX_SYMS);
}
break;
}
const int eob_offset_bits = av1_eob_offset_bits[eob_pt];
if (eob_offset_bits > 0) {
int eob_ctx = eob_pt - 3;
int eob_shift = eob_offset_bits - 1;
int bit = (eob_extra & (1 << eob_shift)) ? 1 : 0;
#if CONFIG_ENTROPY_STATS
counts->eob_extra[cdf_idx][txs_ctx][plane][eob_ctx][bit]++;
#endif // CONFIG_ENTROPY_STATS
if (allow_update_cdf)
update_cdf(ec_ctx->eob_extra_cdf[txs_ctx][plane][eob_ctx], bit, 2);
}
}
static int get_eob_cost(int eob, const LV_MAP_EOB_COST *txb_eob_costs,
const LV_MAP_COEFF_COST *txb_costs
#if CONFIG_EOB_POS_LUMA
,
const int is_inter
#endif // CONFIG_EOB_POS_LUMA
) {
int eob_cost = 0;
int eob_extra;
const int eob_pt = get_eob_pos_token(eob, &eob_extra);
eob_cost += txb_eob_costs->eob_cost
#if CONFIG_EOB_POS_LUMA
[is_inter]
#endif // CONFIG_EOB_POS_LUMA
[eob_pt - 1];
const int eob_offset_bits = av1_eob_offset_bits[eob_pt];
if (eob_offset_bits > 0) {
const int eob_ctx = eob_pt - 3;
const int eob_shift = eob_offset_bits - 1;
const int bit = (eob_extra & (1 << eob_shift)) ? 1 : 0;
eob_cost += txb_costs->eob_extra_cost[eob_ctx][bit];
if (eob_offset_bits > 1) eob_cost += av1_cost_literal(eob_offset_bits - 1);
}
return eob_cost;
}
static const int golomb_bits_cost[32] = {
0, 512, 512 * 3, 512 * 3, 512 * 5, 512 * 5, 512 * 5, 512 * 5,
512 * 7, 512 * 7, 512 * 7, 512 * 7, 512 * 7, 512 * 7, 512 * 7, 512 * 7,
512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9,
512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9, 512 * 9
};
static const int golomb_cost_diff[32] = {
0, 512, 512 * 2, 0, 512 * 2, 0, 0, 0, 512 * 2, 0, 0, 0, 0, 0, 0, 0,
512 * 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};
static INLINE int get_golomb_cost(int abs_qc) {
#if NEWHR
if (abs_qc >= NUM_BASE_LEVELS + COEFF_BASE_RANGE) {
const int r = 1 + get_high_range(abs_qc, 0);
const int length = get_msb(r) + 1;
return av1_cost_literal(2 * length - 1);
}
#else
if (abs_qc >= 1 + NUM_BASE_LEVELS + COEFF_BASE_RANGE) {
const int r = abs_qc - COEFF_BASE_RANGE - NUM_BASE_LEVELS;
const int length = get_msb(r) + 1;
return av1_cost_literal(2 * length - 1);
}
#endif
return 0;
}
// Golomb cost of coding bypass coded level values in the
// low-frequency region.
static INLINE int get_golomb_cost_lf(int abs_qc) {
#if NEWHR
if (abs_qc >= LF_NUM_BASE_LEVELS + COEFF_BASE_RANGE) {
const int r = 1 + get_high_range(abs_qc, 1);
const int length = get_msb(r) + 1;
return av1_cost_literal(2 * length - 1);
}
#else
if (abs_qc >= 1 + LF_NUM_BASE_LEVELS + COEFF_BASE_RANGE) {
const int r = abs_qc - COEFF_BASE_RANGE - LF_NUM_BASE_LEVELS;
const int length = get_msb(r) + 1;
return av1_cost_literal(2 * length - 1);
}
#endif
return 0;
}
// Base range cost of coding level values in the
// low-frequency region, includes the bypass cost.
static INLINE int get_br_lf_cost(tran_low_t level, const int *coeff_lps) {
const int base_range =
AOMMIN(level - 1 - LF_NUM_BASE_LEVELS, COEFF_BASE_RANGE);
return coeff_lps[base_range] + get_golomb_cost_lf(level);
}
// Calculates differential cost for base range coding in the low-frequency
// region for encoder coefficient level optimization.
static INLINE int get_br_lf_cost_with_diff(tran_low_t level,
const int *coeff_lps, int *diff) {
const int base_range =
AOMMIN(level - 1 - LF_NUM_BASE_LEVELS, COEFF_BASE_RANGE);
int golomb_bits = 0;
if (level <= COEFF_BASE_RANGE + 1 + LF_NUM_BASE_LEVELS)
*diff += coeff_lps[base_range + COEFF_BASE_RANGE + 1];
if (level >= COEFF_BASE_RANGE + 1 + LF_NUM_BASE_LEVELS) {
int r = level - COEFF_BASE_RANGE - LF_NUM_BASE_LEVELS;
if (r < 32) {
golomb_bits = golomb_bits_cost[r];
*diff += golomb_cost_diff[r];
} else {
golomb_bits = get_golomb_cost_lf(level);
*diff += (r & (r - 1)) == 0 ? 1024 : 0;
}
}
return coeff_lps[base_range] + golomb_bits;
}
static INLINE int get_br_cost_with_diff(tran_low_t level, const int *coeff_lps,
int *diff) {
const int base_range = AOMMIN(level - 1 - NUM_BASE_LEVELS, COEFF_BASE_RANGE);
int golomb_bits = 0;
if (level <= COEFF_BASE_RANGE + 1 + NUM_BASE_LEVELS)
*diff += coeff_lps[base_range + COEFF_BASE_RANGE + 1];
#if NEWHR
if (level >= COEFF_BASE_RANGE + NUM_BASE_LEVELS) {
int r = get_high_range(level, 0);
#else
if (level >= COEFF_BASE_RANGE + 1 + NUM_BASE_LEVELS) {
int r = level - COEFF_BASE_RANGE - NUM_BASE_LEVELS;
#endif
if (r < 32) {
golomb_bits = golomb_bits_cost[r];
*diff += golomb_cost_diff[r];
} else {
golomb_bits = get_golomb_cost(level);
*diff += (r & (r - 1)) == 0 ? 1024 : 0;
}
}
return coeff_lps[base_range] + golomb_bits;
}
static INLINE int get_br_cost(tran_low_t level, const int *coeff_lps) {
const int base_range = AOMMIN(level - 1 - NUM_BASE_LEVELS, COEFF_BASE_RANGE);
return coeff_lps[base_range] + get_golomb_cost(level);
}
static INLINE int get_nz_map_ctx(const uint8_t *const levels,
const int coeff_idx, const int bwl,
const int height, const int scan_idx,
const int is_eob, const TX_CLASS tx_class,
const int plane) {
if (is_eob) {
if (scan_idx == 0) return 0;
if (scan_idx <= (height << bwl) / 8) return 1;
if (scan_idx <= (height << bwl) / 4) return 2;
return 3;
}
int stats = 0;
const int row = coeff_idx >> bwl;
const int col = coeff_idx - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
stats =
get_nz_mag_lf(levels + get_padded_idx(coeff_idx, bwl), bwl, tx_class);
return get_nz_map_ctx_from_stats_lf(stats, coeff_idx, bwl, tx_class);
} else {
stats = get_nz_mag(levels + get_padded_idx(coeff_idx, bwl), bwl, tx_class);
return get_nz_map_ctx_from_stats(stats, coeff_idx, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
}
}
static INLINE int get_nz_map_ctx_skip(const uint8_t *const levels,
const int height, const int scan_idx,
const int is_bob, const int coeff_idx,
const int bwl) {
if (is_bob) {
return get_lower_levels_ctx_bob(bwl, height, scan_idx);
}
const int stats =
get_nz_mag_skip(levels + get_padded_idx_left(coeff_idx, bwl), bwl);
return get_nz_map_ctx_from_stats_skip(stats, coeff_idx, bwl);
}
void av1_txb_init_levels_signs_c(const tran_low_t *const coeff, const int width,
const int height, uint8_t *const levels,
int8_t *const signs) {
const int stride = width + TX_PAD_LEFT;
int8_t *si = signs;
uint8_t *ls = levels;
// bottom 4 pad
memset(levels + stride * (height + TX_PAD_TOP), 0,
sizeof(*levels) * (TX_PAD_BOTTOM * stride + TX_PAD_END));
memset(signs + stride * (height + TX_PAD_TOP), 0,
sizeof(*signs) * (TX_PAD_BOTTOM * stride + TX_PAD_END));
// top 4 pad
memset(levels, 0, sizeof(*levels) * (TX_PAD_TOP * stride));
ls += TX_PAD_TOP * stride;
memset(signs, 0, sizeof(*signs) * (TX_PAD_TOP * stride));
si += TX_PAD_TOP * stride;
for (int i = 0; i < height; i++) {
// left 4 pad
for (int j = 0; j < TX_PAD_LEFT; j++) {
*ls++ = 0;
}
for (int j = 0; j < width; j++) {
*si++ = (int8_t)(coeff[i * width + j] > 0) ? 1 : -1;
*ls++ = (uint8_t)clamp(abs(coeff[i * width + j]), 0, INT8_MAX);
}
// right 4 pad
for (int j = 0; j < TX_PAD_RIGHT; j++) {
*si++ = 0;
}
}
}
void av1_txb_init_levels_skip_c(const tran_low_t *const coeff, const int width,
const int height, uint8_t *const levels) {
const int stride = width + TX_PAD_LEFT;
uint8_t *ls = levels;
// bottom 4 padded region
memset(levels + stride * (height + TX_PAD_TOP), 0,
sizeof(*levels) * (TX_PAD_BOTTOM * stride + TX_PAD_END));
// top 4 padded region
memset(levels, 0, sizeof(*levels) * (TX_PAD_TOP * stride));
ls += TX_PAD_TOP * stride;
for (int i = 0; i < height; i++) {
// left 4 padded region for each row
for (int j = 0; j < TX_PAD_LEFT; j++) {
*ls++ = 0;
}
for (int j = 0; j < width; j++) {
*ls++ = (uint8_t)clamp(abs(coeff[i * width + j]), 0, INT8_MAX);
}
}
}
void av1_txb_init_levels_c(const tran_low_t *const coeff, const int width,
const int height, uint8_t *const levels) {
const int stride = width + TX_PAD_HOR;
uint8_t *ls = levels;
memset(levels + stride * height, 0,
sizeof(*levels) * (TX_PAD_BOTTOM * stride + TX_PAD_END));
for (int i = 0; i < height; i++) {
for (int j = 0; j < width; j++) {
*ls++ = (uint8_t)clamp(abs(coeff[i * width + j]), 0, INT8_MAX);
}
for (int j = 0; j < TX_PAD_HOR; j++) {
*ls++ = 0;
}
}
}
void av1_get_nz_map_contexts_c(const uint8_t *const levels,
const int16_t *const scan, const uint16_t eob,
const TX_SIZE tx_size, const TX_CLASS tx_class,
int8_t *const coeff_contexts, const int plane) {
const int bwl = get_txb_bwl(tx_size);
const int height = get_txb_high(tx_size);
for (int i = 0; i < eob; ++i) {
const int pos = scan[i];
coeff_contexts[pos] = get_nz_map_ctx(levels, pos, bwl, height, i,
i == eob - 1, tx_class, plane);
}
}
// Encodes the EOB syntax in the bitstream.
static INLINE void code_eob(MACROBLOCK *const x, aom_writer *w, int plane,
TX_SIZE tx_size, int eob) {
MACROBLOCKD *xd = &x->e_mbd;
FRAME_CONTEXT *ec_ctx = xd->tile_ctx;
const PLANE_TYPE plane_type = get_plane_type(plane);
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
#if CONFIG_EOB_POS_LUMA
const int is_inter = is_inter_block(xd->mi[0], xd->tree_type);
const int pl_ctx = get_eob_plane_ctx(plane, is_inter);
#else
const int pl_ctx = plane_type;
#endif // CONFIG_EOB_POS_LUMA
// test
int eob_multi_size = txsize_log2_minus4[tx_size];
int eob_extra;
const int eob_pt = get_eob_pos_token(eob, &eob_extra);
switch (eob_multi_size) {
case 0:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf16[pl_ctx],
EOB_MAX_SYMS - 6);
break;
case 1:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf32[pl_ctx],
EOB_MAX_SYMS - 5);
break;
case 2:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf64[pl_ctx],
EOB_MAX_SYMS - 4);
break;
case 3:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf128[pl_ctx],
EOB_MAX_SYMS - 3);
break;
case 4:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf256[pl_ctx],
EOB_MAX_SYMS - 2);
break;
case 5:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf512[pl_ctx],
EOB_MAX_SYMS - 1);
break;
default:
aom_write_symbol(w, eob_pt - 1, ec_ctx->eob_flag_cdf1024[pl_ctx],
EOB_MAX_SYMS);
break;
}
const int eob_offset_bits = av1_eob_offset_bits[eob_pt];
if (eob_offset_bits > 0) {
const int eob_ctx = eob_pt - 3;
int eob_shift = eob_offset_bits - 1;
int bit = (eob_extra & (1 << eob_shift)) ? 1 : 0;
aom_write_symbol(w, bit,
ec_ctx->eob_extra_cdf[txs_ctx][plane_type][eob_ctx], 2);
#if CONFIG_BYPASS_IMPROVEMENT
// Zero out top bit; write (eob_offset_bits - 1) lsb bits.
eob_extra &= (1 << (eob_offset_bits - 1)) - 1;
aom_write_literal(w, eob_extra, eob_offset_bits - 1);
#else
for (int i = 1; i < eob_offset_bits; i++) {
eob_shift = eob_offset_bits - 1 - i;
bit = (eob_extra & (1 << eob_shift)) ? 1 : 0;
aom_write_bit(w, bit);
}
#endif
}
}
void av1_get_nz_map_contexts_skip_c(const uint8_t *const levels,
const int16_t *const scan,
const uint16_t bob, const uint16_t eob,
const TX_SIZE tx_size,
int8_t *const coeff_contexts) {
const int bwl = get_txb_bwl(tx_size);
const int height = get_txb_high(tx_size);
for (int i = bob; i < eob; ++i) {
const int pos = scan[i];
coeff_contexts[pos] =
get_nz_map_ctx_skip(levels, height, i, bob == i, pos, bwl);
}
}
int av1_write_sig_txtype(const AV1_COMMON *const cm, MACROBLOCK *const x,
aom_writer *w, int blk_row, int blk_col, int plane,
int block, TX_SIZE tx_size) {
MACROBLOCKD *xd = &x->e_mbd;
const CB_COEFF_BUFFER *cb_coef_buff = x->cb_coef_buff;
const int txb_offset =
x->mbmi_ext_frame->cb_offset[plane] / (TX_SIZE_W_MIN * TX_SIZE_H_MIN);
#if CONFIG_CONTEXT_DERIVATION
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
if (plane == AOM_PLANE_U)
memset(xd->tmp_sign, 0, width * height * sizeof(int32_t));
#endif // CONFIG_CONTEXT_DERIVATION
const uint16_t *eob_txb = cb_coef_buff->eobs[plane] + txb_offset;
const uint16_t eob = eob_txb[block];
const uint16_t *bob_txb = cb_coef_buff->bobs[plane] + txb_offset;
const uint16_t bob_code = bob_txb[block];
const uint8_t *entropy_ctx = cb_coef_buff->entropy_ctx[plane] + txb_offset;
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = (entropy_ctx[block] & TXB_SKIP_CTX_MASK);
if (plane == AOM_PLANE_V) {
txb_skip_ctx += (xd->eob_u_flag ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
}
#else
const int txb_skip_ctx = entropy_ctx[block] & TXB_SKIP_CTX_MASK;
#endif // CONFIG_CONTEXT_DERIVATION
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
FRAME_CONTEXT *ec_ctx = xd->tile_ctx;
const PLANE_TYPE plane_type = get_plane_type(plane);
const TX_TYPE tx_type =
av1_get_tx_type(xd, plane_type, blk_row, blk_col, tx_size,
cm->features.reduced_tx_set_used);
const int is_inter = is_inter_block(xd->mi[0], xd->tree_type);
const int is_fsc = (xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter);
#if CCTX_C2_DROPPED
if (plane == AOM_PLANE_V && is_cctx_allowed(cm, xd)) {
CctxType cctx_type = av1_get_cctx_type(xd, blk_row, blk_col);
if (!keep_chroma_c2(cctx_type)) return 0;
}
#endif // CCTX_C2_DROPPED
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_U) {
xd->eob_u_flag = eob ? 1 : 0;
}
if (plane == AOM_PLANE_Y || plane == AOM_PLANE_U) {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const int pred_mode_ctx =
(is_inter || xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
aom_write_symbol(w, eob == 0,
ec_ctx->txb_skip_cdf[pred_mode_ctx][txs_ctx][txb_skip_ctx],
2);
#else
aom_write_symbol(w, eob == 0, ec_ctx->txb_skip_cdf[txs_ctx][txb_skip_ctx],
2);
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
} else {
aom_write_symbol(w, eob == 0, ec_ctx->v_txb_skip_cdf[txb_skip_ctx], 2);
}
#else
aom_write_symbol(w, eob == 0, ec_ctx->txb_skip_cdf[txs_ctx][txb_skip_ctx], 2);
#endif // CONFIG_CONTEXT_DERIVATION
if (eob == 0) return 0;
int esc_eob = is_fsc ? bob_code : eob;
const int dc_skip = (eob == 1) && !is_inter;
code_eob(x, w, plane, tx_size, esc_eob);
av1_write_tx_type(cm, xd, tx_type, tx_size, w, plane, esc_eob, dc_skip);
if (plane == AOM_PLANE_U && is_cctx_allowed(cm, xd)) {
const int skip_cctx = is_inter ? 0 : (eob == 1);
CctxType cctx_type = av1_get_cctx_type(xd, blk_row, blk_col);
if (eob > 0 && !skip_cctx)
av1_write_cctx_type(cm, xd, cctx_type, tx_size, w);
}
return 1;
}
void av1_write_coeffs_txb_skip(const AV1_COMMON *const cm, MACROBLOCK *const x,
aom_writer *w, int blk_row, int blk_col,
int plane, int block, TX_SIZE tx_size) {
MACROBLOCKD *xd = &x->e_mbd;
const CB_COEFF_BUFFER *cb_coef_buff = x->cb_coef_buff;
const uint16_t eob = av1_get_max_eob(tx_size);
FRAME_CONTEXT *ec_ctx = xd->tile_ctx;
const PLANE_TYPE plane_type = get_plane_type(plane);
const TX_TYPE tx_type =
av1_get_tx_type(xd, plane_type, blk_row, blk_col, tx_size,
cm->features.reduced_tx_set_used);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
uint8_t levels_buf[TX_PAD_2D];
int8_t signs_buf[TX_PAD_2D];
const tran_low_t *tcoeff_txb =
cb_coef_buff->tcoeff[plane] + x->mbmi_ext_frame->cb_offset[plane];
const tran_low_t *tcoeff = tcoeff_txb + BLOCK_OFFSET(block);
av1_txb_init_levels_signs(tcoeff, width, height, levels_buf, signs_buf);
uint8_t *const levels = set_levels(levels_buf, width);
int8_t *const signs = set_signs(signs_buf, width);
const SCAN_ORDER *const scan_order = get_scan(tx_size, tx_type);
const int16_t *const scan = scan_order->scan;
const int bwl = get_txb_bwl(tx_size);
DECLARE_ALIGNED(16, int8_t, coeff_contexts[MAX_TX_SQUARE]);
const int txb_offset =
x->mbmi_ext_frame->cb_offset[plane] / (TX_SIZE_W_MIN * TX_SIZE_H_MIN);
const uint16_t *bob_txb = cb_coef_buff->bobs[plane] + txb_offset;
const int bob_code = bob_txb[block];
int bob = av1_get_max_eob(tx_size) - bob_code;
av1_get_nz_map_contexts_skip_c(levels, scan, bob, eob, tx_size,
coeff_contexts);
for (int c = bob; c < eob; ++c) {
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = tcoeff[pos];
const tran_low_t level = abs(v);
if (c == bob) {
aom_write_symbol(w, AOMMIN(level, 3) - 1,
ec_ctx->coeff_base_bob_cdf[coeff_ctx], 3);
} else {
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf_idtx[coeff_ctx], 4);
}
if (level > NUM_BASE_LEVELS) {
// level is above 1.
const int br_ctx = get_br_ctx_skip(levels, pos, bwl);
aom_cdf_prob *cdf = ec_ctx->coeff_br_cdf_idtx[br_ctx];
#if NEWHR
int br = get_low_range(level, 0);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(br, BR_CDF_SIZE - 1);
br -= k;
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#else
const int base_range = level - 1 - NUM_BASE_LEVELS;
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#endif
}
}
// Loop to code all signs, bypass levels in the transform block
for (int c = eob - 1; c >= 0; --c) {
const int pos = scan[c];
const tran_low_t v = tcoeff[pos];
const tran_low_t level = abs(v);
const int sign = (v < 0) ? 1 : 0;
if (level) {
int idtx_sign_ctx = get_sign_ctx_skip(signs, levels, pos, bwl);
aom_write_symbol(w, sign, ec_ctx->idtx_sign_cdf[idtx_sign_ctx], 2);
#if NEWHR
if (level > COEFF_BASE_RANGE + NUM_BASE_LEVELS - 1)
write_golomb(w, get_high_range(level, 0));
#else
if (level > COEFF_BASE_RANGE + NUM_BASE_LEVELS)
write_golomb(w, level - COEFF_BASE_RANGE - 1 - NUM_BASE_LEVELS);
#endif
}
}
}
static INLINE void write_coeff_hidden(aom_writer *w, TX_CLASS tx_class,
const int16_t *scan, int bwl,
uint8_t *levels, const int level,
base_cdf_arr base_cdf_ph,
br_cdf_arr br_cdf_ph) {
const int q_index = (level >> 1);
const int pos = scan[0];
int ctx_id = get_base_ctx_ph(levels, pos, bwl, tx_class);
aom_write_symbol(w, AOMMIN(q_index, 3), base_cdf_ph[ctx_id], 4);
if (q_index > NUM_BASE_LEVELS) {
ctx_id = get_par_br_ctx(levels, pos, bwl, tx_class);
aom_cdf_prob *cdf_br = br_cdf_ph[ctx_id];
const int base_range = q_index - 1 - NUM_BASE_LEVELS;
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
aom_write_symbol(w, k, cdf_br, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
}
}
void av1_write_coeffs_txb(const AV1_COMMON *const cm, MACROBLOCK *const x,
aom_writer *w, int blk_row, int blk_col, int plane,
int block, TX_SIZE tx_size) {
MACROBLOCKD *xd = &x->e_mbd;
const CB_COEFF_BUFFER *cb_coef_buff = x->cb_coef_buff;
#if CONFIG_CONTEXT_DERIVATION
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
if (plane == AOM_PLANE_U)
memset(xd->tmp_sign, 0, width * height * sizeof(int32_t));
#endif // CONFIG_CONTEXT_DERIVATION
const int txb_offset =
x->mbmi_ext_frame->cb_offset[plane] / (TX_SIZE_W_MIN * TX_SIZE_H_MIN);
const uint16_t *eob_txb = cb_coef_buff->eobs[plane] + txb_offset;
const uint16_t eob = eob_txb[block];
const uint8_t *entropy_ctx = cb_coef_buff->entropy_ctx[plane] + txb_offset;
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
FRAME_CONTEXT *ec_ctx = xd->tile_ctx;
#if CONFIG_LR_IMPROVEMENTS
if (!is_global_intrabc_allowed(cm) && !cm->features.coded_lossless) {
// Assert only when LR is enabled.
assert((eob == 0) == av1_get_txk_skip(cm, xd->mi_row, xd->mi_col, plane,
blk_row, blk_col));
}
#endif // CONFIG_LR_IMPROVEMENTS
if (eob == 0) return;
const PLANE_TYPE plane_type = get_plane_type(plane);
const TX_TYPE tx_type =
av1_get_tx_type(xd, plane_type, blk_row, blk_col, tx_size,
cm->features.reduced_tx_set_used);
#if DEBUG_EXTQUANT
fprintf(cm->fEncCoeffLog, "\nblk_row=%d,blk_col=%d,plane=%d,tx_size=%d",
blk_row, blk_col, plane, tx_size);
#endif
const TX_CLASS tx_class = tx_type_to_class[get_primary_tx_type(tx_type)];
// write sec_tx_type here
// Only y plane's sec_tx_type is transmitted
if ((plane == AOM_PLANE_Y) && (cm->seq_params.enable_ist) && eob != 1) {
av1_write_sec_tx_type(cm, xd, tx_type, tx_size, eob, w);
}
#if DEBUG_EXTQUANT
fprintf(cm->fEncCoeffLog, "tx_type=%d, eob=%d\n", tx_type, eob);
#endif
#if !CONFIG_CONTEXT_DERIVATION
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
#endif // CONFIG_CONTEXT_DERIVATION
uint8_t levels_buf[TX_PAD_2D];
uint8_t *const levels = set_levels(levels_buf, width);
const tran_low_t *tcoeff_txb =
cb_coef_buff->tcoeff[plane] + x->mbmi_ext_frame->cb_offset[plane];
const tran_low_t *tcoeff = tcoeff_txb + BLOCK_OFFSET(block);
av1_txb_init_levels(tcoeff, width, height, levels);
const SCAN_ORDER *const scan_order = get_scan(tx_size, tx_type);
const int16_t *const scan = scan_order->scan;
DECLARE_ALIGNED(16, int8_t, coeff_contexts[MAX_TX_SQUARE]);
av1_get_nz_map_contexts(levels, scan, eob, tx_size, tx_class, coeff_contexts,
plane);
const int bwl = get_txb_bwl(tx_size);
bool enable_parity_hiding = cm->features.allow_parity_hiding &&
!xd->lossless[xd->mi[0]->segment_id] &&
plane == PLANE_TYPE_Y &&
get_primary_tx_type(tx_type) < IDTX;
#if NEWCTX
int state = 0;
#endif
for (int c = eob - 1; c > 0; --c) {
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = tcoeff[pos];
const tran_low_t level = abs(v);
int dbg = 0;
if (c == eob - 1) {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1,
ec_ctx->coeff_base_lf_eob_cdf[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS - 1);
} else {
aom_write_symbol(
w, AOMMIN(level, 3) - 1,
ec_ctx->coeff_base_eob_cdf[txs_ctx][plane_type][coeff_ctx], 3);
}
} else {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
#if NEWCTX
if (state >> 1) // Q1
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1),
ec_ctx->coeff_base_lf_cdf_tcq[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS);
else
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1),
ec_ctx->coeff_base_lf_cdf[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS);
#else
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1),
ec_ctx->coeff_base_lf_cdf[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS);
#endif
} else {
#if NEWCTX
if (state >> 1) // Q1
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf_tcq[txs_ctx][plane_type][coeff_ctx],
4);
else
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf[txs_ctx][plane_type][coeff_ctx],
4);
#else
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf[txs_ctx][plane_type][coeff_ctx],
4);
#endif
}
}
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int br_ctx = get_br_lf_ctx(levels, pos, bwl, tx_class);
aom_cdf_prob *cdf = ec_ctx->coeff_br_lf_cdf[plane_type][br_ctx];
#if NEWHR
int br = get_low_range(level, 1);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(br, BR_CDF_SIZE - 1);
br -= k;
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#else
const int base_range =
level - 1 - LF_NUM_BASE_LEVELS; // level is above 1.
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#endif
}
} else {
if (level > NUM_BASE_LEVELS) {
const int br_ctx = get_br_ctx(levels, pos, bwl, tx_class);
aom_cdf_prob *cdf = ec_ctx->coeff_br_cdf[plane_type][br_ctx];
#if NEWHR
int br = get_low_range(level, 0);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(br, BR_CDF_SIZE - 1);
br -= k;
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#else
const int base_range = level - 1 - NUM_BASE_LEVELS;
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#endif
}
}
#if NEWCTX
#if DQENABLE
if (dq_enable(tx_size, plane))
#endif //DQENABLE
if (dbg) {
printf("ENC state %d level %d %d\n", state, level, limits);
}
state = tcq_next_state(state, level, limits);
if (dbg) {
printf("ENC new state %d\n", state);
}
#endif
}
int num_nz = 0;
bool is_hidden = false;
if (enable_parity_hiding) {
for (int c = eob - 1; c > 0; --c) {
const int pos = scan[c];
num_nz += !!tcoeff[pos];
}
is_hidden = num_nz >= PHTHRESH;
}
if (is_hidden) {
const int pos = scan[0];
const tran_low_t v = tcoeff[pos];
const tran_low_t level = abs(v);
write_coeff_hidden(w, tx_class, scan, bwl, levels, level,
ec_ctx->coeff_base_ph_cdf, ec_ctx->coeff_br_ph_cdf);
} else {
const int c = 0;
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = tcoeff[pos];
const tran_low_t level = abs(v);
if (c == eob - 1) {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1,
ec_ctx->coeff_base_lf_eob_cdf[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS - 1);
} else {
aom_write_symbol(
w, AOMMIN(level, 3) - 1,
ec_ctx->coeff_base_eob_cdf[txs_ctx][plane_type][coeff_ctx], 3);
}
} else {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
#if NEWCTX
if (state >> 1) // Q1
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1),
ec_ctx->coeff_base_lf_cdf_tcq[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS);
else
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1),
ec_ctx->coeff_base_lf_cdf[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS);
#else
aom_write_symbol(
w, AOMMIN(level, LF_BASE_SYMBOLS - 1),
ec_ctx->coeff_base_lf_cdf[txs_ctx][plane_type][coeff_ctx],
LF_BASE_SYMBOLS);
#endif
} else {
#if NEWCTX
if (state >> 1) // Q1
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf_tcq[txs_ctx][plane_type][coeff_ctx],
4);
else
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf[txs_ctx][plane_type][coeff_ctx],
4);
#else
aom_write_symbol(w, AOMMIN(level, 3),
ec_ctx->coeff_base_cdf[txs_ctx][plane_type][coeff_ctx],
4);
#endif
}
}
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int br_ctx = get_br_lf_ctx(levels, pos, bwl, tx_class);
aom_cdf_prob *cdf = ec_ctx->coeff_br_lf_cdf[plane_type][br_ctx];
#if NEWHR
int br = get_low_range(level, 1);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(br, BR_CDF_SIZE - 1);
br -= k;
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#else
const int base_range = level - 1 - LF_NUM_BASE_LEVELS;
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#endif
}
} else {
if (level > NUM_BASE_LEVELS) {
const int br_ctx = get_br_ctx(levels, pos, bwl, tx_class);
aom_cdf_prob *cdf = ec_ctx->coeff_br_cdf[plane_type][br_ctx];
#if NEWHR
int br = get_low_range(level, 0);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(br, BR_CDF_SIZE - 1);
br -= k;
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#else
const int base_range = level - 1 - NUM_BASE_LEVELS;
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
aom_write_symbol(w, k, cdf, BR_CDF_SIZE);
if (k < BR_CDF_SIZE - 1) break;
}
#endif
}
}
}
#if DEBUG_EXTQUANT
for (int c = 0; c < eob; ++c) {
const tran_low_t v = tcoeff[scan[c]];
const tran_low_t level = abs(v);
printf("ENC blk c=%d,pos=%d,level=%d,dq_coeff=%d\n", c, scan[c], level, v);
}
#endif
// Loop to code all signs in the transform block,
// starting with the sign of DC (if applicable)
for (int c = 0; c < eob; ++c) {
const tran_low_t v = tcoeff[scan[c]];
const tran_low_t level = abs(v);
const int sign = (v < 0) ? 1 : 0;
if (level) {
if (c == 0) {
const int dc_sign_ctx =
(entropy_ctx[block] >> DC_SIGN_CTX_SHIFT) & DC_SIGN_CTX_MASK;
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_U) xd->tmp_sign[0] = (sign ? 2 : 1);
if (plane == AOM_PLANE_V) {
aom_write_symbol(
w, sign, ec_ctx->v_dc_sign_cdf[xd->tmp_sign[0]][dc_sign_ctx], 2);
} else {
aom_write_symbol(w, sign,
ec_ctx->dc_sign_cdf[plane_type][dc_sign_ctx], 2);
}
#else
aom_write_symbol(w, sign, ec_ctx->dc_sign_cdf[plane_type][dc_sign_ctx],
2);
#endif // CONFIG_CONTEXT_DERIVATION
} else {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_U) xd->tmp_sign[scan[c]] = (sign ? 2 : 1);
if (plane == AOM_PLANE_Y || plane == AOM_PLANE_U)
aom_write_bit(w, sign);
else
aom_write_symbol(w, sign,
ec_ctx->v_ac_sign_cdf[xd->tmp_sign[scan[c]]], 2);
#else
aom_write_bit(w, sign);
#endif // CONFIG_CONTEXT_DERIVATION
}
if (is_hidden && c == 0) {
int q_index = level >> 1;
if (q_index > COEFF_BASE_RANGE + NUM_BASE_LEVELS)
write_golomb(w, q_index - COEFF_BASE_RANGE - 1 - NUM_BASE_LEVELS);
} else {
const int pos = scan[c];
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
#if NEWHR
if (limits) {
if (level > COEFF_BASE_RANGE + LF_NUM_BASE_LEVELS - 1)
write_golomb(w, get_high_range(level, 1));
} else {
if (level > COEFF_BASE_RANGE + NUM_BASE_LEVELS - 1)
write_golomb(w, get_high_range(level, 0));
}
#else
if (limits) {
if (level > COEFF_BASE_RANGE + LF_NUM_BASE_LEVELS)
write_golomb(w, level - COEFF_BASE_RANGE - 1 - LF_NUM_BASE_LEVELS);
} else {
if (level > COEFF_BASE_RANGE + NUM_BASE_LEVELS)
write_golomb(w, level - COEFF_BASE_RANGE - 1 - NUM_BASE_LEVELS);
}
#endif
}
}
}
}
typedef struct encode_txb_args {
const AV1_COMMON *cm;
MACROBLOCK *x;
aom_writer *w;
} ENCODE_TXB_ARGS;
void av1_write_intra_coeffs_mb(const AV1_COMMON *const cm, MACROBLOCK *x,
aom_writer *w, BLOCK_SIZE bsize) {
MACROBLOCKD *xd = &x->e_mbd;
const MB_MODE_INFO *const mbmi = xd->mi[0];
const int is_inter = is_inter_block(mbmi, xd->tree_type);
int block[MAX_MB_PLANE] = { 0 };
int row, col;
assert(bsize == get_plane_block_size(bsize, xd->plane[0].subsampling_x,
xd->plane[0].subsampling_y));
const int max_blocks_wide = max_block_wide(xd, bsize, 0);
const int max_blocks_high = max_block_high(xd, bsize, 0);
const BLOCK_SIZE max_unit_bsize = BLOCK_64X64;
int mu_blocks_wide = mi_size_wide[max_unit_bsize];
int mu_blocks_high = mi_size_high[max_unit_bsize];
mu_blocks_wide = AOMMIN(max_blocks_wide, mu_blocks_wide);
mu_blocks_high = AOMMIN(max_blocks_high, mu_blocks_high);
for (row = 0; row < max_blocks_high; row += mu_blocks_high) {
for (col = 0; col < max_blocks_wide; col += mu_blocks_wide) {
const int plane_start = get_partition_plane_start(xd->tree_type);
const int plane_end =
get_partition_plane_end(xd->tree_type, av1_num_planes(cm));
for (int plane = plane_start; plane < plane_end; ++plane) {
if (plane && !xd->is_chroma_ref) break;
if (plane == AOM_PLANE_U && is_cctx_allowed(cm, xd)) continue;
const TX_SIZE tx_size = av1_get_tx_size(plane, xd);
const int stepr = tx_size_high_unit[tx_size];
const int stepc = tx_size_wide_unit[tx_size];
const int step = stepr * stepc;
const struct macroblockd_plane *const pd = &xd->plane[plane];
const int ss_x = pd->subsampling_x;
const int ss_y = pd->subsampling_y;
const BLOCK_SIZE plane_bsize =
get_mb_plane_block_size(xd, mbmi, plane, ss_x, ss_y);
const int plane_unit_height =
get_plane_tx_unit_height(xd, plane_bsize, plane, row, ss_y);
const int plane_unit_width =
get_plane_tx_unit_width(xd, plane_bsize, plane, col, ss_x);
for (int blk_row = row >> ss_y; blk_row < plane_unit_height;
blk_row += stepr) {
for (int blk_col = col >> ss_x; blk_col < plane_unit_width;
blk_col += stepc) {
// Loop order for the two chroma planes is changed for CCTX
// because the transform information for both planes are needed at
// once at the decoder side.
if (plane == AOM_PLANE_V && is_cctx_allowed(cm, xd)) {
const int code_rest =
av1_write_sig_txtype(cm, x, w, blk_row, blk_col, AOM_PLANE_U,
block[AOM_PLANE_U], tx_size);
if (code_rest)
av1_write_coeffs_txb(cm, x, w, blk_row, blk_col, AOM_PLANE_U,
block[AOM_PLANE_U], tx_size);
block[AOM_PLANE_U] += step;
}
const int code_rest = av1_write_sig_txtype(
cm, x, w, blk_row, blk_col, plane, block[plane], tx_size);
const TX_TYPE tx_type =
av1_get_tx_type(xd, get_plane_type(plane), blk_row, blk_col,
tx_size, cm->features.reduced_tx_set_used);
if (code_rest) {
if ((mbmi->fsc_mode[xd->tree_type == CHROMA_PART] &&
get_primary_tx_type(tx_type) == IDTX &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter)) {
av1_write_coeffs_txb_skip(cm, x, w, blk_row, blk_col, plane,
block[plane], tx_size);
} else {
av1_write_coeffs_txb(cm, x, w, blk_row, blk_col, plane,
block[plane], tx_size);
}
}
block[plane] += step;
}
}
}
}
}
}
int get_cctx_type_cost(const AV1_COMMON *cm, const MACROBLOCK *x,
const MACROBLOCKD *xd, int plane, TX_SIZE tx_size,
int block, CctxType cctx_type) {
const int skip_cctx = is_inter_block(xd->mi[0], xd->tree_type)
? 0
: (x->plane[plane].eobs[block] == 1);
if (plane == AOM_PLANE_U && x->plane[plane].eobs[block] &&
is_cctx_allowed(cm, xd) && !skip_cctx) {
const TX_SIZE square_tx_size = txsize_sqr_map[tx_size];
int above_cctx, left_cctx;
#if CONFIG_EXT_RECUR_PARTITIONS
get_above_and_left_cctx_type(cm, xd, &above_cctx, &left_cctx);
#else
get_above_and_left_cctx_type(cm, xd, tx_size, &above_cctx, &left_cctx);
#endif // CONFIG_EXT_RECUR_PARTITIONS
const int cctx_ctx = get_cctx_context(xd, &above_cctx, &left_cctx);
return x->mode_costs.cctx_type_cost[square_tx_size][cctx_ctx][cctx_type];
} else {
return 0;
}
}
// This function gets the estimated bit cost for a 'secondary tx set'
static int get_sec_tx_set_cost(const MACROBLOCK *x, const MB_MODE_INFO *mbmi,
TX_TYPE tx_type) {
uint8_t stx_set_flag = get_secondary_tx_set(tx_type);
if (get_primary_tx_type(tx_type) == ADST_ADST) stx_set_flag -= IST_DIR_SIZE;
assert(stx_set_flag < IST_DIR_SIZE);
uint8_t intra_mode = mbmi->mode;
uint8_t stx_set_ctx = stx_transpose_mapping[intra_mode];
assert(stx_set_ctx < IST_DIR_SIZE);
return x->mode_costs.stx_set_flag_cost[stx_set_ctx][stx_set_flag];
}
// TODO(angiebird): use this function whenever it's possible
static int get_tx_type_cost(const MACROBLOCK *x, const MACROBLOCKD *xd,
int plane, TX_SIZE tx_size, TX_TYPE tx_type,
int reduced_tx_set_used, int eob, int bob_code,
int is_fsc) {
if (plane > 0) return 0;
const TX_SIZE square_tx_size = txsize_sqr_map[tx_size];
const MB_MODE_INFO *mbmi = xd->mi[0];
if (mbmi->fsc_mode[xd->tree_type == CHROMA_PART] &&
!is_inter_block(mbmi, xd->tree_type) && plane == PLANE_TYPE_Y) {
return 0;
}
const int is_inter = is_inter_block(mbmi, xd->tree_type);
if (get_ext_tx_types(tx_size, is_inter, reduced_tx_set_used) > 1 &&
!xd->lossless[xd->mi[0]->segment_id]) {
const int ext_tx_set =
get_ext_tx_set(tx_size, is_inter, reduced_tx_set_used);
if (is_inter) {
if (ext_tx_set > 0) {
const int esc_eob = is_fsc ? bob_code : eob;
const int eob_tx_ctx =
get_lp2tx_ctx(tx_size, get_txb_bwl(tx_size), esc_eob);
return x->mode_costs.inter_tx_type_costs[ext_tx_set][eob_tx_ctx]
[square_tx_size][tx_type];
}
} else {
if (ext_tx_set > 0) {
PREDICTION_MODE intra_dir;
if (mbmi->filter_intra_mode_info.use_filter_intra)
intra_dir = fimode_to_intradir[mbmi->filter_intra_mode_info
.filter_intra_mode];
else
intra_dir = mbmi->mode;
TX_TYPE primary_tx_type = get_primary_tx_type(tx_type);
int tx_type_cost = 0;
if (eob != 1) {
tx_type_cost +=
x->mode_costs.intra_tx_type_costs[ext_tx_set][square_tx_size]
[intra_dir][primary_tx_type];
} else {
return tx_type_cost;
}
if (block_signals_sec_tx_type(xd, tx_size, tx_type, eob) &&
xd->enable_ist) {
tx_type_cost +=
x->mode_costs.stx_flag_cost[square_tx_size]
[get_secondary_tx_type(tx_type)];
#if CONFIG_IST_SET_FLAG
if (get_secondary_tx_type(tx_type) > 0)
tx_type_cost += get_sec_tx_set_cost(x, mbmi, tx_type);
#endif // CONFIG_IST_SET_FLAG
}
return tx_type_cost;
}
}
} else if (!is_inter && !xd->lossless[xd->mi[0]->segment_id]) {
if (block_signals_sec_tx_type(xd, tx_size, tx_type, eob) &&
xd->enable_ist) {
int tx_type_cost =
x->mode_costs
.stx_flag_cost[square_tx_size][get_secondary_tx_type(tx_type)];
#if CONFIG_IST_SET_FLAG
if (get_secondary_tx_type(tx_type) > 0)
tx_type_cost += get_sec_tx_set_cost(x, mbmi, tx_type);
#endif // CONFIG_IST_SET_FLAG
return tx_type_cost;
}
}
return 0;
}
static INLINE void update_coeff_eob_fast(int *eob, int shift,
const int32_t *dequant_ptr,
const int16_t *scan,
const tran_low_t *coeff_ptr,
tran_low_t *qcoeff_ptr,
tran_low_t *dqcoeff_ptr) {
// TODO(sarahparker) make this work for aomqm
int eob_out = *eob;
int zbin[2] = { dequant_ptr[0] + ROUND_POWER_OF_TWO(dequant_ptr[0] * 70,
7 + QUANT_TABLE_BITS),
dequant_ptr[1] + ROUND_POWER_OF_TWO(dequant_ptr[1] * 70,
7 + QUANT_TABLE_BITS) };
for (int i = *eob - 1; i >= 0; i--) {
const int rc = scan[i];
const int qcoeff = qcoeff_ptr[rc];
const int coeff = coeff_ptr[rc];
const int coeff_sign = AOMSIGN(coeff);
int64_t abs_coeff = (coeff ^ coeff_sign) - coeff_sign;
if (((abs_coeff << (1 + shift)) < zbin[rc != 0]) || (qcoeff == 0)) {
eob_out--;
qcoeff_ptr[rc] = 0;
dqcoeff_ptr[rc] = 0;
} else {
break;
}
}
*eob = eob_out;
}
static AOM_FORCE_INLINE int warehouse_efficients_txb_skip(
const AV1_COMMON *cm, const MACROBLOCK *x, const int plane, const int block,
const TX_SIZE tx_size, const TXB_CTX *const txb_ctx,
const struct macroblock_plane *p, const int eob,
const LV_MAP_COEFF_COST *const coeff_costs, const MACROBLOCKD *const xd,
const TX_TYPE tx_type, const CctxType cctx_type, int reduced_tx_set_used) {
const tran_low_t *const qcoeff = p->qcoeff + BLOCK_OFFSET(block);
const int txb_skip_ctx = txb_ctx->txb_skip_ctx;
const int bwl = get_txb_bwl(tx_size);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
const SCAN_ORDER *const scan_order = get_scan(tx_size, tx_type);
const int16_t *const scan = scan_order->scan;
uint8_t levels_buf[TX_PAD_2D];
uint8_t *const levels = set_levels(levels_buf, width);
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const MB_MODE_INFO *mbmi = xd->mi[0];
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int pred_mode_ctx =
(is_inter || xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
int cost = coeff_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][0];
#else
int cost = coeff_costs->txb_skip_cost[txb_skip_ctx][0];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
int8_t signs_buf[TX_PAD_2D];
int8_t *const signs = set_signs(signs_buf, width);
av1_txb_init_levels_signs(qcoeff, width, height, levels_buf, signs_buf);
const int bob_code = p->bobs[block];
const int bob = av1_get_max_eob(tx_size) - bob_code;
#if !CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const int is_inter = is_inter_block(xd->mi[0], xd->tree_type);
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const int is_fsc = (xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter);
cost += get_tx_type_cost(x, xd, plane, tx_size, tx_type, reduced_tx_set_used,
eob, bob_code, is_fsc);
const int eob_multi_size = txsize_log2_minus4[tx_size];
const LV_MAP_EOB_COST *const eob_costs =
&x->coeff_costs.eob_costs[eob_multi_size][PLANE_TYPE_Y];
cost += get_eob_cost(bob_code, eob_costs, coeff_costs
#if CONFIG_EOB_POS_LUMA
,
is_inter
#endif // CONFIG_EOB_POS_LUMA
);
cost += get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
DECLARE_ALIGNED(16, int8_t, coeff_contexts[MAX_TX_SQUARE]);
av1_get_nz_map_contexts_skip_c(levels, scan, bob, eob, tx_size,
coeff_contexts);
const int(*lps_cost)[COEFF_BASE_RANGE + 1 + COEFF_BASE_RANGE + 1] =
coeff_costs->lps_cost_skip;
const int(*base_cost)[8] = coeff_costs->idtx_base_cost;
for (int c = bob; c < eob; c++) {
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = qcoeff[pos];
const int level = abs(v);
if (c == bob) {
cost += coeff_costs->base_bob_cost[coeff_ctx][AOMMIN(level, 3) - 1];
} else {
cost += base_cost[coeff_ctx][AOMMIN(level, 3)];
}
if (v) {
if (level > NUM_BASE_LEVELS) {
const int ctx = get_br_ctx_skip(levels, pos, bwl);
cost += get_br_cost(level, lps_cost[ctx]);
}
}
}
for (int c = eob - 1; c >= bob; --c) {
const int pos = scan[c];
const tran_low_t v = qcoeff[pos];
const tran_low_t level = abs(v);
const int sign = (v < 0) ? 1 : 0;
if (level) {
int idtx_sign_ctx = get_sign_ctx_skip(signs, levels, pos, bwl);
cost += coeff_costs->idtx_sign_cost[idtx_sign_ctx][sign];
}
}
return cost;
}
static AOM_FORCE_INLINE int warehouse_efficients_txb(
const AV1_COMMON *cm, const MACROBLOCK *x, const int plane, const int block,
const TX_SIZE tx_size, const TXB_CTX *const txb_ctx,
const struct macroblock_plane *p, const int eob,
const PLANE_TYPE plane_type, const LV_MAP_COEFF_COST *const coeff_costs,
const MACROBLOCKD *const xd, const TX_TYPE tx_type,
const CctxType cctx_type, const TX_CLASS tx_class, int reduced_tx_set_used,
bool enable_parity_hiding, const LV_MAP_COEFF_COST *const coeff_costs_ph) {
const tran_low_t *const qcoeff = p->qcoeff + BLOCK_OFFSET(block);
#if CONFIG_CONTEXT_DERIVATION
const struct macroblock_plane *pu = &x->plane[AOM_PLANE_U];
int txb_skip_ctx = txb_ctx->txb_skip_ctx;
if (plane == AOM_PLANE_V) {
txb_skip_ctx += (pu->eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
}
#else
const int txb_skip_ctx = txb_ctx->txb_skip_ctx;
#endif // CONFIG_CONTEXT_DERIVATION
const int bwl = get_txb_bwl(tx_size);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
const SCAN_ORDER *const scan_order = get_scan(tx_size, tx_type);
const int16_t *const scan = scan_order->scan;
uint8_t levels_buf[TX_PAD_2D];
uint8_t *const levels = set_levels(levels_buf, width);
DECLARE_ALIGNED(16, int8_t, coeff_contexts[MAX_TX_SQUARE]);
const int eob_multi_size = txsize_log2_minus4[tx_size];
const LV_MAP_EOB_COST *const eob_costs =
&x->coeff_costs.eob_costs[eob_multi_size][plane_type];
#if CONFIG_CONTEXT_DERIVATION
int cost;
if (plane == AOM_PLANE_V) {
cost = coeff_costs->v_txb_skip_cost[txb_skip_ctx][0];
} else {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const MB_MODE_INFO *mbmi = xd->mi[0];
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
cost = coeff_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][0];
#else
cost = coeff_costs->txb_skip_cost[txb_skip_ctx][0];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
}
#else
int cost = coeff_costs->txb_skip_cost[txb_skip_ctx][0];
#endif // CONFIG_CONTEXT_DERIVATION
av1_txb_init_levels(qcoeff, width, height, levels);
const int bob_code = p->bobs[block];
const int is_inter = is_inter_block(xd->mi[0], xd->tree_type);
const int is_fsc = (xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter);
cost += get_tx_type_cost(x, xd, plane, tx_size, tx_type, reduced_tx_set_used,
eob, bob_code, is_fsc);
cost += get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
cost += get_eob_cost(eob, eob_costs, coeff_costs
#if CONFIG_EOB_POS_LUMA
,
is_inter
#endif // CONFIG_EOB_POS_LUMA
);
av1_get_nz_map_contexts(levels, scan, eob, tx_size, tx_class, coeff_contexts,
plane);
const int(*lps_cost)[COEFF_BASE_RANGE + 1 + COEFF_BASE_RANGE + 1] =
coeff_costs->lps_cost;
const int(*lps_lf_cost)[COEFF_BASE_RANGE + 1 + COEFF_BASE_RANGE + 1] =
coeff_costs->lps_lf_cost;
int c = eob - 1;
{
const int pos = scan[c];
const tran_low_t v = qcoeff[pos];
const int sign = AOMSIGN(v);
const int level = (v ^ sign) - sign;
const int coeff_ctx = coeff_contexts[pos];
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost +=
coeff_costs->base_lf_eob_cost[coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1];
} else {
cost += coeff_costs->base_eob_cost[coeff_ctx][AOMMIN(level, 3) - 1];
}
if (v) {
// sign bit cost
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int ctx = get_br_ctx_lf_eob(pos, tx_class);
cost += get_br_lf_cost(level, lps_lf_cost[ctx]);
}
} else {
if (level > NUM_BASE_LEVELS) {
const int ctx = 7; /* get_lf_ctx_eob */
cost += get_br_cost(level, lps_cost[ctx]);
}
}
if (c) {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V) {
const int sign01 = (sign ^ sign) - sign;
cost += coeff_costs->v_ac_sign_cost[xd->tmp_sign[pos]][sign01];
} else {
cost += av1_cost_literal(1);
}
#else
cost += av1_cost_literal(1);
#endif // CONFIG_CONTEXT_DERIVATION
} else {
const int sign01 = (sign ^ sign) - sign;
const int dc_sign_ctx = txb_ctx->dc_sign_ctx;
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V) {
cost +=
coeff_costs->v_dc_sign_cost[xd->tmp_sign[0]][dc_sign_ctx][sign01];
} else {
cost += coeff_costs->dc_sign_cost[dc_sign_ctx][sign01];
}
#else
cost += coeff_costs->dc_sign_cost[dc_sign_ctx][sign01];
#endif // CONFIG_CONTEXT_DERIVATION
return cost;
}
}
}
const int(*base_lf_cost)[LF_BASE_SYMBOLS * 2] = coeff_costs->base_lf_cost;
const int(*base_cost)[8] = coeff_costs->base_cost;
for (c = eob - 2; c >= 1; --c) {
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = qcoeff[pos];
const int level = abs(v);
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost += base_lf_cost[coeff_ctx][AOMMIN(level, LF_BASE_SYMBOLS - 1)];
} else {
cost += base_cost[coeff_ctx][AOMMIN(level, 3)];
}
if (v) {
// sign bit cost
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V) {
const int sign = AOMSIGN(v);
const int sign01 = (sign ^ sign) - sign;
cost += coeff_costs->v_ac_sign_cost[xd->tmp_sign[pos]][sign01];
} else {
cost += av1_cost_literal(1);
}
#else
cost += av1_cost_literal(1);
#endif // CONFIG_CONTEXT_DERIVATION
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int ctx = get_br_lf_ctx(levels, pos, bwl, tx_class);
cost += get_br_lf_cost(level, lps_lf_cost[ctx]);
}
} else {
if (level > NUM_BASE_LEVELS) {
const int ctx = get_br_ctx(levels, pos, bwl, tx_class);
cost += get_br_cost(level, lps_cost[ctx]);
}
}
}
}
// c == 0 after previous loop
int num_nz = 0;
for (c = eob - 1; c > 0; --c) {
const int pos = scan[c];
num_nz += !!qcoeff[pos];
}
c = 0;
if (num_nz >= PHTHRESH && enable_parity_hiding) {
const int pos = scan[c];
const tran_low_t v = qcoeff[pos];
const int level = abs(v);
const int q_index = level >> 1;
cost += coeff_costs_ph->base_ph_cost[get_base_ctx_ph(
levels, pos, bwl, tx_class)][AOMMIN(q_index, 3)];
if (v) {
const int dc_sign_ctx = txb_ctx->dc_sign_ctx;
cost += coeff_costs->dc_sign_cost[dc_sign_ctx][v < 0];
if (q_index > NUM_BASE_LEVELS) {
const int ctx = get_par_br_ctx(levels, pos, bwl, tx_class);
cost += get_br_cost(q_index, coeff_costs_ph->lps_ph_cost[ctx]);
}
}
return cost;
}
{
const int pos = scan[c];
const tran_low_t v = qcoeff[pos];
const int coeff_ctx = coeff_contexts[pos];
const int sign = AOMSIGN(v);
const int level = (v ^ sign) - sign;
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost += base_lf_cost[coeff_ctx][AOMMIN(level, LF_BASE_SYMBOLS - 1)];
} else {
cost += base_cost[coeff_ctx][AOMMIN(level, 3)];
}
if (v) {
// sign bit cost
const int sign01 = (sign ^ sign) - sign;
const int dc_sign_ctx = txb_ctx->dc_sign_ctx;
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V) {
cost +=
coeff_costs->v_dc_sign_cost[xd->tmp_sign[0]][dc_sign_ctx][sign01];
} else {
cost += coeff_costs->dc_sign_cost[dc_sign_ctx][sign01];
}
#else
cost += coeff_costs->dc_sign_cost[dc_sign_ctx][sign01];
#endif // CONFIG_CONTEXT_DERIVATION
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int ctx = get_br_lf_ctx(levels, pos, bwl, tx_class);
cost += get_br_lf_cost(level, lps_lf_cost[ctx]);
}
} else {
if (level > NUM_BASE_LEVELS) {
const int ctx = get_br_ctx(levels, pos, bwl, tx_class);
cost += get_br_cost(level, lps_cost[ctx]);
}
}
}
}
return cost;
}
static AOM_FORCE_INLINE int warehouse_efficients_txb_laplacian(
const AV1_COMMON *cm, const MACROBLOCK *x, const int plane, const int block,
const TX_SIZE tx_size, const TXB_CTX *const txb_ctx, const int eob,
const PLANE_TYPE plane_type, const LV_MAP_COEFF_COST *const coeff_costs,
const MACROBLOCKD *const xd, const TX_TYPE tx_type,
const CctxType cctx_type, int reduced_tx_set_used) {
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = txb_ctx->txb_skip_ctx;
if (plane == AOM_PLANE_V) {
txb_skip_ctx +=
(x->plane[AOM_PLANE_U].eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
}
#else
const int txb_skip_ctx = txb_ctx->txb_skip_ctx;
#endif // CONFIG_CONTEXT_DERIVATION
const int eob_multi_size = txsize_log2_minus4[tx_size];
const LV_MAP_EOB_COST *const eob_costs =
&x->coeff_costs.eob_costs[eob_multi_size][plane_type];
#if CONFIG_CONTEXT_DERIVATION
int cost;
if (plane == AOM_PLANE_V) {
cost = coeff_costs->v_txb_skip_cost[txb_skip_ctx][0];
} else {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const MB_MODE_INFO *mbmi = xd->mi[0];
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
cost = coeff_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][0];
#else
cost = coeff_costs->txb_skip_cost[txb_skip_ctx][0];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
}
#else
int cost = coeff_costs->txb_skip_cost[txb_skip_ctx][0];
#endif // CONFIG_CONTEXT_DERIVATION
const int bob_code = x->plane[plane].bobs[block];
const int is_inter = is_inter_block(xd->mi[0], xd->tree_type);
const int is_fsc = (xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter);
cost += get_tx_type_cost(x, xd, plane, tx_size, tx_type, reduced_tx_set_used,
eob, bob_code, is_fsc);
cost += get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
const MB_MODE_INFO *mbmi = xd->mi[0];
if ((mbmi->fsc_mode[xd->tree_type == CHROMA_PART] &&
get_primary_tx_type(tx_type) == IDTX && plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter_block(mbmi, xd->tree_type))) {
cost +=
av1_cost_coeffs_txb_skip_estimate(x, plane, block, tx_size, tx_type);
} else {
#if CONFIG_EOB_POS_LUMA
cost += get_eob_cost(eob, eob_costs, coeff_costs, is_inter);
#else
cost += get_eob_cost(eob, eob_costs, coeff_costs);
#endif // CONFIG_EOB_POS_LUMA
cost += av1_cost_coeffs_txb_estimate(x, plane, block, tx_size, tx_type);
}
return cost;
}
// Look up table of individual cost of coefficient by its quantization level.
// determined based on Laplacian distribution conditioned on estimated context
static const int costLUT[15] = { -1143, 53, 545, 825, 1031,
1209, 1393, 1577, 1763, 1947,
2132, 2317, 2501, 2686, 2871 };
static const int const_term = (1 << AV1_PROB_COST_SHIFT);
static const int loge_par = ((14427 << AV1_PROB_COST_SHIFT) + 5000) / 10000;
int av1_cost_coeffs_txb_estimate(const MACROBLOCK *x, const int plane,
const int block, const TX_SIZE tx_size,
const TX_TYPE tx_type) {
assert(plane == 0);
int cost = 0;
const struct macroblock_plane *p = &x->plane[plane];
const SCAN_ORDER *scan_order = get_scan(tx_size, tx_type);
const int16_t *scan = scan_order->scan;
tran_low_t *qcoeff = p->qcoeff + BLOCK_OFFSET(block);
int eob = p->eobs[block];
// coeffs
int c = eob - 1;
// eob
{
const int pos = scan[c];
const tran_low_t v = abs(qcoeff[pos]) - 1;
cost += (v << (AV1_PROB_COST_SHIFT + 2));
}
// other coeffs
for (c = eob - 2; c >= 0; c--) {
const int pos = scan[c];
const tran_low_t v = abs(qcoeff[pos]);
const int idx = AOMMIN(v, 14);
cost += costLUT[idx];
}
// const_term does not contain DC, and log(e) does not contain eob, so both
// (eob-1)
cost += (const_term + loge_par) * (eob - 1);
return cost;
}
int av1_cost_coeffs_txb_skip_estimate(const MACROBLOCK *x, const int plane,
const int block, const TX_SIZE tx_size,
const TX_TYPE tx_type) {
assert(plane == PLANE_TYPE_Y);
int cost = 0;
const struct macroblock_plane *p = &x->plane[plane];
const SCAN_ORDER *scan_order = get_scan(tx_size, tx_type);
const int16_t *scan = scan_order->scan;
tran_low_t *qcoeff = p->qcoeff + BLOCK_OFFSET(block);
int eob = p->eobs[block];
assert(eob == av1_get_max_eob(tx_size));
// coeffs
for (int c = 0; c < eob; c++) {
const int pos = scan[c];
const tran_low_t v = abs(qcoeff[pos]);
const int idx = AOMMIN(v, 14);
cost += costLUT[idx];
}
cost += (const_term + loge_par) * (eob - 1);
return cost;
}
int av1_cost_coeffs_txb(const AV1_COMMON *cm, const MACROBLOCK *x,
const int plane, const int block, const TX_SIZE tx_size,
const TX_TYPE tx_type, const CctxType cctx_type,
const TXB_CTX *const txb_ctx, int reduced_tx_set_used) {
const struct macroblock_plane *p = &x->plane[plane];
const int eob = p->eobs[block];
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
const PLANE_TYPE plane_type = get_plane_type(plane);
const LV_MAP_COEFF_COST *const coeff_costs =
&x->coeff_costs.coeff_costs[txs_ctx][plane_type];
const MACROBLOCKD *const xd = &x->e_mbd;
if (eob == 0) {
int skip_cost = 0;
#if CCTX_C2_DROPPED
if (plane == AOM_PLANE_V && !keep_chroma_c2(cctx_type) &&
is_cctx_allowed(cm, xd))
return 0;
#endif // CCTX_C2_DROPPED
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = txb_ctx->txb_skip_ctx;
if (plane == AOM_PLANE_Y || plane == AOM_PLANE_U) {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const MB_MODE_INFO *mbmi = xd->mi[0];
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
skip_cost += coeff_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][1];
#else
skip_cost += coeff_costs->txb_skip_cost[txb_skip_ctx][1];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
} else {
txb_skip_ctx +=
(x->plane[AOM_PLANE_U].eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
skip_cost += coeff_costs->v_txb_skip_cost[txb_skip_ctx][1];
}
#else
skip_cost += coeff_costs->txb_skip_cost[txb_ctx->txb_skip_ctx][1];
#endif // CONFIG_CONTEXT_DERIVATION
skip_cost +=
get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
return skip_cost;
}
const TX_CLASS tx_class = tx_type_to_class[get_primary_tx_type(tx_type)];
bool enable_parity_hiding = cm->features.allow_parity_hiding &&
!xd->lossless[xd->mi[0]->segment_id] &&
plane == PLANE_TYPE_Y &&
get_primary_tx_type(tx_type) < IDTX;
const MB_MODE_INFO *mbmi = xd->mi[0];
if ((mbmi->fsc_mode[xd->tree_type == CHROMA_PART] &&
get_primary_tx_type(tx_type) == IDTX && plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type, is_inter_block(mbmi, xd->tree_type))) {
return warehouse_efficients_txb_skip(cm, x, plane, block, tx_size, txb_ctx,
p, eob, coeff_costs, xd, tx_type,
cctx_type, reduced_tx_set_used);
} else {
return warehouse_efficients_txb(
cm, x, plane, block, tx_size, txb_ctx, p, eob, plane_type, coeff_costs,
xd, tx_type, cctx_type, tx_class, reduced_tx_set_used,
enable_parity_hiding, &x->coeff_costs.coeff_costs[0][0]);
}
}
int av1_cost_coeffs_txb_laplacian(const AV1_COMMON *cm, const MACROBLOCK *x,
const int plane, const int block,
const TX_SIZE tx_size, const TX_TYPE tx_type,
const CctxType cctx_type,
const TXB_CTX *const txb_ctx,
const int reduced_tx_set_used,
const int adjust_eob) {
const struct macroblock_plane *p = &x->plane[plane];
int eob = p->eobs[block];
if (adjust_eob) {
const SCAN_ORDER *scan_order = get_scan(tx_size, tx_type);
const int16_t *scan = scan_order->scan;
tran_low_t *tcoeff = p->coeff + BLOCK_OFFSET(block);
tran_low_t *qcoeff = p->qcoeff + BLOCK_OFFSET(block);
tran_low_t *dqcoeff = p->dqcoeff + BLOCK_OFFSET(block);
update_coeff_eob_fast(&eob, av1_get_tx_scale(tx_size), p->dequant_QTX, scan,
tcoeff, qcoeff, dqcoeff);
p->eobs[block] = eob;
}
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
const PLANE_TYPE plane_type = get_plane_type(plane);
const LV_MAP_COEFF_COST *const coeff_costs =
&x->coeff_costs.coeff_costs[txs_ctx][plane_type];
const MACROBLOCKD *const xd = &x->e_mbd;
if (eob == 0) {
int skip_cost = 0;
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = txb_ctx->txb_skip_ctx;
if (plane == AOM_PLANE_Y || plane == AOM_PLANE_U) {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const MB_MODE_INFO *mbmi = xd->mi[0];
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
skip_cost += coeff_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][1];
#else
skip_cost += coeff_costs->txb_skip_cost[txb_skip_ctx][1];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
} else {
txb_skip_ctx +=
(x->plane[AOM_PLANE_U].eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
skip_cost += coeff_costs->v_txb_skip_cost[txb_skip_ctx][1];
}
#else
skip_cost += coeff_costs->txb_skip_cost[txb_ctx->txb_skip_ctx][1];
#endif // CONFIG_CONTEXT_DERIVATION
skip_cost +=
get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
return skip_cost;
}
return warehouse_efficients_txb_laplacian(
cm, x, plane, block, tx_size, txb_ctx, eob, plane_type, coeff_costs, xd,
tx_type, cctx_type, reduced_tx_set_used);
}
static AOM_FORCE_INLINE int get_two_coeff_cost_simple(
int plane, int ci, tran_low_t abs_qc, int coeff_ctx,
const LV_MAP_COEFF_COST *txb_costs, int bwl, TX_CLASS tx_class,
const uint8_t *levels, int *cost_low) {
// this simple version assumes the coeff's scan_idx is not DC (scan_idx != 0)
// and not the last (scan_idx != eob - 1)
assert(ci > 0);
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int cost = 0;
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost +=
txb_costs->base_lf_cost[coeff_ctx][AOMMIN(abs_qc, LF_BASE_SYMBOLS - 1)];
} else {
cost += txb_costs->base_cost[coeff_ctx][AOMMIN(abs_qc, 3)];
}
int diff = 0;
if (limits) {
if (abs_qc <= (LF_BASE_SYMBOLS - 1)) {
if (abs_qc == 0) {
diff = 0;
} else if (abs_qc == 1) {
diff = txb_costs->base_lf_cost[coeff_ctx][1] + av1_cost_literal(1) -
txb_costs->base_lf_cost[coeff_ctx][0];
} else if (abs_qc == 2) {
diff = txb_costs->base_lf_cost[coeff_ctx][2] -
txb_costs->base_lf_cost[coeff_ctx][1];
} else if (abs_qc == 3) {
diff = txb_costs->base_lf_cost[coeff_ctx][3] -
txb_costs->base_lf_cost[coeff_ctx][2];
} else if (abs_qc == 4) {
diff = txb_costs->base_lf_cost[coeff_ctx][4] -
txb_costs->base_lf_cost[coeff_ctx][3];
} else {
diff = txb_costs->base_lf_cost[coeff_ctx][5] -
txb_costs->base_lf_cost[coeff_ctx][4];
}
}
} else {
if (abs_qc <= 3) {
if (abs_qc == 0) {
diff = 0;
} else if (abs_qc == 1) {
diff = txb_costs->base_cost[coeff_ctx][1] + av1_cost_literal(1) -
txb_costs->base_cost[coeff_ctx][0];
} else if (abs_qc == 2) {
diff = txb_costs->base_cost[coeff_ctx][2] -
txb_costs->base_cost[coeff_ctx][1];
} else {
diff = txb_costs->base_cost[coeff_ctx][3] -
txb_costs->base_cost[coeff_ctx][2];
}
}
}
if (abs_qc) {
cost += av1_cost_literal(1);
if (limits) {
if (abs_qc > LF_NUM_BASE_LEVELS) {
const int br_ctx = get_br_lf_ctx(levels, ci, bwl, tx_class);
int brcost_diff = 0;
cost += get_br_lf_cost_with_diff(abs_qc, txb_costs->lps_lf_cost[br_ctx],
&brcost_diff);
diff += brcost_diff;
}
} else {
if (abs_qc > NUM_BASE_LEVELS) {
const int br_ctx = get_br_ctx(levels, ci, bwl, tx_class);
int brcost_diff = 0;
cost += get_br_cost_with_diff(abs_qc, txb_costs->lps_cost[br_ctx],
&brcost_diff);
diff += brcost_diff;
}
}
}
*cost_low = cost - diff;
return cost;
}
static INLINE int get_coeff_cost_eob(int ci, tran_low_t abs_qc, int sign,
int coeff_ctx, int dc_sign_ctx,
const LV_MAP_COEFF_COST *txb_costs,
int bwl, TX_CLASS tx_class
#if CONFIG_CONTEXT_DERIVATION
,
int32_t *tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
int plane) {
int cost = 0;
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost +=
txb_costs->base_lf_eob_cost[coeff_ctx]
[AOMMIN(abs_qc, LF_BASE_SYMBOLS - 1) - 1];
} else {
cost += txb_costs->base_eob_cost[coeff_ctx][AOMMIN(abs_qc, 3) - 1];
}
if (abs_qc != 0) {
if (ci == 0) {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V)
cost += txb_costs->v_dc_sign_cost[tmp_sign[0]][dc_sign_ctx][sign];
else
cost += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
#else
cost += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
#endif // CONFIG_CONTEXT_DERIVATION
} else {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V)
cost += txb_costs->v_ac_sign_cost[tmp_sign[ci]][sign];
else
cost += av1_cost_literal(1);
#else
cost += av1_cost_literal(1);
#endif // CONFIG_CONTEXT_DERIVATION
}
if (limits) {
if (abs_qc > LF_NUM_BASE_LEVELS) {
int br_ctx;
br_ctx = get_br_ctx_lf_eob(ci, tx_class);
cost += get_br_lf_cost(abs_qc, txb_costs->lps_lf_cost[br_ctx]);
}
} else {
if (abs_qc > NUM_BASE_LEVELS) {
int br_ctx = 7; /* get_br_ctx_eob */
cost += get_br_cost(abs_qc, txb_costs->lps_cost[br_ctx]);
}
}
}
return cost;
}
static INLINE int get_coeff_cost_general(int is_last, int ci, tran_low_t abs_qc,
int sign, int coeff_ctx,
int dc_sign_ctx,
const LV_MAP_COEFF_COST *txb_costs,
int bwl, TX_CLASS tx_class,
const uint8_t *levels
#if CONFIG_CONTEXT_DERIVATION
,
int32_t *tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
int plane) {
int cost = 0;
if (is_last) {
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost +=
txb_costs->base_lf_eob_cost[coeff_ctx]
[AOMMIN(abs_qc, LF_BASE_SYMBOLS - 1) - 1];
} else {
cost += txb_costs->base_eob_cost[coeff_ctx][AOMMIN(abs_qc, 3) - 1];
}
} else {
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost +=
txb_costs
->base_lf_cost[coeff_ctx][AOMMIN(abs_qc, LF_BASE_SYMBOLS - 1)];
} else {
cost += txb_costs->base_cost[coeff_ctx][AOMMIN(abs_qc, 3)];
}
}
if (abs_qc != 0) {
if (ci == 0) {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V)
cost += txb_costs->v_dc_sign_cost[tmp_sign[0]][dc_sign_ctx][sign];
else
cost += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
#else
cost += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
#endif // CONFIG_CONTEXT_DERIVATION
} else {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V)
cost += txb_costs->v_ac_sign_cost[tmp_sign[ci]][sign];
else
cost += av1_cost_literal(1);
#else
cost += av1_cost_literal(1);
#endif // CONFIG_CONTEXT_DERIVATION
}
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
if (abs_qc > LF_NUM_BASE_LEVELS) {
int br_ctx;
if (is_last)
br_ctx = get_br_ctx_lf_eob(ci, tx_class);
else
br_ctx = get_br_lf_ctx(levels, ci, bwl, tx_class);
cost += get_br_lf_cost(abs_qc, txb_costs->lps_lf_cost[br_ctx]);
}
} else {
if (abs_qc > NUM_BASE_LEVELS) {
int br_ctx;
if (is_last)
br_ctx = 0; /*get_br_ctx_eob*/
else
br_ctx = get_br_ctx(levels, ci, bwl, tx_class);
cost += get_br_cost(abs_qc, txb_costs->lps_cost[br_ctx]);
}
}
}
return cost;
}
static INLINE void get_qc_dqc_low(tran_low_t abs_qc, int sign, int dqv,
int shift, tran_low_t *qc_low,
tran_low_t *dqc_low) {
tran_low_t abs_qc_low = abs_qc - 1;
*qc_low = (-sign ^ abs_qc_low) + sign;
assert((sign ? -abs_qc_low : abs_qc_low) == *qc_low);
tran_low_t abs_dqc_low =
(tran_low_t)(ROUND_POWER_OF_TWO_64((tran_high_t)abs_qc_low * dqv,
QUANT_TABLE_BITS) >>
shift);
*dqc_low = (-sign ^ abs_dqc_low) + sign;
assert((sign ? -abs_dqc_low : abs_dqc_low) == *dqc_low);
}
static INLINE void update_coeff_general(
int *accu_rate, int64_t *accu_dist, int si, int eob, TX_CLASS tx_class,
int bwl, int height, int64_t rdmult, int shift, int dc_sign_ctx,
const int32_t *dequant, const int16_t *scan,
const LV_MAP_COEFF_COST *txb_costs, const tran_low_t *tcoeff,
tran_low_t *qcoeff, tran_low_t *dqcoeff, uint8_t *levels,
const qm_val_t *iqmatrix
#if CONFIG_CONTEXT_DERIVATION
,
int32_t *tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
int plane, coeff_info *coef_info, bool enable_parity_hiding) {
const int dqv = get_dqv(dequant, scan[si], iqmatrix);
const int ci = scan[si];
const tran_low_t qc = qcoeff[ci];
const int is_last = si == (eob - 1);
const int coeff_ctx = get_lower_levels_ctx_general(
is_last, si, bwl, height, levels, ci, tx_class, plane);
if (qc == 0) {
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
*accu_rate += txb_costs->base_lf_cost[coeff_ctx][0];
} else {
*accu_rate += txb_costs->base_cost[coeff_ctx][0];
}
} else {
const int sign = (qc < 0) ? 1 : 0;
const tran_low_t abs_qc = abs(qc);
const tran_low_t tqc = tcoeff[ci];
const tran_low_t dqc = dqcoeff[ci];
const int64_t dist = get_coeff_dist(tqc, dqc, shift);
const int64_t dist0 = get_coeff_dist(tqc, 0, shift);
const int rate =
get_coeff_cost_general(is_last, ci, abs_qc, sign, coeff_ctx,
dc_sign_ctx, txb_costs, bwl, tx_class, levels
#if CONFIG_CONTEXT_DERIVATION
,
tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
const int64_t rd = RDCOST(rdmult, rate, dist);
tran_low_t qc_low, dqc_low;
tran_low_t abs_qc_low;
int64_t dist_low, rd_low;
int rate_low;
if (abs_qc == 1) {
abs_qc_low = qc_low = dqc_low = 0;
dist_low = dist0;
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
rate_low = txb_costs->base_lf_cost[coeff_ctx][0];
} else {
rate_low = txb_costs->base_cost[coeff_ctx][0];
}
} else {
get_qc_dqc_low(abs_qc, sign, dqv, shift, &qc_low, &dqc_low);
abs_qc_low = abs_qc - 1;
dist_low = get_coeff_dist(tqc, dqc_low, shift);
rate_low =
get_coeff_cost_general(is_last, ci, abs_qc_low, sign, coeff_ctx,
dc_sign_ctx, txb_costs, bwl, tx_class, levels
#if CONFIG_CONTEXT_DERIVATION
,
tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
}
rd_low = RDCOST(rdmult, rate_low, dist_low);
if (rd_low < rd) {
qcoeff[ci] = qc_low;
dqcoeff[ci] = dqc_low;
levels[get_padded_idx(ci, bwl)] = AOMMIN(abs_qc_low, INT8_MAX);
*accu_rate += rate_low;
*accu_dist += dist_low - dist0;
if (enable_parity_hiding)
set_coeff_info(qc_low, dqc_low, qc, dqc, rd_low, rd, rate_low, rate,
false, coef_info, si);
} else {
*accu_rate += rate;
*accu_dist += dist - dist0;
if (enable_parity_hiding)
set_coeff_info(qc_low, dqc_low, qc, dqc, rd_low, rd, rate_low, rate,
true, coef_info, si);
}
}
}
static AOM_FORCE_INLINE void update_coeff_simple(
int *accu_rate, int si, int eob, TX_CLASS tx_class, int bwl, int64_t rdmult,
int shift, const int32_t *dequant, const int16_t *scan,
const LV_MAP_COEFF_COST *txb_costs, const tran_low_t *tcoeff,
tran_low_t *qcoeff, tran_low_t *dqcoeff, uint8_t *levels,
const qm_val_t *iqmatrix, coeff_info *coef_info, bool enable_parity_hiding,
int plane) {
const int dqv = get_dqv(dequant, scan[si], iqmatrix);
(void)eob;
// this simple version assumes the coeff's scan_idx is not DC (scan_idx != 0)
// and not the last (scan_idx != eob - 1)
assert(si != eob - 1);
assert(si > 0);
const int ci = scan[si];
const tran_low_t qc = qcoeff[ci];
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
int coeff_ctx = 0;
if (limits) {
coeff_ctx = get_lower_levels_lf_ctx(levels, ci, bwl, tx_class);
} else {
coeff_ctx = get_lower_levels_ctx(levels, ci, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
}
if (qc == 0) {
if (limits) {
*accu_rate += txb_costs->base_lf_cost[coeff_ctx][0];
} else {
*accu_rate += txb_costs->base_cost[coeff_ctx][0];
}
} else {
const tran_low_t abs_qc = abs(qc);
const tran_low_t abs_tqc = abs(tcoeff[ci]);
const tran_low_t abs_dqc = abs(dqcoeff[ci]);
int rate_low = 0;
const int rate =
get_two_coeff_cost_simple(plane, ci, abs_qc, coeff_ctx, txb_costs, bwl,
tx_class, levels, &rate_low);
if (abs_dqc < abs_tqc) {
*accu_rate += rate;
return;
}
const int64_t dist = get_coeff_dist(abs_tqc, abs_dqc, shift);
const int64_t rd = RDCOST(rdmult, rate, dist);
const tran_low_t abs_qc_low = abs_qc - 1;
const tran_low_t abs_dqc_low =
(tran_low_t)ROUND_POWER_OF_TWO_64((tran_high_t)abs_qc_low * dqv,
QUANT_TABLE_BITS) >>
shift;
const int64_t dist_low = get_coeff_dist(abs_tqc, abs_dqc_low, shift);
const int64_t rd_low = RDCOST(rdmult, rate_low, dist_low);
if (rd_low < rd) {
tran_low_t qc_low = qc < 0 ? -abs_qc_low : abs_qc_low;
tran_low_t dqc_low = qc < 0 ? -abs_dqc_low : abs_dqc_low;
if (enable_parity_hiding)
set_coeff_info(qc_low, dqc_low, qc, dqcoeff[ci], rd_low, rd, rate_low,
rate, false, coef_info, si);
qcoeff[ci] = qc_low;
dqcoeff[ci] = dqc_low;
levels[get_padded_idx(ci, bwl)] = AOMMIN(abs_qc_low, INT8_MAX);
*accu_rate += rate_low;
} else {
*accu_rate += rate;
tran_low_t qc_low = qc < 0 ? -abs_qc_low : abs_qc_low;
tran_low_t dqc_low = qc < 0 ? -abs_dqc_low : abs_dqc_low;
if (enable_parity_hiding)
set_coeff_info(qc_low, dqc_low, qc, dqcoeff[ci], rd_low, rd, rate_low,
rate, true, coef_info, si);
}
}
}
static AOM_FORCE_INLINE void update_coeff_eob(
int *accu_rate, int64_t *accu_dist, int *eob, int *nz_num, int *nz_ci,
int si, TX_SIZE tx_size,
#if CONFIG_EOB_POS_LUMA
int is_inter,
#endif // CONFIG_EOB_POS_LUMA
TX_CLASS tx_class, int dc_sign_ctx, int64_t rdmult, int shift,
const int32_t *dequant, const int16_t *scan,
const LV_MAP_EOB_COST *txb_eob_costs, const LV_MAP_COEFF_COST *txb_costs,
const tran_low_t *tcoeff, tran_low_t *qcoeff, tran_low_t *dqcoeff,
uint8_t *levels, int sharpness, const qm_val_t *iqmatrix
#if CONFIG_CONTEXT_DERIVATION
,
int32_t *tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
int plane, coeff_info *coef_info, bool enable_parity_hiding) {
const int bwl = get_txb_bwl(tx_size);
const int height = get_txb_high(tx_size);
const int dqv = get_dqv(dequant, scan[si], iqmatrix);
assert(si != *eob - 1);
const int ci = scan[si];
const tran_low_t qc = qcoeff[ci];
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
int coeff_ctx = 0;
if (limits) {
coeff_ctx = get_lower_levels_lf_ctx(levels, ci, bwl, tx_class);
} else {
coeff_ctx = get_lower_levels_ctx(levels, ci, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
}
if (qc == 0) {
if (limits) {
*accu_rate += txb_costs->base_lf_cost[coeff_ctx][0];
} else {
*accu_rate += txb_costs->base_cost[coeff_ctx][0];
}
} else {
int64_t rd_eob_low = INT64_MAX >> 1;
int rate_eob_low = INT32_MAX >> 1;
int lower_level = 0;
const tran_low_t abs_qc = abs(qc);
const tran_low_t tqc = tcoeff[ci];
const tran_low_t dqc = dqcoeff[ci];
const int sign = (qc < 0) ? 1 : 0;
const int64_t dist0 = get_coeff_dist(tqc, 0, shift);
int64_t dist = get_coeff_dist(tqc, dqc, shift) - dist0;
int rate =
get_coeff_cost_general(0, ci, abs_qc, sign, coeff_ctx, dc_sign_ctx,
txb_costs, bwl, tx_class, levels
#if CONFIG_CONTEXT_DERIVATION
,
tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
int64_t rd = RDCOST(rdmult, *accu_rate + rate, *accu_dist + dist);
tran_low_t qc_low, dqc_low;
tran_low_t abs_qc_low;
int64_t dist_low, rd_low;
int rate_low;
if (abs_qc == 1) {
abs_qc_low = 0;
dqc_low = qc_low = 0;
dist_low = 0;
if (limits) {
rate_low = txb_costs->base_lf_cost[coeff_ctx][0];
} else {
rate_low = txb_costs->base_cost[coeff_ctx][0];
}
rd_low = RDCOST(rdmult, *accu_rate + rate_low, *accu_dist);
} else {
get_qc_dqc_low(abs_qc, sign, dqv, shift, &qc_low, &dqc_low);
abs_qc_low = abs_qc - 1;
dist_low = get_coeff_dist(tqc, dqc_low, shift) - dist0;
rate_low =
get_coeff_cost_general(0, ci, abs_qc_low, sign, coeff_ctx,
dc_sign_ctx, txb_costs, bwl, tx_class, levels
#if CONFIG_CONTEXT_DERIVATION
,
tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
rd_low = RDCOST(rdmult, *accu_rate + rate_low, *accu_dist + dist_low);
}
int rate_up_backup = rate;
int64_t rd_up_backup = rd;
int lower_level_new_eob = 0;
const int new_eob = si + 1;
const int coeff_ctx_new_eob = get_lower_levels_ctx_eob(bwl, height, si);
const int new_eob_cost =
#if CONFIG_EOB_POS_LUMA
get_eob_cost(new_eob, txb_eob_costs, txb_costs, is_inter);
#else
get_eob_cost(new_eob, txb_eob_costs, txb_costs);
#endif // CONFIG_EOB_POS_LUMA
int rate_coeff_eob =
new_eob_cost + get_coeff_cost_eob(ci, abs_qc, sign, coeff_ctx_new_eob,
dc_sign_ctx, txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
int64_t dist_new_eob = dist;
int64_t rd_new_eob = RDCOST(rdmult, rate_coeff_eob, dist_new_eob);
int rateeobup = rate_coeff_eob;
int64_t rdeobup = rd_new_eob;
if (abs_qc_low > 0) {
const int rate_coeff_eob_low =
new_eob_cost + get_coeff_cost_eob(ci, abs_qc_low, sign,
coeff_ctx_new_eob, dc_sign_ctx,
txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
const int64_t dist_new_eob_low = dist_low;
const int64_t rd_new_eob_low =
RDCOST(rdmult, rate_coeff_eob_low, dist_new_eob_low);
rate_eob_low = rate_coeff_eob_low;
rd_eob_low = rd_new_eob_low;
if (rd_new_eob_low < rd_new_eob) {
lower_level_new_eob = 1;
rd_new_eob = rd_new_eob_low;
rate_coeff_eob = rate_coeff_eob_low;
dist_new_eob = dist_new_eob_low;
}
}
if (rd_low < rd) {
lower_level = 1;
rd = rd_low;
rate = rate_low;
dist = dist_low;
}
if (sharpness == 0 && rd_new_eob < rd) {
for (int ni = 0; ni < *nz_num; ++ni) {
int last_ci = nz_ci[ni];
levels[get_padded_idx(last_ci, bwl)] = 0;
qcoeff[last_ci] = 0;
dqcoeff[last_ci] = 0;
}
*eob = new_eob;
*nz_num = 0;
*accu_rate = rate_coeff_eob;
*accu_dist = dist_new_eob;
lower_level = lower_level_new_eob;
if (abs_qc > 1 && enable_parity_hiding) {
set_coeff_info(qc_low, dqc_low, qc, dqc, rd_eob_low, rdeobup,
rate_eob_low, rateeobup, !lower_level, coef_info, si);
}
} else {
*accu_rate += rate;
*accu_dist += dist;
if (enable_parity_hiding)
set_coeff_info(qc_low, dqc_low, qc, dqc, rd_low, rd_up_backup, rate_low,
rate_up_backup, !lower_level, coef_info, si);
}
if (lower_level) {
qcoeff[ci] = qc_low;
dqcoeff[ci] = dqc_low;
levels[get_padded_idx(ci, bwl)] = AOMMIN(abs_qc_low, INT8_MAX);
}
if (qcoeff[ci]) {
nz_ci[*nz_num] = ci;
++*nz_num;
}
}
}
static INLINE void update_skip(int *accu_rate, int64_t accu_dist, int *eob,
int nz_num, int *nz_ci, int64_t rdmult,
int skip_cost, int non_skip_cost,
tran_low_t *qcoeff, tran_low_t *dqcoeff,
int sharpness) {
const int64_t rd = RDCOST(rdmult, *accu_rate + non_skip_cost, accu_dist);
const int64_t rd_new_eob = RDCOST(rdmult, skip_cost, 0);
if (sharpness == 0 && rd_new_eob < rd) {
for (int i = 0; i < nz_num; ++i) {
const int ci = nz_ci[i];
qcoeff[ci] = 0;
dqcoeff[ci] = 0;
// no need to set up levels because this is the last step
// levels[get_padded_idx(ci, bwl)] = 0;
}
*accu_rate = 0;
*eob = 0;
}
}
#if CONFIG_DQ
typedef struct _PQData
{
tran_low_t absLevel;
int64_t deltaDist;
tran_low_t dqc;
} PQData;
static AOM_INLINE void set_levels_buf(DECISION* decision, uint8_t* levels, const int16_t* scan, const int eob_minus1, const int scan_pos, const int bwl, const int sharpness)
{
if (decision->prevId == -2)
{
return;
}
// update current abs level
levels[get_padded_idx(scan[scan_pos], bwl)] = AOMMIN(decision->absLevel, INT8_MAX);
// check current node is a new start position? if so, set all previous position to 0.
// prevId == -1 means a new start, prevId == -2 ?
bool new_eob = decision->prevId < 0 && scan_pos + 1 <= eob_minus1 && sharpness == 0;
if (new_eob)
{
for (int si = scan_pos + 1; si <= eob_minus1; si++)
{
levels[get_padded_idx(scan[si], bwl)] = 0;
}
}
}
#if NEWCTX
static INLINE int get_coeff_cost_general_q1(int is_last, int ci, tran_low_t abs_qc,
int sign, int coeff_ctx,
int dc_sign_ctx,
const LV_MAP_COEFF_COST *txb_costs,
int bwl, TX_CLASS tx_class,
const uint8_t *levels
#if CONFIG_CONTEXT_DERIVATION
,
int32_t *tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
int plane) {
int cost = 0;
if (is_last) {
// quantizer 1 can not be used for last non-zero coeff.
assert(0);
} else {
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
cost +=
txb_costs
->base_lf_cost_tcq[coeff_ctx][AOMMIN(abs_qc, LF_BASE_SYMBOLS - 1)];
} else {
cost += txb_costs->base_cost_tcq[coeff_ctx][AOMMIN(abs_qc, 3)];
}
}
if (abs_qc != 0) {
if (ci == 0) {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V)
cost += txb_costs->v_dc_sign_cost[tmp_sign[0]][dc_sign_ctx][sign];
else
cost += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
#else
cost += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
#endif // CONFIG_CONTEXT_DERIVATION
} else {
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_V)
cost += txb_costs->v_ac_sign_cost[tmp_sign[ci]][sign];
else
cost += av1_cost_literal(1);
#else
cost += av1_cost_literal(1);
#endif // CONFIG_CONTEXT_DERIVATION
}
const int row = ci >> bwl;
const int col = ci - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
if (abs_qc > LF_NUM_BASE_LEVELS) {
int br_ctx;
if (is_last)
br_ctx = get_br_ctx_lf_eob(ci, tx_class);
else
br_ctx = get_br_lf_ctx(levels, ci, bwl, tx_class);
cost += get_br_lf_cost(abs_qc, txb_costs->lps_lf_cost[br_ctx]);
}
} else {
if (abs_qc > NUM_BASE_LEVELS) {
int br_ctx;
if (is_last)
br_ctx = 0; /*get_br_ctx_eob*/
else
br_ctx = get_br_ctx(levels, ci, bwl, tx_class);
cost += get_br_cost(abs_qc, txb_costs->lps_cost[br_ctx]);
}
}
}
return cost;
}
#endif
static void decide(int64_t distA, int64_t distB, int64_t distzero, int64_t rdmult, int rateA, int rateB, int rate_zero, PQData* pqDataA, PQData* pqDataB, int limits, int prev_state, DECISION* decision_02, DECISION* decision_1)
{
int t1 = tcq_parity(pqDataA->absLevel, limits);
int t2 = tcq_parity(pqDataB->absLevel, limits);
int64_t costA = RDCOST(rdmult, rateA, distA);
int64_t costB = RDCOST(rdmult, rateB, distB);
int64_t cost_zero = RDCOST(rdmult, rate_zero, distzero);
if (t1)
{
if (costA < decision_1->rdCost)
{
decision_1->rdCost = costA;
decision_1->dist = distA;
decision_1->rate = rateA;
decision_1->prevId = prev_state;
decision_1->absLevel = pqDataA->absLevel;
decision_1->dqc = pqDataA->dqc;
}
}
else
{
if (costA < decision_02->rdCost)
{
decision_02->rdCost = costA;
decision_02->dist = distA;
decision_02->rate = rateA;
decision_02->prevId = prev_state;
decision_02->absLevel = pqDataA->absLevel;
decision_02->dqc = pqDataA->dqc;
}
}
if (t2)
{
if (costB < decision_1->rdCost)
{
decision_1->rdCost = costB;
decision_1->dist = distB;
decision_1->rate = rateB;
decision_1->prevId = prev_state;
decision_1->absLevel = pqDataB->absLevel;
decision_1->dqc = pqDataB->dqc;
}
}
else
{
if (costB < decision_02->rdCost)
{
decision_02->rdCost = costB;
decision_02->dist = distB;
decision_02->rate = rateB;
decision_02->prevId = prev_state;
decision_02->absLevel = pqDataB->absLevel;
decision_02->dqc = pqDataB->dqc;
}
}
if (cost_zero < decision_02->rdCost)
{
decision_02->rdCost = cost_zero;
decision_02->dist = distzero;
decision_02->rate = rate_zero;
decision_02->prevId = prev_state;
decision_02->absLevel = 0;
decision_02->dqc = 0;
}
}
void pre_quant(tran_low_t tqc, PQData* pqData, const int32_t* quant_ptr, int dqv, int log_scale, int scan_pos)
{
//calculate qIdx
const int shift = 16 - log_scale + QUANT_FP_BITS;
tran_low_t add = -((3 << shift) >> 1);
tran_low_t abs_tqc = abs(tqc);
tran_low_t qIdx = (int)AOMMAX(1, AOMMIN(((1 << 16) - 1), (abs_tqc * quant_ptr[scan_pos != 0] + add) >> shift));
const int64_t dist0 = get_coeff_dist(abs_tqc, 0, log_scale - 1);
int Idx_a = qIdx & 3;
tran_low_t dqca = (tran_low_t)ROUND_POWER_OF_TWO_64((tran_high_t)qIdx * dqv, QUANT_TABLE_BITS) >> log_scale;
pqData[Idx_a].absLevel = (++qIdx) >> 1;
pqData[Idx_a].deltaDist = get_coeff_dist(abs_tqc, dqca, log_scale - 1) - dist0;
pqData[Idx_a].dqc = dqca;
int Idx_b = qIdx & 3;
tran_low_t dqcb = (tran_low_t)ROUND_POWER_OF_TWO_64((tran_high_t)qIdx * dqv, QUANT_TABLE_BITS) >> log_scale;
pqData[Idx_b].absLevel = (++qIdx) >> 1;
pqData[Idx_b].deltaDist = get_coeff_dist(abs_tqc, dqcb, log_scale - 1) - dist0;
pqData[Idx_b].dqc = dqcb;
int Idx_c = qIdx & 3;
tran_low_t dqcc = (tran_low_t)ROUND_POWER_OF_TWO_64((tran_high_t)qIdx * dqv, QUANT_TABLE_BITS) >> log_scale;
pqData[Idx_c].absLevel = (++qIdx) >> 1;
pqData[Idx_c].deltaDist = get_coeff_dist(abs_tqc, dqcc, log_scale - 1) - dist0;
pqData[Idx_c].dqc = dqcc;
int Idx_d = qIdx & 3;
tran_low_t dqcd = (tran_low_t)ROUND_POWER_OF_TWO_64((tran_high_t)qIdx * dqv, QUANT_TABLE_BITS) >> log_scale;
pqData[Idx_d].absLevel = (++qIdx) >> 1;
pqData[Idx_d].deltaDist = get_coeff_dist(abs_tqc, dqcd, log_scale - 1) - dist0;
pqData[Idx_d].dqc = dqcd;
}
int av1_dep_quant(const struct AV1_COMP* cpi, MACROBLOCK* x, int plane,
int block, TX_SIZE tx_size, TX_TYPE tx_type, CctxType cctx_type,
const TXB_CTX* const txb_ctx, int* rate_cost,
int sharpness) {
MACROBLOCKD* xd = &x->e_mbd;
const struct macroblock_plane* p = &x->plane[plane];
const SCAN_ORDER* scan_order =
get_scan(tx_size, get_primary_tx_type(tx_type));
const int16_t* scan = scan_order->scan;
const int shift = av1_get_tx_scale(tx_size);
int eob = p->eobs[block];
const int32_t* dequant = p->dequant_QTX;
const int32_t* quant = p->quant_fp_QTX; // quant_QTX
const qm_val_t* iqmatrix =
av1_get_iqmatrix(&cpi->common.quant_params, xd, plane, tx_size, tx_type);
const int block_offset = BLOCK_OFFSET(block);
tran_low_t* qcoeff = p->qcoeff + block_offset;
tran_low_t* dqcoeff = p->dqcoeff + block_offset;
const tran_low_t* tcoeff = p->coeff + block_offset;
const CoeffCosts* coeff_costs = &x->coeff_costs;
// This function is not called if eob = 0.
assert(eob > 0);
const AV1_COMMON* cm = &cpi->common;
const PLANE_TYPE plane_type = get_plane_type(plane);
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
const TX_CLASS tx_class = tx_type_to_class[get_primary_tx_type(tx_type)];
const MB_MODE_INFO* mbmi = xd->mi[0];
const int bwl = get_txb_bwl(tx_size);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
assert(width == (1 << bwl));
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int bob_code = p->bobs[block];
const int is_fsc = (xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(&cpi->common, plane, tx_type, is_inter);
const LV_MAP_COEFF_COST* txb_costs =
&coeff_costs->coeff_costs[txs_ctx][plane_type];
const int eob_multi_size = txsize_log2_minus4[tx_size];
const LV_MAP_EOB_COST* txb_eob_costs =
&coeff_costs->eob_costs[eob_multi_size][plane_type];
const int rshift =
(sharpness +
(cpi->oxcf.q_cfg.aq_mode == VARIANCE_AQ && mbmi->segment_id < 4
? 7 - mbmi->segment_id
: 2) +
(cpi->oxcf.q_cfg.aq_mode != VARIANCE_AQ &&
cpi->oxcf.q_cfg.deltaq_mode == DELTA_Q_PERCEPTUAL &&
cm->delta_q_info.delta_q_present_flag && x->sb_energy_level < 0
? (3 - x->sb_energy_level)
: 0));
int64_t rdmult =
(((int64_t)x->rdmult *
(plane_rd_mult[is_inter][plane_type] << (2 * (xd->bd - 8)))) +
2) >>
rshift;
// getting context from previous level buf, updating levels on current level buf.
// initialization all value by 0, since we update every position.
int bufsize = (width + 4) * (height + 4) + TX_PAD_END;
uint8_t** const levels = (uint8_t**)malloc(sizeof(uint8_t*) * 4);
uint8_t** const prev_levels = (uint8_t**)malloc(sizeof(uint8_t*) * 4);
for (int i = 0; i < 4; i++)
{
levels[i] = (uint8_t*)malloc(sizeof(uint8_t) * bufsize);
prev_levels[i] = (uint8_t*)malloc(sizeof(uint8_t) * bufsize);
if (eob > 1) memset(levels[i], 0, sizeof(uint8_t) * bufsize);
}
int si = eob - 1;
//populate trellis
DECISION trellis[MAX_TRELLIS][4];
int first_test_pos = si;
for (int scan_pos = first_test_pos; scan_pos >= 0; scan_pos--)
{
int blk_pos = scan[scan_pos];
DECISION* decision = trellis[scan_pos];
PQData pqData[4];
int tempdqv = get_dqv(dequant, scan[scan_pos], iqmatrix);
//preQuant tcoeff qcoeff dqcoeff shift dequant, scan[si], iqmatrix
pre_quant(tcoeff[blk_pos], pqData, quant, tempdqv, shift + 1, scan_pos);
// pre_quant(tcoeff[blk_pos], pqData, quant, tempdqv, shift, scan_pos);
//init state
for (int state = 0; state < 4; state++)
{
decision[state].rdCost = INT64_MAX >> 1;
decision[state].dist = INT64_MAX >> 1;
decision[state].rate = INT32_MAX >> 1;
decision[state].absLevel = -1;
decision[state].prevId = -2;
decision[state].dqc = INT32_MAX;
}
const int coeff_sign = tcoeff[blk_pos] < 0;
const int row = blk_pos >> bwl;
const int col = blk_pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
//calculate rate distortion
bool is_first_pos = (scan_pos == first_test_pos);
if (is_first_pos)
{
//try to quantize first coeff to nzcoeff
int coeff_ctx = get_lower_levels_ctx_eob(bwl, height, si);
#if CONFIG_EOB_POS_LUMA
const int eob_rate = get_eob_cost(si + 1, txb_eob_costs, txb_costs, is_inter);
#else
const int eob_rate = get_eob_cost(si + 1, txb_eob_costs, txb_costs);
#endif // CONFIG_EOB_POS_LUMA
int rate_Q0_a = get_coeff_cost_eob(blk_pos, pqData[0].absLevel, (qcoeff[blk_pos] < 0), coeff_ctx, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
, plane) + eob_rate;
int rate_Q0_b = get_coeff_cost_eob(blk_pos, pqData[2].absLevel, (qcoeff[blk_pos] < 0), coeff_ctx, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
, plane) + eob_rate;
decide(pqData[0].deltaDist, pqData[2].deltaDist, INT32_MAX >> 1, rdmult, rate_Q0_a, rate_Q0_b, INT32_MAX >> 1, &pqData[0], &pqData[2], limits, -1, &decision[0], &decision[2]);
}
else
{
int coeff_ctx0 = 0;
int coeff_ctx1 = 0;
int coeff_ctx2 = 0;
int coeff_ctx3 = 0;
if (limits) {
coeff_ctx0 = get_lower_levels_lf_ctx( prev_levels[0], blk_pos, bwl, tx_class);
coeff_ctx1 = get_lower_levels_lf_ctx( prev_levels[1], blk_pos, bwl, tx_class);
coeff_ctx2 = get_lower_levels_lf_ctx( prev_levels[2], blk_pos, bwl, tx_class);
coeff_ctx3 = get_lower_levels_lf_ctx( prev_levels[3], blk_pos, bwl, tx_class);
} else {
coeff_ctx0 = get_lower_levels_ctx( prev_levels[0], blk_pos, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
coeff_ctx1 = get_lower_levels_ctx( prev_levels[1], blk_pos, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
coeff_ctx2 = get_lower_levels_ctx( prev_levels[2], blk_pos, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
coeff_ctx3 = get_lower_levels_ctx( prev_levels[3], blk_pos, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
}
// calculate RDcost
int rate_zero_0 = 0;
int rate_zero_1 = 0;
int rate_zero_2 = 0;
int rate_zero_3 = 0;
if (limits) {
rate_zero_0 = txb_costs->base_lf_cost[coeff_ctx0][0];
rate_zero_1 = txb_costs->base_lf_cost[coeff_ctx1][0];
#if NEWCTX
rate_zero_2 = txb_costs->base_lf_cost_tcq[coeff_ctx2][0];
rate_zero_3 = txb_costs->base_lf_cost_tcq[coeff_ctx3][0];
#else
rate_zero_2 = txb_costs->base_lf_cost[coeff_ctx2][0];
rate_zero_3 = txb_costs->base_lf_cost[coeff_ctx3][0];
#endif
} else {
rate_zero_0 = txb_costs->base_cost[coeff_ctx0][0];
rate_zero_1 = txb_costs->base_cost[coeff_ctx1][0];
#if NEWCTX
rate_zero_2 = txb_costs->base_cost_tcq[coeff_ctx2][0];
rate_zero_3 = txb_costs->base_cost_tcq[coeff_ctx3][0];
#else
rate_zero_2 = txb_costs->base_cost[coeff_ctx2][0];
rate_zero_3 = txb_costs->base_cost[coeff_ctx3][0];
#endif
}
DECISION* prd = trellis[scan_pos + 1];
int rate_Q0_a_prd0 = get_coeff_cost_general(0, blk_pos, pqData[0].absLevel, coeff_sign, coeff_ctx0, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[0]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q0_a_prd1 = get_coeff_cost_general(0, blk_pos, pqData[0].absLevel, coeff_sign, coeff_ctx1, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[1]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q0_b_prd0 = get_coeff_cost_general(0, blk_pos, pqData[2].absLevel, coeff_sign, coeff_ctx0, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[0]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q0_b_prd1 = get_coeff_cost_general(0, blk_pos, pqData[2].absLevel, coeff_sign, coeff_ctx1, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[1]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
#if NEWCTX
int rate_Q1_a_prd2 = get_coeff_cost_general_q1(0, blk_pos, pqData[1].absLevel, coeff_sign, coeff_ctx2, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[2]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q1_a_prd3 = get_coeff_cost_general_q1(0, blk_pos, pqData[1].absLevel, coeff_sign, coeff_ctx3, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[3]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q1_b_prd2 = get_coeff_cost_general_q1(0, blk_pos, pqData[3].absLevel, coeff_sign, coeff_ctx2, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[2]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q1_b_prd3 = get_coeff_cost_general_q1(0, blk_pos, pqData[3].absLevel, coeff_sign, coeff_ctx3, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[3]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
#else
int rate_Q1_a_prd2 = get_coeff_cost_general(0, blk_pos, pqData[1].absLevel, coeff_sign, coeff_ctx2, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[2]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q1_a_prd3 = get_coeff_cost_general(0, blk_pos, pqData[1].absLevel, coeff_sign, coeff_ctx3, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[3]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q1_b_prd2 = get_coeff_cost_general(0, blk_pos, pqData[3].absLevel, coeff_sign, coeff_ctx2, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[2]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
int rate_Q1_b_prd3 = get_coeff_cost_general(0, blk_pos, pqData[3].absLevel, coeff_sign, coeff_ctx3, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class, prev_levels[3]
#if CONFIG_CONTEXT_DERIVATION
, xd->tmp_sign
#endif
, plane);
#endif
rate_Q0_a_prd0 += prd[0].rate;
rate_Q0_a_prd1 += prd[1].rate;
rate_Q0_b_prd0 += prd[0].rate;
rate_Q0_b_prd1 += prd[1].rate;
rate_Q1_a_prd2 += prd[2].rate;
rate_Q1_a_prd3 += prd[3].rate;
rate_Q1_b_prd2 += prd[2].rate;
rate_Q1_b_prd3 += prd[3].rate;
{
int64_t dist_Q0_a_prd0 = pqData[0].deltaDist + prd[0].dist;
int64_t dist_Q0_a_prd1 = pqData[0].deltaDist + prd[1].dist;
int64_t dist_Q0_b_prd0 = pqData[2].deltaDist + prd[0].dist;
int64_t dist_Q0_b_prd1 = pqData[2].deltaDist + prd[1].dist;
int64_t dist_Q1_a_prd2 = pqData[1].deltaDist + prd[2].dist;
int64_t dist_Q1_a_prd3 = pqData[1].deltaDist + prd[3].dist;
int64_t dist_Q1_b_prd2 = pqData[3].deltaDist + prd[2].dist;
int64_t dist_Q1_b_prd3 = pqData[3].deltaDist + prd[3].dist;
// todo: Q0 can skip the sig_flag or skip some another flag. This is not included in the calculation of RDcost now.
//pre_state is 0
decide(dist_Q0_a_prd0, dist_Q0_b_prd0, prd[0].dist, rdmult, rate_Q0_a_prd0, rate_Q0_b_prd0, rate_zero_0 + prd[0].rate, &pqData[0], &pqData[2], limits, 0, &decision[0], &decision[2]);
//pre_state is 1
decide(dist_Q0_a_prd1, dist_Q0_b_prd1, prd[1].dist, rdmult, rate_Q0_a_prd1, rate_Q0_b_prd1, rate_zero_1 + prd[1].rate, &pqData[0], &pqData[2], limits, 1, &decision[2], &decision[0]);
//pre_state is 2
decide(dist_Q1_a_prd2, dist_Q1_b_prd2, prd[2].dist, rdmult, rate_Q1_a_prd2, rate_Q1_b_prd2, rate_zero_2 + prd[2].rate, &pqData[1], &pqData[3], limits, 2, &decision[1], &decision[3]);
//pre_state is 3
decide(dist_Q1_a_prd3, dist_Q1_b_prd3, prd[3].dist, rdmult, rate_Q1_a_prd3, rate_Q1_b_prd3, rate_zero_3 + prd[3].rate, &pqData[1], &pqData[3], limits, 3, &decision[3], &decision[1]);
//assume current state is 0, current coeff is new eob.
//input: scan_pos,pqData[0],pqData[2], decison[0] and decision[2]
// update eob if better use current position as eob
if (sharpness == 0)
{
#if CONFIG_EOB_POS_LUMA
const int new_eob_rate = get_eob_cost(scan_pos + 1, txb_eob_costs, txb_costs, is_inter);
#else
const int new_eob_rate = get_eob_cost(scan_pos + 1, txb_eob_costs, txb_costs);
#endif // CONFIG_EOB_POS_LUMA
int new_eob_ctx = get_lower_levels_ctx_eob(bwl, height, scan_pos);
int rate_Q0_a = get_coeff_cost_eob(blk_pos, pqData[0].absLevel, (qcoeff[blk_pos] < 0), new_eob_ctx, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
, plane) + new_eob_rate;
int rate_Q0_b = get_coeff_cost_eob(blk_pos, pqData[2].absLevel, (qcoeff[blk_pos] < 0), new_eob_ctx, txb_ctx->dc_sign_ctx, txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
, plane) + new_eob_rate;
decide(pqData[0].deltaDist, pqData[2].deltaDist, INT32_MAX >> 1, rdmult, rate_Q0_a, rate_Q0_b, INT32_MAX >> 1, &pqData[0], &pqData[2], limits, -1, &decision[0], &decision[2]);
}
}
}
// copy corresponding context from previous level buffer
for (int state = 0; state < 4 && scan_pos != si; state++)
{
if (decision[state].prevId >= 0)
memcpy(levels[state], prev_levels[decision[state].prevId], sizeof(uint8_t) * bufsize);
}
// update levels_buf
for (int state = 0; state < 4 && scan_pos != 0; state++)
{
set_levels_buf(&decision[state], levels[state], scan, si, scan_pos, bwl, sharpness);
}
if (scan_pos != 0)
{
for(int state = 0; state < 4; state++)
memcpy(prev_levels[state], levels[state], sizeof(uint8_t) * bufsize);
}
}
for (int i = 0; i < 4; i++)
{
free(levels[i]);
free(prev_levels[i]);
}
free(levels);
free(prev_levels);
//find best path
int64_t min_path_cost = INT64_MAX;
int min_rate = INT32_MAX;
int64_t min_dist = INT32_MAX;
DECISION decision;
decision.prevId = -2;
for (int state = 0; state < 4; state++)
{
if (trellis[0][state].rdCost < min_path_cost)
{
decision.prevId = state;
min_path_cost = trellis[0][state].rdCost;
min_rate = trellis[0][state].rate;
min_dist = trellis[0][state].dist;
}
}
//backward scannig dqc,tqc,qc,level
int scan_pos = 0;
//printf("quantization... \n");
for (; decision.prevId >= 0; scan_pos++)
{
decision = trellis[scan_pos][decision.prevId];
//printf("%d ", decision.absLevel);
int blk_pos = scan[scan_pos];
qcoeff[blk_pos] = (tcoeff[blk_pos] < 0 ? -decision.absLevel : decision.absLevel);
dqcoeff[blk_pos] = (tcoeff[blk_pos] < 0 ? -decision.dqc : decision.dqc);
}
eob = scan_pos;
for (; scan_pos <= first_test_pos; scan_pos++)
{
int blk_pos = scan[scan_pos];
qcoeff[blk_pos] = 0;
dqcoeff[blk_pos] = 0;
}
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = txb_ctx->txb_skip_ctx;
int non_skip_cost = 0;
int skip_cost = 0;
if (plane == AOM_PLANE_V) {
txb_skip_ctx +=
(x->plane[AOM_PLANE_U].eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
non_skip_cost = txb_costs->v_txb_skip_cost[txb_skip_ctx][0];
skip_cost = txb_costs->v_txb_skip_cost[txb_skip_ctx][1];
} else {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
non_skip_cost = txb_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][0];
skip_cost = txb_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][1];
#else
non_skip_cost = txb_costs->txb_skip_cost[txb_skip_ctx][0];
skip_cost = txb_costs->txb_skip_cost[txb_skip_ctx][1];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
}
#else
const int non_skip_cost = txb_costs->txb_skip_cost[txb_ctx->txb_skip_ctx][0];
const int skip_cost = txb_costs->txb_skip_cost[txb_ctx->txb_skip_ctx][1];
#endif // CONFIG_CONTEXT_DERIVATION
int accu_rate = 0;
set_bob(x, plane, block, tx_size, tx_type);
if (eob == 0) assert(0); // in current implementation, this could not happen.
else
{
const int tx_type_cost = get_tx_type_cost(x, xd, plane, tx_size, tx_type,
cm->features.reduced_tx_set_used,
eob, bob_code, is_fsc
);
accu_rate += non_skip_cost + tx_type_cost + min_rate;
// skip block
if (RDCOST(rdmult, accu_rate, min_dist) > RDCOST(rdmult, skip_cost, 0) && sharpness == 0)
{
for (int scan_idx = 0; scan_idx <= first_test_pos; scan_idx++)
{
int blk_idx = scan[scan_idx];
qcoeff[blk_idx] = 0;
dqcoeff[blk_idx] = 0;
}
accu_rate = skip_cost;
eob = 0;
}
}
p->eobs[block] = eob;
p->txb_entropy_ctx[block] =
av1_get_txb_entropy_context(qcoeff, scan_order, p->eobs[block]);
accu_rate += get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
*rate_cost = accu_rate;
return eob;
}
#endif
// This funtion returns the rate saving if the parity of current
// DC coefficient is hidden.
static AOM_FORCE_INLINE int rate_save(const LV_MAP_COEFF_COST *txb_costs,
const LV_MAP_COEFF_COST *txb_costs_ph,
tran_low_t level, int bwl, int pos,
uint8_t *levels, int dc_sign_ctx,
TX_CLASS tx_class, int *rate
#if CONFIG_CHROMA_TX_COEFF_CODING
,
int plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
) {
tran_low_t abslevel = abs(level), q_index = abslevel >> 1;
int sign = level < 0;
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, 0);
int coeff_ctx = 0;
if (limits) {
coeff_ctx = get_lower_levels_lf_ctx(levels, pos, bwl, tx_class);
} else {
coeff_ctx = get_lower_levels_ctx(levels, pos, bwl, tx_class
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
}
*rate = get_coeff_cost_general(0, pos, abslevel, level < 0, coeff_ctx,
dc_sign_ctx, txb_costs, bwl, tx_class, levels
#if CONFIG_CONTEXT_DERIVATION
,
0
#endif // CONFIG_CONTEXT_DERIVATION
,
0);
const int base_ctx_ph = get_base_ctx_ph(levels, pos, bwl, tx_class);
int rate_ph = txb_costs_ph->base_ph_cost[base_ctx_ph][AOMMIN(q_index, 3)];
if (q_index > NUM_BASE_LEVELS) {
rate_ph += get_br_cost(
q_index,
txb_costs_ph->lps_ph_cost[get_par_br_ctx(levels, pos, bwl, tx_class)]);
}
if (abslevel) rate_ph += txb_costs->dc_sign_cost[dc_sign_ctx][sign];
return rate_ph - *rate;
}
typedef struct {
int rate;
int64_t cost;
tran_low_t qcoeff;
tran_low_t dqcoeff;
int scan_idx;
} tune_cand;
// This funtion calculates the cost change if the parity of DC position
// is tuned and hidden.
static AOM_FORCE_INLINE void cost_hide_par(
const tran_low_t qcoeff, const tran_low_t dqcoeff, const tran_low_t tcoeff,
const int shift, const LV_MAP_COEFF_COST *txb_costs, const int pos,
const LV_MAP_COEFF_COST *txb_costs_ph, int dc_sign_ctx, TX_CLASS tx_class,
uint8_t *levels, const int bwl, const int64_t rdmult,
const int32_t *dequant, const qm_val_t *iqmatrix, tune_cand *t_cand,
int rate_cur) {
const int dqv = get_dqv(dequant, pos, iqmatrix);
tran_low_t abslevel = abs(qcoeff), abstqc = abs(tcoeff);
int64_t dist = get_coeff_dist(tcoeff, dqcoeff, shift);
int rate = rate_cur;
int64_t cost = RDCOST(rdmult, rate, dist);
tran_low_t abslevel_cand =
abs(dqcoeff) > abstqc ? abslevel - 1 : abslevel + 1;
tran_low_t absdqc_cand =
(tran_low_t)(ROUND_POWER_OF_TWO_64((tran_high_t)abslevel_cand * dqv,
QUANT_TABLE_BITS) >>
shift);
int64_t dist_cand = get_coeff_dist(abs(tcoeff), absdqc_cand, shift);
int q_index = abslevel_cand >> 1;
int rate_cand = txb_costs_ph->base_ph_cost[get_base_ctx_ph(
levels, pos, bwl, tx_class)][AOMMIN(q_index, 3)];
if (abslevel_cand) {
rate_cand += txb_costs->dc_sign_cost[dc_sign_ctx][tcoeff < 0];
if (q_index > NUM_BASE_LEVELS) {
rate_cand +=
get_br_cost(q_index, txb_costs_ph->lps_ph_cost[get_par_br_ctx(
levels, pos, bwl, tx_class)]);
}
}
int64_t cost_cand = RDCOST(rdmult, rate_cand, dist_cand);
const int sign = tcoeff < 0 ? -1 : 1;
t_cand->cost = cost_cand - cost;
t_cand->qcoeff = abslevel_cand * sign;
t_cand->dqcoeff = absdqc_cand * sign;
t_cand->rate = rate_cand - rate;
t_cand->scan_idx = 0;
}
// This function finds best candidate for tuning among non-DC
// positions when current region has PHTHRESH - 1 non-zero
// coefficients.
static AOM_FORCE_INLINE bool region_nz_minus(
const int eob, tran_low_t *qcoeff, int ratesaving, const int16_t *scan,
coeff_info *coef_info, tune_cand *t_cand, const int64_t rdmult) {
int64_t cost = INT64_MAX >> 1;
int find_si = -1;
for (int scan_idx = eob - 1; scan_idx > 0; --scan_idx) {
int blkpos = scan[scan_idx];
if (abs(qcoeff[blkpos]) == 0 && !coef_info[scan_idx].upround &&
coef_info[scan_idx].tunable) // from 0 to 1
{
if (coef_info[scan_idx].delta_cost < cost) {
cost = coef_info[scan_idx].delta_cost;
find_si = scan_idx;
}
}
}
if (find_si == -1) {
return false;
}
t_cand->qcoeff = coef_info[find_si].qc;
t_cand->dqcoeff = coef_info[find_si].dqc;
t_cand->rate = coef_info[find_si].delta_rate + ratesaving;
t_cand->cost = coef_info[find_si].delta_cost + RDCOST(rdmult, ratesaving, 0);
t_cand->scan_idx = find_si;
return true;
}
// This function finds best candidate for tuning among non-DC
// positions when current region has PHTHRESH non-zero coefficients.
static AOM_FORCE_INLINE bool region_nz_equal(const int eob, tran_low_t *qcoeff,
const int ratesaving,
const int16_t *scan,
coeff_info *coef_info,
tune_cand *t_cand,
const int64_t rdmult) {
int64_t cost = INT64_MAX >> 1, cost_up0 = INT64_MAX >> 1,
cost_tune = INT64_MAX >> 1;
int si = -1, si_up0 = -1, si_tune = -1;
for (int scan_idx = eob - 1; scan_idx > 0; --scan_idx) {
if (coef_info[scan_idx].tunable) {
if (!(abs(qcoeff[scan[scan_idx]]) == 1 && coef_info[scan_idx].upround)) {
if (coef_info[scan_idx].delta_cost < cost) {
cost = coef_info[scan_idx].delta_cost;
si = scan_idx;
}
} else // from 1 to 0
{
if (coef_info[scan_idx].delta_cost < cost_up0) {
cost_up0 = coef_info[scan_idx].delta_cost;
si_up0 = scan_idx;
}
}
}
}
int64_t costsaving = RDCOST(rdmult, ratesaving, 0);
if (cost + costsaving < cost_tune) {
cost_tune = cost + costsaving;
si_tune = si;
}
bool disable = false;
if (cost_up0 < cost_tune) // no extra saving for sig
{
si_tune = si_up0;
cost_tune = cost_up0;
disable = true;
}
// modify
if (si_tune == -1) // not find any tunable position.
{
return false;
} else {
t_cand->scan_idx = si_tune;
t_cand->qcoeff = coef_info[si_tune].qc;
t_cand->dqcoeff = coef_info[si_tune].dqc;
t_cand->rate = coef_info[si_tune].delta_rate;
t_cand->cost = cost_tune;
if (!disable) {
t_cand->rate += ratesaving;
}
return true;
}
}
// This function finds best candidate for tuning among non-DC
// positions when current region has more than PHTHRESH non-zero
// coefficients.
static AOM_FORCE_INLINE bool region_nz_plus(const int eob, const int ratesaving,
coeff_info *coef_info,
tune_cand *t_cand,
const int64_t rdmult) {
int64_t cost = INT64_MAX >> 1;
int find_si = -1;
for (int scan_idx = eob - 1; scan_idx > 0; --scan_idx) {
if (coef_info[scan_idx].tunable && coef_info[scan_idx].delta_cost < cost) {
cost = coef_info[scan_idx].delta_cost;
find_si = scan_idx;
}
}
if (find_si == -1) {
return false;
}
t_cand->scan_idx = find_si;
t_cand->qcoeff = coef_info[find_si].qc;
t_cand->dqcoeff = coef_info[find_si].dqc;
t_cand->rate = coef_info[find_si].delta_rate + ratesaving;
t_cand->cost = cost + RDCOST(rdmult, ratesaving, 0);
return true;
}
static AOM_FORCE_INLINE bool parity_hide_tb(
const int eob, const int16_t *scan, uint8_t *levels, const int bwl,
const int64_t rdmult, const int shift, const LV_MAP_COEFF_COST *txb_costs,
const LV_MAP_COEFF_COST *txb_costs_ph, const int32_t *dequant,
const qm_val_t *iqmatrix, int dc_sign_ctx, const TX_CLASS tx_class,
tran_low_t *qcoeff, tran_low_t *dqcoeff, const tran_low_t *tcoeff,
coeff_info *coef_info, int *accu_rate
#if CONFIG_CHROMA_TX_COEFF_CODING
,
int plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
) {
int nzsbb = 0, sum_abs1 = 0;
for (int scan_idx = eob - 1; scan_idx > 0; --scan_idx) {
const int blkpos = scan[scan_idx];
if (qcoeff[blkpos]) {
++nzsbb;
sum_abs1 += AOMMIN(abs(qcoeff[blkpos]), MAX_BASE_BR_RANGE);
}
}
int hidepos = scan[0], rate_cur = 0;
bool needtune = (qcoeff[hidepos] & 1) != (sum_abs1 & 1);
if (nzsbb < PHTHRESH - 1 ||
(!needtune && nzsbb == PHTHRESH - 1)) // disable coef_info for this sbb
{
return false; // not hide
}
const int ratesaving =
rate_save(txb_costs, txb_costs_ph, qcoeff[hidepos], bwl, hidepos, levels,
dc_sign_ctx, tx_class, &rate_cur
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
if (!needtune && nzsbb >= PHTHRESH) {
*accu_rate += ratesaving;
return true; // hide
}
tune_cand t_cand_dc = { 0 }, t_cand_non_dc = { 0 };
t_cand_dc.cost = INT64_MAX;
t_cand_non_dc.cost = INT64_MAX;
// we change the quantized level's parity to check the rate change.
if (nzsbb >= PHTHRESH) {
cost_hide_par(qcoeff[hidepos], dqcoeff[hidepos], tcoeff[hidepos], shift,
txb_costs, hidepos, txb_costs_ph, dc_sign_ctx, tx_class,
levels, bwl, rdmult, dequant, iqmatrix, &t_cand_dc, rate_cur);
}
// we change the level candidates to check the cost change.
if (nzsbb == PHTHRESH - 1) {
region_nz_minus(eob, qcoeff, ratesaving, scan, coef_info, &t_cand_non_dc,
rdmult);
}
if (nzsbb == PHTHRESH) {
region_nz_equal(eob, qcoeff, ratesaving, scan, coef_info, &t_cand_non_dc,
rdmult);
}
if (nzsbb > PHTHRESH) {
region_nz_plus(eob, ratesaving, coef_info, &t_cand_non_dc, rdmult);
}
tune_cand *best =
t_cand_dc.cost < t_cand_non_dc.cost ? &t_cand_dc : &t_cand_non_dc;
if (nzsbb == PHTHRESH - 1 && best->cost > 0) {
assert(nzsbb == PHTHRESH - 1);
return false;
} else {
int tune_pos = scan[best->scan_idx];
qcoeff[tune_pos] = best->qcoeff;
dqcoeff[tune_pos] = best->dqcoeff;
*accu_rate += best->rate;
levels[get_padded_idx(tune_pos, bwl)] = AOMMIN(abs(best->qcoeff), INT8_MAX);
return true;
}
}
int av1_optimize_txb_new(const struct AV1_COMP *cpi, MACROBLOCK *x, int plane,
int block, TX_SIZE tx_size, TX_TYPE tx_type,
CctxType cctx_type, const TXB_CTX *const txb_ctx,
int *rate_cost, int sharpness) {
MACROBLOCKD *xd = &x->e_mbd;
const struct macroblock_plane *p = &x->plane[plane];
const SCAN_ORDER *scan_order =
get_scan(tx_size, get_primary_tx_type(tx_type));
const int16_t *scan = scan_order->scan;
const int shift = av1_get_tx_scale(tx_size);
int eob = p->eobs[block];
const int32_t *dequant = p->dequant_QTX;
const qm_val_t *iqmatrix =
av1_get_iqmatrix(&cpi->common.quant_params, xd, plane, tx_size, tx_type);
const int block_offset = BLOCK_OFFSET(block);
tran_low_t *qcoeff = p->qcoeff + block_offset;
tran_low_t *dqcoeff = p->dqcoeff + block_offset;
const tran_low_t *tcoeff = p->coeff + block_offset;
const CoeffCosts *coeff_costs = &x->coeff_costs;
// This function is not called if eob = 0.
assert(eob > 0);
const AV1_COMMON *cm = &cpi->common;
const PLANE_TYPE plane_type = get_plane_type(plane);
const TX_SIZE txs_ctx = get_txsize_entropy_ctx(tx_size);
const TX_CLASS tx_class = tx_type_to_class[get_primary_tx_type(tx_type)];
const MB_MODE_INFO *mbmi = xd->mi[0];
const int bwl = get_txb_bwl(tx_size);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
assert(width == (1 << bwl));
const int is_inter = is_inter_block(mbmi, xd->tree_type);
const int bob_code = p->bobs[block];
const int is_fsc = (xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
plane == PLANE_TYPE_Y) ||
use_inter_fsc(&cpi->common, plane, tx_type, is_inter);
const LV_MAP_COEFF_COST *txb_costs =
&coeff_costs->coeff_costs[txs_ctx][plane_type];
const int eob_multi_size = txsize_log2_minus4[tx_size];
const LV_MAP_EOB_COST *txb_eob_costs =
&coeff_costs->eob_costs[eob_multi_size][plane_type];
const LV_MAP_COEFF_COST *txb_costs_ph =
&coeff_costs->coeff_costs[0][plane_type];
bool enable_parity_hiding = cm->features.allow_parity_hiding &&
!xd->lossless[xd->mi[0]->segment_id] &&
plane == PLANE_TYPE_Y &&
get_primary_tx_type(tx_type) < IDTX;
coeff_info *coef_info = aom_malloc(width * height * sizeof(coeff_info));
for (int scan_idx = 0; scan_idx < eob; scan_idx++) {
coef_info[scan_idx].tunable = false;
coef_info[scan_idx].upround = false;
}
const int rshift =
(sharpness +
(cpi->oxcf.q_cfg.aq_mode == VARIANCE_AQ && mbmi->segment_id < 4
? 7 - mbmi->segment_id
: 2) +
(cpi->oxcf.q_cfg.aq_mode != VARIANCE_AQ &&
cpi->oxcf.q_cfg.deltaq_mode == DELTA_Q_PERCEPTUAL &&
cm->delta_q_info.delta_q_present_flag && x->sb_energy_level < 0
? (3 - x->sb_energy_level)
: 0));
const int64_t rdmult =
(((int64_t)x->rdmult *
(plane_rd_mult[is_inter][plane_type] << (2 * (xd->bd - 8)))) +
2) >>
rshift;
uint8_t levels_buf[TX_PAD_2D];
uint8_t *const levels = set_levels(levels_buf, width);
if (eob > 1) av1_txb_init_levels(qcoeff, width, height, levels);
// TODO(angirbird): check iqmatrix
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = txb_ctx->txb_skip_ctx;
int non_skip_cost = 0;
int skip_cost = 0;
if (plane == AOM_PLANE_V) {
txb_skip_ctx +=
(x->plane[AOM_PLANE_U].eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
non_skip_cost = txb_costs->v_txb_skip_cost[txb_skip_ctx][0];
skip_cost = txb_costs->v_txb_skip_cost[txb_skip_ctx][1];
} else {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
non_skip_cost = txb_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][0];
skip_cost = txb_costs->txb_skip_cost[pred_mode_ctx][txb_skip_ctx][1];
#else
non_skip_cost = txb_costs->txb_skip_cost[txb_skip_ctx][0];
skip_cost = txb_costs->txb_skip_cost[txb_skip_ctx][1];
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
}
#else
const int non_skip_cost = txb_costs->txb_skip_cost[txb_ctx->txb_skip_ctx][0];
const int skip_cost = txb_costs->txb_skip_cost[txb_ctx->txb_skip_ctx][1];
#endif // CONFIG_CONTEXT_DERIVATION
#if CONFIG_EOB_POS_LUMA
const int eob_cost = get_eob_cost(eob, txb_eob_costs, txb_costs, is_inter);
#else
const int eob_cost = get_eob_cost(eob, txb_eob_costs, txb_costs);
#endif // CONFIG_EOB_POS_LUMA
int accu_rate = eob_cost;
int64_t accu_dist = 0;
int si = eob - 1;
const int ci = scan[si];
const tran_low_t qc = qcoeff[ci];
const tran_low_t abs_qc = abs(qc);
const int sign = qc < 0;
const int max_nz_num = 2;
int nz_num = 1;
int nz_ci[3] = { ci, 0, 0 };
if (abs_qc >= 2) {
update_coeff_general(&accu_rate, &accu_dist, si, eob, tx_class, bwl, height,
rdmult, shift, txb_ctx->dc_sign_ctx, dequant, scan,
txb_costs, tcoeff, qcoeff, dqcoeff, levels, iqmatrix
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane, coef_info, enable_parity_hiding);
--si;
} else {
assert(abs_qc == 1);
const int coeff_ctx = get_lower_levels_ctx_eob(bwl, height, si);
accu_rate +=
get_coeff_cost_eob(ci, abs_qc, sign, coeff_ctx, txb_ctx->dc_sign_ctx,
txb_costs, bwl, tx_class
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane);
const tran_low_t tqc = tcoeff[ci];
const tran_low_t dqc = dqcoeff[ci];
const int64_t dist = get_coeff_dist(tqc, dqc, shift);
const int64_t dist0 = get_coeff_dist(tqc, 0, shift);
accu_dist += dist - dist0;
--si;
}
for (; si >= 0 && nz_num <= max_nz_num; --si) {
update_coeff_eob(&accu_rate, &accu_dist, &eob, &nz_num, nz_ci, si, tx_size,
#if CONFIG_EOB_POS_LUMA
is_inter,
#endif // CONFIG_EOB_POS_LUMA
tx_class, txb_ctx->dc_sign_ctx, rdmult, shift, dequant,
scan, txb_eob_costs, txb_costs, tcoeff, qcoeff, dqcoeff,
levels, sharpness, iqmatrix
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif
,
plane, coef_info, enable_parity_hiding);
}
if (si == -1 && nz_num <= max_nz_num) {
update_skip(&accu_rate, accu_dist, &eob, nz_num, nz_ci, rdmult, skip_cost,
non_skip_cost, qcoeff, dqcoeff, sharpness);
}
for (; si >= 1; --si) {
update_coeff_simple(&accu_rate, si, eob, tx_class, bwl, rdmult, shift,
dequant, scan, txb_costs, tcoeff, qcoeff, dqcoeff,
levels, iqmatrix, coef_info, enable_parity_hiding,
plane);
}
// DC position
if (si == 0) {
// no need to update accu_dist because it's not used after this point
int64_t dummy_dist = 0;
update_coeff_general(&accu_rate, &dummy_dist, si, eob, tx_class, bwl,
height, rdmult, shift, txb_ctx->dc_sign_ctx, dequant,
scan, txb_costs, tcoeff, qcoeff, dqcoeff, levels,
iqmatrix
#if CONFIG_CONTEXT_DERIVATION
,
xd->tmp_sign
#endif // CONFIG_CONTEXT_DERIVATION
,
plane, coef_info, enable_parity_hiding);
}
if (enable_parity_hiding) {
parity_hide_tb(eob, scan, levels, bwl, rdmult, shift, txb_costs,
txb_costs_ph, dequant, iqmatrix, txb_ctx->dc_sign_ctx,
tx_class, qcoeff, dqcoeff, tcoeff, coef_info, &accu_rate
#if CONFIG_CHROMA_TX_COEFF_CODING
,
plane
#endif // CONFIG_CHROMA_TX_COEFF_CODING
);
}
aom_free(coef_info);
set_bob(x, plane, block, tx_size, tx_type);
if (eob == 0) {
accu_rate += skip_cost;
} else {
const int tx_type_cost = get_tx_type_cost(x, xd, plane, tx_size, tx_type,
cm->features.reduced_tx_set_used,
eob, bob_code, is_fsc);
accu_rate += non_skip_cost + tx_type_cost;
}
p->eobs[block] = eob;
p->txb_entropy_ctx[block] =
av1_get_txb_entropy_context(qcoeff, scan_order, p->eobs[block]);
accu_rate += get_cctx_type_cost(cm, x, xd, plane, tx_size, block, cctx_type);
*rate_cost = accu_rate;
return eob;
}
uint8_t av1_get_txb_entropy_context(const tran_low_t *qcoeff,
const SCAN_ORDER *scan_order, int eob) {
const int16_t *const scan = scan_order->scan;
int cul_level = 0;
int c;
if (eob == 0) return 0;
for (c = 0; c < eob; ++c) {
cul_level += abs(qcoeff[scan[c]]);
if (cul_level > COEFF_CONTEXT_MASK) break;
}
cul_level = AOMMIN(COEFF_CONTEXT_MASK, cul_level);
set_dc_sign(&cul_level, qcoeff[0]);
return (uint8_t)cul_level;
}
// Update counts of cctx types
static void update_cctx_type_count(const AV1_COMMON *cm, MACROBLOCKD *xd,
int blk_row, int blk_col, TX_SIZE tx_size,
FRAME_COUNTS *counts,
uint8_t allow_update_cdf) {
const MB_MODE_INFO *mbmi = xd->mi[0];
FRAME_CONTEXT *fc = xd->tile_ctx;
#if !CONFIG_ENTROPY_STATS
(void)counts;
#endif // !CONFIG_ENTROPY_STATS
if (cm->quant_params.base_qindex > 0 &&
!mbmi->skip_txfm[xd->tree_type == CHROMA_PART] &&
!segfeature_active(&cm->seg, mbmi->segment_id, SEG_LVL_SKIP)) {
const CctxType cctx_type = av1_get_cctx_type(xd, blk_row, blk_col);
int above_cctx, left_cctx;
#if CONFIG_EXT_RECUR_PARTITIONS
get_above_and_left_cctx_type(cm, xd, &above_cctx, &left_cctx);
#else
get_above_and_left_cctx_type(cm, xd, tx_size, &above_cctx, &left_cctx);
#endif // CONFIG_EXT_RECUR_PARTITIONS
const int cctx_ctx = get_cctx_context(xd, &above_cctx, &left_cctx);
if (allow_update_cdf)
update_cdf(fc->cctx_type_cdf[txsize_sqr_map[tx_size]][cctx_ctx],
cctx_type, CCTX_TYPES);
#if CONFIG_ENTROPY_STATS
++counts->cctx_type[txsize_sqr_map[tx_size]][cctx_ctx][cctx_type];
#endif // CONFIG_ENTROPY_STATS
}
}
// This function updates the cdf for a 'secondary tx set'
static void update_sec_tx_set_cdf(FRAME_CONTEXT *fc, MB_MODE_INFO *mbmi,
TX_TYPE tx_type) {
uint8_t stx_set_flag = get_secondary_tx_set(tx_type);
if (get_primary_tx_type(tx_type) == ADST_ADST) stx_set_flag -= IST_DIR_SIZE;
assert(stx_set_flag < IST_DIR_SIZE);
uint8_t intra_mode = mbmi->mode;
uint8_t stx_set_ctx = stx_transpose_mapping[intra_mode];
assert(stx_set_ctx < IST_DIR_SIZE);
update_cdf(fc->stx_set_cdf[stx_set_ctx], (int8_t)stx_set_flag, IST_DIR_SIZE);
}
static void update_tx_type_count(const AV1_COMP *cpi, const AV1_COMMON *cm,
MACROBLOCKD *xd, int blk_row, int blk_col,
int plane, TX_SIZE tx_size,
FRAME_COUNTS *counts, uint8_t allow_update_cdf,
int eob, int bob_code, int is_fsc) {
MB_MODE_INFO *mbmi = xd->mi[0];
int is_inter = is_inter_block(mbmi, xd->tree_type);
const int reduced_tx_set_used = cm->features.reduced_tx_set_used;
FRAME_CONTEXT *fc = xd->tile_ctx;
#if !CONFIG_ENTROPY_STATS
(void)counts;
#endif // !CONFIG_ENTROPY_STATS
// Only y plane's tx_type is updated
if (plane > 0) return;
const TX_TYPE tx_type = av1_get_tx_type(xd, PLANE_TYPE_Y, blk_row, blk_col,
tx_size, reduced_tx_set_used);
if (is_inter) {
if (cpi->oxcf.txfm_cfg.use_inter_dct_only) {
assert(tx_type == DCT_DCT);
}
} else {
if (cpi->oxcf.txfm_cfg.use_intra_dct_only) {
assert(get_primary_tx_type(tx_type) == DCT_DCT);
} else if (cpi->oxcf.txfm_cfg.use_intra_default_tx_only) {
const TX_TYPE default_type = get_default_tx_type(
PLANE_TYPE_Y, xd, tx_size, cpi->is_screen_content_type);
(void)default_type;
assert(get_primary_tx_type(tx_type) == default_type);
}
}
if (get_ext_tx_types(tx_size, is_inter, reduced_tx_set_used) > 1 &&
cm->quant_params.base_qindex > 0 &&
!mbmi->skip_txfm[xd->tree_type == CHROMA_PART] &&
!segfeature_active(&cm->seg, mbmi->segment_id, SEG_LVL_SKIP)) {
const int eset = get_ext_tx_set(tx_size, is_inter, reduced_tx_set_used);
if (eset > 0) {
const TxSetType tx_set_type =
av1_get_ext_tx_set_type(tx_size, is_inter, reduced_tx_set_used);
if (is_inter) {
const int esc_eob = is_fsc ? bob_code : eob;
const int eob_tx_ctx =
get_lp2tx_ctx(tx_size, get_txb_bwl(tx_size), esc_eob);
if (allow_update_cdf) {
update_cdf(
fc->inter_ext_tx_cdf[eset][eob_tx_ctx][txsize_sqr_map[tx_size]],
av1_ext_tx_ind[tx_set_type][tx_type],
av1_num_ext_tx_set[tx_set_type]);
}
#if CONFIG_ENTROPY_STATS
++counts->inter_ext_tx[eset][eob_tx_ctx][txsize_sqr_map[tx_size]]
[av1_ext_tx_ind[tx_set_type][tx_type]];
#endif // CONFIG_ENTROPY_STATS
} else {
if (mbmi->fsc_mode[xd->tree_type == CHROMA_PART] && allow_update_cdf) {
return;
}
if (eob == 1 && allow_update_cdf) return;
PREDICTION_MODE intra_dir;
if (mbmi->filter_intra_mode_info.use_filter_intra)
intra_dir = fimode_to_intradir[mbmi->filter_intra_mode_info
.filter_intra_mode];
else
intra_dir = mbmi->mode;
#if CONFIG_ENTROPY_STATS
const TX_TYPE primary_tx_type = get_primary_tx_type(tx_type);
++counts->intra_ext_tx[eset][txsize_sqr_map[tx_size]][intra_dir]
[av1_tx_type_to_idx(primary_tx_type, tx_set_type,
intra_dir,
av1_size_class[tx_size])];
#endif // CONFIG_ENTROPY_STATS
if (allow_update_cdf) {
update_cdf(
fc->intra_ext_tx_cdf[eset + cm->features.reduced_tx_set_used]
[txsize_sqr_map[tx_size]][intra_dir],
av1_tx_type_to_idx(get_primary_tx_type(tx_type), tx_set_type,
intra_dir, av1_size_class[tx_size]),
cm->features.reduced_tx_set_used
? av1_num_reduced_tx_set
: av1_num_ext_tx_set_intra[tx_set_type]);
// Modified condition for CDF update
if (cm->seq_params.enable_ist &&
block_signals_sec_tx_type(xd, tx_size, tx_type, eob)) {
update_cdf(fc->stx_cdf[txsize_sqr_map[tx_size]],
(int8_t)get_secondary_tx_type(tx_type), STX_TYPES);
#if CONFIG_IST_SET_FLAG
if (get_secondary_tx_type(tx_type) > 0)
update_sec_tx_set_cdf(fc, mbmi, tx_type);
#endif // CONFIG_IST_SET_FLAG
}
}
}
}
}
// CDF update for txsize_sqr_up_map[tx_size] >= TX_32X32
else if (!is_inter && cm->quant_params.base_qindex > 0 &&
!mbmi->skip_txfm[xd->tree_type == CHROMA_PART] &&
!segfeature_active(&cm->seg, mbmi->segment_id, SEG_LVL_SKIP) &&
cm->seq_params.enable_ist &&
block_signals_sec_tx_type(xd, tx_size, tx_type, eob)) {
if (eob == 1 && allow_update_cdf) return;
if (allow_update_cdf) {
update_cdf(fc->stx_cdf[txsize_sqr_map[tx_size]],
(int8_t)get_secondary_tx_type(tx_type), STX_TYPES);
#if CONFIG_IST_SET_FLAG
if (get_secondary_tx_type(tx_type) > 0)
update_sec_tx_set_cdf(fc, mbmi, tx_type);
#endif // CONFIG_IST_SET_FLAG
}
}
}
void av1_update_and_record_txb_skip_context(int plane, int block, int blk_row,
int blk_col, BLOCK_SIZE plane_bsize,
TX_SIZE tx_size, void *arg) {
struct tokenize_b_args *const args = arg;
const AV1_COMP *cpi = args->cpi;
const AV1_COMMON *cm = &cpi->common;
ThreadData *const td = args->td;
MACROBLOCK *const x = &td->mb;
MACROBLOCKD *const xd = &x->e_mbd;
struct macroblock_plane *p = &x->plane[plane];
struct macroblockd_plane *pd = &xd->plane[plane];
const int eob = p->eobs[block];
const int bob_code = p->bobs[block];
const int block_offset = BLOCK_OFFSET(block);
tran_low_t *qcoeff = p->qcoeff + block_offset;
const PLANE_TYPE plane_type = pd->plane_type;
const TX_TYPE tx_type =
av1_get_tx_type(xd, plane_type, blk_row, blk_col, tx_size,
cm->features.reduced_tx_set_used);
const SCAN_ORDER *const scan_order = get_scan(tx_size, tx_type);
tran_low_t *tcoeff;
MB_MODE_INFO *mbmi = xd->mi[0];
#if CONFIG_EOB_POS_LUMA
int is_inter = is_inter_block(mbmi, xd->tree_type);
#endif // CONFIG_EOB_POS_LUMA
assert(args->dry_run != DRY_RUN_COSTCOEFFS);
if (args->dry_run == OUTPUT_ENABLED) {
TXB_CTX txb_ctx;
get_txb_ctx(plane_bsize, tx_size, plane,
pd->above_entropy_context + blk_col,
pd->left_entropy_context + blk_row, &txb_ctx,
mbmi->fsc_mode[xd->tree_type == CHROMA_PART]);
const int bwl = get_txb_bwl(tx_size);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
const uint8_t allow_update_cdf = args->allow_update_cdf;
const TX_SIZE txsize_ctx = get_txsize_entropy_ctx(tx_size);
FRAME_CONTEXT *ec_ctx = xd->tile_ctx;
#if CONFIG_ENTROPY_STATS
int cdf_idx = cm->coef_cdf_category;
++td->counts->txb_skip[cdf_idx][txsize_ctx][txb_ctx.txb_skip_ctx][eob == 0];
#endif // CONFIG_ENTROPY_STATS
if (allow_update_cdf) {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
#if !CONFIG_EOB_POS_LUMA
const int is_inter = is_inter_block(mbmi, xd->tree_type);
#endif // CONFIG_EOB_POS_LUMA
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
update_cdf(
ec_ctx->txb_skip_cdf[pred_mode_ctx][txsize_ctx][txb_ctx.txb_skip_ctx],
eob == 0, 2);
#else
update_cdf(ec_ctx->txb_skip_cdf[txsize_ctx][txb_ctx.txb_skip_ctx],
eob == 0, 2);
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
}
CB_COEFF_BUFFER *cb_coef_buff = x->cb_coef_buff;
const int txb_offset =
x->mbmi_ext_frame->cb_offset[plane] / (TX_SIZE_W_MIN * TX_SIZE_H_MIN);
uint16_t *eob_txb = cb_coef_buff->eobs[plane] + txb_offset;
uint8_t *const entropy_ctx = cb_coef_buff->entropy_ctx[plane] + txb_offset;
entropy_ctx[block] = txb_ctx.txb_skip_ctx;
eob_txb[block] = eob;
uint16_t *bob_txb = cb_coef_buff->bobs[plane] + txb_offset;
bob_txb[block] = bob_code;
if (eob == 0) {
av1_set_entropy_contexts(xd, pd, plane, plane_bsize, tx_size, 0, blk_col,
blk_row);
return;
}
assert(eob == av1_get_max_eob(tx_size));
const int segment_id = mbmi->segment_id;
const int seg_eob = av1_get_tx_eob(&cpi->common.seg, segment_id, tx_size);
tran_low_t *tcoeff_txb =
cb_coef_buff->tcoeff[plane] + x->mbmi_ext_frame->cb_offset[plane];
tcoeff = tcoeff_txb + block_offset;
memcpy(tcoeff, qcoeff, sizeof(*tcoeff) * seg_eob);
uint8_t levels_buf[TX_PAD_2D];
uint8_t *const levels = set_levels(levels_buf, width);
int8_t signs_buf[TX_PAD_2D];
int8_t *const signs = set_signs(signs_buf, width);
av1_txb_init_levels_signs(tcoeff, width, height, levels_buf, signs_buf);
update_tx_type_count(cpi, cm, xd, blk_row, blk_col, plane, tx_size,
td->counts, allow_update_cdf, eob, bob_code,
1 /* is_fsc */);
const int16_t *const scan = scan_order->scan;
// record tx type usage
td->rd_counts.tx_type_used[tx_size][get_primary_tx_type(tx_type)]++;
int bob = av1_get_max_eob(tx_size) - bob_code;
#if CONFIG_ENTROPY_STATS
av1_update_eob_context(cdf_idx, bob_code, tx_size,
#if CONFIG_EOB_POS_LUMA
is_inter,
#endif // CONFIG_EOB_POS_LUMA
plane_type, ec_ctx, td->counts, allow_update_cdf);
#else
av1_update_eob_context(bob_code, tx_size,
#if CONFIG_EOB_POS_LUMA
is_inter,
#endif // CONFIG_EOB_POS_LUMA
plane_type, ec_ctx, allow_update_cdf);
#endif
DECLARE_ALIGNED(16, int8_t, coeff_contexts[MAX_TX_SQUARE]);
av1_get_nz_map_contexts_skip_c(levels, scan, bob, eob, tx_size,
coeff_contexts);
for (int c = bob; c < eob; ++c) {
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = qcoeff[pos];
const tran_low_t level = abs(v);
if (allow_update_cdf) {
if (c == bob) {
update_cdf(ec_ctx->coeff_base_bob_cdf[coeff_ctx],
AOMMIN(level, 3) - 1, 3);
} else {
update_cdf(ec_ctx->coeff_base_cdf_idtx[coeff_ctx], AOMMIN(level, 3),
4);
}
}
#if CONFIG_ENTROPY_STATS
if (c == bob) {
++td->counts
->coeff_base_bob_multi[cdf_idx][coeff_ctx][AOMMIN(level, 3) - 1];
} else {
++td->counts
->coeff_base_multi_skip[cdf_idx][coeff_ctx][AOMMIN(level, 3)];
}
#endif
if (level > NUM_BASE_LEVELS) {
const int base_range = level - 1 - NUM_BASE_LEVELS;
const int br_ctx = get_br_ctx_skip(levels, pos, bwl);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
if (allow_update_cdf) {
update_cdf(ec_ctx->coeff_br_cdf_idtx[br_ctx], k, BR_CDF_SIZE);
}
for (int lps = 0; lps < BR_CDF_SIZE - 1; lps++) {
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_skip[lps][br_ctx][lps == k];
#endif // CONFIG_ENTROPY_STATS
if (lps == k) break;
}
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_multi_skip[cdf_idx][br_ctx][k];
#endif // CONFIG_ENTROPY_STATS
if (k < BR_CDF_SIZE - 1) break;
}
}
}
for (int c = eob - 1; c >= 0; --c) {
const int pos = scan[c];
const tran_low_t v = qcoeff[pos];
const tran_low_t level = abs(v);
const int idtx_sign = (v < 0) ? 1 : 0;
if (level) {
int idtx_sign_ctx = get_sign_ctx_skip(signs, levels, pos, bwl);
#if CONFIG_ENTROPY_STATS
++td->counts->idtx_sign[cdf_idx][idtx_sign_ctx][idtx_sign];
#endif // CONFIG_ENTROPY_STATS
if (allow_update_cdf)
update_cdf(ec_ctx->idtx_sign_cdf[idtx_sign_ctx], idtx_sign, 2);
}
}
} else {
tcoeff = qcoeff;
}
const uint8_t cul_level =
av1_get_txb_entropy_context(tcoeff, scan_order, eob);
av1_set_entropy_contexts(xd, pd, plane, plane_bsize, tx_size, cul_level,
blk_col, blk_row);
}
void update_coeff_ctx_hiden(TX_CLASS tx_class, const int16_t *scan, int bwl,
uint8_t *levels, int level,
base_cdf_arr base_cdf_ph, br_cdf_arr br_cdf_ph
#if CONFIG_ENTROPY_STATS
,
ThreadData *const td, int cdf_idx
#endif // CONFIG_ENTROPY_STATS
) {
const int q_index = (level >> 1);
const int pos = scan[0];
int coeff_ctx = get_base_ctx_ph(levels, pos, bwl, tx_class);
update_cdf(base_cdf_ph[coeff_ctx], AOMMIN(q_index, 3), 4);
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_base_ph_multi[cdf_idx][coeff_ctx][AOMMIN(level, 3)];
#endif // CONFIG_ENTROPY_STATS
if (q_index > NUM_BASE_LEVELS) {
int br_ctx = get_par_br_ctx(levels, pos, bwl, tx_class);
aom_cdf_prob *cdf_br = br_cdf_ph[br_ctx];
const int base_range = q_index - 1 - NUM_BASE_LEVELS;
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
update_cdf(cdf_br, k, BR_CDF_SIZE);
for (int lps = 0; lps < BR_CDF_SIZE - 1; lps++) {
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_ph[lps][br_ctx][lps == k];
#endif // CONFIG_ENTROPY_STATS
if (lps == k) break;
}
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_ph_multi[cdf_idx][br_ctx][k];
#endif // // CONFIG_ENTROPY_STATS
if (k < BR_CDF_SIZE - 1) break;
}
}
}
void av1_update_and_record_txb_context(int plane, int block, int blk_row,
int blk_col, BLOCK_SIZE plane_bsize,
TX_SIZE tx_size, void *arg) {
struct tokenize_b_args *const args = arg;
const AV1_COMP *cpi = args->cpi;
const AV1_COMMON *cm = &cpi->common;
ThreadData *const td = args->td;
MACROBLOCK *const x = &td->mb;
MACROBLOCKD *const xd = &x->e_mbd;
struct macroblock_plane *p = &x->plane[plane];
struct macroblockd_plane *pd = &xd->plane[plane];
const int eob = p->eobs[block];
const int bob_code = p->bobs[block];
const int block_offset = BLOCK_OFFSET(block);
tran_low_t *qcoeff = p->qcoeff + block_offset;
const PLANE_TYPE plane_type = pd->plane_type;
const int is_inter = is_inter_block(xd->mi[0], xd->tree_type);
if (eob == 1 && plane_type == 0 &&
!xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] && !is_inter) {
update_txk_array(xd, blk_row, blk_col, tx_size, DCT_DCT);
}
const TX_TYPE tx_type =
av1_get_tx_type(xd, plane_type, blk_row, blk_col, tx_size,
cm->features.reduced_tx_set_used);
if ((xd->mi[0]->fsc_mode[xd->tree_type == CHROMA_PART] &&
get_primary_tx_type(tx_type) == IDTX && plane == PLANE_TYPE_Y) ||
use_inter_fsc(cm, plane, tx_type,
is_inter_block(xd->mi[0], xd->tree_type))) {
av1_update_and_record_txb_skip_context(plane, block, blk_row, blk_col,
plane_bsize, tx_size, arg);
return;
}
const SCAN_ORDER *const scan_order = get_scan(tx_size, tx_type);
tran_low_t *tcoeff;
assert(args->dry_run != DRY_RUN_COSTCOEFFS);
if (args->dry_run == OUTPUT_ENABLED) {
MB_MODE_INFO *mbmi = xd->mi[0];
TXB_CTX txb_ctx;
get_txb_ctx(plane_bsize, tx_size, plane,
pd->above_entropy_context + blk_col,
pd->left_entropy_context + blk_row, &txb_ctx, 0);
#if CCTX_C2_DROPPED
if (plane == AOM_PLANE_V && is_cctx_allowed(cm, xd)) {
CctxType cctx_type = av1_get_cctx_type(xd, blk_row, blk_col);
if (!keep_chroma_c2(cctx_type)) {
assert(eob == 0);
CB_COEFF_BUFFER *cb_coef_buff = x->cb_coef_buff;
const int txb_offset =
x->mbmi_ext_frame
->cb_offset[(plane > 0 && xd->tree_type == CHROMA_PART) ? 1
: 0] >>
(MIN_TX_SIZE_LOG2 * 2);
uint16_t *eob_txb = cb_coef_buff->eobs[plane] + txb_offset;
uint8_t *const entropy_ctx =
cb_coef_buff->entropy_ctx[plane] + txb_offset;
entropy_ctx[block] = txb_ctx.txb_skip_ctx;
eob_txb[block] = 0;
av1_set_entropy_contexts(xd, pd, plane, plane_bsize, tx_size, 0,
blk_col, blk_row);
return;
}
}
#endif // CCTX_C2_DROPPED
const int bwl = get_txb_bwl(tx_size);
const int width = get_txb_wide(tx_size);
const int height = get_txb_high(tx_size);
const uint8_t allow_update_cdf = args->allow_update_cdf;
const TX_SIZE txsize_ctx = get_txsize_entropy_ctx(tx_size);
FRAME_CONTEXT *ec_ctx = xd->tile_ctx;
#if CONFIG_ENTROPY_STATS
int cdf_idx = cm->coef_cdf_category;
#if CONFIG_CONTEXT_DERIVATION
if (plane == AOM_PLANE_Y || plane == AOM_PLANE_U) {
++td->counts
->txb_skip[cdf_idx][txsize_ctx][txb_ctx.txb_skip_ctx][eob == 0];
} else {
++td->counts->v_txb_skip[cdf_idx][txb_ctx.txb_skip_ctx][eob == 0];
}
#else
++td->counts->txb_skip[cdf_idx][txsize_ctx][txb_ctx.txb_skip_ctx][eob == 0];
#endif
#endif // CONFIG_ENTROPY_STATS
if (allow_update_cdf) {
#if CONFIG_CONTEXT_DERIVATION
int txb_skip_ctx = txb_ctx.txb_skip_ctx;
if (plane == AOM_PLANE_Y || plane == AOM_PLANE_U) {
#if CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
const int pred_mode_ctx =
(is_inter || mbmi->fsc_mode[xd->tree_type == CHROMA_PART]) ? 1 : 0;
update_cdf(
ec_ctx->txb_skip_cdf[pred_mode_ctx][txsize_ctx][txb_skip_ctx],
eob == 0, 2);
#else
update_cdf(ec_ctx->txb_skip_cdf[txsize_ctx][txb_skip_ctx], eob == 0, 2);
#endif // CONFIG_TX_SKIP_FLAG_MODE_DEP_CTX
} else {
txb_skip_ctx +=
(x->plane[AOM_PLANE_U].eobs[block] ? V_TXB_SKIP_CONTEXT_OFFSET : 0);
update_cdf(ec_ctx->v_txb_skip_cdf[txb_skip_ctx], eob == 0, 2);
}
#else
update_cdf(ec_ctx->txb_skip_cdf[txsize_ctx][txb_ctx.txb_skip_ctx],
eob == 0, 2);
#endif // CONFIG_CONTEXT_DERIVATION
}
CB_COEFF_BUFFER *cb_coef_buff = x->cb_coef_buff;
const int txb_offset =
x->mbmi_ext_frame->cb_offset[plane] / (TX_SIZE_W_MIN * TX_SIZE_H_MIN);
uint16_t *eob_txb = cb_coef_buff->eobs[plane] + txb_offset;
uint8_t *const entropy_ctx = cb_coef_buff->entropy_ctx[plane] + txb_offset;
entropy_ctx[block] = txb_ctx.txb_skip_ctx;
eob_txb[block] = eob;
uint16_t *bob_txb = cb_coef_buff->bobs[plane] + txb_offset;
bob_txb[block] = bob_code;
const int skip_cctx = is_inter ? 0 : (eob == 1);
if (is_cctx_allowed(cm, xd) && plane == AOM_PLANE_U && !skip_cctx &&
eob > 0)
update_cctx_type_count(cm, xd, blk_row, blk_col, tx_size, td->counts,
allow_update_cdf);
if (eob == 0) {
av1_set_entropy_contexts(xd, pd, plane, plane_bsize, tx_size, 0, blk_col,
blk_row);
return;
}
const int segment_id = mbmi->segment_id;
const int seg_eob = av1_get_tx_eob(&cpi->common.seg, segment_id, tx_size);
tran_low_t *tcoeff_txb =
cb_coef_buff->tcoeff[plane] + x->mbmi_ext_frame->cb_offset[plane];
tcoeff = tcoeff_txb + block_offset;
memcpy(tcoeff, qcoeff, sizeof(*tcoeff) * seg_eob);
uint8_t levels_buf[TX_PAD_2D];
uint8_t *const levels = set_levels(levels_buf, width);
av1_txb_init_levels(tcoeff, width, height, levels);
update_tx_type_count(cpi, cm, xd, blk_row, blk_col, plane, tx_size,
td->counts, allow_update_cdf, eob, bob_code,
0 /* is_fsc */);
const TX_CLASS tx_class = tx_type_to_class[get_primary_tx_type(tx_type)];
const int16_t *const scan = scan_order->scan;
// record tx type usage
td->rd_counts.tx_type_used[tx_size][get_primary_tx_type(tx_type)]++;
#if CONFIG_ENTROPY_STATS
av1_update_eob_context(cdf_idx, eob, tx_size,
#if CONFIG_EOB_POS_LUMA
is_inter,
#endif // CONFIG_EOB_POS_LUMA
plane_type, ec_ctx, td->counts, allow_update_cdf);
#else
av1_update_eob_context(eob, tx_size,
#if CONFIG_EOB_POS_LUMA
is_inter,
#endif // CONFIG_EOB_POS_LUMA
plane_type, ec_ctx, allow_update_cdf);
#endif
DECLARE_ALIGNED(16, int8_t, coeff_contexts[MAX_TX_SQUARE]);
av1_get_nz_map_contexts(levels, scan, eob, tx_size, tx_class,
coeff_contexts, plane);
bool enable_parity_hiding = cm->features.allow_parity_hiding &&
!xd->lossless[xd->mi[0]->segment_id] &&
plane == PLANE_TYPE_Y &&
get_primary_tx_type(tx_type) < IDTX;
#if NEWCTX
int state = 0;
#endif
for (int c = eob - 1; c > 0; --c) {
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = qcoeff[pos];
const tran_low_t level = abs(v);
if (allow_update_cdf) {
if (c == eob - 1) {
assert(coeff_ctx < 4);
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
update_cdf(
ec_ctx
->coeff_base_lf_eob_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1, LF_BASE_SYMBOLS - 1);
} else {
update_cdf(
ec_ctx->coeff_base_eob_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3) - 1, 3);
}
} else {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
update_cdf(
ec_ctx->coeff_base_lf_cdf_tcq[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1), LF_BASE_SYMBOLS);
else
update_cdf(
ec_ctx->coeff_base_lf_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1), LF_BASE_SYMBOLS);
#else
update_cdf(
ec_ctx->coeff_base_lf_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1), LF_BASE_SYMBOLS);
#endif
} else {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
update_cdf(
ec_ctx->coeff_base_cdf_tcq[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3), 4);
else
update_cdf(
ec_ctx->coeff_base_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3), 4);
#else
update_cdf(
ec_ctx->coeff_base_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3), 4);
#endif
}
}
}
if (c == eob - 1) {
assert(coeff_ctx < 4);
assert(level > 0);
#if CONFIG_ENTROPY_STATS
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
++td->counts->coeff_base_lf_eob_multi
[cdf_idx][txsize_ctx][plane_type][coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1];
} else {
++td->counts->coeff_base_eob_multi[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx][AOMMIN(level, 3) - 1];
}
} else {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
++td->counts->coeff_base_lf_multi_tcq[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1)];
else
#endif
++td->counts->coeff_base_lf_multi[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1)];
} else {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
++td->counts->coeff_base_multi_tcq[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx][AOMMIN(level, 3)];
else
#endif
++td->counts->coeff_base_multi[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx][AOMMIN(level, 3)];
}
#endif
}
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int base_range = level - 1 - LF_NUM_BASE_LEVELS;
const int br_ctx = get_br_lf_ctx(levels, pos, bwl, tx_class);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
if (allow_update_cdf) {
update_cdf(ec_ctx->coeff_br_lf_cdf[plane_type][br_ctx], k,
BR_CDF_SIZE);
}
for (int lps = 0; lps < BR_CDF_SIZE - 1; lps++) {
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_lf[plane_type][lps][br_ctx][lps == k];
#endif // CONFIG_ENTROPY_STATS
if (lps == k) break;
}
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_lf_multi[cdf_idx][plane_type][br_ctx][k];
#endif
if (k < BR_CDF_SIZE - 1) break;
}
}
} else {
if (level > NUM_BASE_LEVELS) {
const int base_range = level - 1 - NUM_BASE_LEVELS;
const int br_ctx = get_br_ctx(levels, pos, bwl, tx_class);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
if (allow_update_cdf) {
update_cdf(ec_ctx->coeff_br_cdf[plane_type][br_ctx], k,
BR_CDF_SIZE);
}
for (int lps = 0; lps < BR_CDF_SIZE - 1; lps++) {
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps[AOMMIN(txsize_ctx, TX_32X32)][plane_type]
[lps][br_ctx][lps == k];
#endif // CONFIG_ENTROPY_STATS
if (lps == k) break;
}
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_multi[cdf_idx][plane_type][br_ctx][k];
#endif
if (k < BR_CDF_SIZE - 1) break;
}
}
}
#if NEWCTX
#if DQENABLE
if (dq_enable(tx_size, plane))
#endif //DQENABLE
state = tcq_next_state(state, level, limits);
#endif
}
bool is_hidden = false;
int num_nz = 0;
for (int c = eob - 1; c > 0; --c) {
const int pos = scan[c];
num_nz += !!qcoeff[pos];
}
is_hidden = enable_parity_hiding && num_nz >= PHTHRESH;
if (is_hidden) {
if (allow_update_cdf) {
const int level = abs(qcoeff[scan[0]]);
update_coeff_ctx_hiden(tx_class, scan, bwl, levels, level,
ec_ctx->coeff_base_ph_cdf,
ec_ctx->coeff_br_ph_cdf
#if CONFIG_ENTROPY_STATS
,
td, cdf_idx
#endif // CONFIG_ENTROPY_STATS
);
}
} else {
int c = 0;
const int pos = scan[c];
const int coeff_ctx = coeff_contexts[pos];
const tran_low_t v = qcoeff[pos];
const tran_low_t level = abs(v);
if (allow_update_cdf) {
if (c == eob - 1) {
assert(coeff_ctx < 4);
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
update_cdf(
ec_ctx
->coeff_base_lf_eob_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1, LF_BASE_SYMBOLS - 1);
} else {
update_cdf(
ec_ctx->coeff_base_eob_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3) - 1, 3);
}
} else {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
update_cdf(
ec_ctx->coeff_base_lf_cdf_tcq[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1), LF_BASE_SYMBOLS);
else
update_cdf(
ec_ctx->coeff_base_lf_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1), LF_BASE_SYMBOLS);
#else
update_cdf(
ec_ctx->coeff_base_lf_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, LF_BASE_SYMBOLS - 1), LF_BASE_SYMBOLS);
#endif
} else {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
update_cdf(
ec_ctx->coeff_base_cdf_tcq[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3), 4);
else
update_cdf(
ec_ctx->coeff_base_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3), 4);
#else
update_cdf(
ec_ctx->coeff_base_cdf[txsize_ctx][plane_type][coeff_ctx],
AOMMIN(level, 3), 4);
#endif
}
}
}
if (c == eob - 1) {
assert(coeff_ctx < 4);
#if CONFIG_ENTROPY_STATS
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
++td->counts->coeff_base_lf_eob_multi
[cdf_idx][txsize_ctx][plane_type][coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1) - 1];
} else {
++td->counts->coeff_base_eob_multi[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx][AOMMIN(level, 3) - 1];
}
} else {
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
++td->counts->coeff_base_lf_multi_tcq[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1)];
else
#endif
++td->counts->coeff_base_lf_multi[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx]
[AOMMIN(level, LF_BASE_SYMBOLS - 1)];
} else {
#if NEWCTX
if (state >> 1) // quantizer 1, update its ctx
++td->counts->coeff_base_multi_tcq[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx][AOMMIN(level, 3)];
else
#endif
++td->counts->coeff_base_multi[cdf_idx][txsize_ctx][plane_type]
[coeff_ctx][AOMMIN(level, 3)];
}
#endif
}
const int row = pos >> bwl;
const int col = pos - (row << bwl);
int limits = get_lf_limits(row, col, tx_class, plane);
if (limits) {
if (level > LF_NUM_BASE_LEVELS) {
const int base_range = level - 1 - LF_NUM_BASE_LEVELS;
const int br_ctx = get_br_lf_ctx(levels, pos, bwl, tx_class);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
if (allow_update_cdf) {
update_cdf(ec_ctx->coeff_br_lf_cdf[plane_type][br_ctx], k,
BR_CDF_SIZE);
}
for (int lps = 0; lps < BR_CDF_SIZE - 1; lps++) {
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_lf[plane_type][lps][br_ctx][lps == k];
#endif // CONFIG_ENTROPY_STATS
if (lps == k) break;
}
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_lf_multi[cdf_idx][plane_type][br_ctx][k];
#endif
if (k < BR_CDF_SIZE - 1) break;
}
}
} else {
if (level > NUM_BASE_LEVELS) {
const int base_range = level - 1 - NUM_BASE_LEVELS;
const int br_ctx = get_br_ctx(levels, pos, bwl, tx_class);
for (int idx = 0; idx < COEFF_BASE_RANGE; idx += BR_CDF_SIZE - 1) {
const int k = AOMMIN(base_range - idx, BR_CDF_SIZE - 1);
if (allow_update_cdf) {
update_cdf(ec_ctx->coeff_br_cdf[plane_type][br_ctx], k,
BR_CDF_SIZE);
}
for (int lps = 0; lps < BR_CDF_SIZE - 1; lps++) {
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps[AOMMIN(txsize_ctx, TX_32X32)][plane_type]
[lps][br_ctx][lps == k];
#endif // CONFIG_ENTROPY_STATS
if (lps == k) break;
}
#if CONFIG_ENTROPY_STATS
++td->counts->coeff_lps_multi[cdf_idx][plane_type][br_ctx][k];
#endif
if (k < BR_CDF_SIZE - 1) break;
}
}
}
}
// Update the context needed to code the DC sign (if applicable)
if (tcoeff[0] != 0) {
const int dc_sign = (tcoeff[0] < 0) ? 1 : 0;
const int dc_sign_ctx = txb_ctx.dc_sign_ctx;
#if CONFIG_ENTROPY_STATS
#if CONFIG_CONTEXT_DERIVATION
if (allow_update_cdf) {
if (plane == AOM_PLANE_V) {
++td->counts
->v_dc_sign[cdf_idx][xd->tmp_sign[0]][dc_sign_ctx][dc_sign];
} else {
++td->counts->dc_sign[cdf_idx][plane_type][dc_sign_ctx][dc_sign];
}
}
#else
if (allow_update_cdf)
++td->counts->dc_sign[cdf_idx][plane_type][dc_sign_ctx][dc_sign];
#endif // CONFIG_CONTEXT_DERIVATION
#endif // CONFIG_ENTROPY_STATS
#if CONFIG_CONTEXT_DERIVATION
if (allow_update_cdf) {
if (plane == AOM_PLANE_V) {
update_cdf(ec_ctx->v_dc_sign_cdf[xd->tmp_sign[0]][dc_sign_ctx],
dc_sign, 2);
} else {
update_cdf(ec_ctx->dc_sign_cdf[plane_type][dc_sign_ctx], dc_sign, 2);
}
}
#else
if (allow_update_cdf)
update_cdf(ec_ctx->dc_sign_cdf[plane_type][dc_sign_ctx], dc_sign, 2);
#endif // CONFIG_CONTEXT_DERIVATION
entropy_ctx[block] |= dc_sign_ctx << DC_SIGN_CTX_SHIFT;
}
#if CONFIG_CONTEXT_DERIVATION
if (allow_update_cdf && plane == AOM_PLANE_V) {
for (int c = eob - 1; c >= 1; --c) {
int pos = scan[c];
if (tcoeff[pos] != 0) {
int ac_sign = (tcoeff[pos] < 0) ? 1 : 0;
#if CONFIG_ENTROPY_STATS
++td->counts->v_ac_sign[cdf_idx][xd->tmp_sign[pos]][ac_sign];
#endif // CONFIG_ENTROPY_STATS
update_cdf(ec_ctx->v_ac_sign_cdf[xd->tmp_sign[pos]], ac_sign, 2);
}
}
}
#endif // CONFIG_CONTEXT_DERIVATION
} else {
tcoeff = qcoeff;
}
const uint8_t cul_level =
av1_get_txb_entropy_context(tcoeff, scan_order, eob);
av1_set_entropy_contexts(xd, pd, plane, plane_bsize, tx_size, cul_level,
blk_col, blk_row);
}
void av1_update_intra_mb_txb_context(const AV1_COMP *cpi, ThreadData *td,
RUN_TYPE dry_run, BLOCK_SIZE bsize,
uint8_t allow_update_cdf) {
const AV1_COMMON *const cm = &cpi->common;
const int num_planes = av1_num_planes(cm);
MACROBLOCK *const x = &td->mb;
MACROBLOCKD *const xd = &x->e_mbd;
MB_MODE_INFO *const mbmi = xd->mi[0];
struct tokenize_b_args arg = { cpi, td, 0, allow_update_cdf, dry_run };
if (mbmi->skip_txfm[xd->tree_type == CHROMA_PART]) {
assert(bsize == mbmi->sb_type[av1_get_sdp_idx(xd->tree_type)]);
av1_reset_entropy_context(xd, bsize, num_planes);
return;
}
const int plane_start = get_partition_plane_start(xd->tree_type);
const int plane_end = get_partition_plane_end(xd->tree_type, num_planes);
for (int plane = plane_start; plane < plane_end; ++plane) {
if (plane && !xd->is_chroma_ref) break;
const struct macroblockd_plane *const pd = &xd->plane[plane];
const int ss_x = pd->subsampling_x;
const int ss_y = pd->subsampling_y;
const BLOCK_SIZE plane_bsize =
get_mb_plane_block_size(xd, mbmi, plane, ss_x, ss_y);
#if !CONFIG_EXT_RECUR_PARTITIONS
assert(plane_bsize == get_plane_block_size(bsize, ss_x, ss_y));
#endif // !CONFIG_EXT_RECUR_PARTITIONS
av1_foreach_transformed_block_in_plane(
xd, plane_bsize, plane, av1_update_and_record_txb_context, &arg);
}
}
CB_COEFF_BUFFER *av1_get_cb_coeff_buffer(const struct AV1_COMP *cpi, int mi_row,
int mi_col) {
const AV1_COMMON *const cm = &cpi->common;
const int mib_size_log2 = cm->mib_size_log2;
const int stride = (cm->mi_params.mi_cols >> mib_size_log2) + 1;
const int offset =
(mi_row >> mib_size_log2) * stride + (mi_col >> mib_size_log2);
return cpi->coeff_buffer_base + offset;
}