diff options
Diffstat (limited to 'libavcodec/hevc_cabac.c')
-rw-r--r-- | libavcodec/hevc_cabac.c | 675 |
1 files changed, 611 insertions, 64 deletions
diff --git a/libavcodec/hevc_cabac.c b/libavcodec/hevc_cabac.c index f2531d41bd..da7a21373d 100644 --- a/libavcodec/hevc_cabac.c +++ b/libavcodec/hevc_cabac.c @@ -4,20 +4,20 @@ * Copyright (C) 2012 - 2013 Guillaume Martres * Copyright (C) 2012 - 2013 Gildas Cocherel * - * This file is part of Libav. + * This file is part of FFmpeg. * - * Libav is free software; you can redistribute it and/or + * FFmpeg is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either * version 2.1 of the License, or (at your option) any later version. * - * Libav is distributed in the hope that it will be useful, + * FFmpeg is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public - * License along with Libav; if not, write to the Free Software + * License along with FFmpeg; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA */ @@ -323,19 +323,191 @@ static const uint8_t init_values[3][HEVC_CONTEXTS] = { 107, 167, 91, 107, 107, 167, }, }; +static const uint8_t scan_1x1[1] = { + 0, +}; + +static const uint8_t horiz_scan2x2_x[4] = { + 0, 1, 0, 1, +}; + +static const uint8_t horiz_scan2x2_y[4] = { + 0, 0, 1, 1 +}; + +static const uint8_t horiz_scan4x4_x[16] = { + 0, 1, 2, 3, + 0, 1, 2, 3, + 0, 1, 2, 3, + 0, 1, 2, 3, +}; + +static const uint8_t horiz_scan4x4_y[16] = { + 0, 0, 0, 0, + 1, 1, 1, 1, + 2, 2, 2, 2, + 3, 3, 3, 3, +}; + +static const uint8_t horiz_scan8x8_inv[8][8] = { + { 0, 1, 2, 3, 16, 17, 18, 19, }, + { 4, 5, 6, 7, 20, 21, 22, 23, }, + { 8, 9, 10, 11, 24, 25, 26, 27, }, + { 12, 13, 14, 15, 28, 29, 30, 31, }, + { 32, 33, 34, 35, 48, 49, 50, 51, }, + { 36, 37, 38, 39, 52, 53, 54, 55, }, + { 40, 41, 42, 43, 56, 57, 58, 59, }, + { 44, 45, 46, 47, 60, 61, 62, 63, }, +}; + +static const uint8_t diag_scan4x1_x[4] = { + 0, 1, 2, 3, +}; + +static const uint8_t diag_scan1x4_y[4] = { + 0, 1, 2, 3, +}; + +static const uint8_t diag_scan2x2_x[4] = { + 0, 0, 1, 1, +}; + +static const uint8_t diag_scan2x2_y[4] = { + 0, 1, 0, 1, +}; + +static const uint8_t diag_scan2x2_inv[2][2] = { + { 0, 2, }, + { 1, 3, }, +}; + +static const uint8_t diag_scan8x2_x[16] = { + 0, 0, 1, 1, + 2, 2, 3, 3, + 4, 4, 5, 5, + 6, 6, 7, 7, +}; + +static const uint8_t diag_scan8x2_y[16] = { + 0, 1, 0, 1, + 0, 1, 0, 1, + 0, 1, 0, 1, + 0, 1, 0, 1, +}; + +static const uint8_t diag_scan8x2_inv[2][8] = { + { 0, 2, 4, 6, 8, 10, 12, 14, }, + { 1, 3, 5, 7, 9, 11, 13, 15, }, +}; + +static const uint8_t diag_scan2x8_x[16] = { + 0, 0, 1, 0, + 1, 0, 1, 0, + 1, 0, 1, 0, + 1, 0, 1, 1, +}; + +static const uint8_t diag_scan2x8_y[16] = { + 0, 1, 0, 2, + 1, 3, 2, 4, + 3, 5, 4, 6, + 5, 7, 6, 7, +}; + +static const uint8_t diag_scan2x8_inv[8][2] = { + { 0, 2, }, + { 1, 4, }, + { 3, 6, }, + { 5, 8, }, + { 7, 10, }, + { 9, 12, }, + { 11, 14, }, + { 13, 15, }, +}; + +const uint8_t ff_hevc_diag_scan4x4_x[16] = { + 0, 0, 1, 0, + 1, 2, 0, 1, + 2, 3, 1, 2, + 3, 2, 3, 3, +}; + +const uint8_t ff_hevc_diag_scan4x4_y[16] = { + 0, 1, 0, 2, + 1, 0, 3, 2, + 1, 0, 3, 2, + 1, 3, 2, 3, +}; + +static const uint8_t diag_scan4x4_inv[4][4] = { + { 0, 2, 5, 9, }, + { 1, 4, 8, 12, }, + { 3, 7, 11, 14, }, + { 6, 10, 13, 15, }, +}; + +const uint8_t ff_hevc_diag_scan8x8_x[64] = { + 0, 0, 1, 0, + 1, 2, 0, 1, + 2, 3, 0, 1, + 2, 3, 4, 0, + 1, 2, 3, 4, + 5, 0, 1, 2, + 3, 4, 5, 6, + 0, 1, 2, 3, + 4, 5, 6, 7, + 1, 2, 3, 4, + 5, 6, 7, 2, + 3, 4, 5, 6, + 7, 3, 4, 5, + 6, 7, 4, 5, + 6, 7, 5, 6, + 7, 6, 7, 7, +}; + +const uint8_t ff_hevc_diag_scan8x8_y[64] = { + 0, 1, 0, 2, + 1, 0, 3, 2, + 1, 0, 4, 3, + 2, 1, 0, 5, + 4, 3, 2, 1, + 0, 6, 5, 4, + 3, 2, 1, 0, + 7, 6, 5, 4, + 3, 2, 1, 0, + 7, 6, 5, 4, + 3, 2, 1, 7, + 6, 5, 4, 3, + 2, 7, 6, 5, + 4, 3, 7, 6, + 5, 4, 7, 6, + 5, 7, 6, 7, +}; + +static const uint8_t diag_scan8x8_inv[8][8] = { + { 0, 2, 5, 9, 14, 20, 27, 35, }, + { 1, 4, 8, 13, 19, 26, 34, 42, }, + { 3, 7, 12, 18, 25, 33, 41, 48, }, + { 6, 11, 17, 24, 32, 40, 47, 53, }, + { 10, 16, 23, 31, 39, 46, 52, 57, }, + { 15, 22, 30, 38, 45, 51, 56, 60, }, + { 21, 29, 37, 44, 50, 55, 59, 62, }, + { 28, 36, 43, 49, 54, 58, 61, 63, }, +}; + void ff_hevc_save_states(HEVCContext *s, int ctb_addr_ts) { if (s->pps->entropy_coding_sync_enabled_flag && (ctb_addr_ts % s->sps->ctb_width == 2 || (s->sps->ctb_width == 2 && ctb_addr_ts % s->sps->ctb_width == 0))) { - memcpy(s->cabac_state, s->HEVClc.cabac_state, HEVC_CONTEXTS); + memcpy(s->cabac_state, s->HEVClc->cabac_state, HEVC_CONTEXTS); } } static void load_states(HEVCContext *s) { - memcpy(s->HEVClc.cabac_state, s->cabac_state, HEVC_CONTEXTS); + memcpy(s->HEVClc->cabac_state, s->cabac_state, HEVC_CONTEXTS); } static void cabac_reinit(HEVCLocalContext *lc) @@ -345,10 +517,10 @@ static void cabac_reinit(HEVCLocalContext *lc) static void cabac_init_decoder(HEVCContext *s) { - GetBitContext *gb = &s->HEVClc.gb; + GetBitContext *gb = &s->HEVClc->gb; skip_bits(gb, 1); align_get_bits(gb); - ff_init_cabac_decoder(&s->HEVClc.cc, + ff_init_cabac_decoder(&s->HEVClc->cc, gb->buffer + get_bits_count(gb) / 8, (get_bits_left(gb) + 7) / 8); } @@ -370,7 +542,7 @@ static void cabac_init_state(HEVCContext *s) pre ^= pre >> 31; if (pre > 124) pre = 124 + (pre & 1); - s->HEVClc.cabac_state[i] = pre; + s->HEVClc->cabac_state[i] = pre; } } @@ -395,13 +567,19 @@ void ff_hevc_cabac_init(HEVCContext *s, int ctb_addr_ts) } else { if (s->pps->tiles_enabled_flag && s->pps->tile_id[ctb_addr_ts] != s->pps->tile_id[ctb_addr_ts - 1]) { - cabac_reinit(&s->HEVClc); + if (s->threads_number == 1) + cabac_reinit(s->HEVClc); + else + cabac_init_decoder(s); cabac_init_state(s); } if (s->pps->entropy_coding_sync_enabled_flag) { if (ctb_addr_ts % s->sps->ctb_width == 0) { - get_cabac_terminate(&s->HEVClc.cc); - cabac_reinit(&s->HEVClc); + get_cabac_terminate(&s->HEVClc->cc); + if (s->threads_number == 1) + cabac_reinit(s->HEVClc); + else + cabac_init_decoder(s); if (s->sps->ctb_width == 1) cabac_init_state(s); @@ -412,7 +590,7 @@ void ff_hevc_cabac_init(HEVCContext *s, int ctb_addr_ts) } } -#define GET_CABAC(ctx) get_cabac(&s->HEVClc.cc, &s->HEVClc.cabac_state[ctx]) +#define GET_CABAC(ctx) get_cabac(&s->HEVClc->cc, &s->HEVClc->cabac_state[ctx]) int ff_hevc_sao_merge_flag_decode(HEVCContext *s) { @@ -424,7 +602,7 @@ int ff_hevc_sao_type_idx_decode(HEVCContext *s) if (!GET_CABAC(elem_offset[SAO_TYPE_IDX])) return 0; - if (!get_cabac_bypass(&s->HEVClc.cc)) + if (!get_cabac_bypass(&s->HEVClc->cc)) return SAO_BAND; return SAO_EDGE; } @@ -432,10 +610,10 @@ int ff_hevc_sao_type_idx_decode(HEVCContext *s) int ff_hevc_sao_band_position_decode(HEVCContext *s) { int i; - int value = get_cabac_bypass(&s->HEVClc.cc); + int value = get_cabac_bypass(&s->HEVClc->cc); for (i = 0; i < 4; i++) - value = (value << 1) | get_cabac_bypass(&s->HEVClc.cc); + value = (value << 1) | get_cabac_bypass(&s->HEVClc->cc); return value; } @@ -444,26 +622,26 @@ int ff_hevc_sao_offset_abs_decode(HEVCContext *s) int i = 0; int length = (1 << (FFMIN(s->sps->bit_depth, 10) - 5)) - 1; - while (i < length && get_cabac_bypass(&s->HEVClc.cc)) + while (i < length && get_cabac_bypass(&s->HEVClc->cc)) i++; return i; } int ff_hevc_sao_offset_sign_decode(HEVCContext *s) { - return get_cabac_bypass(&s->HEVClc.cc); + return get_cabac_bypass(&s->HEVClc->cc); } int ff_hevc_sao_eo_class_decode(HEVCContext *s) { - int ret = get_cabac_bypass(&s->HEVClc.cc) << 1; - ret |= get_cabac_bypass(&s->HEVClc.cc); + int ret = get_cabac_bypass(&s->HEVClc->cc) << 1; + ret |= get_cabac_bypass(&s->HEVClc->cc); return ret; } int ff_hevc_end_of_slice_flag_decode(HEVCContext *s) { - return get_cabac_terminate(&s->HEVClc.cc); + return get_cabac_terminate(&s->HEVClc->cc); } int ff_hevc_cu_transquant_bypass_flag_decode(HEVCContext *s) @@ -478,9 +656,9 @@ int ff_hevc_skip_flag_decode(HEVCContext *s, int x0, int y0, int x_cb, int y_cb) int x0b = x0 & ((1 << s->sps->log2_ctb_size) - 1); int y0b = y0 & ((1 << s->sps->log2_ctb_size) - 1); - if (s->HEVClc.ctb_left_flag || x0b) + if (s->HEVClc->ctb_left_flag || x0b) inc = !!SAMPLE_CTB(s->skip_flag, x_cb - 1, y_cb); - if (s->HEVClc.ctb_up_flag || y0b) + if (s->HEVClc->ctb_up_flag || y0b) inc += !!SAMPLE_CTB(s->skip_flag, x_cb, y_cb - 1); return GET_CABAC(elem_offset[SKIP_FLAG] + inc); @@ -498,7 +676,7 @@ int ff_hevc_cu_qp_delta_abs(HEVCContext *s) } if (prefix_val >= 5) { int k = 0; - while (k < CABAC_MAX_BIN && get_cabac_bypass(&s->HEVClc.cc)) { + while (k < CABAC_MAX_BIN && get_cabac_bypass(&s->HEVClc->cc)) { suffix_val += 1 << k; k++; } @@ -506,14 +684,14 @@ int ff_hevc_cu_qp_delta_abs(HEVCContext *s) av_log(s->avctx, AV_LOG_ERROR, "CABAC_MAX_BIN : %d\n", k); while (k--) - suffix_val += get_cabac_bypass(&s->HEVClc.cc) << k; + suffix_val += get_cabac_bypass(&s->HEVClc->cc) << k; } return prefix_val + suffix_val; } int ff_hevc_cu_qp_delta_sign_flag(HEVCContext *s) { - return get_cabac_bypass(&s->HEVClc.cc); + return get_cabac_bypass(&s->HEVClc->cc); } int ff_hevc_pred_mode_decode(HEVCContext *s) @@ -529,9 +707,9 @@ int ff_hevc_split_coding_unit_flag_decode(HEVCContext *s, int ct_depth, int x0, int x_cb = x0 >> s->sps->log2_min_cb_size; int y_cb = y0 >> s->sps->log2_min_cb_size; - if (s->HEVClc.ctb_left_flag || x0b) + if (s->HEVClc->ctb_left_flag || x0b) depth_left = s->tab_ct_depth[(y_cb) * s->sps->min_cb_width + x_cb - 1]; - if (s->HEVClc.ctb_up_flag || y0b) + if (s->HEVClc->ctb_up_flag || y0b) depth_top = s->tab_ct_depth[(y_cb - 1) * s->sps->min_cb_width + x_cb]; inc += (depth_left > ct_depth); @@ -545,7 +723,7 @@ int ff_hevc_part_mode_decode(HEVCContext *s, int log2_cb_size) if (GET_CABAC(elem_offset[PART_MODE])) // 1 return PART_2Nx2N; if (log2_cb_size == s->sps->log2_min_cb_size) { - if (s->HEVClc.cu.pred_mode == MODE_INTRA) // 0 + if (s->HEVClc->cu.pred_mode == MODE_INTRA) // 0 return PART_NxN; if (GET_CABAC(elem_offset[PART_MODE] + 1)) // 01 return PART_2NxN; @@ -565,21 +743,21 @@ int ff_hevc_part_mode_decode(HEVCContext *s, int log2_cb_size) if (GET_CABAC(elem_offset[PART_MODE] + 1)) { // 01X, 01XX if (GET_CABAC(elem_offset[PART_MODE] + 3)) // 011 return PART_2NxN; - if (get_cabac_bypass(&s->HEVClc.cc)) // 0101 + if (get_cabac_bypass(&s->HEVClc->cc)) // 0101 return PART_2NxnD; return PART_2NxnU; // 0100 } if (GET_CABAC(elem_offset[PART_MODE] + 3)) // 001 return PART_Nx2N; - if (get_cabac_bypass(&s->HEVClc.cc)) // 0001 + if (get_cabac_bypass(&s->HEVClc->cc)) // 0001 return PART_nRx2N; return PART_nLx2N; // 0000 } int ff_hevc_pcm_flag_decode(HEVCContext *s) { - return get_cabac_terminate(&s->HEVClc.cc); + return get_cabac_terminate(&s->HEVClc->cc); } int ff_hevc_prev_intra_luma_pred_flag_decode(HEVCContext *s) @@ -590,7 +768,7 @@ int ff_hevc_prev_intra_luma_pred_flag_decode(HEVCContext *s) int ff_hevc_mpm_idx_decode(HEVCContext *s) { int i = 0; - while (i < 2 && get_cabac_bypass(&s->HEVClc.cc)) + while (i < 2 && get_cabac_bypass(&s->HEVClc->cc)) i++; return i; } @@ -598,10 +776,10 @@ int ff_hevc_mpm_idx_decode(HEVCContext *s) int ff_hevc_rem_intra_luma_pred_mode_decode(HEVCContext *s) { int i; - int value = get_cabac_bypass(&s->HEVClc.cc); + int value = get_cabac_bypass(&s->HEVClc->cc); for (i = 0; i < 4; i++) - value = (value << 1) | get_cabac_bypass(&s->HEVClc.cc); + value = (value << 1) | get_cabac_bypass(&s->HEVClc->cc); return value; } @@ -611,8 +789,8 @@ int ff_hevc_intra_chroma_pred_mode_decode(HEVCContext *s) if (!GET_CABAC(elem_offset[INTRA_CHROMA_PRED_MODE])) return 4; - ret = get_cabac_bypass(&s->HEVClc.cc) << 1; - ret |= get_cabac_bypass(&s->HEVClc.cc); + ret = get_cabac_bypass(&s->HEVClc->cc) << 1; + ret |= get_cabac_bypass(&s->HEVClc->cc); return ret; } @@ -621,7 +799,7 @@ int ff_hevc_merge_idx_decode(HEVCContext *s) int i = GET_CABAC(elem_offset[MERGE_IDX]); if (i != 0) { - while (i < s->sh.max_num_merge_cand-1 && get_cabac_bypass(&s->HEVClc.cc)) + while (i < s->sh.max_num_merge_cand-1 && get_cabac_bypass(&s->HEVClc->cc)) i++; } return i; @@ -636,7 +814,7 @@ int ff_hevc_inter_pred_idc_decode(HEVCContext *s, int nPbW, int nPbH) { if (nPbW + nPbH == 12) return GET_CABAC(elem_offset[INTER_PRED_IDC] + 4); - if (GET_CABAC(elem_offset[INTER_PRED_IDC] + s->HEVClc.ct.depth)) + if (GET_CABAC(elem_offset[INTER_PRED_IDC] + s->HEVClc->ct.depth)) return PRED_BI; return GET_CABAC(elem_offset[INTER_PRED_IDC] + 4); @@ -651,7 +829,7 @@ int ff_hevc_ref_idx_lx_decode(HEVCContext *s, int num_ref_idx_lx) while (i < max_ctx && GET_CABAC(elem_offset[REF_IDX_L0] + i)) i++; if (i == 2) { - while (i < max && get_cabac_bypass(&s->HEVClc.cc)) + while (i < max && get_cabac_bypass(&s->HEVClc->cc)) i++; } @@ -668,35 +846,35 @@ int ff_hevc_no_residual_syntax_flag_decode(HEVCContext *s) return GET_CABAC(elem_offset[NO_RESIDUAL_DATA_FLAG]); } -int ff_hevc_abs_mvd_greater0_flag_decode(HEVCContext *s) +static av_always_inline int abs_mvd_greater0_flag_decode(HEVCContext *s) { return GET_CABAC(elem_offset[ABS_MVD_GREATER0_FLAG]); } -int ff_hevc_abs_mvd_greater1_flag_decode(HEVCContext *s) +static av_always_inline int abs_mvd_greater1_flag_decode(HEVCContext *s) { return GET_CABAC(elem_offset[ABS_MVD_GREATER1_FLAG] + 1); } -int ff_hevc_mvd_decode(HEVCContext *s) +static av_always_inline int mvd_decode(HEVCContext *s) { int ret = 2; int k = 1; - while (k < CABAC_MAX_BIN && get_cabac_bypass(&s->HEVClc.cc)) { + while (k < CABAC_MAX_BIN && get_cabac_bypass(&s->HEVClc->cc)) { ret += 1 << k; k++; } if (k == CABAC_MAX_BIN) av_log(s->avctx, AV_LOG_ERROR, "CABAC_MAX_BIN : %d\n", k); while (k--) - ret += get_cabac_bypass(&s->HEVClc.cc) << k; - return get_cabac_bypass_sign(&s->HEVClc.cc, -ret); + ret += get_cabac_bypass(&s->HEVClc->cc) << k; + return get_cabac_bypass_sign(&s->HEVClc->cc, -ret); } -int ff_hevc_mvd_sign_flag_decode(HEVCContext *s) +static av_always_inline int mvd_sign_flag_decode(HEVCContext *s) { - return get_cabac_bypass_sign(&s->HEVClc.cc, -1); + return get_cabac_bypass_sign(&s->HEVClc->cc, -1); } int ff_hevc_split_transform_flag_decode(HEVCContext *s, int log2_trafo_size) @@ -736,31 +914,31 @@ int ff_hevc_transform_skip_flag_decode(HEVCContext *s, int c_idx) i++; \ return i; -int ff_hevc_last_significant_coeff_x_prefix_decode(HEVCContext *s, int c_idx, +static av_always_inline int last_significant_coeff_x_prefix_decode(HEVCContext *s, int c_idx, int log2_size) { LAST_SIG_COEFF(LAST_SIGNIFICANT_COEFF_X_PREFIX) } -int ff_hevc_last_significant_coeff_y_prefix_decode(HEVCContext *s, int c_idx, +static av_always_inline int last_significant_coeff_y_prefix_decode(HEVCContext *s, int c_idx, int log2_size) { LAST_SIG_COEFF(LAST_SIGNIFICANT_COEFF_Y_PREFIX) } -int ff_hevc_last_significant_coeff_suffix_decode(HEVCContext *s, +static av_always_inline int last_significant_coeff_suffix_decode(HEVCContext *s, int last_significant_coeff_prefix) { int i; int length = (last_significant_coeff_prefix >> 1) - 1; - int value = get_cabac_bypass(&s->HEVClc.cc); + int value = get_cabac_bypass(&s->HEVClc->cc); for (i = 1; i < length; i++) - value = (value << 1) | get_cabac_bypass(&s->HEVClc.cc); + value = (value << 1) | get_cabac_bypass(&s->HEVClc->cc); return value; } -int ff_hevc_significant_coeff_group_flag_decode(HEVCContext *s, int c_idx, int ctx_cg) +static av_always_inline int significant_coeff_group_flag_decode(HEVCContext *s, int c_idx, int ctx_cg) { int inc; @@ -769,7 +947,7 @@ int ff_hevc_significant_coeff_group_flag_decode(HEVCContext *s, int c_idx, int c return GET_CABAC(elem_offset[SIGNIFICANT_COEFF_GROUP_FLAG] + inc); } -int ff_hevc_significant_coeff_flag_decode(HEVCContext *s, int c_idx, int x_c, int y_c, +static av_always_inline int significant_coeff_flag_decode(HEVCContext *s, int c_idx, int x_c, int y_c, int log2_trafo_size, int scan_idx, int prev_sig) { static const uint8_t ctx_idx_map[] = { @@ -777,7 +955,8 @@ int ff_hevc_significant_coeff_flag_decode(HEVCContext *s, int c_idx, int x_c, in }; int x_cg = x_c >> 2; int y_cg = y_c >> 2; - int sig_ctx, inc; + int sig_ctx; + int inc; if (x_c + y_c == 0) { sig_ctx = 0; @@ -819,7 +998,7 @@ int ff_hevc_significant_coeff_flag_decode(HEVCContext *s, int c_idx, int x_c, in return GET_CABAC(elem_offset[SIGNIFICANT_COEFF_FLAG] + inc); } -int ff_hevc_coeff_abs_level_greater1_flag_decode(HEVCContext *s, int c_idx, int inc) +static av_always_inline int coeff_abs_level_greater1_flag_decode(HEVCContext *s, int c_idx, int inc) { if (c_idx > 0) @@ -828,7 +1007,7 @@ int ff_hevc_coeff_abs_level_greater1_flag_decode(HEVCContext *s, int c_idx, int return GET_CABAC(elem_offset[COEFF_ABS_LEVEL_GREATER1_FLAG] + inc); } -int ff_hevc_coeff_abs_level_greater2_flag_decode(HEVCContext *s, int c_idx, int inc) +static av_always_inline int coeff_abs_level_greater2_flag_decode(HEVCContext *s, int c_idx, int inc) { if (c_idx > 0) inc += 4; @@ -836,37 +1015,405 @@ int ff_hevc_coeff_abs_level_greater2_flag_decode(HEVCContext *s, int c_idx, int return GET_CABAC(elem_offset[COEFF_ABS_LEVEL_GREATER2_FLAG] + inc); } -int ff_hevc_coeff_abs_level_remaining(HEVCContext *s, int base_level, int rc_rice_param) +static av_always_inline int coeff_abs_level_remaining_decode(HEVCContext *s, int base_level, int rc_rice_param) { int prefix = 0; int suffix = 0; int last_coeff_abs_level_remaining; int i; - while (prefix < CABAC_MAX_BIN && get_cabac_bypass(&s->HEVClc.cc)) + while (prefix < CABAC_MAX_BIN && get_cabac_bypass(&s->HEVClc->cc)) prefix++; if (prefix == CABAC_MAX_BIN) av_log(s->avctx, AV_LOG_ERROR, "CABAC_MAX_BIN : %d\n", prefix); if (prefix < 3) { for (i = 0; i < rc_rice_param; i++) - suffix = (suffix << 1) | get_cabac_bypass(&s->HEVClc.cc); + suffix = (suffix << 1) | get_cabac_bypass(&s->HEVClc->cc); last_coeff_abs_level_remaining = (prefix << rc_rice_param) + suffix; } else { int prefix_minus3 = prefix - 3; for (i = 0; i < prefix_minus3 + rc_rice_param; i++) - suffix = (suffix << 1) | get_cabac_bypass(&s->HEVClc.cc); + suffix = (suffix << 1) | get_cabac_bypass(&s->HEVClc->cc); last_coeff_abs_level_remaining = (((1 << prefix_minus3) + 3 - 1) << rc_rice_param) + suffix; } return last_coeff_abs_level_remaining; } -int ff_hevc_coeff_sign_flag(HEVCContext *s, uint8_t nb) +static av_always_inline int coeff_sign_flag_decode(HEVCContext *s, uint8_t nb) { int i; int ret = 0; for (i = 0; i < nb; i++) - ret = (ret << 1) | get_cabac_bypass(&s->HEVClc.cc); + ret = (ret << 1) | get_cabac_bypass(&s->HEVClc->cc); return ret; } + +void ff_hevc_hls_residual_coding(HEVCContext *s, int x0, int y0, + int log2_trafo_size, enum ScanType scan_idx, + int c_idx) +{ +#define GET_COORD(offset, n) \ + do { \ + x_c = (scan_x_cg[offset >> 4] << 2) + scan_x_off[n]; \ + y_c = (scan_y_cg[offset >> 4] << 2) + scan_y_off[n]; \ + } while (0) + HEVCLocalContext *lc = s->HEVClc; + int transform_skip_flag = 0; + + int last_significant_coeff_x, last_significant_coeff_y; + int last_scan_pos; + int n_end; + int num_coeff = 0; + int greater1_ctx = 1; + + int num_last_subset; + int x_cg_last_sig, y_cg_last_sig; + + const uint8_t *scan_x_cg, *scan_y_cg, *scan_x_off, *scan_y_off; + + ptrdiff_t stride = s->frame->linesize[c_idx]; + int hshift = s->sps->hshift[c_idx]; + int vshift = s->sps->vshift[c_idx]; + uint8_t *dst = &s->frame->data[c_idx][(y0 >> vshift) * stride + + ((x0 >> hshift) << s->sps->pixel_shift)]; + DECLARE_ALIGNED(16, int16_t, coeffs[MAX_TB_SIZE * MAX_TB_SIZE]) = {0}; + DECLARE_ALIGNED(8, uint8_t, significant_coeff_group_flag[8][8]) = {{0}}; + + int trafo_size = 1 << log2_trafo_size; + int i; + int qp,shift,add,scale,scale_m; + const uint8_t level_scale[] = { 40, 45, 51, 57, 64, 72 }; + const uint8_t *scale_matrix; + uint8_t dc_scale; + + // Derive QP for dequant + if (!lc->cu.cu_transquant_bypass_flag) { + static const int qp_c[] = { 29, 30, 31, 32, 33, 33, 34, 34, 35, 35, 36, 36, 37, 37 }; + static const uint8_t rem6[51 + 2 * 6 + 1] = { + 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, + 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, + 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, 4, 5, 0, 1, 2, 3, + }; + + static const uint8_t div6[51 + 2 * 6 + 1] = { + 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 3, 3, 3, + 3, 3, 3, 4, 4, 4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 6, 6, 6, 6, 6, 6, + 7, 7, 7, 7, 7, 7, 8, 8, 8, 8, 8, 8, 9, 9, 9, 9, 9, 9, 10, 10, 10, 10, + }; + int qp_y = lc->qp_y; + + if (c_idx == 0) { + qp = qp_y + s->sps->qp_bd_offset; + } else { + int qp_i, offset; + + if (c_idx == 1) + offset = s->pps->cb_qp_offset + s->sh.slice_cb_qp_offset; + else + offset = s->pps->cr_qp_offset + s->sh.slice_cr_qp_offset; + + qp_i = av_clip_c(qp_y + offset, - s->sps->qp_bd_offset, 57); + if (qp_i < 30) + qp = qp_i; + else if (qp_i > 43) + qp = qp_i - 6; + else + qp = qp_c[qp_i - 30]; + + qp += s->sps->qp_bd_offset; + } + + shift = s->sps->bit_depth + log2_trafo_size - 5; + add = 1 << (shift-1); + scale = level_scale[rem6[qp]] << (div6[qp]); + scale_m = 16; // default when no custom scaling lists. + dc_scale = 16; + + if (s->sps->scaling_list_enable_flag) { + const ScalingList *sl = s->pps->scaling_list_data_present_flag ? + &s->pps->scaling_list : &s->sps->scaling_list; + int matrix_id = lc->cu.pred_mode != MODE_INTRA; + + if (log2_trafo_size != 5) + matrix_id = 3 * matrix_id + c_idx; + + scale_matrix = sl->sl[log2_trafo_size - 2][matrix_id]; + if (log2_trafo_size >= 4) + dc_scale = sl->sl_dc[log2_trafo_size - 4][matrix_id]; + } + } + + if (s->pps->transform_skip_enabled_flag && !lc->cu.cu_transquant_bypass_flag && + log2_trafo_size == 2) { + transform_skip_flag = ff_hevc_transform_skip_flag_decode(s, c_idx); + } + + last_significant_coeff_x = + last_significant_coeff_x_prefix_decode(s, c_idx, log2_trafo_size); + last_significant_coeff_y = + last_significant_coeff_y_prefix_decode(s, c_idx, log2_trafo_size); + + if (last_significant_coeff_x > 3) { + int suffix = last_significant_coeff_suffix_decode(s, last_significant_coeff_x); + last_significant_coeff_x = (1 << ((last_significant_coeff_x >> 1) - 1)) * + (2 + (last_significant_coeff_x & 1)) + + suffix; + } + + if (last_significant_coeff_y > 3) { + int suffix = last_significant_coeff_suffix_decode(s, last_significant_coeff_y); + last_significant_coeff_y = (1 << ((last_significant_coeff_y >> 1) - 1)) * + (2 + (last_significant_coeff_y & 1)) + + suffix; + } + + if (scan_idx == SCAN_VERT) + FFSWAP(int, last_significant_coeff_x, last_significant_coeff_y); + + x_cg_last_sig = last_significant_coeff_x >> 2; + y_cg_last_sig = last_significant_coeff_y >> 2; + + switch (scan_idx) { + case SCAN_DIAG: { + int last_x_c = last_significant_coeff_x & 3; + int last_y_c = last_significant_coeff_y & 3; + + scan_x_off = ff_hevc_diag_scan4x4_x; + scan_y_off = ff_hevc_diag_scan4x4_y; + num_coeff = diag_scan4x4_inv[last_y_c][last_x_c]; + if (trafo_size == 4) { + scan_x_cg = scan_1x1; + scan_y_cg = scan_1x1; + } else if (trafo_size == 8) { + num_coeff += diag_scan2x2_inv[y_cg_last_sig][x_cg_last_sig] << 4; + scan_x_cg = diag_scan2x2_x; + scan_y_cg = diag_scan2x2_y; + } else if (trafo_size == 16) { + num_coeff += diag_scan4x4_inv[y_cg_last_sig][x_cg_last_sig] << 4; + scan_x_cg = ff_hevc_diag_scan4x4_x; + scan_y_cg = ff_hevc_diag_scan4x4_y; + } else { // trafo_size == 32 + num_coeff += diag_scan8x8_inv[y_cg_last_sig][x_cg_last_sig] << 4; + scan_x_cg = ff_hevc_diag_scan8x8_x; + scan_y_cg = ff_hevc_diag_scan8x8_y; + } + break; + } + case SCAN_HORIZ: + scan_x_cg = horiz_scan2x2_x; + scan_y_cg = horiz_scan2x2_y; + scan_x_off = horiz_scan4x4_x; + scan_y_off = horiz_scan4x4_y; + num_coeff = horiz_scan8x8_inv[last_significant_coeff_y][last_significant_coeff_x]; + break; + default: //SCAN_VERT + scan_x_cg = horiz_scan2x2_y; + scan_y_cg = horiz_scan2x2_x; + scan_x_off = horiz_scan4x4_y; + scan_y_off = horiz_scan4x4_x; + num_coeff = horiz_scan8x8_inv[last_significant_coeff_x][last_significant_coeff_y]; + break; + } + num_coeff++; + num_last_subset = (num_coeff - 1) >> 4; + + for (i = num_last_subset; i >= 0; i--) { + int n, m; + int x_cg, y_cg, x_c, y_c, pos; + int implicit_non_zero_coeff = 0; + int64_t trans_coeff_level; + int prev_sig = 0; + int offset = i << 4; + + uint8_t significant_coeff_flag_idx[16]; + uint8_t nb_significant_coeff_flag = 0; + + x_cg = scan_x_cg[i]; + y_cg = scan_y_cg[i]; + + if ((i < num_last_subset) && (i > 0)) { + int ctx_cg = 0; + if (x_cg < (1 << (log2_trafo_size - 2)) - 1) + ctx_cg += significant_coeff_group_flag[x_cg + 1][y_cg]; + if (y_cg < (1 << (log2_trafo_size - 2)) - 1) + ctx_cg += significant_coeff_group_flag[x_cg][y_cg + 1]; + + significant_coeff_group_flag[x_cg][y_cg] = + significant_coeff_group_flag_decode(s, c_idx, ctx_cg); + implicit_non_zero_coeff = 1; + } else { + significant_coeff_group_flag[x_cg][y_cg] = + ((x_cg == x_cg_last_sig && y_cg == y_cg_last_sig) || + (x_cg == 0 && y_cg == 0)); + } + + last_scan_pos = num_coeff - offset - 1; + + if (i == num_last_subset) { + n_end = last_scan_pos - 1; + significant_coeff_flag_idx[0] = last_scan_pos; + nb_significant_coeff_flag = 1; + } else { + n_end = 15; + } + + if (x_cg < ((1 << log2_trafo_size) - 1) >> 2) + prev_sig = significant_coeff_group_flag[x_cg + 1][y_cg]; + if (y_cg < ((1 << log2_trafo_size) - 1) >> 2) + prev_sig += (significant_coeff_group_flag[x_cg][y_cg + 1] << 1); + + for (n = n_end; n >= 0; n--) { + GET_COORD(offset, n); + + if (significant_coeff_group_flag[x_cg][y_cg] && + (n > 0 || implicit_non_zero_coeff == 0)) { + if (significant_coeff_flag_decode(s, c_idx, x_c, y_c, log2_trafo_size, scan_idx, prev_sig) == 1) { + significant_coeff_flag_idx[nb_significant_coeff_flag] = n; + nb_significant_coeff_flag++; + implicit_non_zero_coeff = 0; + } + } else { + int last_cg = (x_c == (x_cg << 2) && y_c == (y_cg << 2)); + if (last_cg && implicit_non_zero_coeff && significant_coeff_group_flag[x_cg][y_cg]) { + significant_coeff_flag_idx[nb_significant_coeff_flag] = n; + nb_significant_coeff_flag++; + } + } + } + + n_end = nb_significant_coeff_flag; + + + if (n_end) { + int first_nz_pos_in_cg = 16; + int last_nz_pos_in_cg = -1; + int c_rice_param = 0; + int first_greater1_coeff_idx = -1; + uint8_t coeff_abs_level_greater1_flag[16] = {0}; + uint16_t coeff_sign_flag; + int sum_abs = 0; + int sign_hidden = 0; + + // initialize first elem of coeff_bas_level_greater1_flag + int ctx_set = (i > 0 && c_idx == 0) ? 2 : 0; + + if (!(i == num_last_subset) && greater1_ctx == 0) + ctx_set++; + greater1_ctx = 1; + last_nz_pos_in_cg = significant_coeff_flag_idx[0]; + + for (m = 0; m < (n_end > 8 ? 8 : n_end); m++) { + int n_idx = significant_coeff_flag_idx[m]; + int inc = (ctx_set << 2) + greater1_ctx; + coeff_abs_level_greater1_flag[n_idx] = + coeff_abs_level_greater1_flag_decode(s, c_idx, inc); + if (coeff_abs_level_greater1_flag[n_idx]) { + greater1_ctx = 0; + } else if (greater1_ctx > 0 && greater1_ctx < 3) { + greater1_ctx++; + } + + if (coeff_abs_level_greater1_flag[n_idx] && + first_greater1_coeff_idx == -1) + first_greater1_coeff_idx = n_idx; + } + first_nz_pos_in_cg = significant_coeff_flag_idx[n_end - 1]; + sign_hidden = (last_nz_pos_in_cg - first_nz_pos_in_cg >= 4 && + !lc->cu.cu_transquant_bypass_flag); + + if (first_greater1_coeff_idx != -1) { + coeff_abs_level_greater1_flag[first_greater1_coeff_idx] += coeff_abs_level_greater2_flag_decode(s, c_idx, ctx_set); + } + if (!s->pps->sign_data_hiding_flag || !sign_hidden ) { + coeff_sign_flag = coeff_sign_flag_decode(s, nb_significant_coeff_flag) << (16 - nb_significant_coeff_flag); + } else { + coeff_sign_flag = coeff_sign_flag_decode(s, nb_significant_coeff_flag - 1) << (16 - (nb_significant_coeff_flag - 1)); + } + + for (m = 0; m < n_end; m++) { + n = significant_coeff_flag_idx[m]; + GET_COORD(offset, n); + trans_coeff_level = 1 + coeff_abs_level_greater1_flag[n]; + if (trans_coeff_level == ((m < 8) ? + ((n == first_greater1_coeff_idx) ? 3 : 2) : 1)) { + int last_coeff_abs_level_remaining = coeff_abs_level_remaining_decode(s, trans_coeff_level, c_rice_param); + + trans_coeff_level += last_coeff_abs_level_remaining; + if (trans_coeff_level > (3 << c_rice_param)) + c_rice_param = FFMIN(c_rice_param + 1, 4); + + } + if (s->pps->sign_data_hiding_flag && sign_hidden) { + sum_abs += trans_coeff_level; + if (n == first_nz_pos_in_cg && (sum_abs&1)) + trans_coeff_level = -trans_coeff_level; + } + if (coeff_sign_flag >> 15) + trans_coeff_level = -trans_coeff_level; + coeff_sign_flag <<= 1; + if(!lc->cu.cu_transquant_bypass_flag) { + if(s->sps->scaling_list_enable_flag) { + if(y_c || x_c || log2_trafo_size < 4) { + switch(log2_trafo_size) { + case 3: pos = (y_c << 3) + x_c; break; + case 4: pos = ((y_c >> 1) << 3) + (x_c >> 1); break; + case 5: pos = ((y_c >> 2) << 3) + (x_c >> 2); break; + default: pos = (y_c << 2) + x_c; + } + scale_m = scale_matrix[pos]; + } else { + scale_m = dc_scale; + } + } + trans_coeff_level = (trans_coeff_level * (int64_t)scale * (int64_t)scale_m + add) >> shift; + if(trans_coeff_level < 0) { + if((~trans_coeff_level) & 0xFffffffffff8000) + trans_coeff_level = -32768; + } else { + if(trans_coeff_level & 0xffffffffffff8000) + trans_coeff_level = 32767; + } + } + coeffs[y_c * trafo_size + x_c] = trans_coeff_level; + } + } + } + + if (lc->cu.cu_transquant_bypass_flag) { + s->hevcdsp.transquant_bypass[log2_trafo_size-2](dst, coeffs, stride); + } else { + if (transform_skip_flag) + s->hevcdsp.transform_skip(dst, coeffs, stride); + else if (lc->cu.pred_mode == MODE_INTRA && c_idx == 0 && log2_trafo_size == 2) + s->hevcdsp.transform_4x4_luma_add(dst, coeffs, stride); + else + s->hevcdsp.transform_add[log2_trafo_size-2](dst, coeffs, stride); + } +} + +void ff_hevc_hls_mvd_coding(HEVCContext *s, int x0, int y0, int log2_cb_size) +{ + HEVCLocalContext *lc = s->HEVClc; + int x = abs_mvd_greater0_flag_decode(s); + int y = abs_mvd_greater0_flag_decode(s); + + if (x) + x += abs_mvd_greater1_flag_decode(s); + if (y) + y += abs_mvd_greater1_flag_decode(s); + + switch (x) { + case 2: lc->pu.mvd.x = mvd_decode(s); break; + case 1: lc->pu.mvd.x = mvd_sign_flag_decode(s); break; + case 0: lc->pu.mvd.x = 0; break; + } + + switch (y) { + case 2: lc->pu.mvd.y = mvd_decode(s); break; + case 1: lc->pu.mvd.y = mvd_sign_flag_decode(s); break; + case 0: lc->pu.mvd.y = 0; break; + } +} + |