| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | /* | ||
| 2 | * H.266/VVC MP4 to Annex B byte stream format filter | ||
| 3 | * Copyright (c) 2022, Thomas Siedel | ||
| 4 | * | ||
| 5 | * This file is part of FFmpeg. | ||
| 6 | * | ||
| 7 | * FFmpeg is free software; you can redistribute it and/or | ||
| 8 | * modify it under the terms of the GNU Lesser General Public | ||
| 9 | * License as published by the Free Software Foundation; either | ||
| 10 | * version 2.1 of the License, or (at your option) any later version. | ||
| 11 | * | ||
| 12 | * FFmpeg is distributed in the hope that it will be useful, | ||
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | ||
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | ||
| 15 | * Lesser General Public License for more details. | ||
| 16 | * | ||
| 17 | * You should have received a copy of the GNU Lesser General Public | ||
| 18 | * License along with FFmpeg; if not, write to the Free Software | ||
| 19 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA | ||
| 20 | */ | ||
| 21 | |||
| 22 | #include <string.h> | ||
| 23 | |||
| 24 | #include "libavutil/intreadwrite.h" | ||
| 25 | #include "libavutil/mem.h" | ||
| 26 | |||
| 27 | #include "bsf.h" | ||
| 28 | #include "bsf_internal.h" | ||
| 29 | #include "bytestream.h" | ||
| 30 | #include "defs.h" | ||
| 31 | #include "vvc.h" | ||
| 32 | |||
| 33 | #define MIN_VVCC_LENGTH 23 | ||
| 34 | |||
| 35 | typedef struct VVCBSFContext { | ||
| 36 | uint8_t length_size; | ||
| 37 | int extradata_parsed; | ||
| 38 | } VVCBSFContext; | ||
| 39 | |||
| 40 | ✗ | static int vvc_extradata_to_annexb(AVBSFContext *ctx) | |
| 41 | { | ||
| 42 | GetByteContext gb; | ||
| 43 | int length_size, num_arrays, i, j; | ||
| 44 | ✗ | int ret = 0; | |
| 45 | ✗ | int temp = 0; | |
| 46 | int ptl_present; | ||
| 47 | |||
| 48 | ✗ | uint8_t *new_extradata = NULL; | |
| 49 | ✗ | size_t new_extradata_size = 0; | |
| 50 | |||
| 51 | ✗ | int max_picture_width = 0; | |
| 52 | ✗ | int max_picture_height = 0; | |
| 53 | ✗ | int avg_frame_rate = 0; | |
| 54 | |||
| 55 | ✗ | bytestream2_init(&gb, ctx->par_in->extradata, ctx->par_in->extradata_size); | |
| 56 | ✗ | temp = bytestream2_get_byte(&gb); | |
| 57 | ✗ | length_size = ((temp & 6) >> 1) + 1; | |
| 58 | ✗ | ptl_present = temp & 1; | |
| 59 | ✗ | if (ptl_present) { | |
| 60 | int num_bytes_constraint_info; | ||
| 61 | int general_profile_idc; | ||
| 62 | int general_tier_flag; | ||
| 63 | int general_level_idc; | ||
| 64 | int ptl_frame_only_constraint_flag; | ||
| 65 | int ptl_multi_layer_enabled_flag; | ||
| 66 | int ptl_num_sub_profiles; | ||
| 67 | int temp3, temp4, temp5; | ||
| 68 | ✗ | int temp2 = bytestream2_get_be16(&gb); | |
| 69 | ✗ | int ols_idx = (temp2 >> 7) & 0x1ff; | |
| 70 | ✗ | int num_sublayers = (temp2 >> 4) & 0x7; | |
| 71 | ✗ | int constant_frame_rate = (temp2 >> 2) & 0x3; | |
| 72 | ✗ | int chroma_format_idc = temp2 & 0x3; | |
| 73 | ✗ | int bit_depth_minus8 = (bytestream2_get_byte(&gb) >> 5) & 0x7; | |
| 74 | ✗ | av_log(ctx, AV_LOG_DEBUG, | |
| 75 | "bit_depth_minus8 %d chroma_format_idc %d\n", bit_depth_minus8, | ||
| 76 | chroma_format_idc); | ||
| 77 | ✗ | av_log(ctx, AV_LOG_DEBUG, "constant_frame_rate %d, ols_idx %d\n", | |
| 78 | constant_frame_rate, ols_idx); | ||
| 79 | // VvcPTLRecord(num_sublayers) native_ptl | ||
| 80 | ✗ | temp3 = bytestream2_get_byte(&gb); | |
| 81 | ✗ | num_bytes_constraint_info = (temp3) & 0x3f; | |
| 82 | ✗ | temp4 = bytestream2_get_byte(&gb); | |
| 83 | ✗ | general_profile_idc = (temp4 >> 1) & 0x7f; | |
| 84 | ✗ | general_tier_flag = (temp4) & 1; | |
| 85 | ✗ | general_level_idc = bytestream2_get_byte(&gb); | |
| 86 | ✗ | av_log(ctx, AV_LOG_DEBUG, | |
| 87 | "general_profile_idc %d, general_tier_flag %d, general_level_idc %d, num_sublayers %d num_bytes_constraint_info %d\n", | ||
| 88 | general_profile_idc, general_tier_flag, general_level_idc, | ||
| 89 | num_sublayers, num_bytes_constraint_info); | ||
| 90 | |||
| 91 | ✗ | temp5 = bytestream2_get_byte(&gb); | |
| 92 | ✗ | ptl_frame_only_constraint_flag = (temp5 >> 7) & 0x1; | |
| 93 | ✗ | ptl_multi_layer_enabled_flag = (temp5 >> 6) & 0x1; | |
| 94 | ✗ | for (i = 0; i < num_bytes_constraint_info - 1; i++) { | |
| 95 | // unsigned int(8*num_bytes_constraint_info - 2) general_constraint_info; | ||
| 96 | ✗ | bytestream2_get_byte(&gb); | |
| 97 | } | ||
| 98 | |||
| 99 | ✗ | av_log(ctx, AV_LOG_DEBUG, | |
| 100 | "ptl_multi_layer_enabled_flag %d, ptl_frame_only_constraint_flag %d\n", | ||
| 101 | ptl_multi_layer_enabled_flag, ptl_frame_only_constraint_flag); | ||
| 102 | |||
| 103 | ✗ | if (num_sublayers > 1) { | |
| 104 | ✗ | int temp6 = bytestream2_get_byte(&gb); | |
| 105 | ✗ | uint8_t ptl_sublayer_level_present_flag[8] = { 0 }; | |
| 106 | //uint8_t sublayer_level_idc[8] = {0}; | ||
| 107 | ✗ | for (i = num_sublayers - 2; i >= 0; i--) { | |
| 108 | ✗ | ptl_sublayer_level_present_flag[i] = | |
| 109 | ✗ | (temp6 >> (7 - (num_sublayers - 2 - i))) & 0x01; | |
| 110 | } | ||
| 111 | // for (j=num_sublayers; j<=8 && num_sublayers > 1; j++) | ||
| 112 | // bit(1) ptl_reserved_zero_bit = 0; | ||
| 113 | ✗ | for (i = num_sublayers - 2; i >= 0; i--) { | |
| 114 | if (ptl_sublayer_level_present_flag[i]) { | ||
| 115 | //sublayer_level_idc[i] = bytestream2_get_byte(&gb); | ||
| 116 | } | ||
| 117 | } | ||
| 118 | } | ||
| 119 | |||
| 120 | ✗ | ptl_num_sub_profiles = bytestream2_get_byte(&gb); | |
| 121 | ✗ | for (j = 0; j < ptl_num_sub_profiles; j++) { | |
| 122 | // unsigned int(32) general_sub_profile_idc[j]; | ||
| 123 | ✗ | bytestream2_get_be16(&gb); | |
| 124 | ✗ | bytestream2_get_be16(&gb); | |
| 125 | } | ||
| 126 | |||
| 127 | ✗ | max_picture_width = bytestream2_get_be16(&gb); // unsigned_int(16) max_picture_width; | |
| 128 | ✗ | max_picture_height = bytestream2_get_be16(&gb); // unsigned_int(16) max_picture_height; | |
| 129 | ✗ | avg_frame_rate = bytestream2_get_be16(&gb); // unsigned int(16) avg_frame_rate; } | |
| 130 | ✗ | av_log(ctx, AV_LOG_DEBUG, | |
| 131 | "max_picture_width %d, max_picture_height %d, avg_frame_rate %d\n", | ||
| 132 | max_picture_width, max_picture_height, avg_frame_rate); | ||
| 133 | } | ||
| 134 | |||
| 135 | ✗ | num_arrays = bytestream2_get_byte(&gb); | |
| 136 | |||
| 137 | ✗ | for (i = 0; i < num_arrays; i++) { | |
| 138 | int cnt; | ||
| 139 | ✗ | int type = bytestream2_get_byte(&gb) & 0x1f; | |
| 140 | |||
| 141 | ✗ | if (type == VVC_OPI_NUT || type == VVC_DCI_NUT) | |
| 142 | ✗ | cnt = 1; | |
| 143 | else | ||
| 144 | ✗ | cnt = bytestream2_get_be16(&gb); | |
| 145 | |||
| 146 | ✗ | av_log(ctx, AV_LOG_DEBUG, "nalu_type %d cnt %d\n", type, cnt); | |
| 147 | |||
| 148 | ✗ | if (!(type == VVC_OPI_NUT || type == VVC_DCI_NUT || | |
| 149 | ✗ | type == VVC_VPS_NUT || type == VVC_SPS_NUT || type == VVC_PPS_NUT | |
| 150 | ✗ | || type == VVC_PREFIX_SEI_NUT || type == VVC_SUFFIX_SEI_NUT)) { | |
| 151 | ✗ | av_log(ctx, AV_LOG_ERROR, | |
| 152 | "Invalid NAL unit type in extradata: %d\n", type); | ||
| 153 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 154 | ✗ | goto fail; | |
| 155 | } | ||
| 156 | |||
| 157 | ✗ | for (j = 0; j < cnt; j++) { | |
| 158 | ✗ | const int nalu_len = bytestream2_get_be16(&gb); | |
| 159 | |||
| 160 | ✗ | if (!nalu_len || | |
| 161 | ✗ | nalu_len > bytestream2_get_bytes_left(&gb) || | |
| 162 | ✗ | 4 + AV_INPUT_BUFFER_PADDING_SIZE + nalu_len > SIZE_MAX - new_extradata_size) { | |
| 163 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 164 | ✗ | goto fail; | |
| 165 | } | ||
| 166 | ✗ | ret = av_reallocp(&new_extradata, new_extradata_size + nalu_len + 4 | |
| 167 | + AV_INPUT_BUFFER_PADDING_SIZE); | ||
| 168 | ✗ | if (ret < 0) | |
| 169 | ✗ | goto fail; | |
| 170 | |||
| 171 | ✗ | AV_WB32(new_extradata + new_extradata_size, 1); // add the startcode | |
| 172 | ✗ | bytestream2_get_buffer(&gb, new_extradata + new_extradata_size + 4, | |
| 173 | nalu_len); | ||
| 174 | ✗ | new_extradata_size += 4 + nalu_len; | |
| 175 | ✗ | memset(new_extradata + new_extradata_size, 0, | |
| 176 | AV_INPUT_BUFFER_PADDING_SIZE); | ||
| 177 | } | ||
| 178 | } | ||
| 179 | |||
| 180 | ✗ | av_freep(&ctx->par_out->extradata); | |
| 181 | ✗ | ctx->par_out->extradata = new_extradata; | |
| 182 | ✗ | ctx->par_out->extradata_size = new_extradata_size; | |
| 183 | |||
| 184 | ✗ | if (!new_extradata_size) | |
| 185 | ✗ | av_log(ctx, AV_LOG_WARNING, "No parameter sets in the extradata\n"); | |
| 186 | |||
| 187 | ✗ | return length_size; | |
| 188 | ✗ | fail: | |
| 189 | ✗ | av_freep(&new_extradata); | |
| 190 | ✗ | return ret; | |
| 191 | } | ||
| 192 | |||
| 193 | ✗ | static int vvc_mp4toannexb_init(AVBSFContext *ctx) | |
| 194 | { | ||
| 195 | ✗ | VVCBSFContext *s = ctx->priv_data; | |
| 196 | int ret; | ||
| 197 | |||
| 198 | ✗ | if (ctx->par_in->extradata_size < MIN_VVCC_LENGTH || | |
| 199 | ✗ | AV_RB24(ctx->par_in->extradata) == 1 || | |
| 200 | ✗ | AV_RB32(ctx->par_in->extradata) == 1) { | |
| 201 | ✗ | av_log(ctx, AV_LOG_VERBOSE, | |
| 202 | "The input looks like it is Annex B already\n"); | ||
| 203 | } else { | ||
| 204 | ✗ | ret = vvc_extradata_to_annexb(ctx); | |
| 205 | ✗ | if (ret < 0) | |
| 206 | ✗ | return ret; | |
| 207 | ✗ | s->length_size = ret; | |
| 208 | ✗ | s->extradata_parsed = 1; | |
| 209 | } | ||
| 210 | |||
| 211 | ✗ | return 0; | |
| 212 | } | ||
| 213 | |||
| 214 | ✗ | static int vvc_mp4toannexb_filter(AVBSFContext *ctx, AVPacket *out) | |
| 215 | { | ||
| 216 | ✗ | VVCBSFContext *s = ctx->priv_data; | |
| 217 | AVPacket *in; | ||
| 218 | GetByteContext gb; | ||
| 219 | |||
| 220 | ✗ | int is_irap = 0; | |
| 221 | ✗ | int added_extra = 0; | |
| 222 | ✗ | int i, ret = 0; | |
| 223 | |||
| 224 | ✗ | ret = ff_bsf_get_packet(ctx, &in); | |
| 225 | ✗ | if (ret < 0) | |
| 226 | ✗ | return ret; | |
| 227 | |||
| 228 | ✗ | if (!s->extradata_parsed) { | |
| 229 | ✗ | av_packet_move_ref(out, in); | |
| 230 | ✗ | av_packet_free(&in); | |
| 231 | ✗ | return 0; | |
| 232 | } | ||
| 233 | |||
| 234 | ✗ | bytestream2_init(&gb, in->data, in->size); | |
| 235 | |||
| 236 | /* check if this packet contains an IRAP. The extradata will need to be added before any potential PH_NUT */ | ||
| 237 | ✗ | while (bytestream2_get_bytes_left(&gb)) { | |
| 238 | ✗ | uint32_t nalu_size = 0; | |
| 239 | int nalu_type; | ||
| 240 | |||
| 241 | ✗ | if (bytestream2_get_bytes_left(&gb) < s->length_size) { | |
| 242 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 243 | ✗ | goto fail; | |
| 244 | } | ||
| 245 | |||
| 246 | ✗ | for (i = 0; i < s->length_size; i++) | |
| 247 | ✗ | nalu_size = (nalu_size << 8) | bytestream2_get_byte(&gb); | |
| 248 | |||
| 249 | ✗ | if (nalu_size < 2 || nalu_size > bytestream2_get_bytes_left(&gb)) { | |
| 250 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 251 | ✗ | goto fail; | |
| 252 | } | ||
| 253 | |||
| 254 | ✗ | nalu_type = (bytestream2_peek_be16(&gb) >> 3) & 0x1f; | |
| 255 | ✗ | is_irap = nalu_type >= VVC_IDR_W_RADL && nalu_type <= VVC_RSV_IRAP_11; | |
| 256 | ✗ | if (is_irap) { | |
| 257 | ✗ | break; | |
| 258 | } | ||
| 259 | ✗ | bytestream2_seek(&gb, nalu_size, SEEK_CUR); | |
| 260 | } | ||
| 261 | |||
| 262 | ✗ | bytestream2_seek(&gb, 0, SEEK_SET); | |
| 263 | ✗ | while (bytestream2_get_bytes_left(&gb)) { | |
| 264 | ✗ | uint32_t nalu_size = 0; | |
| 265 | int nalu_type; | ||
| 266 | int add_extradata, extra_size, prev_size; | ||
| 267 | |||
| 268 | ✗ | if (bytestream2_get_bytes_left(&gb) < s->length_size) { | |
| 269 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 270 | ✗ | goto fail; | |
| 271 | } | ||
| 272 | |||
| 273 | ✗ | for (i = 0; i < s->length_size; i++) | |
| 274 | ✗ | nalu_size = (nalu_size << 8) | bytestream2_get_byte(&gb); | |
| 275 | |||
| 276 | ✗ | if (nalu_size < 2 || nalu_size > bytestream2_get_bytes_left(&gb)) { | |
| 277 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 278 | ✗ | goto fail; | |
| 279 | } | ||
| 280 | |||
| 281 | ✗ | nalu_type = (bytestream2_peek_be16(&gb) >> 3) & 0x1f; | |
| 282 | |||
| 283 | /* prepend extradata to IRAP frames */ | ||
| 284 | ✗ | add_extradata = is_irap && nalu_type != VVC_AUD_NUT && !added_extra; | |
| 285 | ✗ | extra_size = add_extradata * ctx->par_out->extradata_size; | |
| 286 | ✗ | added_extra |= add_extradata; | |
| 287 | |||
| 288 | ✗ | if (FFMIN(INT_MAX, SIZE_MAX) < 4ULL + nalu_size + extra_size) { | |
| 289 | ✗ | ret = AVERROR_INVALIDDATA; | |
| 290 | ✗ | goto fail; | |
| 291 | } | ||
| 292 | |||
| 293 | ✗ | prev_size = out->size; | |
| 294 | |||
| 295 | ✗ | ret = av_grow_packet(out, 4 + nalu_size + extra_size); | |
| 296 | ✗ | if (ret < 0) | |
| 297 | ✗ | goto fail; | |
| 298 | |||
| 299 | ✗ | if (extra_size) | |
| 300 | ✗ | memcpy(out->data + prev_size, ctx->par_out->extradata, extra_size); | |
| 301 | ✗ | AV_WB32(out->data + prev_size + extra_size, 1); | |
| 302 | ✗ | bytestream2_get_buffer(&gb, out->data + prev_size + 4 + extra_size, | |
| 303 | nalu_size); | ||
| 304 | } | ||
| 305 | |||
| 306 | ✗ | ret = av_packet_copy_props(out, in); | |
| 307 | ✗ | if (ret < 0) | |
| 308 | ✗ | goto fail; | |
| 309 | |||
| 310 | ✗ | fail: | |
| 311 | ✗ | if (ret < 0) | |
| 312 | ✗ | av_packet_unref(out); | |
| 313 | ✗ | av_packet_free(&in); | |
| 314 | |||
| 315 | ✗ | return ret; | |
| 316 | } | ||
| 317 | |||
| 318 | static const enum AVCodecID codec_ids[] = { | ||
| 319 | AV_CODEC_ID_VVC, AV_CODEC_ID_NONE, | ||
| 320 | }; | ||
| 321 | |||
| 322 | const FFBitStreamFilter ff_vvc_mp4toannexb_bsf = { | ||
| 323 | .p.name = "vvc_mp4toannexb", | ||
| 324 | .p.codec_ids = codec_ids, | ||
| 325 | .priv_data_size = sizeof(VVCBSFContext), | ||
| 326 | .init = vvc_mp4toannexb_init, | ||
| 327 | .filter = vvc_mp4toannexb_filter, | ||
| 328 | }; | ||
| 329 |