ijkplayer框架簡析 -- av_read_frame
av_read_frame()
的作用是讀取碼流中的音訊若干幀或者視訊一幀
av_read_frame
int av_read_frame(AVFormatContext *s, AVPacket *pkt) { const int genpts = s->flags & AVFMT_FLAG_GENPTS; int eof = 0; int ret; AVStream *st; if (!genpts) { ret = s->internal->packet_buffer ? read_from_packet_buffer(&s->internal->packet_buffer, // 從快取中讀取 &s->internal->packet_buffer_end, pkt) : read_frame_internal(s, pkt); // 從檔案,或探測快取中讀取 if (ret < 0) return ret; goto return_packet; } for (;;) { AVPacketList *pktl = s->internal->packet_buffer; if (pktl) { AVPacket *next_pkt = &pktl->pkt; if (next_pkt->dts != AV_NOPTS_VALUE) { int wrap_bits = s->streams[next_pkt->stream_index]->pts_wrap_bits; // last dts seen for this stream. if any of packets following // current one had no dts, we will set this to AV_NOPTS_VALUE. int64_t last_dts = next_pkt->dts; while (pktl && next_pkt->pts == AV_NOPTS_VALUE) { if (pktl->pkt.stream_index == next_pkt->stream_index && (av_compare_mod(next_pkt->dts, pktl->pkt.dts, 2LL << (wrap_bits - 1)) < 0)) { if (av_compare_mod(pktl->pkt.pts, pktl->pkt.dts, 2LL << (wrap_bits - 1))) { // not B-frame next_pkt->pts = pktl->pkt.dts; } if (last_dts != AV_NOPTS_VALUE) { // Once last dts was set to AV_NOPTS_VALUE, we don't change it. last_dts = pktl->pkt.dts; } } pktl = pktl->next; } if (eof && next_pkt->pts == AV_NOPTS_VALUE && last_dts != AV_NOPTS_VALUE) { // Fixing the last reference frame had none pts issue (For MXF etc). // We only do this when // 1. eof. // 2. we are not able to resolve a pts value for current packet. // 3. the packets for this stream at the end of the files had valid dts. next_pkt->pts = last_dts + next_pkt->duration; } pktl = s->internal->packet_buffer; } /* read packet from packet buffer, if there is data */ st = s->streams[next_pkt->stream_index]; if (!(next_pkt->pts == AV_NOPTS_VALUE && st->discard < AVDISCARD_ALL && next_pkt->dts != AV_NOPTS_VALUE && !eof)) { ret = read_from_packet_buffer(&s->internal->packet_buffer, &s->internal->packet_buffer_end, pkt); goto return_packet; } } ret = read_frame_internal(s, pkt); if (ret < 0) { if (pktl && ret != AVERROR(EAGAIN)) { eof = 1; continue; } else return ret; } ret = add_to_pktbuf(&s->internal->packet_buffer, pkt, &s->internal->packet_buffer_end, 1); av_packet_unref(pkt); if (ret < 0) return ret; } return_packet: st = s->streams[pkt->stream_index]; if ((s->iformat->flags & AVFMT_GENERIC_INDEX) && pkt->flags & AV_PKT_FLAG_KEY) { ff_reduce_index(s, st->index); av_add_index_entry(st, pkt->pos, pkt->dts, 0, 0, AVINDEX_KEYFRAME); } if (is_relative(pkt->dts)) pkt->dts -= RELATIVE_TS_BASE; if (is_relative(pkt->pts)) pkt->pts -= RELATIVE_TS_BASE; return ret; }
av_read_frame()
會先判斷未解碼快取中是否又資料,如果有的話則呼叫read_from_packet_buffer()
,否則就呼叫read_frame_internal()
read_frame_internal
static int read_frame_internal(AVFormatContext *s, AVPacket *pkt) { int ret = 0, i, got_packet = 0; AVDictionary *metadata = NULL; av_init_packet(pkt); while (!got_packet && !s->internal->parse_queue) { AVStream *st; AVPacket cur_pkt; /* read next packet */ ret = ff_read_packet(s, &cur_pkt); if (ret < 0) { if (ret == AVERROR(EAGAIN)) return ret; /* flush the parsers */ for (i = 0; i < s->nb_streams; i++) { st = s->streams[i]; if (st->parser && st->need_parsing) parse_packet(s, NULL, st->index); } /* all remaining packets are now in parse_queue => * really terminate parsing */ break; } ret = 0; st= s->streams[cur_pkt.stream_index]; /* update context if required */ if (st->internal->need_context_update) { if (avcodec_is_open(st->internal->avctx)) { av_log(s, AV_LOG_DEBUG, "Demuxer context update while decoder is open, closing and trying to re-open\n"); avcodec_close(st->internal->avctx); st->info->found_decoder = 0; } /* close parser, because it depends on the codec */ if (st->parser && st->internal->avctx->codec_id != st->codecpar->codec_id) { av_parser_close(st->parser); st->parser = NULL; } ret = avcodec_parameters_to_context(st->internal->avctx, st->codecpar); if (ret < 0) return ret; #if FF_API_LAVF_AVCTX FF_DISABLE_DEPRECATION_WARNINGS /* update deprecated public codec context */ ret = avcodec_parameters_to_context(st->codec, st->codecpar); if (ret < 0) return ret; FF_ENABLE_DEPRECATION_WARNINGS #endif st->internal->need_context_update = 0; } if (cur_pkt.pts != AV_NOPTS_VALUE && cur_pkt.dts != AV_NOPTS_VALUE && cur_pkt.pts < cur_pkt.dts) { av_log(s, AV_LOG_WARNING, "Invalid timestamps stream=%d, pts=%s, dts=%s, size=%d\n", cur_pkt.stream_index, av_ts2str(cur_pkt.pts), av_ts2str(cur_pkt.dts), cur_pkt.size); } if (s->debug & FF_FDEBUG_TS) av_log(s, AV_LOG_DEBUG, "ff_read_packet stream=%d, pts=%s, dts=%s, size=%d, duration=%"PRId64", flags=%d\n", cur_pkt.stream_index, av_ts2str(cur_pkt.pts), av_ts2str(cur_pkt.dts), cur_pkt.size, cur_pkt.duration, cur_pkt.flags); if (st->need_parsing && !st->parser && !(s->flags & AVFMT_FLAG_NOPARSE)) { st->parser = av_parser_init(st->codecpar->codec_id); if (!st->parser) { av_log(s, AV_LOG_VERBOSE, "parser not found for codec " "%s, packets or times may be invalid.\n", avcodec_get_name(st->codecpar->codec_id)); /* no parser available: just output the raw packets */ st->need_parsing = AVSTREAM_PARSE_NONE; } else if (st->need_parsing == AVSTREAM_PARSE_HEADERS) st->parser->flags |= PARSER_FLAG_COMPLETE_FRAMES; else if (st->need_parsing == AVSTREAM_PARSE_FULL_ONCE) st->parser->flags |= PARSER_FLAG_ONCE; else if (st->need_parsing == AVSTREAM_PARSE_FULL_RAW) st->parser->flags |= PARSER_FLAG_USE_CODEC_TS; } if (!st->need_parsing || !st->parser) { /* no parsing needed: we just output the packet as is */ *pkt = cur_pkt; compute_pkt_fields(s, st, NULL, pkt, AV_NOPTS_VALUE, AV_NOPTS_VALUE); if ((s->iformat->flags & AVFMT_GENERIC_INDEX) && (pkt->flags & AV_PKT_FLAG_KEY) && pkt->dts != AV_NOPTS_VALUE) { ff_reduce_index(s, st->index); av_add_index_entry(st, pkt->pos, pkt->dts, 0, 0, AVINDEX_KEYFRAME); } got_packet = 1; } else if (st->discard < AVDISCARD_ALL) { if ((ret = parse_packet(s, &cur_pkt, cur_pkt.stream_index)) < 0) return ret; st->codecpar->sample_rate = st->internal->avctx->sample_rate; st->codecpar->bit_rate = st->internal->avctx->bit_rate; st->codecpar->channels = st->internal->avctx->channels; st->codecpar->channel_layout = st->internal->avctx->channel_layout; st->codecpar->codec_id = st->internal->avctx->codec_id; } else { /* free packet */ av_packet_unref(&cur_pkt); } if (pkt->flags & AV_PKT_FLAG_KEY) st->skip_to_keyframe = 0; if (st->skip_to_keyframe) { av_packet_unref(&cur_pkt); if (got_packet) { *pkt = cur_pkt; } got_packet = 0; } } if (!got_packet && s->internal->parse_queue) ret = read_from_packet_buffer(&s->internal->parse_queue, &s->internal->parse_queue_end, pkt); if (ret >= 0) { AVStream *st = s->streams[pkt->stream_index]; int discard_padding = 0; if (st->first_discard_sample && pkt->pts != AV_NOPTS_VALUE) { int64_t pts = pkt->pts - (is_relative(pkt->pts) ? RELATIVE_TS_BASE : 0); int64_t sample = ts_to_samples(st, pts); int duration = ts_to_samples(st, pkt->duration); int64_t end_sample = sample + duration; if (duration > 0 && end_sample >= st->first_discard_sample && sample < st->last_discard_sample) discard_padding = FFMIN(end_sample - st->first_discard_sample, duration); } if (st->start_skip_samples && (pkt->pts == 0 || pkt->pts == RELATIVE_TS_BASE)) st->skip_samples = st->start_skip_samples; if (st->skip_samples || discard_padding) { uint8_t *p = av_packet_new_side_data(pkt, AV_PKT_DATA_SKIP_SAMPLES, 10); if (p) { AV_WL32(p, st->skip_samples); AV_WL32(p + 4, discard_padding); av_log(s, AV_LOG_DEBUG, "demuxer injecting skip %d / discard %d\n", st->skip_samples, discard_padding); } st->skip_samples = 0; } if (st->inject_global_side_data) { for (i = 0; i < st->nb_side_data; i++) { AVPacketSideData *src_sd = &st->side_data[i]; uint8_t *dst_data; if (av_packet_get_side_data(pkt, src_sd->type, NULL)) continue; dst_data = av_packet_new_side_data(pkt, src_sd->type, src_sd->size); if (!dst_data) { av_log(s, AV_LOG_WARNING, "Could not inject global side data\n"); continue; } memcpy(dst_data, src_sd->data, src_sd->size); } st->inject_global_side_data = 0; } #if FF_API_LAVF_MERGE_SD FF_DISABLE_DEPRECATION_WARNINGS if (!(s->flags & AVFMT_FLAG_KEEP_SIDE_DATA)) av_packet_merge_side_data(pkt); FF_ENABLE_DEPRECATION_WARNINGS #endif } av_opt_get_dict_val(s, "metadata", AV_OPT_SEARCH_CHILDREN, &metadata); if (metadata) { s->event_flags |= AVFMT_EVENT_FLAG_METADATA_UPDATED; av_dict_copy(&s->metadata, metadata, 0); av_dict_free(&metadata); av_opt_set_dict_val(s, "metadata", NULL, AV_OPT_SEARCH_CHILDREN); } #if FF_API_LAVF_AVCTX update_stream_avctx(s); #endif if (s->debug & FF_FDEBUG_TS) av_log(s, AV_LOG_DEBUG, "read_frame_internal stream=%d, pts=%s, dts=%s, " "size=%d, duration=%"PRId64", flags=%d\n", pkt->stream_index, av_ts2str(pkt->pts), av_ts2str(pkt->dts), pkt->size, pkt->duration, pkt->flags); return ret; }
ff_read_packet() parse_packet()
ff_read_packet
int ff_read_packet(AVFormatContext *s, AVPacket *pkt) { int ret, i, err; AVStream *st; for (;;) { AVPacketList *pktl = s->internal->raw_packet_buffer; // 如果快取中有資料,則到快取中取 if (pktl) { *pkt = pktl->pkt; st= s->streams[pkt->stream_index]; // raw_packet_buffer 剩下的有效buf大小小於0 if (s->internal->raw_packet_buffer_remaining_size <= 0) if ((err = probe_codec(s, st, NULL)) < 0) return err; // 如果 探測結束,則更新資訊,並返回 if (st->request_probe <= 0) { s->internal->raw_packet_buffer= pktl->next; s->internal->raw_packet_buffer_remaining_size += pkt->size; av_free(pktl); return 0; } } pkt->data = NULL; pkt->size = 0; // 用 AVInputFormat 的 read_packet 函式指標從檔案讀取資料 av_init_packet(pkt); ret = s->iformat->read_packet(s, pkt); if (ret < 0) { /* Some demuxers return FFERROR_REDO when they consume data and discard it (ignored streams, junk, extradata). We must re-call the demuxer to get the real packet. */ if (ret == FFERROR_REDO) continue; if (!pktl || ret == AVERROR(EAGAIN)) return ret; for (i = 0; i < s->nb_streams; i++) { st = s->streams[i]; if (st->probe_packets || st->request_probe > 0) if ((err = probe_codec(s, st, NULL)) < 0) return err; av_assert0(st->request_probe <= 0); } continue; } if (!pkt->buf) { AVPacket tmp = { 0 }; ret = av_packet_ref(&tmp, pkt); if (ret < 0) return ret; *pkt = tmp; } // 如果 AVFormatContext 的 flags 設定了丟棄損壞包,並且此包是損壞的,則丟棄 if ((s->flags & AVFMT_FLAG_DISCARD_CORRUPT) && (pkt->flags & AV_PKT_FLAG_CORRUPT)) { av_log(s, AV_LOG_WARNING, "Dropped corrupted packet (stream = %d)\n", pkt->stream_index); av_packet_unref(pkt); continue; } // 讀取到的 pkt 的 stream_index 不合法,拋棄 if (pkt->stream_index >= (unsigned)s->nb_streams) { av_log(s, AV_LOG_ERROR, "Invalid stream index %d\n", pkt->stream_index); continue; } st = s->streams[pkt->stream_index]; if (update_wrap_reference(s, st, pkt->stream_index, pkt) && st->pts_wrap_behavior == AV_PTS_WRAP_SUB_OFFSET) { // correct first time stamps to negative values if (!is_relative(st->first_dts)) st->first_dts = wrap_timestamp(st, st->first_dts); if (!is_relative(st->start_time)) st->start_time = wrap_timestamp(st, st->start_time); if (!is_relative(st->cur_dts)) st->cur_dts = wrap_timestamp(st, st->cur_dts); } pkt->dts = wrap_timestamp(st, pkt->dts); pkt->pts = wrap_timestamp(st, pkt->pts); force_codec_ids(s, st); /* TODO: audio: time filter; video: frame reordering (pts != dts) */ if (s->use_wallclock_as_timestamps) pkt->dts = pkt->pts = av_rescale_q(av_gettime(), AV_TIME_BASE_Q, st->time_base); // 如果不需要探測,則返回,如果需要探測,且 raw_packet_buffer 為空,則把剛讀到的包加進去 if (!pktl && st->request_probe <= 0) return ret; // 把剛讀的包,加入raw_packet_buffe err = add_to_pktbuf(&s->internal->raw_packet_buffer, pkt, &s->internal->raw_packet_buffer_end, 0); if (err) return err; s->internal->raw_packet_buffer_remaining_size -= pkt->size; if ((err = probe_codec(s, st, pkt)) < 0) return err; } }
在 for 迴圈中呼叫了iformat->read_packet(s, pkt)
,是 AVInputFormat 的指標函式,以 mp4 為例對應的是
AVInputFormat ff_mov_demuxer = { .name= "mov,mp4,m4a,3gp,3g2,mj2", .long_name= NULL_IF_CONFIG_SMALL("QuickTime / MOV"), .priv_class= &mov_class, .priv_data_size = sizeof(MOVContext), .extensions= "mov,mp4,m4a,3gp,3g2,mj2", .read_probe= mov_probe, .read_header= mov_read_header, .read_packet= mov_read_packet, .read_close= mov_read_close, .read_seek= mov_read_seek, .flags= AVFMT_NO_BYTE_SEEK, }; static int mov_read_packet(AVFormatContext *s, AVPacket *pkt) { MOVContext *mov = s->priv_data; MOVStreamContext *sc; AVIndexEntry *sample; AVStream *st = NULL; int64_t current_index; int ret; mov->fc = s; retry: // 查詢下一個sample sample = mov_find_next_sample(s, &st); if (!sample || (mov->next_root_atom && sample->pos > mov->next_root_atom)) { if (!mov->next_root_atom) return AVERROR_EOF; if ((ret = mov_switch_root(s, mov->next_root_atom)) < 0) return ret; goto retry; } sc = st->priv_data; /* must be done just before reading, to avoid infinite loop on sample */ current_index = sc->current_index; mov_current_sample_inc(sc); if (mov->next_root_atom) { sample->pos = FFMIN(sample->pos, mov->next_root_atom); sample->size = FFMIN(sample->size, (mov->next_root_atom - sample->pos)); } if (st->discard != AVDISCARD_ALL) { // 通過 sample 中的 pos 找到媒體檔案中的 sample 起始點與大小,然後根據這個去讀取媒體檔案的內容 int64_t ret64 = avio_seek(sc->pb, sample->pos, SEEK_SET); if (ret64 != sample->pos) { av_log(mov->fc, AV_LOG_ERROR, "stream %d, offset 0x%"PRIx64": partial file\n", sc->ffindex, sample->pos); if (should_retry(sc->pb, ret64)) { mov_current_sample_dec(sc); } return AVERROR_INVALIDDATA; } if( st->discard == AVDISCARD_NONKEY && 0==(sample->flags & AVINDEX_KEYFRAME) ) { av_log(mov->fc, AV_LOG_DEBUG, "Nonkey frame from stream %d discarded due to AVDISCARD_NONKEY\n", sc->ffindex); goto retry; } // 讀取媒體檔案的內容,大小是 sample->size ret = av_get_packet(sc->pb, pkt, sample->size); if (ret < 0) { if (should_retry(sc->pb, ret)) { mov_current_sample_dec(sc); } return ret; } if (sc->has_palette) { uint8_t *pal; pal = av_packet_new_side_data(pkt, AV_PKT_DATA_PALETTE, AVPALETTE_SIZE); if (!pal) { av_log(mov->fc, AV_LOG_ERROR, "Cannot append palette to packet\n"); } else { memcpy(pal, sc->palette, AVPALETTE_SIZE); sc->has_palette = 0; } } #if CONFIG_DV_DEMUXER if (mov->dv_demux && sc->dv_audio_container) { avpriv_dv_produce_packet(mov->dv_demux, pkt, pkt->data, pkt->size, pkt->pos); av_freep(&pkt->data); pkt->size = 0; ret = avpriv_dv_get_packet(mov->dv_demux, pkt); if (ret < 0) return ret; } #endif if (st->codecpar->codec_id == AV_CODEC_ID_MP3 && !st->need_parsing && pkt->size > 4) { if (ff_mpa_check_header(AV_RB32(pkt->data)) < 0) st->need_parsing = AVSTREAM_PARSE_FULL; } } pkt->stream_index = sc->ffindex; // dts 是計算出來的 start_time + sampleduration pkt->dts = sample->timestamp; if (sample->flags & AVINDEX_DISCARD_FRAME) { pkt->flags |= AV_PKT_FLAG_DISCARD; } if (sc->ctts_data && sc->ctts_index < sc->ctts_count) { pkt->pts = pkt->dts + sc->dts_shift + sc->ctts_data[sc->ctts_index].duration; /* update ctts context */ sc->ctts_sample++; if (sc->ctts_index < sc->ctts_count && sc->ctts_data[sc->ctts_index].count == sc->ctts_sample) { sc->ctts_index++; sc->ctts_sample = 0; } } else { int64_t next_dts = (sc->current_sample < st->nb_index_entries) ? st->index_entries[sc->current_sample].timestamp : st->duration; pkt->duration = next_dts - pkt->dts; pkt->pts = pkt->dts; } if (st->discard == AVDISCARD_ALL) goto retry; pkt->flags |= sample->flags & AVINDEX_KEYFRAME ? AV_PKT_FLAG_KEY : 0; pkt->pos = sample->pos; /* Multiple stsd handling. */ if (sc->stsc_data) { /* Keep track of the stsc index for the given sample, then check * if the stsd index is different from the last used one. */ sc->stsc_sample++; if (mov_stsc_index_valid(sc->stsc_index, sc->stsc_count) && mov_get_stsc_samples(sc, sc->stsc_index) == sc->stsc_sample) { sc->stsc_index++; sc->stsc_sample = 0; /* Do not check indexes after a switch. */ } else if (sc->stsc_data[sc->stsc_index].id > 0 && sc->stsc_data[sc->stsc_index].id - 1 < sc->stsd_count && sc->stsc_data[sc->stsc_index].id - 1 != sc->last_stsd_index) { ret = mov_change_extradata(sc, pkt); if (ret < 0) return ret; } } if (mov->aax_mode) aax_filter(pkt->data, pkt->size, mov); if (sc->cenc.aes_ctr) { ret = cenc_filter(mov, sc, current_index, pkt->data, pkt->size); if (ret) { return ret; } } return 0; }
mov_find_next_sample() avio_seek(sc->pb, sample->pos, SEEK_SET) av_get_packet(sc->pb, pkt, sample->size)
parse_packet
static int parse_packet(AVFormatContext *s, AVPacket *pkt, int stream_index) { AVPacket out_pkt = { 0 }, flush_pkt = { 0 }; AVStream *st = s->streams[stream_index]; uint8_t *data = pkt ? pkt->data : NULL; int size= pkt ? pkt->size : 0; int ret = 0, got_output = 0; if (!pkt) { av_init_packet(&flush_pkt); pkt= &flush_pkt; got_output = 1; } else if (!size && st->parser->flags & PARSER_FLAG_COMPLETE_FRAMES) { // preserve 0-size sync packets compute_pkt_fields(s, st, st->parser, pkt, AV_NOPTS_VALUE, AV_NOPTS_VALUE); } while (size > 0 || (pkt == &flush_pkt && got_output)) { int len; int64_t next_pts = pkt->pts; int64_t next_dts = pkt->dts; av_init_packet(&out_pkt); len = av_parser_parse2(st->parser, st->internal->avctx, &out_pkt.data, &out_pkt.size, data, size, pkt->pts, pkt->dts, pkt->pos); pkt->pts = pkt->dts = AV_NOPTS_VALUE; pkt->pos = -1; /* increment read pointer */ data += len; size -= len; got_output = !!out_pkt.size; if (!out_pkt.size) continue; if (pkt->side_data) { out_pkt.side_data= pkt->side_data; out_pkt.side_data_elems = pkt->side_data_elems; pkt->side_data= NULL; pkt->side_data_elems= 0; } /* set the duration */ out_pkt.duration = (st->parser->flags & PARSER_FLAG_COMPLETE_FRAMES) ? pkt->duration : 0; if (st->codecpar->codec_type == AVMEDIA_TYPE_AUDIO) { if (st->internal->avctx->sample_rate > 0) { out_pkt.duration = av_rescale_q_rnd(st->parser->duration, (AVRational) { 1, st->internal->avctx->sample_rate }, st->time_base, AV_ROUND_DOWN); } } out_pkt.stream_index = st->index; out_pkt.pts= st->parser->pts; out_pkt.dts= st->parser->dts; out_pkt.pos= st->parser->pos; if (st->need_parsing == AVSTREAM_PARSE_FULL_RAW) out_pkt.pos = st->parser->frame_offset; if (st->parser->key_frame == 1 || (st->parser->key_frame == -1 && st->parser->pict_type == AV_PICTURE_TYPE_I)) out_pkt.flags |= AV_PKT_FLAG_KEY; if (st->parser->key_frame == -1 && st->parser->pict_type ==AV_PICTURE_TYPE_NONE && (pkt->flags&AV_PKT_FLAG_KEY)) out_pkt.flags |= AV_PKT_FLAG_KEY; compute_pkt_fields(s, st, st->parser, &out_pkt, next_dts, next_pts); ret = add_to_pktbuf(&s->internal->parse_queue, &out_pkt, &s->internal->parse_queue_end, 1); av_packet_unref(&out_pkt); if (ret < 0) goto fail; } /* end of the stream => close and free the parser */ if (pkt == &flush_pkt) { av_parser_close(st->parser); st->parser = NULL; } fail: av_packet_unref(pkt); return ret; }
parse_packet()
的作用是給需要 AVCodecParser 的媒體流提供解析 AVPacket 的功能,主要是通過av_parser_parse2()
來解析 AVPacket