* Copyright (C) 2012 - 2013 Gildas Cocherel
* Copyright (C) 2013 Vittorio Giovara
*
- * 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
*/
#include "libavutil/imgutils.h"
-
#include "golomb.h"
- #include "hevcdec.h"
#include "hevc_data.h"
+ #include "hevc_ps.h"
static const uint8_t default_scaling_list_intra[] = {
16, 16, 16, 16, 17, 18, 21, 24,
for (i = 0; i < FF_ARRAY_ELEMS(s->pps_list); i++)
if (s->pps_list[i] && ((HEVCPPS*)s->pps_list[i]->data)->sps_id == id)
remove_pps(s, i);
+
+ av_assert0(!(s->sps_list[id] && s->sps == (HEVCSPS*)s->sps_list[id]->data));
}
av_buffer_unref(&s->sps_list[id]);
}
if (rps_predict) {
const ShortTermRPS *rps_ridx;
- int delta_rps, abs_delta_rps;
+ int delta_rps;
+ unsigned abs_delta_rps;
uint8_t use_delta_flag = 0;
uint8_t delta_rps_sign;
delta_rps_sign = get_bits1(gb);
abs_delta_rps = get_ue_golomb_long(gb) + 1;
+ if (abs_delta_rps < 1 || abs_delta_rps > 32768) {
+ av_log(avctx, AV_LOG_ERROR,
+ "Invalid value of abs_delta_rps: %d\n",
+ abs_delta_rps);
+ return AVERROR_INVALIDDATA;
+ }
delta_rps = (1 - (delta_rps_sign << 1)) * abs_delta_rps;
for (i = 0; i <= rps_ridx->num_delta_pocs; i++) {
int used = rps->used[k] = get_bits1(gb);
rps->num_negative_pics = get_ue_golomb_long(gb);
nb_positive_pics = get_ue_golomb_long(gb);
- if (rps->num_negative_pics >= MAX_REFS ||
- nb_positive_pics >= MAX_REFS) {
+ if (rps->num_negative_pics >= HEVC_MAX_REFS ||
+ nb_positive_pics >= HEVC_MAX_REFS) {
av_log(avctx, AV_LOG_ERROR, "Too many refs in a short term RPS.\n");
return AVERROR_INVALIDDATA;
}
}
-static void decode_profile_tier_level(GetBitContext *gb, AVCodecContext *avctx,
+static int decode_profile_tier_level(GetBitContext *gb, AVCodecContext *avctx,
PTLCommon *ptl)
{
int i;
+ if (get_bits_left(gb) < 2+1+5 + 32 + 4 + 16 + 16 + 12)
+ return -1;
+
ptl->profile_space = get_bits(gb, 2);
ptl->tier_flag = get_bits1(gb);
ptl->profile_idc = get_bits(gb, 5);
av_log(avctx, AV_LOG_DEBUG, "Main 10 profile bitstream\n");
else if (ptl->profile_idc == FF_PROFILE_HEVC_MAIN_STILL_PICTURE)
av_log(avctx, AV_LOG_DEBUG, "Main Still Picture profile bitstream\n");
+ else if (ptl->profile_idc == FF_PROFILE_HEVC_REXT)
+ av_log(avctx, AV_LOG_DEBUG, "Range Extension profile bitstream\n");
else
av_log(avctx, AV_LOG_WARNING, "Unknown HEVC profile: %d\n", ptl->profile_idc);
skip_bits(gb, 16); // XXX_reserved_zero_44bits[0..15]
skip_bits(gb, 16); // XXX_reserved_zero_44bits[16..31]
skip_bits(gb, 12); // XXX_reserved_zero_44bits[32..43]
+
+ return 0;
}
-static void parse_ptl(GetBitContext *gb, AVCodecContext *avctx,
+static int parse_ptl(GetBitContext *gb, AVCodecContext *avctx,
PTL *ptl, int max_num_sub_layers)
{
int i;
- decode_profile_tier_level(gb, avctx, &ptl->general_ptl);
+ if (decode_profile_tier_level(gb, avctx, &ptl->general_ptl) < 0 ||
+ get_bits_left(gb) < 8 + (8*2 * (max_num_sub_layers - 1 > 0))) {
+ av_log(avctx, AV_LOG_ERROR, "PTL information too short\n");
+ return -1;
+ }
+
ptl->general_ptl.level_idc = get_bits(gb, 8);
for (i = 0; i < max_num_sub_layers - 1; i++) {
ptl->sub_layer_profile_present_flag[i] = get_bits1(gb);
ptl->sub_layer_level_present_flag[i] = get_bits1(gb);
}
- if (max_num_sub_layers - 1 > 0)
+
+ if (max_num_sub_layers - 1> 0)
for (i = max_num_sub_layers - 1; i < 8; i++)
skip_bits(gb, 2); // reserved_zero_2bits[i]
for (i = 0; i < max_num_sub_layers - 1; i++) {
- if (ptl->sub_layer_profile_present_flag[i])
- decode_profile_tier_level(gb, avctx, &ptl->sub_layer_ptl[i]);
- if (ptl->sub_layer_level_present_flag[i])
- ptl->sub_layer_ptl[i].level_idc = get_bits(gb, 8);
+ if (ptl->sub_layer_profile_present_flag[i] &&
+ decode_profile_tier_level(gb, avctx, &ptl->sub_layer_ptl[i]) < 0) {
+ av_log(avctx, AV_LOG_ERROR,
+ "PTL information for sublayer %i too short\n", i);
+ return -1;
+ }
+ if (ptl->sub_layer_level_present_flag[i]) {
+ if (get_bits_left(gb) < 8) {
+ av_log(avctx, AV_LOG_ERROR,
+ "Not enough data for sublayer %i level_idc\n", i);
+ return -1;
+ } else
+ ptl->sub_layer_ptl[i].level_idc = get_bits(gb, 8);
+ }
}
+
+ return 0;
}
static void decode_sublayer_hrd(GetBitContext *gb, unsigned int nb_cpb,
}
}
-static void decode_hrd(GetBitContext *gb, int common_inf_present,
+static int decode_hrd(GetBitContext *gb, int common_inf_present,
int max_sublayers)
{
int nal_params_present = 0, vcl_params_present = 0;
else
low_delay = get_bits1(gb);
- if (!low_delay)
+ if (!low_delay) {
nb_cpb = get_ue_golomb_long(gb) + 1;
+ if (nb_cpb < 1 || nb_cpb > 32) {
+ av_log(NULL, AV_LOG_ERROR, "nb_cpb %d invalid\n", nb_cpb);
+ return AVERROR_INVALIDDATA;
+ }
+ }
if (nal_params_present)
decode_sublayer_hrd(gb, nb_cpb, subpic_params_present);
if (vcl_params_present)
decode_sublayer_hrd(gb, nb_cpb, subpic_params_present);
}
+ return 0;
}
int ff_hevc_decode_nal_vps(GetBitContext *gb, AVCodecContext *avctx,
{
int i,j;
int vps_id = 0;
+ ptrdiff_t nal_size;
HEVCVPS *vps;
AVBufferRef *vps_buf = av_buffer_allocz(sizeof(*vps));
av_log(avctx, AV_LOG_DEBUG, "Decoding VPS\n");
+ nal_size = gb->buffer_end - gb->buffer;
+ if (nal_size > sizeof(vps->data)) {
+ av_log(avctx, AV_LOG_WARNING, "Truncating likely oversized VPS "
+ "(%"PTRDIFF_SPECIFIER" > %"SIZE_SPECIFIER")\n",
+ nal_size, sizeof(vps->data));
+ vps->data_size = sizeof(vps->data);
+ } else {
+ vps->data_size = nal_size;
+ }
+ memcpy(vps->data, gb->buffer, vps->data_size);
+
vps_id = get_bits(gb, 4);
if (vps_id >= HEVC_MAX_VPS_COUNT) {
av_log(avctx, AV_LOG_ERROR, "VPS id out of range: %d\n", vps_id);
goto err;
}
- parse_ptl(gb, avctx, &vps->ptl, vps->vps_max_sub_layers);
+ if (parse_ptl(gb, avctx, &vps->ptl, vps->vps_max_sub_layers) < 0)
+ goto err;
vps->vps_sub_layer_ordering_info_present_flag = get_bits1(gb);
vps->vps_num_reorder_pics[i] = get_ue_golomb_long(gb);
vps->vps_max_latency_increase[i] = get_ue_golomb_long(gb) - 1;
- if (vps->vps_max_dec_pic_buffering[i] > MAX_DPB_SIZE || !vps->vps_max_dec_pic_buffering[i]) {
- if (vps->vps_max_dec_pic_buffering[i] > HEVC_MAX_DPB_SIZE) {
++ if (vps->vps_max_dec_pic_buffering[i] > HEVC_MAX_DPB_SIZE || !vps->vps_max_dec_pic_buffering[i]) {
av_log(avctx, AV_LOG_ERROR, "vps_max_dec_pic_buffering_minus1 out of range: %d\n",
vps->vps_max_dec_pic_buffering[i] - 1);
goto err;
vps->vps_max_layer_id = get_bits(gb, 6);
vps->vps_num_layer_sets = get_ue_golomb_long(gb) + 1;
+ if (vps->vps_num_layer_sets < 1 || vps->vps_num_layer_sets > 1024 ||
+ (vps->vps_num_layer_sets - 1LL) * (vps->vps_max_layer_id + 1LL) > get_bits_left(gb)) {
+ av_log(avctx, AV_LOG_ERROR, "too many layer_id_included_flags\n");
+ goto err;
+ }
+
for (i = 1; i < vps->vps_num_layer_sets; i++)
for (j = 0; j <= vps->vps_max_layer_id; j++)
skip_bits(gb, 1); // layer_id_included_flag[i][j]
if (vps->vps_poc_proportional_to_timing_flag)
vps->vps_num_ticks_poc_diff_one = get_ue_golomb_long(gb) + 1;
vps->vps_num_hrd_parameters = get_ue_golomb_long(gb);
+ if (vps->vps_num_hrd_parameters > (unsigned)vps->vps_num_layer_sets) {
+ av_log(avctx, AV_LOG_ERROR,
+ "vps_num_hrd_parameters %d is invalid\n", vps->vps_num_hrd_parameters);
+ goto err;
+ }
for (i = 0; i < vps->vps_num_hrd_parameters; i++) {
int common_inf_present = 1;
}
get_bits1(gb); /* vps_extension_flag */
+ if (get_bits_left(gb) < 0) {
+ av_log(avctx, AV_LOG_ERROR,
+ "Overread VPS by %d bits\n", -get_bits_left(gb));
+ if (ps->vps_list[vps_id])
+ goto err;
+ }
+
if (ps->vps_list[vps_id] &&
!memcmp(ps->vps_list[vps_id]->data, vps_buf->data, vps_buf->size)) {
av_buffer_unref(&vps_buf);
int apply_defdispwin, HEVCSPS *sps)
{
VUI *vui = &sps->vui;
- int sar_present;
+ GetBitContext backup;
+ int sar_present, alt = 0;
av_log(avctx, AV_LOG_DEBUG, "Decoding VUI\n");
vui->transfer_characteristic = AVCOL_TRC_UNSPECIFIED;
if (vui->matrix_coeffs >= AVCOL_SPC_NB)
vui->matrix_coeffs = AVCOL_SPC_UNSPECIFIED;
+ if (vui->matrix_coeffs == AVCOL_SPC_RGB) {
+ switch (sps->pix_fmt) {
+ case AV_PIX_FMT_YUV444P:
+ sps->pix_fmt = AV_PIX_FMT_GBRP;
+ break;
+ case AV_PIX_FMT_YUV444P10:
+ sps->pix_fmt = AV_PIX_FMT_GBRP10;
+ break;
+ case AV_PIX_FMT_YUV444P12:
+ sps->pix_fmt = AV_PIX_FMT_GBRP12;
+ break;
+ }
+ }
}
}
vui->field_seq_flag = get_bits1(gb);
vui->frame_field_info_present_flag = get_bits1(gb);
- vui->default_display_window_flag = get_bits1(gb);
+ if (get_bits_left(gb) >= 68 && show_bits_long(gb, 21) == 0x100000) {
+ vui->default_display_window_flag = 0;
+ av_log(avctx, AV_LOG_WARNING, "Invalid default display window\n");
+ } else
+ vui->default_display_window_flag = get_bits1(gb);
+ // Backup context in case an alternate header is detected
+ memcpy(&backup, gb, sizeof(backup));
+
if (vui->default_display_window_flag) {
- //TODO: * 2 is only valid for 420
- vui->def_disp_win.left_offset = get_ue_golomb_long(gb) * 2;
- vui->def_disp_win.right_offset = get_ue_golomb_long(gb) * 2;
- vui->def_disp_win.top_offset = get_ue_golomb_long(gb) * 2;
- vui->def_disp_win.bottom_offset = get_ue_golomb_long(gb) * 2;
+ int vert_mult = 1 + (sps->chroma_format_idc < 2);
+ int horiz_mult = 1 + (sps->chroma_format_idc < 3);
+ vui->def_disp_win.left_offset = get_ue_golomb_long(gb) * horiz_mult;
+ vui->def_disp_win.right_offset = get_ue_golomb_long(gb) * horiz_mult;
+ vui->def_disp_win.top_offset = get_ue_golomb_long(gb) * vert_mult;
+ vui->def_disp_win.bottom_offset = get_ue_golomb_long(gb) * vert_mult;
if (apply_defdispwin &&
avctx->flags2 & AV_CODEC_FLAG2_IGNORE_CROP) {
}
vui->vui_timing_info_present_flag = get_bits1(gb);
+
if (vui->vui_timing_info_present_flag) {
+ if( get_bits_left(gb) < 66) {
+ // The alternate syntax seem to have timing info located
+ // at where def_disp_win is normally located
+ av_log(avctx, AV_LOG_WARNING,
+ "Strange VUI timing information, retrying...\n");
+ vui->default_display_window_flag = 0;
+ memset(&vui->def_disp_win, 0, sizeof(vui->def_disp_win));
+ memcpy(gb, &backup, sizeof(backup));
+ alt = 1;
+ }
vui->vui_num_units_in_tick = get_bits_long(gb, 32);
vui->vui_time_scale = get_bits_long(gb, 32);
+ if (alt) {
+ av_log(avctx, AV_LOG_INFO, "Retry got %i/%i fps\n",
+ vui->vui_time_scale, vui->vui_num_units_in_tick);
+ }
vui->vui_poc_proportional_to_timing_flag = get_bits1(gb);
if (vui->vui_poc_proportional_to_timing_flag)
vui->vui_num_ticks_poc_diff_one_minus1 = get_ue_golomb_long(gb);
memcpy(sl->sl[2][4], default_scaling_list_inter, 64);
memcpy(sl->sl[2][5], default_scaling_list_inter, 64);
memcpy(sl->sl[3][0], default_scaling_list_intra, 64);
- memcpy(sl->sl[3][1], default_scaling_list_inter, 64);
+ memcpy(sl->sl[3][1], default_scaling_list_intra, 64);
+ memcpy(sl->sl[3][2], default_scaling_list_intra, 64);
+ memcpy(sl->sl[3][3], default_scaling_list_inter, 64);
+ memcpy(sl->sl[3][4], default_scaling_list_inter, 64);
+ memcpy(sl->sl[3][5], default_scaling_list_inter, 64);
}
-static int scaling_list_data(GetBitContext *gb, AVCodecContext *avctx, ScalingList *sl)
+static int scaling_list_data(GetBitContext *gb, AVCodecContext *avctx, ScalingList *sl, HEVCSPS *sps)
{
- uint8_t scaling_list_pred_mode_flag[4][6];
+ uint8_t scaling_list_pred_mode_flag;
int32_t scaling_list_dc_coef[2][6];
- int size_id, matrix_id, i, pos;
+ int size_id, matrix_id, pos;
+ int i;
for (size_id = 0; size_id < 4; size_id++)
- for (matrix_id = 0; matrix_id < (size_id == 3 ? 2 : 6); matrix_id++) {
- scaling_list_pred_mode_flag[size_id][matrix_id] = get_bits1(gb);
- if (!scaling_list_pred_mode_flag[size_id][matrix_id]) {
+ for (matrix_id = 0; matrix_id < 6; matrix_id += ((size_id == 3) ? 3 : 1)) {
+ scaling_list_pred_mode_flag = get_bits1(gb);
+ if (!scaling_list_pred_mode_flag) {
unsigned int delta = get_ue_golomb_long(gb);
/* Only need to handle non-zero delta. Zero means default,
* which should already be in the arrays. */
}
}
+ if (sps->chroma_format_idc == 3) {
+ for (i = 0; i < 64; i++) {
+ sl->sl[3][1][i] = sl->sl[2][1][i];
+ sl->sl[3][2][i] = sl->sl[2][2][i];
+ sl->sl[3][4][i] = sl->sl[2][4][i];
+ sl->sl[3][5][i] = sl->sl[2][5][i];
+ }
+ sl->sl_dc[1][1] = sl->sl_dc[0][1];
+ sl->sl_dc[1][2] = sl->sl_dc[0][2];
+ sl->sl_dc[1][4] = sl->sl_dc[0][4];
+ sl->sl_dc[1][5] = sl->sl_dc[0][5];
+ }
+
+
return 0;
}
static int map_pixel_format(AVCodecContext *avctx, HEVCSPS *sps)
{
const AVPixFmtDescriptor *desc;
- if (sps->chroma_format_idc == 1) {
- switch (sps->bit_depth) {
- case 8: sps->pix_fmt = AV_PIX_FMT_YUV420P; break;
- case 9: sps->pix_fmt = AV_PIX_FMT_YUV420P9; break;
- case 10: sps->pix_fmt = AV_PIX_FMT_YUV420P10; break;
- default:
- av_log(avctx, AV_LOG_ERROR, "Unsupported bit depth: %d\n",
- sps->bit_depth);
- return AVERROR_PATCHWELCOME;
- }
- } else {
+ switch (sps->bit_depth) {
+ case 8:
+ if (sps->chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY8;
+ if (sps->chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P;
+ if (sps->chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P;
+ if (sps->chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P;
+ break;
+ case 9:
+ if (sps->chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY16;
+ if (sps->chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P9;
+ if (sps->chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P9;
+ if (sps->chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P9;
+ break;
+ case 10:
+ if (sps->chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY10;
+ if (sps->chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P10;
+ if (sps->chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P10;
+ if (sps->chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P10;
+ break;
+ case 12:
+ if (sps->chroma_format_idc == 0) sps->pix_fmt = AV_PIX_FMT_GRAY12;
+ if (sps->chroma_format_idc == 1) sps->pix_fmt = AV_PIX_FMT_YUV420P12;
+ if (sps->chroma_format_idc == 2) sps->pix_fmt = AV_PIX_FMT_YUV422P12;
+ if (sps->chroma_format_idc == 3) sps->pix_fmt = AV_PIX_FMT_YUV444P12;
+ break;
+ default:
av_log(avctx, AV_LOG_ERROR,
- "non-4:2:0 support is currently unspecified.\n");
- return AVERROR_PATCHWELCOME;
+ "The following bit-depths are currently specified: 8, 9, 10 and 12 bits, "
+ "chroma_format_idc is %d, depth is %d\n",
+ sps->chroma_format_idc, sps->bit_depth);
+ return AVERROR_INVALIDDATA;
}
desc = av_pix_fmt_desc_get(sps->pix_fmt);
sps->vps_id = get_bits(gb, 4);
if (sps->vps_id >= HEVC_MAX_VPS_COUNT) {
av_log(avctx, AV_LOG_ERROR, "VPS id out of range: %d\n", sps->vps_id);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
if (vps_list && !vps_list[sps->vps_id]) {
av_log(avctx, AV_LOG_ERROR, "VPS %d does not exist\n",
sps->vps_id);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
sps->max_sub_layers = get_bits(gb, 3) + 1;
if (sps->max_sub_layers > HEVC_MAX_SUB_LAYERS) {
av_log(avctx, AV_LOG_ERROR, "sps_max_sub_layers out of range: %d\n",
sps->max_sub_layers);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
skip_bits1(gb); // temporal_id_nesting_flag
- parse_ptl(gb, avctx, &sps->ptl, sps->max_sub_layers);
+ if ((ret = parse_ptl(gb, avctx, &sps->ptl, sps->max_sub_layers)) < 0)
+ return ret;
*sps_id = get_ue_golomb_long(gb);
if (*sps_id >= HEVC_MAX_SPS_COUNT) {
av_log(avctx, AV_LOG_ERROR, "SPS id out of range: %d\n", *sps_id);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
sps->chroma_format_idc = get_ue_golomb_long(gb);
- if (sps->chroma_format_idc != 1) {
- avpriv_report_missing_feature(avctx, "chroma_format_idc %d",
- sps->chroma_format_idc);
- ret = AVERROR_PATCHWELCOME;
- goto err;
+ if (sps->chroma_format_idc > 3U) {
+ av_log(avctx, AV_LOG_ERROR, "chroma_format_idc %d is invalid\n", sps->chroma_format_idc);
+ return AVERROR_INVALIDDATA;
}
if (sps->chroma_format_idc == 3)
sps->separate_colour_plane_flag = get_bits1(gb);
+ if (sps->separate_colour_plane_flag)
+ sps->chroma_format_idc = 0;
+
sps->width = get_ue_golomb_long(gb);
sps->height = get_ue_golomb_long(gb);
if ((ret = av_image_check_size(sps->width,
sps->height, 0, avctx)) < 0)
- goto err;
+ return ret;
if (get_bits1(gb)) { // pic_conformance_flag
- //TODO: * 2 is only valid for 420
- sps->pic_conf_win.left_offset = get_ue_golomb_long(gb) * 2;
- sps->pic_conf_win.right_offset = get_ue_golomb_long(gb) * 2;
- sps->pic_conf_win.top_offset = get_ue_golomb_long(gb) * 2;
- sps->pic_conf_win.bottom_offset = get_ue_golomb_long(gb) * 2;
+ int vert_mult = 1 + (sps->chroma_format_idc < 2);
+ int horiz_mult = 1 + (sps->chroma_format_idc < 3);
+ sps->pic_conf_win.left_offset = get_ue_golomb_long(gb) * horiz_mult;
+ sps->pic_conf_win.right_offset = get_ue_golomb_long(gb) * horiz_mult;
+ sps->pic_conf_win.top_offset = get_ue_golomb_long(gb) * vert_mult;
+ sps->pic_conf_win.bottom_offset = get_ue_golomb_long(gb) * vert_mult;
if (avctx->flags2 & AV_CODEC_FLAG2_IGNORE_CROP) {
av_log(avctx, AV_LOG_DEBUG,
sps->bit_depth = get_ue_golomb_long(gb) + 8;
bit_depth_chroma = get_ue_golomb_long(gb) + 8;
- if (bit_depth_chroma != sps->bit_depth) {
+ if (sps->chroma_format_idc && bit_depth_chroma != sps->bit_depth) {
av_log(avctx, AV_LOG_ERROR,
"Luma bit depth (%d) is different from chroma bit depth (%d), "
"this is unsupported.\n",
sps->bit_depth, bit_depth_chroma);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
-
ret = map_pixel_format(avctx, sps);
if (ret < 0)
- goto err;
+ return ret;
sps->log2_max_poc_lsb = get_ue_golomb_long(gb) + 4;
if (sps->log2_max_poc_lsb > 16) {
av_log(avctx, AV_LOG_ERROR, "log2_max_pic_order_cnt_lsb_minus4 out range: %d\n",
sps->log2_max_poc_lsb - 4);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
sublayer_ordering_info = get_bits1(gb);
sps->temporal_layer[i].max_dec_pic_buffering = get_ue_golomb_long(gb) + 1;
sps->temporal_layer[i].num_reorder_pics = get_ue_golomb_long(gb);
sps->temporal_layer[i].max_latency_increase = get_ue_golomb_long(gb) - 1;
- if (sps->temporal_layer[i].max_dec_pic_buffering > MAX_DPB_SIZE) {
+ if (sps->temporal_layer[i].max_dec_pic_buffering > HEVC_MAX_DPB_SIZE) {
av_log(avctx, AV_LOG_ERROR, "sps_max_dec_pic_buffering_minus1 out of range: %d\n",
sps->temporal_layer[i].max_dec_pic_buffering - 1);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
if (sps->temporal_layer[i].num_reorder_pics > sps->temporal_layer[i].max_dec_pic_buffering - 1) {
av_log(avctx, AV_LOG_WARNING, "sps_max_num_reorder_pics out of range: %d\n",
sps->temporal_layer[i].num_reorder_pics);
if (avctx->err_recognition & AV_EF_EXPLODE ||
- sps->temporal_layer[i].num_reorder_pics > MAX_DPB_SIZE - 1) {
+ sps->temporal_layer[i].num_reorder_pics > HEVC_MAX_DPB_SIZE - 1) {
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
sps->temporal_layer[i].max_dec_pic_buffering = sps->temporal_layer[i].num_reorder_pics + 1;
}
sps->log2_max_trafo_size = log2_diff_max_min_transform_block_size +
sps->log2_min_tb_size;
- if (sps->log2_min_tb_size >= sps->log2_min_cb_size) {
+ if (sps->log2_min_cb_size < 3 || sps->log2_min_cb_size > 30) {
+ av_log(avctx, AV_LOG_ERROR, "Invalid value %d for log2_min_cb_size", sps->log2_min_cb_size);
+ return AVERROR_INVALIDDATA;
+ }
+
+ if (sps->log2_diff_max_min_coding_block_size > 30) {
+ av_log(avctx, AV_LOG_ERROR, "Invalid value %d for log2_diff_max_min_coding_block_size", sps->log2_diff_max_min_coding_block_size);
+ return AVERROR_INVALIDDATA;
+ }
+
+ if (sps->log2_min_tb_size >= sps->log2_min_cb_size || sps->log2_min_tb_size < 2) {
av_log(avctx, AV_LOG_ERROR, "Invalid value for log2_min_tb_size");
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
+
+ if (log2_diff_max_min_transform_block_size < 0 || log2_diff_max_min_transform_block_size > 30) {
+ av_log(avctx, AV_LOG_ERROR, "Invalid value %d for log2_diff_max_min_transform_block_size", log2_diff_max_min_transform_block_size);
+ return AVERROR_INVALIDDATA;
+ }
+
sps->max_transform_hierarchy_depth_inter = get_ue_golomb_long(gb);
sps->max_transform_hierarchy_depth_intra = get_ue_golomb_long(gb);
set_default_scaling_list_data(&sps->scaling_list);
if (get_bits1(gb)) {
- ret = scaling_list_data(gb, avctx, &sps->scaling_list);
+ ret = scaling_list_data(gb, avctx, &sps->scaling_list, sps);
if (ret < 0)
- goto err;
+ return ret;
}
}
av_log(avctx, AV_LOG_ERROR,
"PCM bit depth (%d) is greater than normal bit depth (%d)\n",
sps->pcm.bit_depth, sps->bit_depth);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
sps->pcm.loop_filter_disable_flag = get_bits1(gb);
if (sps->nb_st_rps > HEVC_MAX_SHORT_TERM_RPS_COUNT) {
av_log(avctx, AV_LOG_ERROR, "Too many short term RPS: %d.\n",
sps->nb_st_rps);
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
for (i = 0; i < sps->nb_st_rps; i++) {
if ((ret = ff_hevc_decode_short_term_rps(gb, avctx, &sps->st_rps[i],
sps, 0)) < 0)
- goto err;
+ return ret;
}
sps->long_term_ref_pics_present_flag = get_bits1(gb);
if (sps->long_term_ref_pics_present_flag) {
sps->num_long_term_ref_pics_sps = get_ue_golomb_long(gb);
+ if (sps->num_long_term_ref_pics_sps > 31U) {
+ av_log(avctx, AV_LOG_ERROR, "num_long_term_ref_pics_sps %d is out of range.\n",
+ sps->num_long_term_ref_pics_sps);
+ return AVERROR_INVALIDDATA;
+ }
for (i = 0; i < sps->num_long_term_ref_pics_sps; i++) {
sps->lt_ref_pic_poc_lsb_sps[i] = get_bits(gb, sps->log2_max_poc_lsb);
sps->used_by_curr_pic_lt_sps_flag[i] = get_bits1(gb);
vui_present = get_bits1(gb);
if (vui_present)
decode_vui(gb, avctx, apply_defdispwin, sps);
- skip_bits1(gb); // sps_extension_flag
+ if (get_bits1(gb)) { // sps_extension_flag
+ int sps_extension_flag[1];
+ for (i = 0; i < 1; i++)
+ sps_extension_flag[i] = get_bits1(gb);
+ skip_bits(gb, 7); //sps_extension_7bits = get_bits(gb, 7);
+ if (sps_extension_flag[0]) {
+ int extended_precision_processing_flag;
+ int high_precision_offsets_enabled_flag;
+ int cabac_bypass_alignment_enabled_flag;
+
+ sps->transform_skip_rotation_enabled_flag = get_bits1(gb);
+ sps->transform_skip_context_enabled_flag = get_bits1(gb);
+ sps->implicit_rdpcm_enabled_flag = get_bits1(gb);
+
+ sps->explicit_rdpcm_enabled_flag = get_bits1(gb);
+
+ extended_precision_processing_flag = get_bits1(gb);
+ if (extended_precision_processing_flag)
+ av_log(avctx, AV_LOG_WARNING,
+ "extended_precision_processing_flag not yet implemented\n");
+
+ sps->intra_smoothing_disabled_flag = get_bits1(gb);
+ high_precision_offsets_enabled_flag = get_bits1(gb);
+ if (high_precision_offsets_enabled_flag)
+ av_log(avctx, AV_LOG_WARNING,
+ "high_precision_offsets_enabled_flag not yet implemented\n");
+
+ sps->persistent_rice_adaptation_enabled_flag = get_bits1(gb);
+
+ cabac_bypass_alignment_enabled_flag = get_bits1(gb);
+ if (cabac_bypass_alignment_enabled_flag)
+ av_log(avctx, AV_LOG_WARNING,
+ "cabac_bypass_alignment_enabled_flag not yet implemented\n");
+ }
+ }
if (apply_defdispwin) {
sps->output_window.left_offset += sps->vui.def_disp_win.left_offset;
sps->output_window.right_offset += sps->vui.def_disp_win.right_offset;
(sps->output_window.left_offset + sps->output_window.right_offset);
sps->output_height = sps->height -
(sps->output_window.top_offset + sps->output_window.bottom_offset);
- if (sps->output_width <= 0 || sps->output_height <= 0) {
+ if (sps->width <= sps->output_window.left_offset + (int64_t)sps->output_window.right_offset ||
+ sps->height <= sps->output_window.top_offset + (int64_t)sps->output_window.bottom_offset) {
av_log(avctx, AV_LOG_WARNING, "Invalid visible frame dimensions: %dx%d.\n",
sps->output_width, sps->output_height);
if (avctx->err_recognition & AV_EF_EXPLODE) {
- ret = AVERROR_INVALIDDATA;
- goto err;
+ return AVERROR_INVALIDDATA;
}
av_log(avctx, AV_LOG_WARNING,
"Displaying the whole video surface.\n");
- sps->output_window.left_offset =
- sps->output_window.right_offset =
- sps->output_window.top_offset =
- sps->output_window.bottom_offset = 0;
+ memset(&sps->pic_conf_win, 0, sizeof(sps->pic_conf_win));
+ memset(&sps->output_window, 0, sizeof(sps->output_window));
sps->output_width = sps->width;
sps->output_height = sps->height;
}
sps->log2_diff_max_min_coding_block_size;
sps->log2_min_pu_size = sps->log2_min_cb_size - 1;
- if (sps->log2_ctb_size > MAX_LOG2_CTB_SIZE) {
++ if (sps->log2_ctb_size > HEVC_MAX_LOG2_CTB_SIZE) {
+ av_log(avctx, AV_LOG_ERROR, "CTB size out of range: 2^%d\n", sps->log2_ctb_size);
+ return AVERROR_INVALIDDATA;
+ }
+ if (sps->log2_ctb_size < 4) {
+ av_log(avctx,
+ AV_LOG_ERROR,
+ "log2_ctb_size %d differs from the bounds of any known profile\n",
+ sps->log2_ctb_size);
+ avpriv_request_sample(avctx, "log2_ctb_size %d", sps->log2_ctb_size);
+ return AVERROR_INVALIDDATA;
+ }
+
sps->ctb_width = (sps->width + (1 << sps->log2_ctb_size) - 1) >> sps->log2_ctb_size;
sps->ctb_height = (sps->height + (1 << sps->log2_ctb_size) - 1) >> sps->log2_ctb_size;
sps->ctb_size = sps->ctb_width * sps->ctb_height;
sps->min_tb_height = sps->height >> sps->log2_min_tb_size;
sps->min_pu_width = sps->width >> sps->log2_min_pu_size;
sps->min_pu_height = sps->height >> sps->log2_min_pu_size;
+ sps->tb_mask = (1 << (sps->log2_ctb_size - sps->log2_min_tb_size)) - 1;
sps->qp_bd_offset = 6 * (sps->bit_depth - 8);
- if (sps->width & ((1 << sps->log2_min_cb_size) - 1) ||
- sps->height & ((1 << sps->log2_min_cb_size) - 1)) {
+ if (av_mod_uintp2(sps->width, sps->log2_min_cb_size) ||
+ av_mod_uintp2(sps->height, sps->log2_min_cb_size)) {
av_log(avctx, AV_LOG_ERROR, "Invalid coded frame dimensions.\n");
- goto err;
+ return AVERROR_INVALIDDATA;
}
- if (sps->log2_ctb_size > HEVC_MAX_LOG2_CTB_SIZE) {
- av_log(avctx, AV_LOG_ERROR, "CTB size out of range: 2^%d\n", sps->log2_ctb_size);
- goto err;
- }
if (sps->max_transform_hierarchy_depth_inter > sps->log2_ctb_size - sps->log2_min_tb_size) {
av_log(avctx, AV_LOG_ERROR, "max_transform_hierarchy_depth_inter out of range: %d\n",
sps->max_transform_hierarchy_depth_inter);
- goto err;
+ return AVERROR_INVALIDDATA;
}
if (sps->max_transform_hierarchy_depth_intra > sps->log2_ctb_size - sps->log2_min_tb_size) {
av_log(avctx, AV_LOG_ERROR, "max_transform_hierarchy_depth_intra out of range: %d\n",
sps->max_transform_hierarchy_depth_intra);
- goto err;
+ return AVERROR_INVALIDDATA;
}
if (sps->log2_max_trafo_size > FFMIN(sps->log2_ctb_size, 5)) {
av_log(avctx, AV_LOG_ERROR,
"max transform block size out of range: %d\n",
sps->log2_max_trafo_size);
- goto err;
+ return AVERROR_INVALIDDATA;
}
- return 0;
+ if (get_bits_left(gb) < 0) {
+ av_log(avctx, AV_LOG_ERROR,
+ "Overread SPS by %d bits\n", -get_bits_left(gb));
+ return AVERROR_INVALIDDATA;
+ }
-err:
- return ret < 0 ? ret : AVERROR_INVALIDDATA;
+ return 0;
}
int ff_hevc_decode_nal_sps(GetBitContext *gb, AVCodecContext *avctx,
AVBufferRef *sps_buf = av_buffer_allocz(sizeof(*sps));
unsigned int sps_id;
int ret;
+ ptrdiff_t nal_size;
if (!sps_buf)
return AVERROR(ENOMEM);
av_log(avctx, AV_LOG_DEBUG, "Decoding SPS\n");
+ nal_size = gb->buffer_end - gb->buffer;
+ if (nal_size > sizeof(sps->data)) {
+ av_log(avctx, AV_LOG_WARNING, "Truncating likely oversized SPS "
+ "(%"PTRDIFF_SPECIFIER" > %"SIZE_SPECIFIER")\n",
+ nal_size, sizeof(sps->data));
+ sps->data_size = sizeof(sps->data);
+ } else {
+ sps->data_size = nal_size;
+ }
+ memcpy(sps->data, gb->buffer, sps->data_size);
+
ret = ff_hevc_parse_sps(sps, gb, &sps_id,
apply_defdispwin,
ps->vps_list, avctx);
av_freep(&pps->ctb_addr_ts_to_rs);
av_freep(&pps->tile_pos_rs);
av_freep(&pps->tile_id);
- av_freep(&pps->min_tb_addr_zs);
+ av_freep(&pps->min_tb_addr_zs_tab);
av_freep(&pps);
}
+static int pps_range_extensions(GetBitContext *gb, AVCodecContext *avctx,
+ HEVCPPS *pps, HEVCSPS *sps) {
+ int i;
+
+ if (pps->transform_skip_enabled_flag) {
+ pps->log2_max_transform_skip_block_size = get_ue_golomb_long(gb) + 2;
+ }
+ pps->cross_component_prediction_enabled_flag = get_bits1(gb);
+ pps->chroma_qp_offset_list_enabled_flag = get_bits1(gb);
+ if (pps->chroma_qp_offset_list_enabled_flag) {
+ pps->diff_cu_chroma_qp_offset_depth = get_ue_golomb_long(gb);
+ pps->chroma_qp_offset_list_len_minus1 = get_ue_golomb_long(gb);
+ if (pps->chroma_qp_offset_list_len_minus1 && pps->chroma_qp_offset_list_len_minus1 >= 5) {
+ av_log(avctx, AV_LOG_ERROR,
+ "chroma_qp_offset_list_len_minus1 shall be in the range [0, 5].\n");
+ return AVERROR_INVALIDDATA;
+ }
+ for (i = 0; i <= pps->chroma_qp_offset_list_len_minus1; i++) {
+ pps->cb_qp_offset_list[i] = get_se_golomb_long(gb);
+ if (pps->cb_qp_offset_list[i]) {
+ av_log(avctx, AV_LOG_WARNING,
+ "cb_qp_offset_list not tested yet.\n");
+ }
+ pps->cr_qp_offset_list[i] = get_se_golomb_long(gb);
+ if (pps->cr_qp_offset_list[i]) {
+ av_log(avctx, AV_LOG_WARNING,
+ "cb_qp_offset_list not tested yet.\n");
+ }
+ }
+ }
+ pps->log2_sao_offset_scale_luma = get_ue_golomb_long(gb);
+ pps->log2_sao_offset_scale_chroma = get_ue_golomb_long(gb);
+
+ return(0);
+}
+
static inline int setup_pps(AVCodecContext *avctx, GetBitContext *gb,
HEVCPPS *pps, HEVCSPS *sps)
{
int log2_diff;
- int pic_area_in_ctbs, pic_area_in_min_tbs;
+ int pic_area_in_ctbs;
int i, j, x, y, ctb_addr_rs, tile_id;
// Inferred parameters
* 6.5
*/
pic_area_in_ctbs = sps->ctb_width * sps->ctb_height;
- pic_area_in_min_tbs = sps->min_tb_width * sps->min_tb_height;
pps->ctb_addr_rs_to_ts = av_malloc_array(pic_area_in_ctbs, sizeof(*pps->ctb_addr_rs_to_ts));
pps->ctb_addr_ts_to_rs = av_malloc_array(pic_area_in_ctbs, sizeof(*pps->ctb_addr_ts_to_rs));
pps->tile_id = av_malloc_array(pic_area_in_ctbs, sizeof(*pps->tile_id));
- pps->min_tb_addr_zs = av_malloc_array(pic_area_in_min_tbs, sizeof(*pps->min_tb_addr_zs));
+ pps->min_tb_addr_zs_tab = av_malloc_array((sps->tb_mask+2) * (sps->tb_mask+2), sizeof(*pps->min_tb_addr_zs_tab));
if (!pps->ctb_addr_rs_to_ts || !pps->ctb_addr_ts_to_rs ||
- !pps->tile_id || !pps->min_tb_addr_zs) {
+ !pps->tile_id || !pps->min_tb_addr_zs_tab) {
return AVERROR(ENOMEM);
}
pps->row_bd[j] * sps->ctb_width + pps->col_bd[i];
log2_diff = sps->log2_ctb_size - sps->log2_min_tb_size;
- for (y = 0; y < sps->min_tb_height; y++) {
- for (x = 0; x < sps->min_tb_width; x++) {
+ pps->min_tb_addr_zs = &pps->min_tb_addr_zs_tab[1*(sps->tb_mask+2)+1];
+ for (y = 0; y < sps->tb_mask+2; y++) {
+ pps->min_tb_addr_zs_tab[y*(sps->tb_mask+2)] = -1;
+ pps->min_tb_addr_zs_tab[y] = -1;
+ }
+ for (y = 0; y < sps->tb_mask+1; y++) {
+ for (x = 0; x < sps->tb_mask+1; x++) {
int tb_x = x >> log2_diff;
int tb_y = y >> log2_diff;
int rs = sps->ctb_width * tb_y + tb_x;
int m = 1 << i;
val += (m & x ? m * m : 0) + (m & y ? 2 * m * m : 0);
}
- pps->min_tb_addr_zs[y * sps->min_tb_width + x] = val;
+ pps->min_tb_addr_zs[y * (sps->tb_mask+2) + x] = val;
}
}
HEVCSPS *sps = NULL;
int i, ret = 0;
unsigned int pps_id = 0;
+ ptrdiff_t nal_size;
AVBufferRef *pps_buf;
HEVCPPS *pps = av_mallocz(sizeof(*pps));
av_log(avctx, AV_LOG_DEBUG, "Decoding PPS\n");
+ nal_size = gb->buffer_end - gb->buffer;
+ if (nal_size > sizeof(pps->data)) {
+ av_log(avctx, AV_LOG_WARNING, "Truncating likely oversized PPS "
+ "(%"PTRDIFF_SPECIFIER" > %"SIZE_SPECIFIER")\n",
+ nal_size, sizeof(pps->data));
+ pps->data_size = sizeof(pps->data);
+ } else {
+ pps->data_size = nal_size;
+ }
+ memcpy(pps->data, gb->buffer, pps->data_size);
+
// Default values
pps->loop_filter_across_tiles_enabled_flag = 1;
pps->num_tile_columns = 1;
pps->disable_dbf = 0;
pps->beta_offset = 0;
pps->tc_offset = 0;
+ pps->log2_max_transform_skip_block_size = 2;
// Coded parameters
pps_id = get_ue_golomb_long(gb);
if (pps->cu_qp_delta_enabled_flag)
pps->diff_cu_qp_delta_depth = get_ue_golomb_long(gb);
+ if (pps->diff_cu_qp_delta_depth < 0 ||
+ pps->diff_cu_qp_delta_depth > sps->log2_diff_max_min_coding_block_size) {
+ av_log(avctx, AV_LOG_ERROR, "diff_cu_qp_delta_depth %d is invalid\n",
+ pps->diff_cu_qp_delta_depth);
+ ret = AVERROR_INVALIDDATA;
+ goto err;
+ }
+
pps->cb_qp_offset = get_se_golomb(gb);
if (pps->cb_qp_offset < -12 || pps->cb_qp_offset > 12) {
av_log(avctx, AV_LOG_ERROR, "pps_cb_qp_offset out of range: %d\n",
if (pps->tiles_enabled_flag) {
pps->num_tile_columns = get_ue_golomb_long(gb) + 1;
pps->num_tile_rows = get_ue_golomb_long(gb) + 1;
- if (pps->num_tile_columns == 0 ||
+ if (pps->num_tile_columns <= 0 ||
pps->num_tile_columns >= sps->width) {
av_log(avctx, AV_LOG_ERROR, "num_tile_columns_minus1 out of range: %d\n",
pps->num_tile_columns - 1);
ret = AVERROR_INVALIDDATA;
goto err;
}
- if (pps->num_tile_rows == 0 ||
+ if (pps->num_tile_rows <= 0 ||
pps->num_tile_rows >= sps->height) {
av_log(avctx, AV_LOG_ERROR, "num_tile_rows_minus1 out of range: %d\n",
pps->num_tile_rows - 1);
pps->scaling_list_data_present_flag = get_bits1(gb);
if (pps->scaling_list_data_present_flag) {
set_default_scaling_list_data(&pps->scaling_list);
- ret = scaling_list_data(gb, avctx, &pps->scaling_list);
+ ret = scaling_list_data(gb, avctx, &pps->scaling_list, sps);
if (ret < 0)
goto err;
}
}
pps->slice_header_extension_present_flag = get_bits1(gb);
- skip_bits1(gb); // pps_extension_flag
+
+ if (get_bits1(gb)) { // pps_extension_present_flag
+ int pps_range_extensions_flag = get_bits1(gb);
+ /* int pps_extension_7bits = */ get_bits(gb, 7);
+ if (sps->ptl.general_ptl.profile_idc == FF_PROFILE_HEVC_REXT && pps_range_extensions_flag) {
+ if ((ret = pps_range_extensions(gb, avctx, pps, sps)) < 0)
+ goto err;
+ }
+ }
ret = setup_pps(avctx, gb, pps, sps);
if (ret < 0)
goto err;
+ if (get_bits_left(gb) < 0) {
+ av_log(avctx, AV_LOG_ERROR,
+ "Overread PPS by %d bits\n", -get_bits_left(gb));
+ goto err;
+ }
+
remove_pps(ps, pps_id);
ps->pps_list[pps_id] = pps_buf;
/*
- * HEVC video decoder
+ * HEVC video Decoder
*
* Copyright (C) 2012 - 2013 Guillaume Martres
* Copyright (C) 2012 - 2013 Mickael Raulet
* Copyright (C) 2012 - 2013 Gildas Cocherel
* Copyright (C) 2012 - 2013 Wassim Hamidouche
*
- * 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
*/
#include "libavutil/common.h"
#include "libavutil/display.h"
#include "libavutil/internal.h"
+#include "libavutil/mastering_display_metadata.h"
#include "libavutil/md5.h"
#include "libavutil/opt.h"
#include "libavutil/pixdesc.h"
#include "hevcdec.h"
#include "profiles.h"
-const uint8_t ff_hevc_qpel_extra_before[4] = { 0, 3, 3, 3 };
-const uint8_t ff_hevc_qpel_extra_after[4] = { 0, 4, 4, 4 };
-const uint8_t ff_hevc_qpel_extra[4] = { 0, 7, 7, 7 };
-
-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_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_scan4x4_inv[4][4] = {
- { 0, 2, 5, 9, },
- { 1, 4, 8, 12, },
- { 3, 7, 11, 14, },
- { 6, 10, 13, 15, },
-};
-
-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, },
-};
+const uint8_t ff_hevc_pel_weight[65] = { [2] = 0, [4] = 1, [6] = 2, [8] = 3, [12] = 4, [16] = 5, [24] = 6, [32] = 7, [48] = 8, [64] = 9 };
/**
* NOTE: Each function hls_foo correspond to the function foo in the
av_freep(&s->horizontal_bs);
av_freep(&s->vertical_bs);
+ av_freep(&s->sh.entry_point_offset);
+ av_freep(&s->sh.size);
+ av_freep(&s->sh.offset);
+
av_buffer_pool_uninit(&s->tab_mvf_pool);
av_buffer_pool_uninit(&s->rpl_tab_pool);
}
int ctb_count = sps->ctb_width * sps->ctb_height;
int min_pu_size = sps->min_pu_width * sps->min_pu_height;
- s->bs_width = width >> 3;
- s->bs_height = height >> 3;
+ s->bs_width = (width >> 2) + 1;
+ s->bs_height = (height >> 2) + 1;
s->sao = av_mallocz_array(ctb_count, sizeof(*s->sao));
s->deblock = av_mallocz_array(ctb_count, sizeof(*s->deblock));
if (!s->sao || !s->deblock)
goto fail;
- s->skip_flag = av_malloc(pic_size_in_ctb);
- s->tab_ct_depth = av_malloc(sps->min_cb_height * sps->min_cb_width);
+ s->skip_flag = av_malloc_array(sps->min_cb_height, sps->min_cb_width);
+ s->tab_ct_depth = av_malloc_array(sps->min_cb_height, sps->min_cb_width);
if (!s->skip_flag || !s->tab_ct_depth)
goto fail;
- s->cbf_luma = av_malloc(sps->min_tb_width * sps->min_tb_height);
+ s->cbf_luma = av_malloc_array(sps->min_tb_width, sps->min_tb_height);
s->tab_ipm = av_mallocz(min_pu_size);
- s->is_pcm = av_malloc(min_pu_size);
+ s->is_pcm = av_malloc_array(sps->min_pu_width + 1, sps->min_pu_height + 1);
if (!s->tab_ipm || !s->cbf_luma || !s->is_pcm)
goto fail;
- s->filter_slice_edges = av_malloc(ctb_count);
- s->tab_slice_address = av_malloc(pic_size_in_ctb *
+ s->filter_slice_edges = av_mallocz(ctb_count);
+ s->tab_slice_address = av_malloc_array(pic_size_in_ctb,
sizeof(*s->tab_slice_address));
- s->qp_y_tab = av_malloc(pic_size_in_ctb *
+ s->qp_y_tab = av_malloc_array(pic_size_in_ctb,
sizeof(*s->qp_y_tab));
if (!s->qp_y_tab || !s->filter_slice_edges || !s->tab_slice_address)
goto fail;
- s->horizontal_bs = av_mallocz(2 * s->bs_width * (s->bs_height + 1));
- s->vertical_bs = av_mallocz(2 * s->bs_width * (s->bs_height + 1));
+ s->horizontal_bs = av_mallocz_array(s->bs_width, s->bs_height);
+ s->vertical_bs = av_mallocz_array(s->bs_width, s->bs_height);
if (!s->horizontal_bs || !s->vertical_bs)
goto fail;
s->tab_mvf_pool = av_buffer_pool_init(min_pu_size * sizeof(MvField),
- av_buffer_alloc);
+ av_buffer_allocz);
s->rpl_tab_pool = av_buffer_pool_init(ctb_count * sizeof(RefPicListTab),
av_buffer_allocz);
if (!s->tab_mvf_pool || !s->rpl_tab_pool)
uint8_t chroma_weight_l0_flag[16];
uint8_t luma_weight_l1_flag[16];
uint8_t chroma_weight_l1_flag[16];
+ int luma_log2_weight_denom;
- s->sh.luma_log2_weight_denom = av_clip(get_ue_golomb_long(gb), 0, 7);
+ luma_log2_weight_denom = get_ue_golomb_long(gb);
+ if (luma_log2_weight_denom < 0 || luma_log2_weight_denom > 7)
+ av_log(s->avctx, AV_LOG_ERROR, "luma_log2_weight_denom %d is invalid\n", luma_log2_weight_denom);
+ s->sh.luma_log2_weight_denom = av_clip_uintp2(luma_log2_weight_denom, 3);
if (s->ps.sps->chroma_format_idc != 0) {
int delta = get_se_golomb(gb);
- s->sh.chroma_log2_weight_denom = av_clip(s->sh.luma_log2_weight_denom + delta, 0, 7);
+ s->sh.chroma_log2_weight_denom = av_clip_uintp2(s->sh.luma_log2_weight_denom + delta, 3);
}
for (i = 0; i < s->sh.nb_refs[L0]; i++) {
s->sh.luma_offset_l0[i] = 0;
}
}
- if (s->ps.sps->chroma_format_idc != 0) { // FIXME: invert "if" and "for"
+ if (s->ps.sps->chroma_format_idc != 0) {
for (i = 0; i < s->sh.nb_refs[L0]; i++)
chroma_weight_l0_flag[i] = get_bits1(gb);
} else {
nb_sps = get_ue_golomb_long(gb);
nb_sh = get_ue_golomb_long(gb);
- if (nb_sh + nb_sps > FF_ARRAY_ELEMS(rps->poc))
+ if (nb_sh + (uint64_t)nb_sps > FF_ARRAY_ELEMS(rps->poc))
return AVERROR_INVALIDDATA;
rps->nb_refs = nb_sh + nb_sps;
num, den, 1 << 30);
}
-static int set_sps(HEVCContext *s, const HEVCSPS *sps)
+static int set_sps(HEVCContext *s, const HEVCSPS *sps, enum AVPixelFormat pix_fmt)
{
- #define HWACCEL_MAX (CONFIG_HEVC_DXVA2_HWACCEL + CONFIG_HEVC_D3D11VA_HWACCEL + CONFIG_HEVC_VDPAU_HWACCEL)
+ #define HWACCEL_MAX (CONFIG_HEVC_DXVA2_HWACCEL + CONFIG_HEVC_D3D11VA_HWACCEL + CONFIG_HEVC_VAAPI_HWACCEL + CONFIG_HEVC_VDPAU_HWACCEL)
enum AVPixelFormat pix_fmts[HWACCEL_MAX + 2], *fmt = pix_fmts;
- int ret;
+ int ret, i;
pic_arrays_free(s);
s->ps.sps = NULL;
export_stream_params(s->avctx, &s->ps, sps);
- if (sps->pix_fmt == AV_PIX_FMT_YUV420P || sps->pix_fmt == AV_PIX_FMT_YUVJ420P ||
- sps->pix_fmt == AV_PIX_FMT_YUV420P10) {
+ switch (sps->pix_fmt) {
+ case AV_PIX_FMT_YUV420P:
+ case AV_PIX_FMT_YUVJ420P:
#if CONFIG_HEVC_DXVA2_HWACCEL
*fmt++ = AV_PIX_FMT_DXVA2_VLD;
#endif
- }
- if (sps->pix_fmt == AV_PIX_FMT_YUV420P || sps->pix_fmt == AV_PIX_FMT_YUVJ420P) {
#if CONFIG_HEVC_D3D11VA_HWACCEL
*fmt++ = AV_PIX_FMT_D3D11VA_VLD;
#endif
+#if CONFIG_HEVC_VAAPI_HWACCEL
+ *fmt++ = AV_PIX_FMT_VAAPI;
+#endif
#if CONFIG_HEVC_VDPAU_HWACCEL
*fmt++ = AV_PIX_FMT_VDPAU;
#endif
+ break;
+ case AV_PIX_FMT_YUV420P10:
+#if CONFIG_HEVC_DXVA2_HWACCEL
+ *fmt++ = AV_PIX_FMT_DXVA2_VLD;
+#endif
+#if CONFIG_HEVC_D3D11VA_HWACCEL
+ *fmt++ = AV_PIX_FMT_D3D11VA_VLD;
+#endif
+#if CONFIG_HEVC_VAAPI_HWACCEL
+ *fmt++ = AV_PIX_FMT_VAAPI;
+#endif
+ break;
}
- *fmt++ = sps->pix_fmt;
- *fmt = AV_PIX_FMT_NONE;
+ if (pix_fmt == AV_PIX_FMT_NONE) {
+ *fmt++ = sps->pix_fmt;
+ *fmt = AV_PIX_FMT_NONE;
- ret = ff_get_format(s->avctx, pix_fmts);
- if (ret < 0)
- goto fail;
- s->avctx->pix_fmt = ret;
+ ret = ff_thread_get_format(s->avctx, pix_fmts);
+ if (ret < 0)
+ goto fail;
+ s->avctx->pix_fmt = ret;
+ }
+ else {
+ s->avctx->pix_fmt = pix_fmt;
+ }
ff_hevc_pred_init(&s->hpc, sps->bit_depth);
ff_hevc_dsp_init (&s->hevcdsp, sps->bit_depth);
ff_videodsp_init (&s->vdsp, sps->bit_depth);
+ for (i = 0; i < 3; i++) {
+ av_freep(&s->sao_pixel_buffer_h[i]);
+ av_freep(&s->sao_pixel_buffer_v[i]);
+ }
+
if (sps->sao_enabled && !s->avctx->hwaccel) {
- av_frame_unref(s->tmp_frame);
- ret = ff_get_buffer(s->avctx, s->tmp_frame, AV_GET_BUFFER_FLAG_REF);
- if (ret < 0)
- goto fail;
- s->frame = s->tmp_frame;
+ int c_count = (sps->chroma_format_idc != 0) ? 3 : 1;
+ int c_idx;
+
+ for(c_idx = 0; c_idx < c_count; c_idx++) {
+ int w = sps->width >> sps->hshift[c_idx];
+ int h = sps->height >> sps->vshift[c_idx];
+ s->sao_pixel_buffer_h[c_idx] =
+ av_malloc((w * 2 * sps->ctb_height) <<
+ sps->pixel_shift);
+ s->sao_pixel_buffer_v[c_idx] =
+ av_malloc((h * 2 * sps->ctb_width) <<
+ sps->pixel_shift);
+ }
}
s->ps.sps = sps;
static int hls_slice_header(HEVCContext *s)
{
- GetBitContext *gb = &s->HEVClc.gb;
+ GetBitContext *gb = &s->HEVClc->gb;
SliceHeader *sh = &s->sh;
int i, ret;
if (IS_IDR(s))
ff_hevc_clear_refs(s);
}
+ sh->no_output_of_prior_pics_flag = 0;
if (IS_IRAP(s))
sh->no_output_of_prior_pics_flag = get_bits1(gb);
return AVERROR_INVALIDDATA;
}
s->ps.pps = (HEVCPPS*)s->ps.pps_list[sh->pps_id]->data;
+ if (s->nal_unit_type == HEVC_NAL_CRA_NUT && s->last_eos == 1)
+ sh->no_output_of_prior_pics_flag = 1;
if (s->ps.sps != (HEVCSPS*)s->ps.sps_list[s->ps.pps->sps_id]->data) {
+ const HEVCSPS* last_sps = s->ps.sps;
s->ps.sps = (HEVCSPS*)s->ps.sps_list[s->ps.pps->sps_id]->data;
-
+ if (last_sps && IS_IRAP(s) && s->nal_unit_type != HEVC_NAL_CRA_NUT) {
+ if (s->ps.sps->width != last_sps->width || s->ps.sps->height != last_sps->height ||
+ s->ps.sps->temporal_layer[s->ps.sps->max_sub_layers - 1].max_dec_pic_buffering !=
+ last_sps->temporal_layer[last_sps->max_sub_layers - 1].max_dec_pic_buffering)
+ sh->no_output_of_prior_pics_flag = 0;
+ }
ff_hevc_clear_refs(s);
- ret = set_sps(s, s->ps.sps);
+ ret = set_sps(s, s->ps.sps, AV_PIX_FMT_NONE);
if (ret < 0)
return ret;
slice_address_length = av_ceil_log2(s->ps.sps->ctb_width *
s->ps.sps->ctb_height);
- sh->slice_segment_addr = slice_address_length ? get_bits(gb, slice_address_length) : 0;
+ sh->slice_segment_addr = get_bitsz(gb, slice_address_length);
if (sh->slice_segment_addr >= s->ps.sps->ctb_width * s->ps.sps->ctb_height) {
av_log(s->avctx, AV_LOG_ERROR,
"Invalid slice segment address: %u.\n",
if (s->ps.sps->sao_enabled) {
sh->slice_sample_adaptive_offset_flag[0] = get_bits1(gb);
- sh->slice_sample_adaptive_offset_flag[1] =
- sh->slice_sample_adaptive_offset_flag[2] = get_bits1(gb);
+ if (s->ps.sps->chroma_format_idc) {
+ sh->slice_sample_adaptive_offset_flag[1] =
+ sh->slice_sample_adaptive_offset_flag[2] = get_bits1(gb);
+ }
} else {
sh->slice_sample_adaptive_offset_flag[0] = 0;
sh->slice_sample_adaptive_offset_flag[1] = 0;
if (sh->slice_type == B_SLICE)
sh->nb_refs[L1] = get_ue_golomb_long(gb) + 1;
}
- if (sh->nb_refs[L0] > MAX_REFS || sh->nb_refs[L1] > MAX_REFS) {
+ if (sh->nb_refs[L0] > HEVC_MAX_REFS || sh->nb_refs[L1] > HEVC_MAX_REFS) {
av_log(s->avctx, AV_LOG_ERROR, "Too many refs: %d/%d.\n",
sh->nb_refs[L0], sh->nb_refs[L1]);
return AVERROR_INVALIDDATA;
sh->slice_cr_qp_offset = 0;
}
+ if (s->ps.pps->chroma_qp_offset_list_enabled_flag)
+ sh->cu_chroma_qp_offset_enabled_flag = get_bits1(gb);
+ else
+ sh->cu_chroma_qp_offset_enabled_flag = 0;
+
if (s->ps.pps->deblocking_filter_control_present_flag) {
int deblocking_filter_override_flag = 0;
sh->num_entry_point_offsets = 0;
if (s->ps.pps->tiles_enabled_flag || s->ps.pps->entropy_coding_sync_enabled_flag) {
- sh->num_entry_point_offsets = get_ue_golomb_long(gb);
+ unsigned num_entry_point_offsets = get_ue_golomb_long(gb);
+ // It would be possible to bound this tighter but this here is simpler
+ if (num_entry_point_offsets > get_bits_left(gb)) {
+ av_log(s->avctx, AV_LOG_ERROR, "num_entry_point_offsets %d is invalid\n", num_entry_point_offsets);
+ return AVERROR_INVALIDDATA;
+ }
+
+ sh->num_entry_point_offsets = num_entry_point_offsets;
if (sh->num_entry_point_offsets > 0) {
int offset_len = get_ue_golomb_long(gb) + 1;
- for (i = 0; i < sh->num_entry_point_offsets; i++)
- skip_bits(gb, offset_len);
- }
+ if (offset_len < 1 || offset_len > 32) {
+ sh->num_entry_point_offsets = 0;
+ av_log(s->avctx, AV_LOG_ERROR, "offset_len %d is invalid\n", offset_len);
+ return AVERROR_INVALIDDATA;
+ }
+
+ av_freep(&sh->entry_point_offset);
+ av_freep(&sh->offset);
+ av_freep(&sh->size);
+ sh->entry_point_offset = av_malloc_array(sh->num_entry_point_offsets, sizeof(unsigned));
+ sh->offset = av_malloc_array(sh->num_entry_point_offsets, sizeof(int));
+ sh->size = av_malloc_array(sh->num_entry_point_offsets, sizeof(int));
+ if (!sh->entry_point_offset || !sh->offset || !sh->size) {
+ sh->num_entry_point_offsets = 0;
+ av_log(s->avctx, AV_LOG_ERROR, "Failed to allocate memory\n");
+ return AVERROR(ENOMEM);
+ }
+ for (i = 0; i < sh->num_entry_point_offsets; i++) {
+ unsigned val = get_bits_long(gb, offset_len);
+ sh->entry_point_offset[i] = val + 1; // +1; // +1 to get the size
+ }
+ if (s->threads_number > 1 && (s->ps.pps->num_tile_rows > 1 || s->ps.pps->num_tile_columns > 1)) {
+ s->enable_parallel_tiles = 0; // TODO: you can enable tiles in parallel here
+ s->threads_number = 1;
+ } else
+ s->enable_parallel_tiles = 0;
+ } else
+ s->enable_parallel_tiles = 0;
}
if (s->ps.pps->slice_header_extension_present_flag) {
unsigned int length = get_ue_golomb_long(gb);
+ if (length*8LL > get_bits_left(gb)) {
+ av_log(s->avctx, AV_LOG_ERROR, "too many slice_header_extension_data_bytes\n");
+ return AVERROR_INVALIDDATA;
+ }
for (i = 0; i < length; i++)
skip_bits(gb, 8); // slice_header_extension_data_byte
}
// Inferred parameters
- sh->slice_qp = 26 + s->ps.pps->pic_init_qp_minus26 + sh->slice_qp_delta;
+ sh->slice_qp = 26U + s->ps.pps->pic_init_qp_minus26 + sh->slice_qp_delta;
if (sh->slice_qp > 51 ||
sh->slice_qp < -s->ps.sps->qp_bd_offset) {
av_log(s->avctx, AV_LOG_ERROR,
return AVERROR_INVALIDDATA;
}
- s->HEVClc.first_qp_group = !s->sh.dependent_slice_segment_flag;
+ if (get_bits_left(gb) < 0) {
+ av_log(s->avctx, AV_LOG_ERROR,
+ "Overread slice header by %d bits\n", -get_bits_left(gb));
+ return AVERROR_INVALIDDATA;
+ }
+
+ s->HEVClc->first_qp_group = !s->sh.dependent_slice_segment_flag;
if (!s->ps.pps->cu_qp_delta_enabled_flag)
- s->HEVClc.qp_y = FFUMOD(s->sh.slice_qp + 52 + 2 * s->ps.sps->qp_bd_offset,
- 52 + s->ps.sps->qp_bd_offset) - s->ps.sps->qp_bd_offset;
+ s->HEVClc->qp_y = s->sh.slice_qp;
s->slice_initialized = 1;
+ s->HEVClc->tu.cu_qp_offset_cb = 0;
+ s->HEVClc->tu.cu_qp_offset_cr = 0;
+
+ s->no_rasl_output_flag = IS_IDR(s) || IS_BLA(s) || (s->nal_unit_type == HEVC_NAL_CRA_NUT && s->last_eos);
return 0;
}
static void hls_sao_param(HEVCContext *s, int rx, int ry)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
int sao_merge_left_flag = 0;
int sao_merge_up_flag = 0;
- int shift = s->ps.sps->bit_depth - FFMIN(s->ps.sps->bit_depth, 10);
SAOParams *sao = &CTB(s->sao, rx, ry);
int c_idx, i;
}
}
- for (c_idx = 0; c_idx < 3; c_idx++) {
+ for (c_idx = 0; c_idx < (s->ps.sps->chroma_format_idc ? 3 : 1); c_idx++) {
+ int log2_sao_offset_scale = c_idx == 0 ? s->ps.pps->log2_sao_offset_scale_luma :
+ s->ps.pps->log2_sao_offset_scale_chroma;
+
if (!s->sh.slice_sample_adaptive_offset_flag[c_idx]) {
sao->type_idx[c_idx] = SAO_NOT_APPLIED;
continue;
// Inferred parameters
sao->offset_val[c_idx][0] = 0;
for (i = 0; i < 4; i++) {
- sao->offset_val[c_idx][i + 1] = sao->offset_abs[c_idx][i] << shift;
+ sao->offset_val[c_idx][i + 1] = sao->offset_abs[c_idx][i];
if (sao->type_idx[c_idx] == SAO_EDGE) {
if (i > 1)
sao->offset_val[c_idx][i + 1] = -sao->offset_val[c_idx][i + 1];
} else if (sao->offset_sign[c_idx][i]) {
sao->offset_val[c_idx][i + 1] = -sao->offset_val[c_idx][i + 1];
}
+ sao->offset_val[c_idx][i + 1] *= 1 << log2_sao_offset_scale;
}
}
}
#undef SET_SAO
#undef CTB
-static void 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->ps.sps->hshift[c_idx];
- int vshift = s->ps.sps->vshift[c_idx];
- uint8_t *dst = &s->frame->data[c_idx][(y0 >> vshift) * stride +
- ((x0 >> hshift) << s->ps.sps->pixel_shift)];
- DECLARE_ALIGNED(32, 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, qp, shift, add, scale, scale_m;
- static 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->ps.sps->qp_bd_offset;
- } else {
- int qp_i, offset;
-
- if (c_idx == 1)
- offset = s->ps.pps->cb_qp_offset + s->sh.slice_cb_qp_offset;
- else
- offset = s->ps.pps->cr_qp_offset + s->sh.slice_cr_qp_offset;
-
- qp_i = av_clip(qp_y + offset, -s->ps.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->ps.sps->qp_bd_offset;
- }
-
- shift = s->ps.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->ps.sps->scaling_list_enable_flag) {
- const ScalingList *sl = s->ps.pps->scaling_list_data_present_flag ?
- &s->ps.pps->scaling_list : &s->ps.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->ps.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 =
- ff_hevc_last_significant_coeff_x_prefix_decode(s, c_idx, log2_trafo_size);
- last_significant_coeff_y =
- ff_hevc_last_significant_coeff_y_prefix_decode(s, c_idx, log2_trafo_size);
-
- if (last_significant_coeff_x > 3) {
- int suffix = ff_hevc_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 = ff_hevc_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;
+static int hls_cross_component_pred(HEVCContext *s, int idx) {
+ HEVCLocalContext *lc = s->HEVClc;
+ int log2_res_scale_abs_plus1 = ff_hevc_log2_res_scale_abs(s, idx);
- 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;
+ if (log2_res_scale_abs_plus1 != 0) {
+ int res_scale_sign_flag = ff_hevc_res_scale_sign_flag(s, idx);
+ lc->tu.res_scale_val = (1 << (log2_res_scale_abs_plus1 - 1)) *
+ (1 - 2 * res_scale_sign_flag);
+ } else {
+ lc->tu.res_scale_val = 0;
}
- 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;
- 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] =
- ff_hevc_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 (ff_hevc_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] =
- ff_hevc_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] += ff_hevc_coeff_abs_level_greater2_flag_decode(s, c_idx, ctx_set);
- }
- if (!s->ps.pps->sign_data_hiding_flag || !sign_hidden) {
- coeff_sign_flag = ff_hevc_coeff_sign_flag(s, nb_significant_coeff_flag) << (16 - nb_significant_coeff_flag);
- } else {
- coeff_sign_flag = ff_hevc_coeff_sign_flag(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 = ff_hevc_coeff_abs_level_remaining(s, trans_coeff_level, c_rice_param);
-
- trans_coeff_level += last_coeff_abs_level_remaining;
- if ((trans_coeff_level) > (3 * (1 << c_rice_param)))
- c_rice_param = FFMIN(c_rice_param + 1, 4);
- }
- if (s->ps.pps->sign_data_hiding_flag && sign_hidden) {
- sum_abs += trans_coeff_level;
- if (n == first_nz_pos_in_cg && ((sum_abs & 1) == 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->ps.sps->scaling_list_enable_flag) {
- if (y_c || x_c || log2_trafo_size < 4) {
- int pos;
- 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) {
- if (transform_skip_flag)
- s->hevcdsp.dequant(coeffs);
- else if (lc->cu.pred_mode == MODE_INTRA && c_idx == 0 &&
- log2_trafo_size == 2)
- s->hevcdsp.transform_4x4_luma(coeffs);
- else {
- int max_xy = FFMAX(last_significant_coeff_x, last_significant_coeff_y);
- if (max_xy == 0)
- s->hevcdsp.idct_dc[log2_trafo_size - 2](coeffs);
- else {
- int col_limit = last_significant_coeff_x + last_significant_coeff_y + 4;
- if (max_xy < 4)
- col_limit = FFMIN(4, col_limit);
- else if (max_xy < 8)
- col_limit = FFMIN(8, col_limit);
- else if (max_xy < 12)
- col_limit = FFMIN(24, col_limit);
- s->hevcdsp.idct[log2_trafo_size - 2](coeffs, col_limit);
- }
- }
- }
- s->hevcdsp.add_residual[log2_trafo_size - 2](dst, coeffs, stride);
+ return 0;
}
static int hls_transform_unit(HEVCContext *s, int x0, int y0,
int xBase, int yBase, int cb_xBase, int cb_yBase,
int log2_cb_size, int log2_trafo_size,
- int blk_idx, int cbf_luma, int cbf_cb, int cbf_cr)
+ int blk_idx, int cbf_luma, int *cbf_cb, int *cbf_cr)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
+ const int log2_trafo_size_c = log2_trafo_size - s->ps.sps->hshift[1];
+ int i;
if (lc->cu.pred_mode == MODE_INTRA) {
int trafo_size = 1 << log2_trafo_size;
ff_hevc_set_neighbour_available(s, x0, y0, trafo_size, trafo_size);
s->hpc.intra_pred[log2_trafo_size - 2](s, x0, y0, 0);
- if (log2_trafo_size > 2) {
- trafo_size = trafo_size << (s->ps.sps->hshift[1] - 1);
- ff_hevc_set_neighbour_available(s, x0, y0, trafo_size, trafo_size);
- s->hpc.intra_pred[log2_trafo_size - 3](s, x0, y0, 1);
- s->hpc.intra_pred[log2_trafo_size - 3](s, x0, y0, 2);
- } else if (blk_idx == 3) {
- trafo_size = trafo_size << s->ps.sps->hshift[1];
- ff_hevc_set_neighbour_available(s, xBase, yBase,
- trafo_size, trafo_size);
- s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase, 1);
- s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase, 2);
- }
}
- if (cbf_luma || cbf_cb || cbf_cr) {
+ if (cbf_luma || cbf_cb[0] || cbf_cr[0] ||
+ (s->ps.sps->chroma_format_idc == 2 && (cbf_cb[1] || cbf_cr[1]))) {
int scan_idx = SCAN_DIAG;
int scan_idx_c = SCAN_DIAG;
+ int cbf_chroma = cbf_cb[0] || cbf_cr[0] ||
+ (s->ps.sps->chroma_format_idc == 2 &&
+ (cbf_cb[1] || cbf_cr[1]));
if (s->ps.pps->cu_qp_delta_enabled_flag && !lc->tu.is_cu_qp_delta_coded) {
lc->tu.cu_qp_delta = ff_hevc_cu_qp_delta_abs(s);
return AVERROR_INVALIDDATA;
}
- ff_hevc_set_qPy(s, x0, y0, cb_xBase, cb_yBase, log2_cb_size);
+ ff_hevc_set_qPy(s, cb_xBase, cb_yBase, log2_cb_size);
+ }
+
+ if (s->sh.cu_chroma_qp_offset_enabled_flag && cbf_chroma &&
+ !lc->cu.cu_transquant_bypass_flag && !lc->tu.is_cu_chroma_qp_offset_coded) {
+ int cu_chroma_qp_offset_flag = ff_hevc_cu_chroma_qp_offset_flag(s);
+ if (cu_chroma_qp_offset_flag) {
+ int cu_chroma_qp_offset_idx = 0;
+ if (s->ps.pps->chroma_qp_offset_list_len_minus1 > 0) {
+ cu_chroma_qp_offset_idx = ff_hevc_cu_chroma_qp_offset_idx(s);
+ av_log(s->avctx, AV_LOG_ERROR,
+ "cu_chroma_qp_offset_idx not yet tested.\n");
+ }
+ lc->tu.cu_qp_offset_cb = s->ps.pps->cb_qp_offset_list[cu_chroma_qp_offset_idx];
+ lc->tu.cu_qp_offset_cr = s->ps.pps->cr_qp_offset_list[cu_chroma_qp_offset_idx];
+ } else {
+ lc->tu.cu_qp_offset_cb = 0;
+ lc->tu.cu_qp_offset_cr = 0;
+ }
+ lc->tu.is_cu_chroma_qp_offset_coded = 1;
}
if (lc->cu.pred_mode == MODE_INTRA && log2_trafo_size < 4) {
- if (lc->tu.cur_intra_pred_mode >= 6 &&
- lc->tu.cur_intra_pred_mode <= 14) {
+ if (lc->tu.intra_pred_mode >= 6 &&
+ lc->tu.intra_pred_mode <= 14) {
scan_idx = SCAN_VERT;
- } else if (lc->tu.cur_intra_pred_mode >= 22 &&
- lc->tu.cur_intra_pred_mode <= 30) {
+ } else if (lc->tu.intra_pred_mode >= 22 &&
+ lc->tu.intra_pred_mode <= 30) {
scan_idx = SCAN_HORIZ;
}
- if (lc->pu.intra_pred_mode_c >= 6 &&
- lc->pu.intra_pred_mode_c <= 14) {
+ if (lc->tu.intra_pred_mode_c >= 6 &&
+ lc->tu.intra_pred_mode_c <= 14) {
scan_idx_c = SCAN_VERT;
- } else if (lc->pu.intra_pred_mode_c >= 22 &&
- lc->pu.intra_pred_mode_c <= 30) {
+ } else if (lc->tu.intra_pred_mode_c >= 22 &&
+ lc->tu.intra_pred_mode_c <= 30) {
scan_idx_c = SCAN_HORIZ;
}
}
+ lc->tu.cross_pf = 0;
+
if (cbf_luma)
- hls_residual_coding(s, x0, y0, log2_trafo_size, scan_idx, 0);
- if (log2_trafo_size > 2) {
- if (cbf_cb)
- hls_residual_coding(s, x0, y0, log2_trafo_size - 1, scan_idx_c, 1);
- if (cbf_cr)
- hls_residual_coding(s, x0, y0, log2_trafo_size - 1, scan_idx_c, 2);
+ ff_hevc_hls_residual_coding(s, x0, y0, log2_trafo_size, scan_idx, 0);
+ if (s->ps.sps->chroma_format_idc && (log2_trafo_size > 2 || s->ps.sps->chroma_format_idc == 3)) {
+ int trafo_size_h = 1 << (log2_trafo_size_c + s->ps.sps->hshift[1]);
+ int trafo_size_v = 1 << (log2_trafo_size_c + s->ps.sps->vshift[1]);
+ lc->tu.cross_pf = (s->ps.pps->cross_component_prediction_enabled_flag && cbf_luma &&
+ (lc->cu.pred_mode == MODE_INTER ||
+ (lc->tu.chroma_mode_c == 4)));
+
+ if (lc->tu.cross_pf) {
+ hls_cross_component_pred(s, 0);
+ }
+ for (i = 0; i < (s->ps.sps->chroma_format_idc == 2 ? 2 : 1); i++) {
+ if (lc->cu.pred_mode == MODE_INTRA) {
+ ff_hevc_set_neighbour_available(s, x0, y0 + (i << log2_trafo_size_c), trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size_c - 2](s, x0, y0 + (i << log2_trafo_size_c), 1);
+ }
+ if (cbf_cb[i])
+ ff_hevc_hls_residual_coding(s, x0, y0 + (i << log2_trafo_size_c),
+ log2_trafo_size_c, scan_idx_c, 1);
+ else
+ if (lc->tu.cross_pf) {
+ ptrdiff_t stride = s->frame->linesize[1];
+ int hshift = s->ps.sps->hshift[1];
+ int vshift = s->ps.sps->vshift[1];
+ int16_t *coeffs_y = (int16_t*)lc->edge_emu_buffer;
+ int16_t *coeffs = (int16_t*)lc->edge_emu_buffer2;
+ int size = 1 << log2_trafo_size_c;
+
+ uint8_t *dst = &s->frame->data[1][(y0 >> vshift) * stride +
+ ((x0 >> hshift) << s->ps.sps->pixel_shift)];
+ for (i = 0; i < (size * size); i++) {
+ coeffs[i] = ((lc->tu.res_scale_val * coeffs_y[i]) >> 3);
+ }
+ s->hevcdsp.add_residual[log2_trafo_size_c-2](dst, coeffs, stride);
+ }
+ }
+
+ if (lc->tu.cross_pf) {
+ hls_cross_component_pred(s, 1);
+ }
+ for (i = 0; i < (s->ps.sps->chroma_format_idc == 2 ? 2 : 1); i++) {
+ if (lc->cu.pred_mode == MODE_INTRA) {
+ ff_hevc_set_neighbour_available(s, x0, y0 + (i << log2_trafo_size_c), trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size_c - 2](s, x0, y0 + (i << log2_trafo_size_c), 2);
+ }
+ if (cbf_cr[i])
+ ff_hevc_hls_residual_coding(s, x0, y0 + (i << log2_trafo_size_c),
+ log2_trafo_size_c, scan_idx_c, 2);
+ else
+ if (lc->tu.cross_pf) {
+ ptrdiff_t stride = s->frame->linesize[2];
+ int hshift = s->ps.sps->hshift[2];
+ int vshift = s->ps.sps->vshift[2];
+ int16_t *coeffs_y = (int16_t*)lc->edge_emu_buffer;
+ int16_t *coeffs = (int16_t*)lc->edge_emu_buffer2;
+ int size = 1 << log2_trafo_size_c;
+
+ uint8_t *dst = &s->frame->data[2][(y0 >> vshift) * stride +
+ ((x0 >> hshift) << s->ps.sps->pixel_shift)];
+ for (i = 0; i < (size * size); i++) {
+ coeffs[i] = ((lc->tu.res_scale_val * coeffs_y[i]) >> 3);
+ }
+ s->hevcdsp.add_residual[log2_trafo_size_c-2](dst, coeffs, stride);
+ }
+ }
+ } else if (s->ps.sps->chroma_format_idc && blk_idx == 3) {
+ int trafo_size_h = 1 << (log2_trafo_size + 1);
+ int trafo_size_v = 1 << (log2_trafo_size + s->ps.sps->vshift[1]);
+ for (i = 0; i < (s->ps.sps->chroma_format_idc == 2 ? 2 : 1); i++) {
+ if (lc->cu.pred_mode == MODE_INTRA) {
+ ff_hevc_set_neighbour_available(s, xBase, yBase + (i << log2_trafo_size),
+ trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase + (i << log2_trafo_size), 1);
+ }
+ if (cbf_cb[i])
+ ff_hevc_hls_residual_coding(s, xBase, yBase + (i << log2_trafo_size),
+ log2_trafo_size, scan_idx_c, 1);
+ }
+ for (i = 0; i < (s->ps.sps->chroma_format_idc == 2 ? 2 : 1); i++) {
+ if (lc->cu.pred_mode == MODE_INTRA) {
+ ff_hevc_set_neighbour_available(s, xBase, yBase + (i << log2_trafo_size),
+ trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase + (i << log2_trafo_size), 2);
+ }
+ if (cbf_cr[i])
+ ff_hevc_hls_residual_coding(s, xBase, yBase + (i << log2_trafo_size),
+ log2_trafo_size, scan_idx_c, 2);
+ }
+ }
+ } else if (s->ps.sps->chroma_format_idc && lc->cu.pred_mode == MODE_INTRA) {
+ if (log2_trafo_size > 2 || s->ps.sps->chroma_format_idc == 3) {
+ int trafo_size_h = 1 << (log2_trafo_size_c + s->ps.sps->hshift[1]);
+ int trafo_size_v = 1 << (log2_trafo_size_c + s->ps.sps->vshift[1]);
+ ff_hevc_set_neighbour_available(s, x0, y0, trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size_c - 2](s, x0, y0, 1);
+ s->hpc.intra_pred[log2_trafo_size_c - 2](s, x0, y0, 2);
+ if (s->ps.sps->chroma_format_idc == 2) {
+ ff_hevc_set_neighbour_available(s, x0, y0 + (1 << log2_trafo_size_c),
+ trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size_c - 2](s, x0, y0 + (1 << log2_trafo_size_c), 1);
+ s->hpc.intra_pred[log2_trafo_size_c - 2](s, x0, y0 + (1 << log2_trafo_size_c), 2);
+ }
} else if (blk_idx == 3) {
- if (cbf_cb)
- hls_residual_coding(s, xBase, yBase, log2_trafo_size, scan_idx_c, 1);
- if (cbf_cr)
- hls_residual_coding(s, xBase, yBase, log2_trafo_size, scan_idx_c, 2);
+ int trafo_size_h = 1 << (log2_trafo_size + 1);
+ int trafo_size_v = 1 << (log2_trafo_size + s->ps.sps->vshift[1]);
+ ff_hevc_set_neighbour_available(s, xBase, yBase,
+ trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase, 1);
+ s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase, 2);
+ if (s->ps.sps->chroma_format_idc == 2) {
+ ff_hevc_set_neighbour_available(s, xBase, yBase + (1 << (log2_trafo_size)),
+ trafo_size_h, trafo_size_v);
+ s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase + (1 << (log2_trafo_size)), 1);
+ s->hpc.intra_pred[log2_trafo_size - 2](s, xBase, yBase + (1 << (log2_trafo_size)), 2);
+ }
}
}
+
return 0;
}
int xBase, int yBase, int cb_xBase, int cb_yBase,
int log2_cb_size, int log2_trafo_size,
int trafo_depth, int blk_idx,
- int cbf_cb, int cbf_cr)
+ const int *base_cbf_cb, const int *base_cbf_cr)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
uint8_t split_transform_flag;
+ int cbf_cb[2];
+ int cbf_cr[2];
int ret;
+ cbf_cb[0] = base_cbf_cb[0];
+ cbf_cb[1] = base_cbf_cb[1];
+ cbf_cr[0] = base_cbf_cr[0];
+ cbf_cr[1] = base_cbf_cr[1];
+
if (lc->cu.intra_split_flag) {
- if (trafo_depth == 1)
- lc->tu.cur_intra_pred_mode = lc->pu.intra_pred_mode[blk_idx];
+ if (trafo_depth == 1) {
+ lc->tu.intra_pred_mode = lc->pu.intra_pred_mode[blk_idx];
+ if (s->ps.sps->chroma_format_idc == 3) {
+ lc->tu.intra_pred_mode_c = lc->pu.intra_pred_mode_c[blk_idx];
+ lc->tu.chroma_mode_c = lc->pu.chroma_mode_c[blk_idx];
+ } else {
+ lc->tu.intra_pred_mode_c = lc->pu.intra_pred_mode_c[0];
+ lc->tu.chroma_mode_c = lc->pu.chroma_mode_c[0];
+ }
+ }
} else {
- lc->tu.cur_intra_pred_mode = lc->pu.intra_pred_mode[0];
+ lc->tu.intra_pred_mode = lc->pu.intra_pred_mode[0];
+ lc->tu.intra_pred_mode_c = lc->pu.intra_pred_mode_c[0];
+ lc->tu.chroma_mode_c = lc->pu.chroma_mode_c[0];
}
if (log2_trafo_size <= s->ps.sps->log2_max_trafo_size &&
inter_split;
}
- if (log2_trafo_size > 2 && (trafo_depth == 0 || cbf_cb))
- cbf_cb = ff_hevc_cbf_cb_cr_decode(s, trafo_depth);
- else if (log2_trafo_size > 2 || trafo_depth == 0)
- cbf_cb = 0;
- if (log2_trafo_size > 2 && (trafo_depth == 0 || cbf_cr))
- cbf_cr = ff_hevc_cbf_cb_cr_decode(s, trafo_depth);
- else if (log2_trafo_size > 2 || trafo_depth == 0)
- cbf_cr = 0;
+ if (s->ps.sps->chroma_format_idc && (log2_trafo_size > 2 || s->ps.sps->chroma_format_idc == 3)) {
+ if (trafo_depth == 0 || cbf_cb[0]) {
+ cbf_cb[0] = ff_hevc_cbf_cb_cr_decode(s, trafo_depth);
+ if (s->ps.sps->chroma_format_idc == 2 && (!split_transform_flag || log2_trafo_size == 3)) {
+ cbf_cb[1] = ff_hevc_cbf_cb_cr_decode(s, trafo_depth);
+ }
+ }
+
+ if (trafo_depth == 0 || cbf_cr[0]) {
+ cbf_cr[0] = ff_hevc_cbf_cb_cr_decode(s, trafo_depth);
+ if (s->ps.sps->chroma_format_idc == 2 && (!split_transform_flag || log2_trafo_size == 3)) {
+ cbf_cr[1] = ff_hevc_cbf_cb_cr_decode(s, trafo_depth);
+ }
+ }
+ }
if (split_transform_flag) {
const int trafo_size_split = 1 << (log2_trafo_size - 1);
int cbf_luma = 1;
if (lc->cu.pred_mode == MODE_INTRA || trafo_depth != 0 ||
- cbf_cb || cbf_cr)
+ cbf_cb[0] || cbf_cr[0] ||
+ (s->ps.sps->chroma_format_idc == 2 && (cbf_cb[1] || cbf_cr[1]))) {
cbf_luma = ff_hevc_cbf_luma_decode(s, trafo_depth);
+ }
ret = hls_transform_unit(s, x0, y0, xBase, yBase, cb_xBase, cb_yBase,
log2_cb_size, log2_trafo_size,
static int hls_pcm_sample(HEVCContext *s, int x0, int y0, int log2_cb_size)
{
- //TODO: non-4:2:0 support
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
GetBitContext gb;
int cb_size = 1 << log2_cb_size;
ptrdiff_t stride0 = s->frame->linesize[0];
uint8_t *dst1 = &s->frame->data[1][(y0 >> s->ps.sps->vshift[1]) * stride1 + ((x0 >> s->ps.sps->hshift[1]) << s->ps.sps->pixel_shift)];
uint8_t *dst2 = &s->frame->data[2][(y0 >> s->ps.sps->vshift[2]) * stride2 + ((x0 >> s->ps.sps->hshift[2]) << s->ps.sps->pixel_shift)];
- int length = cb_size * cb_size * s->ps.sps->pcm.bit_depth + ((cb_size * cb_size) >> 1) * s->ps.sps->pcm.bit_depth_chroma;
+ int length = cb_size * cb_size * s->ps.sps->pcm.bit_depth +
+ (((cb_size >> s->ps.sps->hshift[1]) * (cb_size >> s->ps.sps->vshift[1])) +
+ ((cb_size >> s->ps.sps->hshift[2]) * (cb_size >> s->ps.sps->vshift[2]))) *
+ s->ps.sps->pcm.bit_depth_chroma;
const uint8_t *pcm = skip_bytes(&lc->cc, (length + 7) >> 3);
int ret;
if (ret < 0)
return ret;
- s->hevcdsp.put_pcm(dst0, stride0, cb_size, &gb, s->ps.sps->pcm.bit_depth);
- s->hevcdsp.put_pcm(dst1, stride1, cb_size / 2, &gb, s->ps.sps->pcm.bit_depth_chroma);
- s->hevcdsp.put_pcm(dst2, stride2, cb_size / 2, &gb, s->ps.sps->pcm.bit_depth_chroma);
- return 0;
-}
-
-static void hls_mvd_coding(HEVCContext *s, int x0, int y0, int log2_cb_size)
-{
- HEVCLocalContext *lc = &s->HEVClc;
- int x = ff_hevc_abs_mvd_greater0_flag_decode(s);
- int y = ff_hevc_abs_mvd_greater0_flag_decode(s);
-
- if (x)
- x += ff_hevc_abs_mvd_greater1_flag_decode(s);
- if (y)
- y += ff_hevc_abs_mvd_greater1_flag_decode(s);
-
- switch (x) {
- case 2: lc->pu.mvd.x = ff_hevc_mvd_decode(s); break;
- case 1: lc->pu.mvd.x = ff_hevc_mvd_sign_flag_decode(s); break;
- case 0: lc->pu.mvd.x = 0; break;
+ s->hevcdsp.put_pcm(dst0, stride0, cb_size, cb_size, &gb, s->ps.sps->pcm.bit_depth);
+ if (s->ps.sps->chroma_format_idc) {
+ s->hevcdsp.put_pcm(dst1, stride1,
+ cb_size >> s->ps.sps->hshift[1],
+ cb_size >> s->ps.sps->vshift[1],
+ &gb, s->ps.sps->pcm.bit_depth_chroma);
+ s->hevcdsp.put_pcm(dst2, stride2,
+ cb_size >> s->ps.sps->hshift[2],
+ cb_size >> s->ps.sps->vshift[2],
+ &gb, s->ps.sps->pcm.bit_depth_chroma);
}
- switch (y) {
- case 2: lc->pu.mvd.y = ff_hevc_mvd_decode(s); break;
- case 1: lc->pu.mvd.y = ff_hevc_mvd_sign_flag_decode(s); break;
- case 0: lc->pu.mvd.y = 0; break;
- }
+ return 0;
}
/**
- * 8.5.3.2.2.1 Luma sample interpolation process
+ * 8.5.3.2.2.1 Luma sample unidirectional interpolation process
*
* @param s HEVC decoding context
* @param dst target buffer for block data at block position
* @param y_off vertical position of block from origin (0, 0)
* @param block_w width of block
* @param block_h height of block
+ * @param luma_weight weighting factor applied to the luma prediction
+ * @param luma_offset additive offset applied to the luma prediction value
*/
-static void luma_mc(HEVCContext *s, int16_t *dst, ptrdiff_t dststride,
- AVFrame *ref, const Mv *mv, int x_off, int y_off,
- int block_w, int block_h, int pred_idx)
+
+static void luma_mc_uni(HEVCContext *s, uint8_t *dst, ptrdiff_t dststride,
+ AVFrame *ref, const Mv *mv, int x_off, int y_off,
+ int block_w, int block_h, int luma_weight, int luma_offset)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
uint8_t *src = ref->data[0];
ptrdiff_t srcstride = ref->linesize[0];
int pic_width = s->ps.sps->width;
int pic_height = s->ps.sps->height;
-
- int mx = mv->x & 3;
- int my = mv->y & 3;
- int extra_left = ff_hevc_qpel_extra_before[mx];
- int extra_top = ff_hevc_qpel_extra_before[my];
+ int mx = mv->x & 3;
+ int my = mv->y & 3;
+ int weight_flag = (s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
+ (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag);
+ int idx = ff_hevc_pel_weight[block_w];
x_off += mv->x >> 2;
y_off += mv->y >> 2;
src += y_off * srcstride + (x_off * (1 << s->ps.sps->pixel_shift));
- if (x_off < extra_left || y_off < extra_top ||
- x_off >= pic_width - block_w - ff_hevc_qpel_extra_after[mx] ||
- y_off >= pic_height - block_h - ff_hevc_qpel_extra_after[my]) {
+ if (x_off < QPEL_EXTRA_BEFORE || y_off < QPEL_EXTRA_AFTER ||
+ x_off >= pic_width - block_w - QPEL_EXTRA_AFTER ||
+ y_off >= pic_height - block_h - QPEL_EXTRA_AFTER) {
const ptrdiff_t edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
- int offset = extra_top * srcstride + (extra_left << s->ps.sps->pixel_shift);
- int buf_offset = extra_top *
- edge_emu_stride + (extra_left << s->ps.sps->pixel_shift);
+ int offset = QPEL_EXTRA_BEFORE * srcstride + (QPEL_EXTRA_BEFORE << s->ps.sps->pixel_shift);
+ int buf_offset = QPEL_EXTRA_BEFORE * edge_emu_stride + (QPEL_EXTRA_BEFORE << s->ps.sps->pixel_shift);
s->vdsp.emulated_edge_mc(lc->edge_emu_buffer, src - offset,
edge_emu_stride, srcstride,
- block_w + ff_hevc_qpel_extra[mx],
- block_h + ff_hevc_qpel_extra[my],
- x_off - extra_left, y_off - extra_top,
+ block_w + QPEL_EXTRA,
+ block_h + QPEL_EXTRA,
+ x_off - QPEL_EXTRA_BEFORE, y_off - QPEL_EXTRA_BEFORE,
pic_width, pic_height);
src = lc->edge_emu_buffer + buf_offset;
srcstride = edge_emu_stride;
}
- s->hevcdsp.put_hevc_qpel[!!my][!!mx][pred_idx](dst, dststride, src, srcstride,
- block_h, mx, my, lc->mc_buffer);
+
+ if (!weight_flag)
+ s->hevcdsp.put_hevc_qpel_uni[idx][!!my][!!mx](dst, dststride, src, srcstride,
+ block_h, mx, my, block_w);
+ else
+ s->hevcdsp.put_hevc_qpel_uni_w[idx][!!my][!!mx](dst, dststride, src, srcstride,
+ block_h, s->sh.luma_log2_weight_denom,
+ luma_weight, luma_offset, mx, my, block_w);
+}
+
+/**
+ * 8.5.3.2.2.1 Luma sample bidirectional interpolation process
+ *
+ * @param s HEVC decoding context
+ * @param dst target buffer for block data at block position
+ * @param dststride stride of the dst buffer
+ * @param ref0 reference picture0 buffer at origin (0, 0)
+ * @param mv0 motion vector0 (relative to block position) to get pixel data from
+ * @param x_off horizontal position of block from origin (0, 0)
+ * @param y_off vertical position of block from origin (0, 0)
+ * @param block_w width of block
+ * @param block_h height of block
+ * @param ref1 reference picture1 buffer at origin (0, 0)
+ * @param mv1 motion vector1 (relative to block position) to get pixel data from
+ * @param current_mv current motion vector structure
+ */
+ static void luma_mc_bi(HEVCContext *s, uint8_t *dst, ptrdiff_t dststride,
+ AVFrame *ref0, const Mv *mv0, int x_off, int y_off,
+ int block_w, int block_h, AVFrame *ref1, const Mv *mv1, struct MvField *current_mv)
+{
+ HEVCLocalContext *lc = s->HEVClc;
+ ptrdiff_t src0stride = ref0->linesize[0];
+ ptrdiff_t src1stride = ref1->linesize[0];
+ int pic_width = s->ps.sps->width;
+ int pic_height = s->ps.sps->height;
+ int mx0 = mv0->x & 3;
+ int my0 = mv0->y & 3;
+ int mx1 = mv1->x & 3;
+ int my1 = mv1->y & 3;
+ int weight_flag = (s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
+ (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag);
+ int x_off0 = x_off + (mv0->x >> 2);
+ int y_off0 = y_off + (mv0->y >> 2);
+ int x_off1 = x_off + (mv1->x >> 2);
+ int y_off1 = y_off + (mv1->y >> 2);
+ int idx = ff_hevc_pel_weight[block_w];
+
+ uint8_t *src0 = ref0->data[0] + y_off0 * src0stride + (int)((unsigned)x_off0 << s->ps.sps->pixel_shift);
+ uint8_t *src1 = ref1->data[0] + y_off1 * src1stride + (int)((unsigned)x_off1 << s->ps.sps->pixel_shift);
+
+ if (x_off0 < QPEL_EXTRA_BEFORE || y_off0 < QPEL_EXTRA_AFTER ||
+ x_off0 >= pic_width - block_w - QPEL_EXTRA_AFTER ||
+ y_off0 >= pic_height - block_h - QPEL_EXTRA_AFTER) {
+ const ptrdiff_t edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
+ int offset = QPEL_EXTRA_BEFORE * src0stride + (QPEL_EXTRA_BEFORE << s->ps.sps->pixel_shift);
+ int buf_offset = QPEL_EXTRA_BEFORE * edge_emu_stride + (QPEL_EXTRA_BEFORE << s->ps.sps->pixel_shift);
+
+ s->vdsp.emulated_edge_mc(lc->edge_emu_buffer, src0 - offset,
+ edge_emu_stride, src0stride,
+ block_w + QPEL_EXTRA,
+ block_h + QPEL_EXTRA,
+ x_off0 - QPEL_EXTRA_BEFORE, y_off0 - QPEL_EXTRA_BEFORE,
+ pic_width, pic_height);
+ src0 = lc->edge_emu_buffer + buf_offset;
+ src0stride = edge_emu_stride;
+ }
+
+ if (x_off1 < QPEL_EXTRA_BEFORE || y_off1 < QPEL_EXTRA_AFTER ||
+ x_off1 >= pic_width - block_w - QPEL_EXTRA_AFTER ||
+ y_off1 >= pic_height - block_h - QPEL_EXTRA_AFTER) {
+ const ptrdiff_t edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
+ int offset = QPEL_EXTRA_BEFORE * src1stride + (QPEL_EXTRA_BEFORE << s->ps.sps->pixel_shift);
+ int buf_offset = QPEL_EXTRA_BEFORE * edge_emu_stride + (QPEL_EXTRA_BEFORE << s->ps.sps->pixel_shift);
+
+ s->vdsp.emulated_edge_mc(lc->edge_emu_buffer2, src1 - offset,
+ edge_emu_stride, src1stride,
+ block_w + QPEL_EXTRA,
+ block_h + QPEL_EXTRA,
+ x_off1 - QPEL_EXTRA_BEFORE, y_off1 - QPEL_EXTRA_BEFORE,
+ pic_width, pic_height);
+ src1 = lc->edge_emu_buffer2 + buf_offset;
+ src1stride = edge_emu_stride;
+ }
+
+ s->hevcdsp.put_hevc_qpel[idx][!!my0][!!mx0](lc->tmp, src0, src0stride,
+ block_h, mx0, my0, block_w);
+ if (!weight_flag)
+ s->hevcdsp.put_hevc_qpel_bi[idx][!!my1][!!mx1](dst, dststride, src1, src1stride, lc->tmp,
+ block_h, mx1, my1, block_w);
+ else
+ s->hevcdsp.put_hevc_qpel_bi_w[idx][!!my1][!!mx1](dst, dststride, src1, src1stride, lc->tmp,
+ block_h, s->sh.luma_log2_weight_denom,
+ s->sh.luma_weight_l0[current_mv->ref_idx[0]],
+ s->sh.luma_weight_l1[current_mv->ref_idx[1]],
+ s->sh.luma_offset_l0[current_mv->ref_idx[0]],
+ s->sh.luma_offset_l1[current_mv->ref_idx[1]],
+ mx1, my1, block_w);
+
}
/**
- * 8.5.3.2.2.2 Chroma sample interpolation process
+ * 8.5.3.2.2.2 Chroma sample uniprediction interpolation process
*
* @param s HEVC decoding context
* @param dst1 target buffer for block data at block position (U plane)
* @param y_off vertical position of block from origin (0, 0)
* @param block_w width of block
* @param block_h height of block
+ * @param chroma_weight weighting factor applied to the chroma prediction
+ * @param chroma_offset additive offset applied to the chroma prediction value
*/
-static void chroma_mc(HEVCContext *s, int16_t *dst1, int16_t *dst2,
- ptrdiff_t dststride, AVFrame *ref, const Mv *mv,
- int x_off, int y_off, int block_w, int block_h, int pred_idx)
+
+static void chroma_mc_uni(HEVCContext *s, uint8_t *dst0,
+ ptrdiff_t dststride, uint8_t *src0, ptrdiff_t srcstride, int reflist,
+ int x_off, int y_off, int block_w, int block_h, struct MvField *current_mv, int chroma_weight, int chroma_offset)
{
- HEVCLocalContext *lc = &s->HEVClc;
- uint8_t *src1 = ref->data[1];
- uint8_t *src2 = ref->data[2];
- ptrdiff_t src1stride = ref->linesize[1];
- ptrdiff_t src2stride = ref->linesize[2];
- int pic_width = s->ps.sps->width >> 1;
- int pic_height = s->ps.sps->height >> 1;
-
- int mx = mv->x & 7;
- int my = mv->y & 7;
-
- x_off += mv->x >> 3;
- y_off += mv->y >> 3;
- src1 += y_off * src1stride + (x_off * (1 << s->ps.sps->pixel_shift));
- src2 += y_off * src2stride + (x_off * (1 << s->ps.sps->pixel_shift));
+ HEVCLocalContext *lc = s->HEVClc;
+ int pic_width = s->ps.sps->width >> s->ps.sps->hshift[1];
+ int pic_height = s->ps.sps->height >> s->ps.sps->vshift[1];
+ const Mv *mv = ¤t_mv->mv[reflist];
+ int weight_flag = (s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
+ (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag);
+ int idx = ff_hevc_pel_weight[block_w];
+ int hshift = s->ps.sps->hshift[1];
+ int vshift = s->ps.sps->vshift[1];
+ intptr_t mx = av_mod_uintp2(mv->x, 2 + hshift);
+ intptr_t my = av_mod_uintp2(mv->y, 2 + vshift);
+ intptr_t _mx = mx << (1 - hshift);
+ intptr_t _my = my << (1 - vshift);
+
+ x_off += mv->x >> (2 + hshift);
+ y_off += mv->y >> (2 + vshift);
+ src0 += y_off * srcstride + (x_off * (1 << s->ps.sps->pixel_shift));
if (x_off < EPEL_EXTRA_BEFORE || y_off < EPEL_EXTRA_AFTER ||
x_off >= pic_width - block_w - EPEL_EXTRA_AFTER ||
y_off >= pic_height - block_h - EPEL_EXTRA_AFTER) {
- const ptrdiff_t edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
+ const int edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
+ int offset0 = EPEL_EXTRA_BEFORE * (srcstride + (1 << s->ps.sps->pixel_shift));
+ int buf_offset0 = EPEL_EXTRA_BEFORE *
+ (edge_emu_stride + (1 << s->ps.sps->pixel_shift));
+ s->vdsp.emulated_edge_mc(lc->edge_emu_buffer, src0 - offset0,
+ edge_emu_stride, srcstride,
+ block_w + EPEL_EXTRA, block_h + EPEL_EXTRA,
+ x_off - EPEL_EXTRA_BEFORE,
+ y_off - EPEL_EXTRA_BEFORE,
+ pic_width, pic_height);
+
+ src0 = lc->edge_emu_buffer + buf_offset0;
+ srcstride = edge_emu_stride;
+ }
+ if (!weight_flag)
+ s->hevcdsp.put_hevc_epel_uni[idx][!!my][!!mx](dst0, dststride, src0, srcstride,
+ block_h, _mx, _my, block_w);
+ else
+ s->hevcdsp.put_hevc_epel_uni_w[idx][!!my][!!mx](dst0, dststride, src0, srcstride,
+ block_h, s->sh.chroma_log2_weight_denom,
+ chroma_weight, chroma_offset, _mx, _my, block_w);
+}
+
+/**
+ * 8.5.3.2.2.2 Chroma sample bidirectional interpolation process
+ *
+ * @param s HEVC decoding context
+ * @param dst target buffer for block data at block position
+ * @param dststride stride of the dst buffer
+ * @param ref0 reference picture0 buffer at origin (0, 0)
+ * @param mv0 motion vector0 (relative to block position) to get pixel data from
+ * @param x_off horizontal position of block from origin (0, 0)
+ * @param y_off vertical position of block from origin (0, 0)
+ * @param block_w width of block
+ * @param block_h height of block
+ * @param ref1 reference picture1 buffer at origin (0, 0)
+ * @param mv1 motion vector1 (relative to block position) to get pixel data from
+ * @param current_mv current motion vector structure
+ * @param cidx chroma component(cb, cr)
+ */
+static void chroma_mc_bi(HEVCContext *s, uint8_t *dst0, ptrdiff_t dststride, AVFrame *ref0, AVFrame *ref1,
+ int x_off, int y_off, int block_w, int block_h, struct MvField *current_mv, int cidx)
+{
+ HEVCLocalContext *lc = s->HEVClc;
+ uint8_t *src1 = ref0->data[cidx+1];
+ uint8_t *src2 = ref1->data[cidx+1];
+ ptrdiff_t src1stride = ref0->linesize[cidx+1];
+ ptrdiff_t src2stride = ref1->linesize[cidx+1];
+ int weight_flag = (s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
+ (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag);
+ int pic_width = s->ps.sps->width >> s->ps.sps->hshift[1];
+ int pic_height = s->ps.sps->height >> s->ps.sps->vshift[1];
+ Mv *mv0 = ¤t_mv->mv[0];
+ Mv *mv1 = ¤t_mv->mv[1];
+ int hshift = s->ps.sps->hshift[1];
+ int vshift = s->ps.sps->vshift[1];
+
+ intptr_t mx0 = av_mod_uintp2(mv0->x, 2 + hshift);
+ intptr_t my0 = av_mod_uintp2(mv0->y, 2 + vshift);
+ intptr_t mx1 = av_mod_uintp2(mv1->x, 2 + hshift);
+ intptr_t my1 = av_mod_uintp2(mv1->y, 2 + vshift);
+ intptr_t _mx0 = mx0 << (1 - hshift);
+ intptr_t _my0 = my0 << (1 - vshift);
+ intptr_t _mx1 = mx1 << (1 - hshift);
+ intptr_t _my1 = my1 << (1 - vshift);
+
+ int x_off0 = x_off + (mv0->x >> (2 + hshift));
+ int y_off0 = y_off + (mv0->y >> (2 + vshift));
+ int x_off1 = x_off + (mv1->x >> (2 + hshift));
+ int y_off1 = y_off + (mv1->y >> (2 + vshift));
+ int idx = ff_hevc_pel_weight[block_w];
+ src1 += y_off0 * src1stride + (int)((unsigned)x_off0 << s->ps.sps->pixel_shift);
+ src2 += y_off1 * src2stride + (int)((unsigned)x_off1 << s->ps.sps->pixel_shift);
+
+ if (x_off0 < EPEL_EXTRA_BEFORE || y_off0 < EPEL_EXTRA_AFTER ||
+ x_off0 >= pic_width - block_w - EPEL_EXTRA_AFTER ||
+ y_off0 >= pic_height - block_h - EPEL_EXTRA_AFTER) {
+ const int edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
int offset1 = EPEL_EXTRA_BEFORE * (src1stride + (1 << s->ps.sps->pixel_shift));
int buf_offset1 = EPEL_EXTRA_BEFORE *
(edge_emu_stride + (1 << s->ps.sps->pixel_shift));
- int offset2 = EPEL_EXTRA_BEFORE * (src2stride + (1 << s->ps.sps->pixel_shift));
- int buf_offset2 = EPEL_EXTRA_BEFORE *
- (edge_emu_stride + (1 << s->ps.sps->pixel_shift));
s->vdsp.emulated_edge_mc(lc->edge_emu_buffer, src1 - offset1,
edge_emu_stride, src1stride,
block_w + EPEL_EXTRA, block_h + EPEL_EXTRA,
- x_off - EPEL_EXTRA_BEFORE,
- y_off - EPEL_EXTRA_BEFORE,
+ x_off0 - EPEL_EXTRA_BEFORE,
+ y_off0 - EPEL_EXTRA_BEFORE,
pic_width, pic_height);
src1 = lc->edge_emu_buffer + buf_offset1;
src1stride = edge_emu_stride;
- s->hevcdsp.put_hevc_epel[!!my][!!mx][pred_idx](dst1, dststride, src1, src1stride,
- block_h, mx, my, lc->mc_buffer);
+ }
- s->vdsp.emulated_edge_mc(lc->edge_emu_buffer, src2 - offset2,
+ if (x_off1 < EPEL_EXTRA_BEFORE || y_off1 < EPEL_EXTRA_AFTER ||
+ x_off1 >= pic_width - block_w - EPEL_EXTRA_AFTER ||
+ y_off1 >= pic_height - block_h - EPEL_EXTRA_AFTER) {
+ const int edge_emu_stride = EDGE_EMU_BUFFER_STRIDE << s->ps.sps->pixel_shift;
+ int offset1 = EPEL_EXTRA_BEFORE * (src2stride + (1 << s->ps.sps->pixel_shift));
+ int buf_offset1 = EPEL_EXTRA_BEFORE *
+ (edge_emu_stride + (1 << s->ps.sps->pixel_shift));
+
+ s->vdsp.emulated_edge_mc(lc->edge_emu_buffer2, src2 - offset1,
edge_emu_stride, src2stride,
block_w + EPEL_EXTRA, block_h + EPEL_EXTRA,
- x_off - EPEL_EXTRA_BEFORE,
- y_off - EPEL_EXTRA_BEFORE,
+ x_off1 - EPEL_EXTRA_BEFORE,
+ y_off1 - EPEL_EXTRA_BEFORE,
pic_width, pic_height);
- src2 = lc->edge_emu_buffer + buf_offset2;
- src2stride = edge_emu_stride;
- s->hevcdsp.put_hevc_epel[!!my][!!mx][pred_idx](dst2, dststride, src2, src2stride,
- block_h, mx, my, lc->mc_buffer);
- } else {
- s->hevcdsp.put_hevc_epel[!!my][!!mx][pred_idx](dst1, dststride, src1, src1stride,
- block_h, mx, my, lc->mc_buffer);
- s->hevcdsp.put_hevc_epel[!!my][!!mx][pred_idx](dst2, dststride, src2, src2stride,
- block_h, mx, my, lc->mc_buffer);
+ src2 = lc->edge_emu_buffer2 + buf_offset1;
+ src2stride = edge_emu_stride;
}
+
+ s->hevcdsp.put_hevc_epel[idx][!!my0][!!mx0](lc->tmp, src1, src1stride,
+ block_h, _mx0, _my0, block_w);
+ if (!weight_flag)
+ s->hevcdsp.put_hevc_epel_bi[idx][!!my1][!!mx1](dst0, s->frame->linesize[cidx+1],
+ src2, src2stride, lc->tmp,
+ block_h, _mx1, _my1, block_w);
+ else
+ s->hevcdsp.put_hevc_epel_bi_w[idx][!!my1][!!mx1](dst0, s->frame->linesize[cidx+1],
+ src2, src2stride, lc->tmp,
+ block_h,
+ s->sh.chroma_log2_weight_denom,
+ s->sh.chroma_weight_l0[current_mv->ref_idx[0]][cidx],
+ s->sh.chroma_weight_l1[current_mv->ref_idx[1]][cidx],
+ s->sh.chroma_offset_l0[current_mv->ref_idx[0]][cidx],
+ s->sh.chroma_offset_l1[current_mv->ref_idx[1]][cidx],
+ _mx1, _my1, block_w);
}
static void hevc_await_progress(HEVCContext *s, HEVCFrame *ref,
const Mv *mv, int y0, int height)
{
- int y = (mv->y >> 2) + y0 + height + 9;
- ff_thread_await_progress(&ref->tf, y, 0);
+ int y = FFMAX(0, (mv->y >> 2) + y0 + height + 9);
+
+ if (s->threads_type == FF_THREAD_FRAME )
+ ff_thread_await_progress(&ref->tf, y, 0);
}
-static void hevc_luma_mv_mpv_mode(HEVCContext *s, int x0, int y0, int nPbW,
+static void hevc_luma_mv_mvp_mode(HEVCContext *s, int x0, int y0, int nPbW,
int nPbH, int log2_cb_size, int part_idx,
int merge_idx, MvField *mv)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
enum InterPredIdc inter_pred_idc = PRED_L0;
int mvp_flag;
ff_hevc_set_neighbour_available(s, x0, y0, nPbW, nPbH);
+ mv->pred_flag = 0;
if (s->sh.slice_type == B_SLICE)
inter_pred_idc = ff_hevc_inter_pred_idc_decode(s, nPbW, nPbH);
if (s->sh.nb_refs[L0])
mv->ref_idx[0]= ff_hevc_ref_idx_lx_decode(s, s->sh.nb_refs[L0]);
- mv->pred_flag[0] = 1;
- hls_mvd_coding(s, x0, y0, 0);
+ mv->pred_flag = PF_L0;
+ ff_hevc_hls_mvd_coding(s, x0, y0, 0);
mvp_flag = ff_hevc_mvp_lx_flag_decode(s);
ff_hevc_luma_mv_mvp_mode(s, x0, y0, nPbW, nPbH, log2_cb_size,
part_idx, merge_idx, mv, mvp_flag, 0);
if (s->sh.mvd_l1_zero_flag == 1 && inter_pred_idc == PRED_BI) {
AV_ZERO32(&lc->pu.mvd);
} else {
- hls_mvd_coding(s, x0, y0, 1);
+ ff_hevc_hls_mvd_coding(s, x0, y0, 1);
}
- mv->pred_flag[1] = 1;
+ mv->pred_flag += PF_L1;
mvp_flag = ff_hevc_mvp_lx_flag_decode(s);
ff_hevc_luma_mv_mvp_mode(s, x0, y0, nPbW, nPbH, log2_cb_size,
part_idx, merge_idx, mv, mvp_flag, 1);
static void hls_prediction_unit(HEVCContext *s, int x0, int y0,
int nPbW, int nPbH,
- int log2_cb_size, int partIdx)
+ int log2_cb_size, int partIdx, int idx)
{
- static const int pred_indices[] = {
- [4] = 0, [8] = 1, [12] = 2, [16] = 3, [24] = 4, [32] = 5, [48] = 6, [64] = 7,
- };
- const int pred_idx = pred_indices[nPbW];
-
#define POS(c_idx, x, y) \
&s->frame->data[c_idx][((y) >> s->ps.sps->vshift[c_idx]) * s->frame->linesize[c_idx] + \
(((x) >> s->ps.sps->hshift[c_idx]) << s->ps.sps->pixel_shift)]
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
int merge_idx = 0;
struct MvField current_mv = {{{ 0 }}};
MvField *tab_mvf = s->ref->tab_mvf;
RefPicList *refPicList = s->ref->refPicList;
- HEVCFrame *ref0, *ref1;
-
- ptrdiff_t tmpstride = MAX_PB_SIZE * sizeof(int16_t);
-
+ HEVCFrame *ref0 = NULL, *ref1 = NULL;
uint8_t *dst0 = POS(0, x0, y0);
uint8_t *dst1 = POS(1, x0, y0);
uint8_t *dst2 = POS(2, x0, y0);
ff_hevc_luma_mv_merge_mode(s, x0, y0, nPbW, nPbH, log2_cb_size,
partIdx, merge_idx, ¤t_mv);
} else {
- hevc_luma_mv_mpv_mode(s, x0, y0, nPbW, nPbH, log2_cb_size,
+ hevc_luma_mv_mvp_mode(s, x0, y0, nPbW, nPbH, log2_cb_size,
partIdx, merge_idx, ¤t_mv);
}
for (i = 0; i < nPbW >> s->ps.sps->log2_min_pu_size; i++)
tab_mvf[(y_pu + j) * min_pu_width + x_pu + i] = current_mv;
- if (current_mv.pred_flag[0]) {
+ if (current_mv.pred_flag & PF_L0) {
ref0 = refPicList[0].ref[current_mv.ref_idx[0]];
if (!ref0)
return;
hevc_await_progress(s, ref0, ¤t_mv.mv[0], y0, nPbH);
}
- if (current_mv.pred_flag[1]) {
+ if (current_mv.pred_flag & PF_L1) {
ref1 = refPicList[1].ref[current_mv.ref_idx[1]];
if (!ref1)
return;
hevc_await_progress(s, ref1, ¤t_mv.mv[1], y0, nPbH);
}
- if (current_mv.pred_flag[0] && !current_mv.pred_flag[1]) {
- DECLARE_ALIGNED(16, int16_t, tmp[MAX_PB_SIZE * MAX_PB_SIZE]);
- DECLARE_ALIGNED(16, int16_t, tmp2[MAX_PB_SIZE * MAX_PB_SIZE]);
+ if (current_mv.pred_flag == PF_L0) {
+ int x0_c = x0 >> s->ps.sps->hshift[1];
+ int y0_c = y0 >> s->ps.sps->vshift[1];
+ int nPbW_c = nPbW >> s->ps.sps->hshift[1];
+ int nPbH_c = nPbH >> s->ps.sps->vshift[1];
- luma_mc(s, tmp, tmpstride, ref0->frame,
- ¤t_mv.mv[0], x0, y0, nPbW, nPbH, pred_idx);
+ luma_mc_uni(s, dst0, s->frame->linesize[0], ref0->frame,
+ ¤t_mv.mv[0], x0, y0, nPbW, nPbH,
+ s->sh.luma_weight_l0[current_mv.ref_idx[0]],
+ s->sh.luma_offset_l0[current_mv.ref_idx[0]]);
- if ((s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
- (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag)) {
- s->hevcdsp.weighted_pred[pred_idx](s->sh.luma_log2_weight_denom,
- s->sh.luma_weight_l0[current_mv.ref_idx[0]],
- s->sh.luma_offset_l0[current_mv.ref_idx[0]],
- dst0, s->frame->linesize[0], tmp,
- tmpstride, nPbH);
- } else {
- s->hevcdsp.put_unweighted_pred[pred_idx](dst0, s->frame->linesize[0], tmp, tmpstride, nPbH);
- }
- chroma_mc(s, tmp, tmp2, tmpstride, ref0->frame,
- ¤t_mv.mv[0], x0 / 2, y0 / 2, nPbW / 2, nPbH / 2, pred_idx);
-
- if ((s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
- (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag)) {
- s->hevcdsp.weighted_pred_chroma[pred_idx](s->sh.chroma_log2_weight_denom,
- s->sh.chroma_weight_l0[current_mv.ref_idx[0]][0],
- s->sh.chroma_offset_l0[current_mv.ref_idx[0]][0],
- dst1, s->frame->linesize[1], tmp, tmpstride,
- nPbH / 2);
- s->hevcdsp.weighted_pred_chroma[pred_idx](s->sh.chroma_log2_weight_denom,
- s->sh.chroma_weight_l0[current_mv.ref_idx[0]][1],
- s->sh.chroma_offset_l0[current_mv.ref_idx[0]][1],
- dst2, s->frame->linesize[2], tmp2, tmpstride,
- nPbH / 2);
- } else {
- s->hevcdsp.put_unweighted_pred_chroma[pred_idx](dst1, s->frame->linesize[1], tmp, tmpstride, nPbH / 2);
- s->hevcdsp.put_unweighted_pred_chroma[pred_idx](dst2, s->frame->linesize[2], tmp2, tmpstride, nPbH / 2);
- }
- } else if (!current_mv.pred_flag[0] && current_mv.pred_flag[1]) {
- DECLARE_ALIGNED(16, int16_t, tmp [MAX_PB_SIZE * MAX_PB_SIZE]);
- DECLARE_ALIGNED(16, int16_t, tmp2[MAX_PB_SIZE * MAX_PB_SIZE]);
-
- luma_mc(s, tmp, tmpstride, ref1->frame,
- ¤t_mv.mv[1], x0, y0, nPbW, nPbH, pred_idx);
-
- if ((s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
- (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag)) {
- s->hevcdsp.weighted_pred[pred_idx](s->sh.luma_log2_weight_denom,
- s->sh.luma_weight_l1[current_mv.ref_idx[1]],
- s->sh.luma_offset_l1[current_mv.ref_idx[1]],
- dst0, s->frame->linesize[0], tmp, tmpstride,
- nPbH);
- } else {
- s->hevcdsp.put_unweighted_pred[pred_idx](dst0, s->frame->linesize[0], tmp, tmpstride, nPbH);
- }
-
- chroma_mc(s, tmp, tmp2, tmpstride, ref1->frame,
- ¤t_mv.mv[1], x0 / 2, y0 / 2, nPbW / 2, nPbH / 2, pred_idx);
-
- if ((s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
- (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag)) {
- s->hevcdsp.weighted_pred_chroma[pred_idx](s->sh.chroma_log2_weight_denom,
- s->sh.chroma_weight_l1[current_mv.ref_idx[1]][0],
- s->sh.chroma_offset_l1[current_mv.ref_idx[1]][0],
- dst1, s->frame->linesize[1], tmp, tmpstride, nPbH/2);
- s->hevcdsp.weighted_pred_chroma[pred_idx](s->sh.chroma_log2_weight_denom,
- s->sh.chroma_weight_l1[current_mv.ref_idx[1]][1],
- s->sh.chroma_offset_l1[current_mv.ref_idx[1]][1],
- dst2, s->frame->linesize[2], tmp2, tmpstride, nPbH/2);
- } else {
- s->hevcdsp.put_unweighted_pred_chroma[pred_idx](dst1, s->frame->linesize[1], tmp, tmpstride, nPbH / 2);
- s->hevcdsp.put_unweighted_pred_chroma[pred_idx](dst2, s->frame->linesize[2], tmp2, tmpstride, nPbH / 2);
- }
- } else if (current_mv.pred_flag[0] && current_mv.pred_flag[1]) {
- DECLARE_ALIGNED(16, int16_t, tmp [MAX_PB_SIZE * MAX_PB_SIZE]);
- DECLARE_ALIGNED(16, int16_t, tmp2[MAX_PB_SIZE * MAX_PB_SIZE]);
- DECLARE_ALIGNED(16, int16_t, tmp3[MAX_PB_SIZE * MAX_PB_SIZE]);
- DECLARE_ALIGNED(16, int16_t, tmp4[MAX_PB_SIZE * MAX_PB_SIZE]);
-
- luma_mc(s, tmp, tmpstride, ref0->frame,
- ¤t_mv.mv[0], x0, y0, nPbW, nPbH, pred_idx);
- luma_mc(s, tmp2, tmpstride, ref1->frame,
- ¤t_mv.mv[1], x0, y0, nPbW, nPbH, pred_idx);
-
- if ((s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
- (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag)) {
- s->hevcdsp.weighted_pred_avg[pred_idx](s->sh.luma_log2_weight_denom,
- s->sh.luma_weight_l0[current_mv.ref_idx[0]],
- s->sh.luma_weight_l1[current_mv.ref_idx[1]],
- s->sh.luma_offset_l0[current_mv.ref_idx[0]],
- s->sh.luma_offset_l1[current_mv.ref_idx[1]],
- dst0, s->frame->linesize[0],
- tmp, tmp2, tmpstride, nPbH);
- } else {
- s->hevcdsp.put_unweighted_pred_avg[pred_idx](dst0, s->frame->linesize[0],
- tmp, tmp2, tmpstride, nPbH);
- }
-
- chroma_mc(s, tmp, tmp2, tmpstride, ref0->frame,
- ¤t_mv.mv[0], x0 / 2, y0 / 2, nPbW / 2, nPbH / 2, pred_idx);
- chroma_mc(s, tmp3, tmp4, tmpstride, ref1->frame,
- ¤t_mv.mv[1], x0 / 2, y0 / 2, nPbW / 2, nPbH / 2, pred_idx);
-
- if ((s->sh.slice_type == P_SLICE && s->ps.pps->weighted_pred_flag) ||
- (s->sh.slice_type == B_SLICE && s->ps.pps->weighted_bipred_flag)) {
- s->hevcdsp.weighted_pred_avg_chroma[pred_idx](s->sh.chroma_log2_weight_denom,
- s->sh.chroma_weight_l0[current_mv.ref_idx[0]][0],
- s->sh.chroma_weight_l1[current_mv.ref_idx[1]][0],
- s->sh.chroma_offset_l0[current_mv.ref_idx[0]][0],
- s->sh.chroma_offset_l1[current_mv.ref_idx[1]][0],
- dst1, s->frame->linesize[1], tmp, tmp3,
- tmpstride, nPbH / 2);
- s->hevcdsp.weighted_pred_avg_chroma[pred_idx](s->sh.chroma_log2_weight_denom,
- s->sh.chroma_weight_l0[current_mv.ref_idx[0]][1],
- s->sh.chroma_weight_l1[current_mv.ref_idx[1]][1],
- s->sh.chroma_offset_l0[current_mv.ref_idx[0]][1],
- s->sh.chroma_offset_l1[current_mv.ref_idx[1]][1],
- dst2, s->frame->linesize[2], tmp2, tmp4,
- tmpstride, nPbH / 2);
- } else {
- s->hevcdsp.put_unweighted_pred_avg_chroma[pred_idx](dst1, s->frame->linesize[1], tmp, tmp3, tmpstride, nPbH/2);
- s->hevcdsp.put_unweighted_pred_avg_chroma[pred_idx](dst2, s->frame->linesize[2], tmp2, tmp4, tmpstride, nPbH/2);
+ if (s->ps.sps->chroma_format_idc) {
+ chroma_mc_uni(s, dst1, s->frame->linesize[1], ref0->frame->data[1], ref0->frame->linesize[1],
+ 0, x0_c, y0_c, nPbW_c, nPbH_c, ¤t_mv,
+ s->sh.chroma_weight_l0[current_mv.ref_idx[0]][0], s->sh.chroma_offset_l0[current_mv.ref_idx[0]][0]);
+ chroma_mc_uni(s, dst2, s->frame->linesize[2], ref0->frame->data[2], ref0->frame->linesize[2],
+ 0, x0_c, y0_c, nPbW_c, nPbH_c, ¤t_mv,
+ s->sh.chroma_weight_l0[current_mv.ref_idx[0]][1], s->sh.chroma_offset_l0[current_mv.ref_idx[0]][1]);
+ }
+ } else if (current_mv.pred_flag == PF_L1) {
+ int x0_c = x0 >> s->ps.sps->hshift[1];
+ int y0_c = y0 >> s->ps.sps->vshift[1];
+ int nPbW_c = nPbW >> s->ps.sps->hshift[1];
+ int nPbH_c = nPbH >> s->ps.sps->vshift[1];
+
+ luma_mc_uni(s, dst0, s->frame->linesize[0], ref1->frame,
+ ¤t_mv.mv[1], x0, y0, nPbW, nPbH,
+ s->sh.luma_weight_l1[current_mv.ref_idx[1]],
+ s->sh.luma_offset_l1[current_mv.ref_idx[1]]);
+
+ if (s->ps.sps->chroma_format_idc) {
+ chroma_mc_uni(s, dst1, s->frame->linesize[1], ref1->frame->data[1], ref1->frame->linesize[1],
+ 1, x0_c, y0_c, nPbW_c, nPbH_c, ¤t_mv,
+ s->sh.chroma_weight_l1[current_mv.ref_idx[1]][0], s->sh.chroma_offset_l1[current_mv.ref_idx[1]][0]);
+
+ chroma_mc_uni(s, dst2, s->frame->linesize[2], ref1->frame->data[2], ref1->frame->linesize[2],
+ 1, x0_c, y0_c, nPbW_c, nPbH_c, ¤t_mv,
+ s->sh.chroma_weight_l1[current_mv.ref_idx[1]][1], s->sh.chroma_offset_l1[current_mv.ref_idx[1]][1]);
+ }
+ } else if (current_mv.pred_flag == PF_BI) {
+ int x0_c = x0 >> s->ps.sps->hshift[1];
+ int y0_c = y0 >> s->ps.sps->vshift[1];
+ int nPbW_c = nPbW >> s->ps.sps->hshift[1];
+ int nPbH_c = nPbH >> s->ps.sps->vshift[1];
+
+ luma_mc_bi(s, dst0, s->frame->linesize[0], ref0->frame,
+ ¤t_mv.mv[0], x0, y0, nPbW, nPbH,
+ ref1->frame, ¤t_mv.mv[1], ¤t_mv);
+
+ if (s->ps.sps->chroma_format_idc) {
+ chroma_mc_bi(s, dst1, s->frame->linesize[1], ref0->frame, ref1->frame,
+ x0_c, y0_c, nPbW_c, nPbH_c, ¤t_mv, 0);
+
+ chroma_mc_bi(s, dst2, s->frame->linesize[2], ref0->frame, ref1->frame,
+ x0_c, y0_c, nPbW_c, nPbH_c, ¤t_mv, 1);
}
}
}
static int luma_intra_pred_mode(HEVCContext *s, int x0, int y0, int pu_size,
int prev_intra_luma_pred_flag)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
int x_pu = x0 >> s->ps.sps->log2_min_pu_size;
int y_pu = y0 >> s->ps.sps->log2_min_pu_size;
int min_pu_width = s->ps.sps->min_pu_width;
int size_in_pus = pu_size >> s->ps.sps->log2_min_pu_size;
- int x0b = x0 & ((1 << s->ps.sps->log2_ctb_size) - 1);
- int y0b = y0 & ((1 << s->ps.sps->log2_ctb_size) - 1);
+ int x0b = av_mod_uintp2(x0, s->ps.sps->log2_ctb_size);
+ int y0b = av_mod_uintp2(y0, s->ps.sps->log2_ctb_size);
int cand_up = (lc->ctb_up_flag || y0b) ?
s->tab_ipm[(y_pu - 1) * min_pu_width + x_pu] : INTRA_DC;
intra_pred_mode, size_in_pus);
for (j = 0; j < size_in_pus; j++) {
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].is_intra = 1;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].pred_flag[0] = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].pred_flag[1] = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].ref_idx[0] = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].ref_idx[1] = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].mv[0].x = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].mv[0].y = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].mv[1].x = 0;
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].mv[1].y = 0;
+ tab_mvf[(y_pu + j) * min_pu_width + x_pu + i].pred_flag = PF_INTRA;
}
}
ct_depth, length);
}
+static const uint8_t tab_mode_idx[] = {
+ 0, 1, 2, 2, 2, 2, 3, 5, 7, 8, 10, 12, 13, 15, 17, 18, 19, 20,
+ 21, 22, 23, 23, 24, 24, 25, 25, 26, 27, 27, 28, 28, 29, 29, 30, 31};
+
static void intra_prediction_unit(HEVCContext *s, int x0, int y0,
int log2_cb_size)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
static const uint8_t intra_chroma_table[4] = { 0, 26, 10, 1 };
uint8_t prev_intra_luma_pred_flag[4];
int split = lc->cu.part_mode == PART_NxN;
}
}
- chroma_mode = ff_hevc_intra_chroma_pred_mode_decode(s);
- if (chroma_mode != 4) {
- if (lc->pu.intra_pred_mode[0] == intra_chroma_table[chroma_mode])
- lc->pu.intra_pred_mode_c = 34;
- else
- lc->pu.intra_pred_mode_c = intra_chroma_table[chroma_mode];
- } else {
- lc->pu.intra_pred_mode_c = lc->pu.intra_pred_mode[0];
+ if (s->ps.sps->chroma_format_idc == 3) {
+ for (i = 0; i < side; i++) {
+ for (j = 0; j < side; j++) {
+ lc->pu.chroma_mode_c[2 * i + j] = chroma_mode = ff_hevc_intra_chroma_pred_mode_decode(s);
+ if (chroma_mode != 4) {
+ if (lc->pu.intra_pred_mode[2 * i + j] == intra_chroma_table[chroma_mode])
+ lc->pu.intra_pred_mode_c[2 * i + j] = 34;
+ else
+ lc->pu.intra_pred_mode_c[2 * i + j] = intra_chroma_table[chroma_mode];
+ } else {
+ lc->pu.intra_pred_mode_c[2 * i + j] = lc->pu.intra_pred_mode[2 * i + j];
+ }
+ }
+ }
+ } else if (s->ps.sps->chroma_format_idc == 2) {
+ int mode_idx;
+ lc->pu.chroma_mode_c[0] = chroma_mode = ff_hevc_intra_chroma_pred_mode_decode(s);
+ if (chroma_mode != 4) {
+ if (lc->pu.intra_pred_mode[0] == intra_chroma_table[chroma_mode])
+ mode_idx = 34;
+ else
+ mode_idx = intra_chroma_table[chroma_mode];
+ } else {
+ mode_idx = lc->pu.intra_pred_mode[0];
+ }
+ lc->pu.intra_pred_mode_c[0] = tab_mode_idx[mode_idx];
+ } else if (s->ps.sps->chroma_format_idc != 0) {
+ chroma_mode = ff_hevc_intra_chroma_pred_mode_decode(s);
+ if (chroma_mode != 4) {
+ if (lc->pu.intra_pred_mode[0] == intra_chroma_table[chroma_mode])
+ lc->pu.intra_pred_mode_c[0] = 34;
+ else
+ lc->pu.intra_pred_mode_c[0] = intra_chroma_table[chroma_mode];
+ } else {
+ lc->pu.intra_pred_mode_c[0] = lc->pu.intra_pred_mode[0];
+ }
}
}
int x0, int y0,
int log2_cb_size)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
int pb_size = 1 << log2_cb_size;
int size_in_pus = pb_size >> s->ps.sps->log2_min_pu_size;
int min_pu_width = s->ps.sps->min_pu_width;
if (size_in_pus == 0)
size_in_pus = 1;
- for (j = 0; j < size_in_pus; j++) {
+ for (j = 0; j < size_in_pus; j++)
memset(&s->tab_ipm[(y_pu + j) * min_pu_width + x_pu], INTRA_DC, size_in_pus);
- for (k = 0; k < size_in_pus; k++)
- tab_mvf[(y_pu + j) * min_pu_width + x_pu + k].is_intra = lc->cu.pred_mode == MODE_INTRA;
- }
+ if (lc->cu.pred_mode == MODE_INTRA)
+ for (j = 0; j < size_in_pus; j++)
+ for (k = 0; k < size_in_pus; k++)
+ tab_mvf[(y_pu + j) * min_pu_width + x_pu + k].pred_flag = PF_INTRA;
}
static int hls_coding_unit(HEVCContext *s, int x0, int y0, int log2_cb_size)
{
int cb_size = 1 << log2_cb_size;
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
int log2_min_cb_size = s->ps.sps->log2_min_cb_size;
int length = cb_size >> log2_min_cb_size;
int min_cb_width = s->ps.sps->min_cb_width;
int x_cb = x0 >> log2_min_cb_size;
int y_cb = y0 >> log2_min_cb_size;
+ int idx = log2_cb_size - 2;
+ int qp_block_mask = (1<<(s->ps.sps->log2_ctb_size - s->ps.pps->diff_cu_qp_delta_depth)) - 1;
int x, y, ret;
lc->cu.x = x0;
x += min_cb_width;
}
lc->cu.pred_mode = skip_flag ? MODE_SKIP : MODE_INTER;
+ } else {
+ x = y_cb * min_cb_width + x_cb;
+ for (y = 0; y < length; y++) {
+ memset(&s->skip_flag[x], 0, length);
+ x += min_cb_width;
+ }
}
if (SAMPLE_CTB(s->skip_flag, x_cb, y_cb)) {
- hls_prediction_unit(s, x0, y0, cb_size, cb_size, log2_cb_size, 0);
+ hls_prediction_unit(s, x0, y0, cb_size, cb_size, log2_cb_size, 0, idx);
intra_prediction_unit_default_value(s, x0, y0, log2_cb_size);
if (!s->sh.disable_deblocking_filter_flag)
intra_prediction_unit_default_value(s, x0, y0, log2_cb_size);
switch (lc->cu.part_mode) {
case PART_2Nx2N:
- hls_prediction_unit(s, x0, y0, cb_size, cb_size, log2_cb_size, 0);
+ hls_prediction_unit(s, x0, y0, cb_size, cb_size, log2_cb_size, 0, idx);
break;
case PART_2NxN:
- hls_prediction_unit(s, x0, y0, cb_size, cb_size / 2, log2_cb_size, 0);
- hls_prediction_unit(s, x0, y0 + cb_size / 2, cb_size, cb_size / 2, log2_cb_size, 1);
+ hls_prediction_unit(s, x0, y0, cb_size, cb_size / 2, log2_cb_size, 0, idx);
+ hls_prediction_unit(s, x0, y0 + cb_size / 2, cb_size, cb_size / 2, log2_cb_size, 1, idx);
break;
case PART_Nx2N:
- hls_prediction_unit(s, x0, y0, cb_size / 2, cb_size, log2_cb_size, 0);
- hls_prediction_unit(s, x0 + cb_size / 2, y0, cb_size / 2, cb_size, log2_cb_size, 1);
+ hls_prediction_unit(s, x0, y0, cb_size / 2, cb_size, log2_cb_size, 0, idx - 1);
+ hls_prediction_unit(s, x0 + cb_size / 2, y0, cb_size / 2, cb_size, log2_cb_size, 1, idx - 1);
break;
case PART_2NxnU:
- hls_prediction_unit(s, x0, y0, cb_size, cb_size / 4, log2_cb_size, 0);
- hls_prediction_unit(s, x0, y0 + cb_size / 4, cb_size, cb_size * 3 / 4, log2_cb_size, 1);
+ hls_prediction_unit(s, x0, y0, cb_size, cb_size / 4, log2_cb_size, 0, idx);
+ hls_prediction_unit(s, x0, y0 + cb_size / 4, cb_size, cb_size * 3 / 4, log2_cb_size, 1, idx);
break;
case PART_2NxnD:
- hls_prediction_unit(s, x0, y0, cb_size, cb_size * 3 / 4, log2_cb_size, 0);
- hls_prediction_unit(s, x0, y0 + cb_size * 3 / 4, cb_size, cb_size / 4, log2_cb_size, 1);
+ hls_prediction_unit(s, x0, y0, cb_size, cb_size * 3 / 4, log2_cb_size, 0, idx);
+ hls_prediction_unit(s, x0, y0 + cb_size * 3 / 4, cb_size, cb_size / 4, log2_cb_size, 1, idx);
break;
case PART_nLx2N:
- hls_prediction_unit(s, x0, y0, cb_size / 4, cb_size, log2_cb_size, 0);
- hls_prediction_unit(s, x0 + cb_size / 4, y0, cb_size * 3 / 4, cb_size, log2_cb_size, 1);
+ hls_prediction_unit(s, x0, y0, cb_size / 4, cb_size, log2_cb_size, 0, idx - 2);
+ hls_prediction_unit(s, x0 + cb_size / 4, y0, cb_size * 3 / 4, cb_size, log2_cb_size, 1, idx - 2);
break;
case PART_nRx2N:
- hls_prediction_unit(s, x0, y0, cb_size * 3 / 4, cb_size, log2_cb_size, 0);
- hls_prediction_unit(s, x0 + cb_size * 3 / 4, y0, cb_size / 4, cb_size, log2_cb_size, 1);
+ hls_prediction_unit(s, x0, y0, cb_size * 3 / 4, cb_size, log2_cb_size, 0, idx - 2);
+ hls_prediction_unit(s, x0 + cb_size * 3 / 4, y0, cb_size / 4, cb_size, log2_cb_size, 1, idx - 2);
break;
case PART_NxN:
- hls_prediction_unit(s, x0, y0, cb_size / 2, cb_size / 2, log2_cb_size, 0);
- hls_prediction_unit(s, x0 + cb_size / 2, y0, cb_size / 2, cb_size / 2, log2_cb_size, 1);
- hls_prediction_unit(s, x0, y0 + cb_size / 2, cb_size / 2, cb_size / 2, log2_cb_size, 2);
- hls_prediction_unit(s, x0 + cb_size / 2, y0 + cb_size / 2, cb_size / 2, cb_size / 2, log2_cb_size, 3);
+ hls_prediction_unit(s, x0, y0, cb_size / 2, cb_size / 2, log2_cb_size, 0, idx - 1);
+ hls_prediction_unit(s, x0 + cb_size / 2, y0, cb_size / 2, cb_size / 2, log2_cb_size, 1, idx - 1);
+ hls_prediction_unit(s, x0, y0 + cb_size / 2, cb_size / 2, cb_size / 2, log2_cb_size, 2, idx - 1);
+ hls_prediction_unit(s, x0 + cb_size / 2, y0 + cb_size / 2, cb_size / 2, cb_size / 2, log2_cb_size, 3, idx - 1);
break;
}
}
rqt_root_cbf = ff_hevc_no_residual_syntax_flag_decode(s);
}
if (rqt_root_cbf) {
+ const static int cbf[2] = { 0 };
lc->cu.max_trafo_depth = lc->cu.pred_mode == MODE_INTRA ?
s->ps.sps->max_transform_hierarchy_depth_intra + lc->cu.intra_split_flag :
s->ps.sps->max_transform_hierarchy_depth_inter;
ret = hls_transform_tree(s, x0, y0, x0, y0, x0, y0,
log2_cb_size,
- log2_cb_size, 0, 0, 0, 0);
+ log2_cb_size, 0, 0, cbf, cbf);
if (ret < 0)
return ret;
} else {
}
if (s->ps.pps->cu_qp_delta_enabled_flag && lc->tu.is_cu_qp_delta_coded == 0)
- ff_hevc_set_qPy(s, x0, y0, x0, y0, log2_cb_size);
+ ff_hevc_set_qPy(s, x0, y0, log2_cb_size);
x = y_cb * min_cb_width + x_cb;
for (y = 0; y < length; y++) {
x += min_cb_width;
}
- set_ct_depth(s, x0, y0, log2_cb_size, lc->ct.depth);
+ if(((x0 + (1<<log2_cb_size)) & qp_block_mask) == 0 &&
+ ((y0 + (1<<log2_cb_size)) & qp_block_mask) == 0) {
+ lc->qPy_pred = lc->qp_y;
+ }
+
+ set_ct_depth(s, x0, y0, log2_cb_size, lc->ct_depth);
return 0;
}
static int hls_coding_quadtree(HEVCContext *s, int x0, int y0,
int log2_cb_size, int cb_depth)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
const int cb_size = 1 << log2_cb_size;
+ int ret;
int split_cu;
- lc->ct.depth = cb_depth;
+ lc->ct_depth = cb_depth;
if (x0 + cb_size <= s->ps.sps->width &&
y0 + cb_size <= s->ps.sps->height &&
log2_cb_size > s->ps.sps->log2_min_cb_size) {
lc->tu.cu_qp_delta = 0;
}
+ if (s->sh.cu_chroma_qp_offset_enabled_flag &&
+ log2_cb_size >= s->ps.sps->log2_ctb_size - s->ps.pps->diff_cu_chroma_qp_offset_depth) {
+ lc->tu.is_cu_chroma_qp_offset_coded = 0;
+ }
+
if (split_cu) {
+ int qp_block_mask = (1<<(s->ps.sps->log2_ctb_size - s->ps.pps->diff_cu_qp_delta_depth)) - 1;
const int cb_size_split = cb_size >> 1;
const int x1 = x0 + cb_size_split;
const int y1 = y0 + cb_size_split;
- log2_cb_size--;
- cb_depth++;
+ int more_data = 0;
-#define SUBDIVIDE(x, y) \
-do { \
- if (x < s->ps.sps->width && y < s->ps.sps->height) { \
- int ret = hls_coding_quadtree(s, x, y, log2_cb_size, cb_depth);\
- if (ret < 0) \
- return ret; \
- } \
-} while (0)
+ more_data = hls_coding_quadtree(s, x0, y0, log2_cb_size - 1, cb_depth + 1);
+ if (more_data < 0)
+ return more_data;
+
+ if (more_data && x1 < s->ps.sps->width) {
+ more_data = hls_coding_quadtree(s, x1, y0, log2_cb_size - 1, cb_depth + 1);
+ if (more_data < 0)
+ return more_data;
+ }
+ if (more_data && y1 < s->ps.sps->height) {
+ more_data = hls_coding_quadtree(s, x0, y1, log2_cb_size - 1, cb_depth + 1);
+ if (more_data < 0)
+ return more_data;
+ }
+ if (more_data && x1 < s->ps.sps->width &&
+ y1 < s->ps.sps->height) {
+ more_data = hls_coding_quadtree(s, x1, y1, log2_cb_size - 1, cb_depth + 1);
+ if (more_data < 0)
+ return more_data;
+ }
+
+ if(((x0 + (1<<log2_cb_size)) & qp_block_mask) == 0 &&
+ ((y0 + (1<<log2_cb_size)) & qp_block_mask) == 0)
+ lc->qPy_pred = lc->qp_y;
- SUBDIVIDE(x0, y0);
- SUBDIVIDE(x1, y0);
- SUBDIVIDE(x0, y1);
- SUBDIVIDE(x1, y1);
+ if (more_data)
+ return ((x1 + cb_size_split) < s->ps.sps->width ||
+ (y1 + cb_size_split) < s->ps.sps->height);
+ else
+ return 0;
} else {
- int ret = hls_coding_unit(s, x0, y0, log2_cb_size);
+ ret = hls_coding_unit(s, x0, y0, log2_cb_size);
if (ret < 0)
return ret;
+ if ((!((x0 + cb_size) %
+ (1 << (s->ps.sps->log2_ctb_size))) ||
+ (x0 + cb_size >= s->ps.sps->width)) &&
+ (!((y0 + cb_size) %
+ (1 << (s->ps.sps->log2_ctb_size))) ||
+ (y0 + cb_size >= s->ps.sps->height))) {
+ int end_of_slice_flag = ff_hevc_end_of_slice_flag_decode(s);
+ return !end_of_slice_flag;
+ } else {
+ return 1;
+ }
}
return 0;
static void hls_decode_neighbour(HEVCContext *s, int x_ctb, int y_ctb,
int ctb_addr_ts)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
int ctb_size = 1 << s->ps.sps->log2_ctb_size;
int ctb_addr_rs = s->ps.pps->ctb_addr_ts_to_rs[ctb_addr_ts];
int ctb_addr_in_slice = ctb_addr_rs - s->sh.slice_addr;
} else if (s->ps.pps->tiles_enabled_flag) {
if (ctb_addr_ts && s->ps.pps->tile_id[ctb_addr_ts] != s->ps.pps->tile_id[ctb_addr_ts - 1]) {
int idxX = s->ps.pps->col_idxX[x_ctb >> s->ps.sps->log2_ctb_size];
- lc->start_of_tiles_x = x_ctb;
lc->end_of_tiles_x = x_ctb + (s->ps.pps->column_width[idxX] << s->ps.sps->log2_ctb_size);
lc->first_qp_group = 1;
}
if (y_ctb > 0 && s->tab_slice_address[ctb_addr_rs] != s->tab_slice_address[ctb_addr_rs - s->ps.sps->ctb_width])
lc->boundary_flags |= BOUNDARY_UPPER_SLICE;
} else {
- if (!ctb_addr_in_slice)
+ if (ctb_addr_in_slice <= 0)
lc->boundary_flags |= BOUNDARY_LEFT_SLICE;
if (ctb_addr_in_slice < s->ps.sps->ctb_width)
lc->boundary_flags |= BOUNDARY_UPPER_SLICE;
lc->ctb_up_left_flag = ((x_ctb > 0) && (y_ctb > 0) && (ctb_addr_in_slice-1 >= s->ps.sps->ctb_width) && (s->ps.pps->tile_id[ctb_addr_ts] == s->ps.pps->tile_id[s->ps.pps->ctb_addr_rs_to_ts[ctb_addr_rs-1 - s->ps.sps->ctb_width]]));
}
-static int hls_slice_data(HEVCContext *s)
+static int hls_decode_entry(AVCodecContext *avctxt, void *isFilterThread)
{
+ HEVCContext *s = avctxt->priv_data;
int ctb_size = 1 << s->ps.sps->log2_ctb_size;
int more_data = 1;
int x_ctb = 0;
int y_ctb = 0;
int ctb_addr_ts = s->ps.pps->ctb_addr_rs_to_ts[s->sh.slice_ctb_addr_rs];
- int ret;
+
+ if (!ctb_addr_ts && s->sh.dependent_slice_segment_flag) {
+ av_log(s->avctx, AV_LOG_ERROR, "Impossible initial tile.\n");
+ return AVERROR_INVALIDDATA;
+ }
+
+ if (s->sh.dependent_slice_segment_flag) {
+ int prev_rs = s->ps.pps->ctb_addr_ts_to_rs[ctb_addr_ts - 1];
+ if (s->tab_slice_address[prev_rs] != s->sh.slice_addr) {
+ av_log(s->avctx, AV_LOG_ERROR, "Previous slice segment missing\n");
+ return AVERROR_INVALIDDATA;
+ }
+ }
while (more_data && ctb_addr_ts < s->ps.sps->ctb_size) {
int ctb_addr_rs = s->ps.pps->ctb_addr_ts_to_rs[ctb_addr_ts];
s->deblock[ctb_addr_rs].tc_offset = s->sh.tc_offset;
s->filter_slice_edges[ctb_addr_rs] = s->sh.slice_loop_filter_across_slices_enabled_flag;
- ret = hls_coding_quadtree(s, x_ctb, y_ctb, s->ps.sps->log2_ctb_size, 0);
- if (ret < 0)
- return ret;
- more_data = !ff_hevc_end_of_slice_flag_decode(s);
+ more_data = hls_coding_quadtree(s, x_ctb, y_ctb, s->ps.sps->log2_ctb_size, 0);
+ if (more_data < 0) {
+ s->tab_slice_address[ctb_addr_rs] = -1;
+ return more_data;
+ }
+
ctb_addr_ts++;
ff_hevc_save_states(s, ctb_addr_ts);
if (x_ctb + ctb_size >= s->ps.sps->width &&
y_ctb + ctb_size >= s->ps.sps->height)
- ff_hevc_hls_filter(s, x_ctb, y_ctb);
+ ff_hevc_hls_filter(s, x_ctb, y_ctb, ctb_size);
return ctb_addr_ts;
}
-static void restore_tqb_pixels(HEVCContext *s)
+static int hls_slice_data(HEVCContext *s)
{
- int min_pu_size = 1 << s->ps.sps->log2_min_pu_size;
- int x, y, c_idx;
-
- for (c_idx = 0; c_idx < 3; c_idx++) {
- ptrdiff_t stride = s->frame->linesize[c_idx];
- int hshift = s->ps.sps->hshift[c_idx];
- int vshift = s->ps.sps->vshift[c_idx];
- for (y = 0; y < s->ps.sps->min_pu_height; y++) {
- for (x = 0; x < s->ps.sps->min_pu_width; x++) {
- if (s->is_pcm[y * s->ps.sps->min_pu_width + x]) {
- int n;
- int len = min_pu_size >> hshift;
- uint8_t *src = &s->frame->data[c_idx][((y << s->ps.sps->log2_min_pu_size) >> vshift) * stride + (((x << s->ps.sps->log2_min_pu_size) >> hshift) << s->ps.sps->pixel_shift)];
- uint8_t *dst = &s->sao_frame->data[c_idx][((y << s->ps.sps->log2_min_pu_size) >> vshift) * stride + (((x << s->ps.sps->log2_min_pu_size) >> hshift) << s->ps.sps->pixel_shift)];
- for (n = 0; n < (min_pu_size >> vshift); n++) {
- memcpy(dst, src, len);
- src += stride;
- dst += stride;
- }
- }
+ int arg[2];
+ int ret[2];
+
+ arg[0] = 0;
+ arg[1] = 1;
+
+ s->avctx->execute(s->avctx, hls_decode_entry, arg, ret , 1, sizeof(int));
+ return ret[0];
+}
+static int hls_decode_entry_wpp(AVCodecContext *avctxt, void *input_ctb_row, int job, int self_id)
+{
+ HEVCContext *s1 = avctxt->priv_data, *s;
+ HEVCLocalContext *lc;
+ int ctb_size = 1<< s1->ps.sps->log2_ctb_size;
+ int more_data = 1;
+ int *ctb_row_p = input_ctb_row;
+ int ctb_row = ctb_row_p[job];
+ int ctb_addr_rs = s1->sh.slice_ctb_addr_rs + ctb_row * ((s1->ps.sps->width + ctb_size - 1) >> s1->ps.sps->log2_ctb_size);
+ int ctb_addr_ts = s1->ps.pps->ctb_addr_rs_to_ts[ctb_addr_rs];
+ int thread = ctb_row % s1->threads_number;
+ int ret;
+
+ s = s1->sList[self_id];
+ lc = s->HEVClc;
+
+ if(ctb_row) {
+ ret = init_get_bits8(&lc->gb, s->data + s->sh.offset[ctb_row - 1], s->sh.size[ctb_row - 1]);
+
+ if (ret < 0)
+ return ret;
+ ff_init_cabac_decoder(&lc->cc, s->data + s->sh.offset[(ctb_row)-1], s->sh.size[ctb_row - 1]);
+ }
+
+ while(more_data && ctb_addr_ts < s->ps.sps->ctb_size) {
+ int x_ctb = (ctb_addr_rs % s->ps.sps->ctb_width) << s->ps.sps->log2_ctb_size;
+ int y_ctb = (ctb_addr_rs / s->ps.sps->ctb_width) << s->ps.sps->log2_ctb_size;
+
+ hls_decode_neighbour(s, x_ctb, y_ctb, ctb_addr_ts);
+
+ ff_thread_await_progress2(s->avctx, ctb_row, thread, SHIFT_CTB_WPP);
+
+ if (atomic_load(&s1->wpp_err)) {
+ ff_thread_report_progress2(s->avctx, ctb_row , thread, SHIFT_CTB_WPP);
+ return 0;
+ }
+
+ ff_hevc_cabac_init(s, ctb_addr_ts);
+ hls_sao_param(s, x_ctb >> s->ps.sps->log2_ctb_size, y_ctb >> s->ps.sps->log2_ctb_size);
+ more_data = hls_coding_quadtree(s, x_ctb, y_ctb, s->ps.sps->log2_ctb_size, 0);
+
+ if (more_data < 0) {
+ s->tab_slice_address[ctb_addr_rs] = -1;
+ atomic_store(&s1->wpp_err, 1);
+ ff_thread_report_progress2(s->avctx, ctb_row ,thread, SHIFT_CTB_WPP);
+ return more_data;
+ }
+
+ ctb_addr_ts++;
+
+ ff_hevc_save_states(s, ctb_addr_ts);
+ ff_thread_report_progress2(s->avctx, ctb_row, thread, 1);
+ ff_hevc_hls_filters(s, x_ctb, y_ctb, ctb_size);
+
+ if (!more_data && (x_ctb+ctb_size) < s->ps.sps->width && ctb_row != s->sh.num_entry_point_offsets) {
+ atomic_store(&s1->wpp_err, 1);
+ ff_thread_report_progress2(s->avctx, ctb_row ,thread, SHIFT_CTB_WPP);
+ return 0;
+ }
+
+ if ((x_ctb+ctb_size) >= s->ps.sps->width && (y_ctb+ctb_size) >= s->ps.sps->height ) {
+ ff_hevc_hls_filter(s, x_ctb, y_ctb, ctb_size);
+ ff_thread_report_progress2(s->avctx, ctb_row , thread, SHIFT_CTB_WPP);
+ return ctb_addr_ts;
+ }
+ ctb_addr_rs = s->ps.pps->ctb_addr_ts_to_rs[ctb_addr_ts];
+ x_ctb+=ctb_size;
+
+ if(x_ctb >= s->ps.sps->width) {
+ break;
+ }
+ }
+ ff_thread_report_progress2(s->avctx, ctb_row ,thread, SHIFT_CTB_WPP);
+
+ return 0;
+}
+
+static int hls_slice_data_wpp(HEVCContext *s, const H2645NAL *nal)
+{
+ const uint8_t *data = nal->data;
+ int length = nal->size;
+ HEVCLocalContext *lc = s->HEVClc;
+ int *ret = av_malloc_array(s->sh.num_entry_point_offsets + 1, sizeof(int));
+ int *arg = av_malloc_array(s->sh.num_entry_point_offsets + 1, sizeof(int));
+ int64_t offset;
+ int64_t startheader, cmpt = 0;
+ int i, j, res = 0;
+
+ if (!ret || !arg) {
+ av_free(ret);
+ av_free(arg);
+ return AVERROR(ENOMEM);
+ }
+
+ if (s->sh.slice_ctb_addr_rs + s->sh.num_entry_point_offsets * s->ps.sps->ctb_width >= s->ps.sps->ctb_width * s->ps.sps->ctb_height) {
+ av_log(s->avctx, AV_LOG_ERROR, "WPP ctb addresses are wrong (%d %d %d %d)\n",
+ s->sh.slice_ctb_addr_rs, s->sh.num_entry_point_offsets,
+ s->ps.sps->ctb_width, s->ps.sps->ctb_height
+ );
+ res = AVERROR_INVALIDDATA;
+ goto error;
+ }
+
+ ff_alloc_entries(s->avctx, s->sh.num_entry_point_offsets + 1);
+
+ if (!s->sList[1]) {
+ for (i = 1; i < s->threads_number; i++) {
+ s->sList[i] = av_malloc(sizeof(HEVCContext));
+ memcpy(s->sList[i], s, sizeof(HEVCContext));
+ s->HEVClcList[i] = av_mallocz(sizeof(HEVCLocalContext));
+ s->sList[i]->HEVClc = s->HEVClcList[i];
+ }
+ }
+
+ offset = (lc->gb.index >> 3);
+
+ for (j = 0, cmpt = 0, startheader = offset + s->sh.entry_point_offset[0]; j < nal->skipped_bytes; j++) {
+ if (nal->skipped_bytes_pos[j] >= offset && nal->skipped_bytes_pos[j] < startheader) {
+ startheader--;
+ cmpt++;
+ }
+ }
+
+ for (i = 1; i < s->sh.num_entry_point_offsets; i++) {
+ offset += (s->sh.entry_point_offset[i - 1] - cmpt);
+ for (j = 0, cmpt = 0, startheader = offset
+ + s->sh.entry_point_offset[i]; j < nal->skipped_bytes; j++) {
+ if (nal->skipped_bytes_pos[j] >= offset && nal->skipped_bytes_pos[j] < startheader) {
+ startheader--;
+ cmpt++;
}
}
+ s->sh.size[i - 1] = s->sh.entry_point_offset[i] - cmpt;
+ s->sh.offset[i - 1] = offset;
+
+ }
+ if (s->sh.num_entry_point_offsets != 0) {
+ offset += s->sh.entry_point_offset[s->sh.num_entry_point_offsets - 1] - cmpt;
+ if (length < offset) {
+ av_log(s->avctx, AV_LOG_ERROR, "entry_point_offset table is corrupted\n");
+ res = AVERROR_INVALIDDATA;
+ goto error;
+ }
+ s->sh.size[s->sh.num_entry_point_offsets - 1] = length - offset;
+ s->sh.offset[s->sh.num_entry_point_offsets - 1] = offset;
+
+ }
+ s->data = data;
+
+ for (i = 1; i < s->threads_number; i++) {
+ s->sList[i]->HEVClc->first_qp_group = 1;
+ s->sList[i]->HEVClc->qp_y = s->sList[0]->HEVClc->qp_y;
+ memcpy(s->sList[i], s, sizeof(HEVCContext));
+ s->sList[i]->HEVClc = s->HEVClcList[i];
}
+
+ atomic_store(&s->wpp_err, 0);
+ ff_reset_entries(s->avctx);
+
+ for (i = 0; i <= s->sh.num_entry_point_offsets; i++) {
+ arg[i] = i;
+ ret[i] = 0;
+ }
+
+ if (s->ps.pps->entropy_coding_sync_enabled_flag)
+ s->avctx->execute2(s->avctx, hls_decode_entry_wpp, arg, ret, s->sh.num_entry_point_offsets + 1);
+
+ for (i = 0; i <= s->sh.num_entry_point_offsets; i++)
+ res += ret[i];
+error:
+ av_free(ret);
+ av_free(arg);
+ return res;
}
static int set_side_data(HEVCContext *s)
s->sei_hflip, s->sei_vflip);
}
+ // Decrement the mastering display flag when IRAP frame has no_rasl_output_flag=1
+ // so the side data persists for the entire coded video sequence.
+ if (s->sei_mastering_display_info_present > 0 &&
+ IS_IRAP(s) && s->no_rasl_output_flag) {
+ s->sei_mastering_display_info_present--;
+ }
+ if (s->sei_mastering_display_info_present) {
+ // HEVC uses a g,b,r ordering, which we convert to a more natural r,g,b
+ const int mapping[3] = {2, 0, 1};
+ const int chroma_den = 50000;
+ const int luma_den = 10000;
+ int i;
+ AVMasteringDisplayMetadata *metadata =
+ av_mastering_display_metadata_create_side_data(out);
+ if (!metadata)
+ return AVERROR(ENOMEM);
+
+ for (i = 0; i < 3; i++) {
+ const int j = mapping[i];
+ metadata->display_primaries[i][0].num = s->display_primaries[j][0];
+ metadata->display_primaries[i][0].den = chroma_den;
+ metadata->display_primaries[i][1].num = s->display_primaries[j][1];
+ metadata->display_primaries[i][1].den = chroma_den;
+ }
+ metadata->white_point[0].num = s->white_point[0];
+ metadata->white_point[0].den = chroma_den;
+ metadata->white_point[1].num = s->white_point[1];
+ metadata->white_point[1].den = chroma_den;
+
+ metadata->max_luminance.num = s->max_mastering_luminance;
+ metadata->max_luminance.den = luma_den;
+ metadata->min_luminance.num = s->min_mastering_luminance;
+ metadata->min_luminance.den = luma_den;
+ metadata->has_luminance = 1;
+ metadata->has_primaries = 1;
+
+ av_log(s->avctx, AV_LOG_DEBUG, "Mastering Display Metadata:\n");
+ av_log(s->avctx, AV_LOG_DEBUG,
+ "r(%5.4f,%5.4f) g(%5.4f,%5.4f) b(%5.4f %5.4f) wp(%5.4f, %5.4f)\n",
+ av_q2d(metadata->display_primaries[0][0]),
+ av_q2d(metadata->display_primaries[0][1]),
+ av_q2d(metadata->display_primaries[1][0]),
+ av_q2d(metadata->display_primaries[1][1]),
+ av_q2d(metadata->display_primaries[2][0]),
+ av_q2d(metadata->display_primaries[2][1]),
+ av_q2d(metadata->white_point[0]), av_q2d(metadata->white_point[1]));
+ av_log(s->avctx, AV_LOG_DEBUG,
+ "min_luminance=%f, max_luminance=%f\n",
+ av_q2d(metadata->min_luminance), av_q2d(metadata->max_luminance));
+ }
+
+ if (s->a53_caption) {
+ AVFrameSideData* sd = av_frame_new_side_data(out,
+ AV_FRAME_DATA_A53_CC,
+ s->a53_caption_size);
+ if (sd)
+ memcpy(sd->data, s->a53_caption, s->a53_caption_size);
+ av_freep(&s->a53_caption);
+ s->a53_caption_size = 0;
+ s->avctx->properties |= FF_CODEC_PROPERTY_CLOSED_CAPTIONS;
+ }
+
return 0;
}
static int hevc_frame_start(HEVCContext *s)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
+ int pic_size_in_ctb = ((s->ps.sps->width >> s->ps.sps->log2_min_cb_size) + 1) *
+ ((s->ps.sps->height >> s->ps.sps->log2_min_cb_size) + 1);
int ret;
- memset(s->horizontal_bs, 0, 2 * s->bs_width * (s->bs_height + 1));
- memset(s->vertical_bs, 0, 2 * s->bs_width * (s->bs_height + 1));
+ memset(s->horizontal_bs, 0, s->bs_width * s->bs_height);
+ memset(s->vertical_bs, 0, s->bs_width * s->bs_height);
memset(s->cbf_luma, 0, s->ps.sps->min_tb_width * s->ps.sps->min_tb_height);
- memset(s->is_pcm, 0, s->ps.sps->min_pu_width * s->ps.sps->min_pu_height);
+ memset(s->is_pcm, 0, (s->ps.sps->min_pu_width + 1) * (s->ps.sps->min_pu_height + 1));
+ memset(s->tab_slice_address, -1, pic_size_in_ctb * sizeof(*s->tab_slice_address));
- lc->start_of_tiles_x = 0;
s->is_decoded = 0;
s->first_nal_type = s->nal_unit_type;
if (s->ps.pps->tiles_enabled_flag)
lc->end_of_tiles_x = s->ps.pps->column_width[0] << s->ps.sps->log2_ctb_size;
- ret = ff_hevc_set_new_ref(s, s->ps.sps->sao_enabled ? &s->sao_frame : &s->frame,
- s->poc);
+ ret = ff_hevc_set_new_ref(s, &s->frame, s->poc);
if (ret < 0)
goto fail;
if (ret < 0)
goto fail;
+ s->frame->pict_type = 3 - s->sh.slice_type;
+
+ if (!IS_IRAP(s))
+ ff_hevc_bump_frame(s);
+
av_frame_unref(s->output_frame);
ret = ff_hevc_output_frame(s, s->output_frame, 0);
if (ret < 0)
goto fail;
- ff_thread_finish_setup(s->avctx);
+ if (!s->avctx->hwaccel)
+ ff_thread_finish_setup(s->avctx);
return 0;
static int decode_nal_unit(HEVCContext *s, const H2645NAL *nal)
{
- HEVCLocalContext *lc = &s->HEVClc;
+ HEVCLocalContext *lc = s->HEVClc;
GetBitContext *gb = &lc->gb;
int ctb_addr_ts, ret;
if (ret < 0)
goto fail;
} else {
- ctb_addr_ts = hls_slice_data(s);
+ if (s->threads_number > 1 && s->sh.num_entry_point_offsets > 0)
+ ctb_addr_ts = hls_slice_data_wpp(s, nal);
+ else
+ ctb_addr_ts = hls_slice_data(s);
if (ctb_addr_ts >= (s->ps.sps->ctb_width * s->ps.sps->ctb_height)) {
s->is_decoded = 1;
- if ((s->ps.pps->transquant_bypass_enable_flag ||
- (s->ps.sps->pcm.loop_filter_disable_flag && s->ps.sps->pcm_enabled_flag)) &&
- s->ps.sps->sao_enabled)
- restore_tqb_pixels(s);
}
if (ctb_addr_ts < 0) {
int i, ret = 0;
s->ref = NULL;
+ s->last_eos = s->eos;
s->eos = 0;
/* split the input packet into NAL units, so we know the upper bound on the
* number of slices in the frame */
ret = ff_h2645_packet_split(&s->pkt, buf, length, s->avctx, s->is_nalff,
- s->nal_length_size, s->avctx->codec_id);
+ s->nal_length_size, s->avctx->codec_id, 1);
if (ret < 0) {
av_log(s->avctx, AV_LOG_ERROR,
"Error splitting the input into NAL units.\n");
}
fail:
- if (s->ref)
+ if (s->ref && s->threads_type == FF_THREAD_FRAME)
ff_thread_report_progress(&s->ref->tf, INT_MAX, 0);
return ret;
return 0;
}
+static int hevc_decode_extradata(HEVCContext *s, uint8_t *buf, int length)
+{
+ AVCodecContext *avctx = s->avctx;
+ GetByteContext gb;
+ int ret, i;
+
+ bytestream2_init(&gb, buf, length);
+
+ if (length > 3 && (buf[0] || buf[1] || buf[2] > 1)) {
+ /* It seems the extradata is encoded as hvcC format.
+ * Temporarily, we support configurationVersion==0 until 14496-15 3rd
+ * is finalized. When finalized, configurationVersion will be 1 and we
+ * can recognize hvcC by checking if avctx->extradata[0]==1 or not. */
+ int i, j, num_arrays, nal_len_size;
+
+ s->is_nalff = 1;
+
+ bytestream2_skip(&gb, 21);
+ nal_len_size = (bytestream2_get_byte(&gb) & 3) + 1;
+ num_arrays = bytestream2_get_byte(&gb);
+
+ /* nal units in the hvcC always have length coded with 2 bytes,
+ * so put a fake nal_length_size = 2 while parsing them */
+ s->nal_length_size = 2;
+
+ /* Decode nal units from hvcC. */
+ for (i = 0; i < num_arrays; i++) {
+ int type = bytestream2_get_byte(&gb) & 0x3f;
+ int cnt = bytestream2_get_be16(&gb);
+
+ for (j = 0; j < cnt; j++) {
+ // +2 for the nal size field
+ int nalsize = bytestream2_peek_be16(&gb) + 2;
+ if (bytestream2_get_bytes_left(&gb) < nalsize) {
+ av_log(s->avctx, AV_LOG_ERROR,
+ "Invalid NAL unit size in extradata.\n");
+ return AVERROR_INVALIDDATA;
+ }
+
+ ret = decode_nal_units(s, gb.buffer, nalsize);
+ if (ret < 0) {
+ av_log(avctx, AV_LOG_ERROR,
+ "Decoding nal unit %d %d from hvcC failed\n",
+ type, i);
+ return ret;
+ }
+ bytestream2_skip(&gb, nalsize);
+ }
+ }
+
+ /* Now store right nal length size, that will be used to parse
+ * all other nals */
+ s->nal_length_size = nal_len_size;
+ } else {
+ s->is_nalff = 0;
+ ret = decode_nal_units(s, buf, length);
+ if (ret < 0)
+ return ret;
+ }
+
+ /* export stream parameters from the first SPS */
+ for (i = 0; i < FF_ARRAY_ELEMS(s->ps.sps_list); i++) {
+ if (s->ps.sps_list[i]) {
+ const HEVCSPS *sps = (const HEVCSPS*)s->ps.sps_list[i]->data;
+ export_stream_params(s->avctx, &s->ps, sps);
+ break;
+ }
+ }
+
+ return 0;
+}
+
static int hevc_decode_frame(AVCodecContext *avctx, void *data, int *got_output,
AVPacket *avpkt)
{
int ret;
+ int new_extradata_size;
+ uint8_t *new_extradata;
HEVCContext *s = avctx->priv_data;
if (!avpkt->size) {
return 0;
}
+ new_extradata = av_packet_get_side_data(avpkt, AV_PKT_DATA_NEW_EXTRADATA,
+ &new_extradata_size);
+ if (new_extradata && new_extradata_size > 0) {
+ ret = hevc_decode_extradata(s, new_extradata, new_extradata_size);
+ if (ret < 0)
+ return ret;
+ }
+
s->ref = NULL;
ret = decode_nal_units(s, avpkt->data, avpkt->size);
if (ret < 0)
return ret;
if (avctx->hwaccel) {
- if (s->ref && avctx->hwaccel->end_frame(avctx) < 0)
+ if (s->ref && (ret = avctx->hwaccel->end_frame(avctx)) < 0) {
av_log(avctx, AV_LOG_ERROR,
"hardware accelerator failed to decode picture\n");
+ ff_hevc_unref_frame(s, s->ref, ~0);
+ return ret;
+ }
} else {
/* verify the SEI checksum */
if (avctx->err_recognition & AV_EF_CRCCHECK && s->is_decoded &&
static int hevc_ref_frame(HEVCContext *s, HEVCFrame *dst, HEVCFrame *src)
{
- int ret = ff_thread_ref_frame(&dst->tf, &src->tf);
+ int ret;
+
+ ret = ff_thread_ref_frame(&dst->tf, &src->tf);
if (ret < 0)
return ret;
av_freep(&s->md5_ctx);
- av_frame_free(&s->tmp_frame);
+ av_freep(&s->cabac_state);
+
+ for (i = 0; i < 3; i++) {
+ av_freep(&s->sao_pixel_buffer_h[i]);
+ av_freep(&s->sao_pixel_buffer_v[i]);
+ }
av_frame_free(&s->output_frame);
for (i = 0; i < FF_ARRAY_ELEMS(s->DPB); i++) {
av_buffer_unref(&s->ps.sps_list[i]);
for (i = 0; i < FF_ARRAY_ELEMS(s->ps.pps_list); i++)
av_buffer_unref(&s->ps.pps_list[i]);
+ s->ps.sps = NULL;
+ s->ps.pps = NULL;
+ s->ps.vps = NULL;
+
+ av_freep(&s->sh.entry_point_offset);
+ av_freep(&s->sh.offset);
+ av_freep(&s->sh.size);
+
+ for (i = 1; i < s->threads_number; i++) {
+ HEVCLocalContext *lc = s->HEVClcList[i];
+ if (lc) {
+ av_freep(&s->HEVClcList[i]);
+ av_freep(&s->sList[i]);
+ }
+ }
+ if (s->HEVClc == s->HEVClcList[0])
+ s->HEVClc = NULL;
+ av_freep(&s->HEVClcList[0]);
ff_h2645_packet_uninit(&s->pkt);
s->avctx = avctx;
- s->tmp_frame = av_frame_alloc();
- if (!s->tmp_frame)
+ s->HEVClc = av_mallocz(sizeof(HEVCLocalContext));
+ if (!s->HEVClc)
+ goto fail;
+ s->HEVClcList[0] = s->HEVClc;
+ s->sList[0] = s;
+
+ s->cabac_state = av_malloc(HEVC_CONTEXTS);
+ if (!s->cabac_state)
goto fail;
s->output_frame = av_frame_alloc();
ff_bswapdsp_init(&s->bdsp);
s->context_initialized = 1;
+ s->eos = 0;
+
+ ff_hevc_reset_sei(s);
return 0;
}
}
+ if (s->ps.sps != s0->ps.sps)
+ s->ps.sps = NULL;
for (i = 0; i < FF_ARRAY_ELEMS(s->ps.vps_list); i++) {
av_buffer_unref(&s->ps.vps_list[i]);
if (s0->ps.vps_list[i]) {
}
if (s->ps.sps != s0->ps.sps)
- ret = set_sps(s, s0->ps.sps);
+ if ((ret = set_sps(s, s0->ps.sps, src->pix_fmt)) < 0)
+ return ret;
s->seq_decode = s0->seq_decode;
s->seq_output = s0->seq_output;
s->pocTid0 = s0->pocTid0;
s->max_ra = s0->max_ra;
+ s->eos = s0->eos;
+ s->no_rasl_output_flag = s0->no_rasl_output_flag;
s->is_nalff = s0->is_nalff;
s->nal_length_size = s0->nal_length_size;
+ s->threads_number = s0->threads_number;
+ s->threads_type = s0->threads_type;
+
if (s0->eos) {
s->seq_decode = (s->seq_decode + 1) & 0xff;
s->max_ra = INT_MAX;
return 0;
}
-static int hevc_decode_extradata(HEVCContext *s)
-{
- AVCodecContext *avctx = s->avctx;
- GetByteContext gb;
- int ret, i;
-
- bytestream2_init(&gb, avctx->extradata, avctx->extradata_size);
-
- if (avctx->extradata_size > 3 &&
- (avctx->extradata[0] || avctx->extradata[1] ||
- avctx->extradata[2] > 1)) {
- /* It seems the extradata is encoded as hvcC format.
- * Temporarily, we support configurationVersion==0 until 14496-15 3rd
- * is finalized. When finalized, configurationVersion will be 1 and we
- * can recognize hvcC by checking if avctx->extradata[0]==1 or not. */
- int i, j, num_arrays, nal_len_size;
-
- s->is_nalff = 1;
-
- bytestream2_skip(&gb, 21);
- nal_len_size = (bytestream2_get_byte(&gb) & 3) + 1;
- num_arrays = bytestream2_get_byte(&gb);
-
- /* nal units in the hvcC always have length coded with 2 bytes,
- * so put a fake nal_length_size = 2 while parsing them */
- s->nal_length_size = 2;
-
- /* Decode nal units from hvcC. */
- for (i = 0; i < num_arrays; i++) {
- int type = bytestream2_get_byte(&gb) & 0x3f;
- int cnt = bytestream2_get_be16(&gb);
-
- for (j = 0; j < cnt; j++) {
- // +2 for the nal size field
- int nalsize = bytestream2_peek_be16(&gb) + 2;
- if (bytestream2_get_bytes_left(&gb) < nalsize) {
- av_log(s->avctx, AV_LOG_ERROR,
- "Invalid NAL unit size in extradata.\n");
- return AVERROR_INVALIDDATA;
- }
-
- ret = decode_nal_units(s, gb.buffer, nalsize);
- if (ret < 0) {
- av_log(avctx, AV_LOG_ERROR,
- "Decoding nal unit %d %d from hvcC failed\n",
- type, i);
- return ret;
- }
- bytestream2_skip(&gb, nalsize);
- }
- }
-
- /* Now store right nal length size, that will be used to parse
- * all other nals */
- s->nal_length_size = nal_len_size;
- } else {
- s->is_nalff = 0;
- ret = decode_nal_units(s, avctx->extradata, avctx->extradata_size);
- if (ret < 0)
- return ret;
- }
-
- /* export stream parameters from the first SPS */
- for (i = 0; i < FF_ARRAY_ELEMS(s->ps.sps_list); i++) {
- if (s->ps.sps_list[i]) {
- const HEVCSPS *sps = (const HEVCSPS*)s->ps.sps_list[i]->data;
- export_stream_params(s->avctx, &s->ps, sps);
- break;
- }
- }
-
- return 0;
-}
-
static av_cold int hevc_decode_init(AVCodecContext *avctx)
{
HEVCContext *s = avctx->priv_data;
if (ret < 0)
return ret;
+ s->enable_parallel_tiles = 0;
+ s->picture_struct = 0;
+ s->eos = 1;
+
+ atomic_init(&s->wpp_err, 0);
+
+ if(avctx->active_thread_type & FF_THREAD_SLICE)
+ s->threads_number = avctx->thread_count;
+ else
+ s->threads_number = 1;
+
if (avctx->extradata_size > 0 && avctx->extradata) {
- ret = hevc_decode_extradata(s);
+ ret = hevc_decode_extradata(s, avctx->extradata, avctx->extradata_size);
if (ret < 0) {
hevc_decode_free(avctx);
return ret;
}
}
+ if((avctx->active_thread_type & FF_THREAD_FRAME) && avctx->thread_count > 1)
+ s->threads_type = FF_THREAD_FRAME;
+ else
+ s->threads_type = FF_THREAD_SLICE;
+
return 0;
}
HEVCContext *s = avctx->priv_data;
ff_hevc_flush_dpb(s);
s->max_ra = INT_MAX;
+ s->eos = 1;
}
#define OFFSET(x) offsetof(HEVCContext, x)
static const AVOption options[] = {
{ "apply_defdispwin", "Apply default display window from VUI", OFFSET(apply_defdispwin),
- AV_OPT_TYPE_INT, {.i64 = 0}, 0, 1, PAR },
+ AV_OPT_TYPE_BOOL, {.i64 = 0}, 0, 1, PAR },
+ { "strict-displaywin", "stricly apply default display window size", OFFSET(apply_defdispwin),
+ AV_OPT_TYPE_BOOL, {.i64 = 0}, 0, 1, PAR },
{ NULL },
};
.update_thread_context = hevc_update_thread_context,
.init_thread_copy = hevc_init_thread_copy,
.capabilities = AV_CODEC_CAP_DR1 | AV_CODEC_CAP_DELAY |
- AV_CODEC_CAP_FRAME_THREADS,
+ AV_CODEC_CAP_SLICE_THREADS | AV_CODEC_CAP_FRAME_THREADS,
+ .caps_internal = FF_CODEC_CAP_INIT_THREADSAFE,
.profiles = NULL_IF_CONFIG_SMALL(ff_hevc_profiles),
};
projects / ffmpeg / commitdiff