| /* |
| * Copyright (c) 2012 The WebRTC project authors. All Rights Reserved. |
| * |
| * Use of this source code is governed by a BSD-style license |
| * that can be found in the LICENSE file in the root of the source |
| * tree. An additional intellectual property rights grant can be found |
| * in the file PATENTS. All contributing project authors may |
| * be found in the AUTHORS file in the root of the source tree. |
| */ |
| |
| #include "video/rtp_video_stream_receiver.h" |
| |
| #include <algorithm> |
| #include <limits> |
| #include <utility> |
| #include <vector> |
| |
| #include "absl/algorithm/container.h" |
| #include "absl/base/macros.h" |
| #include "absl/memory/memory.h" |
| #include "absl/types/optional.h" |
| #include "media/base/media_constants.h" |
| #include "modules/pacing/packet_router.h" |
| #include "modules/remote_bitrate_estimator/include/remote_bitrate_estimator.h" |
| #include "modules/rtp_rtcp/include/receive_statistics.h" |
| #include "modules/rtp_rtcp/include/rtp_cvo.h" |
| #include "modules/rtp_rtcp/include/rtp_rtcp.h" |
| #include "modules/rtp_rtcp/include/ulpfec_receiver.h" |
| #include "modules/rtp_rtcp/source/create_video_rtp_depacketizer.h" |
| #include "modules/rtp_rtcp/source/rtp_format.h" |
| #include "modules/rtp_rtcp/source/rtp_generic_frame_descriptor_extension.h" |
| #include "modules/rtp_rtcp/source/rtp_header_extensions.h" |
| #include "modules/rtp_rtcp/source/rtp_packet_received.h" |
| #include "modules/rtp_rtcp/source/rtp_rtcp_config.h" |
| #include "modules/rtp_rtcp/source/video_rtp_depacketizer.h" |
| #include "modules/rtp_rtcp/source/video_rtp_depacketizer_raw.h" |
| #include "modules/utility/include/process_thread.h" |
| #include "modules/video_coding/frame_object.h" |
| #include "modules/video_coding/h264_sprop_parameter_sets.h" |
| #include "modules/video_coding/h264_sps_pps_tracker.h" |
| #include "modules/video_coding/nack_module.h" |
| #include "modules/video_coding/packet_buffer.h" |
| #include "rtc_base/checks.h" |
| #include "rtc_base/location.h" |
| #include "rtc_base/logging.h" |
| #include "rtc_base/strings/string_builder.h" |
| #include "system_wrappers/include/field_trial.h" |
| #include "system_wrappers/include/metrics.h" |
| #include "video/receive_statistics_proxy.h" |
| |
| namespace webrtc { |
| |
| namespace { |
| // TODO(philipel): Change kPacketBufferStartSize back to 32 in M63 see: |
| // crbug.com/752886 |
| constexpr int kPacketBufferStartSize = 512; |
| constexpr int kPacketBufferMaxSize = 2048; |
| |
| int PacketBufferMaxSize() { |
| // The group here must be a positive power of 2, in which case that is used as |
| // size. All other values shall result in the default value being used. |
| const std::string group_name = |
| webrtc::field_trial::FindFullName("WebRTC-PacketBufferMaxSize"); |
| int packet_buffer_max_size = kPacketBufferMaxSize; |
| if (!group_name.empty() && |
| (sscanf(group_name.c_str(), "%d", &packet_buffer_max_size) != 1 || |
| packet_buffer_max_size <= 0 || |
| // Verify that the number is a positive power of 2. |
| (packet_buffer_max_size & (packet_buffer_max_size - 1)) != 0)) { |
| RTC_LOG(LS_WARNING) << "Invalid packet buffer max size: " << group_name; |
| packet_buffer_max_size = kPacketBufferMaxSize; |
| } |
| return packet_buffer_max_size; |
| } |
| |
| } // namespace |
| |
| std::unique_ptr<RtpRtcp> CreateRtpRtcpModule( |
| Clock* clock, |
| ReceiveStatistics* receive_statistics, |
| Transport* outgoing_transport, |
| RtcpRttStats* rtt_stats, |
| RtcpPacketTypeCounterObserver* rtcp_packet_type_counter_observer, |
| uint32_t local_ssrc) { |
| RtpRtcp::Configuration configuration; |
| configuration.clock = clock; |
| configuration.audio = false; |
| configuration.receiver_only = true; |
| configuration.receive_statistics = receive_statistics; |
| configuration.outgoing_transport = outgoing_transport; |
| configuration.rtt_stats = rtt_stats; |
| configuration.rtcp_packet_type_counter_observer = |
| rtcp_packet_type_counter_observer; |
| configuration.local_media_ssrc = local_ssrc; |
| |
| std::unique_ptr<RtpRtcp> rtp_rtcp = RtpRtcp::Create(configuration); |
| rtp_rtcp->SetRTCPStatus(RtcpMode::kCompound); |
| |
| return rtp_rtcp; |
| } |
| |
| static const int kPacketLogIntervalMs = 10000; |
| |
| RtpVideoStreamReceiver::RtcpFeedbackBuffer::RtcpFeedbackBuffer( |
| KeyFrameRequestSender* key_frame_request_sender, |
| NackSender* nack_sender, |
| LossNotificationSender* loss_notification_sender) |
| : key_frame_request_sender_(key_frame_request_sender), |
| nack_sender_(nack_sender), |
| loss_notification_sender_(loss_notification_sender), |
| request_key_frame_(false) { |
| RTC_DCHECK(key_frame_request_sender_); |
| RTC_DCHECK(nack_sender_); |
| RTC_DCHECK(loss_notification_sender_); |
| } |
| |
| void RtpVideoStreamReceiver::RtcpFeedbackBuffer::RequestKeyFrame() { |
| rtc::CritScope lock(&cs_); |
| request_key_frame_ = true; |
| } |
| |
| void RtpVideoStreamReceiver::RtcpFeedbackBuffer::SendNack( |
| const std::vector<uint16_t>& sequence_numbers, |
| bool buffering_allowed) { |
| RTC_DCHECK(!sequence_numbers.empty()); |
| rtc::CritScope lock(&cs_); |
| nack_sequence_numbers_.insert(nack_sequence_numbers_.end(), |
| sequence_numbers.cbegin(), |
| sequence_numbers.cend()); |
| if (!buffering_allowed) { |
| // Note that while *buffering* is not allowed, *batching* is, meaning that |
| // previously buffered messages may be sent along with the current message. |
| SendBufferedRtcpFeedback(); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::RtcpFeedbackBuffer::SendLossNotification( |
| uint16_t last_decoded_seq_num, |
| uint16_t last_received_seq_num, |
| bool decodability_flag, |
| bool buffering_allowed) { |
| RTC_DCHECK(buffering_allowed); |
| rtc::CritScope lock(&cs_); |
| RTC_DCHECK(!lntf_state_) |
| << "SendLossNotification() called twice in a row with no call to " |
| "SendBufferedRtcpFeedback() in between."; |
| lntf_state_ = absl::make_optional<LossNotificationState>( |
| last_decoded_seq_num, last_received_seq_num, decodability_flag); |
| } |
| |
| void RtpVideoStreamReceiver::RtcpFeedbackBuffer::SendBufferedRtcpFeedback() { |
| bool request_key_frame = false; |
| std::vector<uint16_t> nack_sequence_numbers; |
| absl::optional<LossNotificationState> lntf_state; |
| |
| { |
| rtc::CritScope lock(&cs_); |
| std::swap(request_key_frame, request_key_frame_); |
| std::swap(nack_sequence_numbers, nack_sequence_numbers_); |
| std::swap(lntf_state, lntf_state_); |
| } |
| |
| if (lntf_state) { |
| // If either a NACK or a key frame request is sent, we should buffer |
| // the LNTF and wait for them (NACK or key frame request) to trigger |
| // the compound feedback message. |
| // Otherwise, the LNTF should be sent out immediately. |
| const bool buffering_allowed = |
| request_key_frame || !nack_sequence_numbers.empty(); |
| |
| loss_notification_sender_->SendLossNotification( |
| lntf_state->last_decoded_seq_num, lntf_state->last_received_seq_num, |
| lntf_state->decodability_flag, buffering_allowed); |
| } |
| |
| if (request_key_frame) { |
| key_frame_request_sender_->RequestKeyFrame(); |
| } else if (!nack_sequence_numbers.empty()) { |
| nack_sender_->SendNack(nack_sequence_numbers, true); |
| } |
| } |
| |
| RtpVideoStreamReceiver::RtpVideoStreamReceiver( |
| Clock* clock, |
| Transport* transport, |
| RtcpRttStats* rtt_stats, |
| PacketRouter* packet_router, |
| const VideoReceiveStream::Config* config, |
| ReceiveStatistics* rtp_receive_statistics, |
| ReceiveStatisticsProxy* receive_stats_proxy, |
| ProcessThread* process_thread, |
| NackSender* nack_sender, |
| KeyFrameRequestSender* keyframe_request_sender, |
| video_coding::OnCompleteFrameCallback* complete_frame_callback, |
| rtc::scoped_refptr<FrameDecryptorInterface> frame_decryptor) |
| : clock_(clock), |
| config_(*config), |
| packet_router_(packet_router), |
| process_thread_(process_thread), |
| ntp_estimator_(clock), |
| rtp_header_extensions_(config_.rtp.extensions), |
| rtp_receive_statistics_(rtp_receive_statistics), |
| ulpfec_receiver_(UlpfecReceiver::Create(config->rtp.remote_ssrc, |
| this, |
| config->rtp.extensions)), |
| receiving_(false), |
| last_packet_log_ms_(-1), |
| rtp_rtcp_(CreateRtpRtcpModule(clock, |
| rtp_receive_statistics_, |
| transport, |
| rtt_stats, |
| receive_stats_proxy, |
| config_.rtp.local_ssrc)), |
| complete_frame_callback_(complete_frame_callback), |
| keyframe_request_sender_(keyframe_request_sender), |
| // TODO(bugs.webrtc.org/10336): Let |rtcp_feedback_buffer_| communicate |
| // directly with |rtp_rtcp_|. |
| rtcp_feedback_buffer_(this, nack_sender, this), |
| packet_buffer_(clock_, kPacketBufferStartSize, PacketBufferMaxSize()), |
| has_received_frame_(false), |
| frames_decryptable_(false), |
| absolute_capture_time_receiver_(clock) { |
| constexpr bool remb_candidate = true; |
| if (packet_router_) |
| packet_router_->AddReceiveRtpModule(rtp_rtcp_.get(), remb_candidate); |
| |
| RTC_DCHECK(config_.rtp.rtcp_mode != RtcpMode::kOff) |
| << "A stream should not be configured with RTCP disabled. This value is " |
| "reserved for internal usage."; |
| // TODO(pbos): What's an appropriate local_ssrc for receive-only streams? |
| RTC_DCHECK(config_.rtp.local_ssrc != 0); |
| RTC_DCHECK(config_.rtp.remote_ssrc != config_.rtp.local_ssrc); |
| |
| rtp_rtcp_->SetRTCPStatus(config_.rtp.rtcp_mode); |
| rtp_rtcp_->SetRemoteSSRC(config_.rtp.remote_ssrc); |
| |
| static const int kMaxPacketAgeToNack = 450; |
| const int max_reordering_threshold = (config_.rtp.nack.rtp_history_ms > 0) |
| ? kMaxPacketAgeToNack |
| : kDefaultMaxReorderingThreshold; |
| rtp_receive_statistics_->SetMaxReorderingThreshold(config_.rtp.remote_ssrc, |
| max_reordering_threshold); |
| // TODO(nisse): For historic reasons, we applied the above |
| // max_reordering_threshold also for RTX stats, which makes little sense since |
| // we don't NACK rtx packets. Consider deleting the below block, and rely on |
| // the default threshold. |
| if (config_.rtp.rtx_ssrc) { |
| rtp_receive_statistics_->SetMaxReorderingThreshold( |
| config_.rtp.rtx_ssrc, max_reordering_threshold); |
| } |
| if (config_.rtp.rtcp_xr.receiver_reference_time_report) |
| rtp_rtcp_->SetRtcpXrRrtrStatus(true); |
| |
| // Stats callback for CNAME changes. |
| rtp_rtcp_->RegisterRtcpCnameCallback(receive_stats_proxy); |
| |
| process_thread_->RegisterModule(rtp_rtcp_.get(), RTC_FROM_HERE); |
| |
| if (config_.rtp.lntf.enabled) { |
| loss_notification_controller_ = |
| std::make_unique<LossNotificationController>(&rtcp_feedback_buffer_, |
| &rtcp_feedback_buffer_); |
| } |
| |
| if (config_.rtp.nack.rtp_history_ms != 0) { |
| nack_module_ = std::make_unique<NackModule>(clock_, &rtcp_feedback_buffer_, |
| &rtcp_feedback_buffer_); |
| process_thread_->RegisterModule(nack_module_.get(), RTC_FROM_HERE); |
| } |
| |
| reference_finder_ = |
| std::make_unique<video_coding::RtpFrameReferenceFinder>(this); |
| |
| // Only construct the encrypted receiver if frame encryption is enabled. |
| if (config_.crypto_options.sframe.require_frame_encryption) { |
| buffered_frame_decryptor_ = |
| std::make_unique<BufferedFrameDecryptor>(this, this); |
| if (frame_decryptor != nullptr) { |
| buffered_frame_decryptor_->SetFrameDecryptor(std::move(frame_decryptor)); |
| } |
| } |
| } |
| |
| RtpVideoStreamReceiver::~RtpVideoStreamReceiver() { |
| RTC_DCHECK(secondary_sinks_.empty()); |
| |
| if (nack_module_) { |
| process_thread_->DeRegisterModule(nack_module_.get()); |
| } |
| |
| process_thread_->DeRegisterModule(rtp_rtcp_.get()); |
| |
| if (packet_router_) |
| packet_router_->RemoveReceiveRtpModule(rtp_rtcp_.get()); |
| UpdateHistograms(); |
| } |
| |
| void RtpVideoStreamReceiver::AddReceiveCodec( |
| const VideoCodec& video_codec, |
| const std::map<std::string, std::string>& codec_params, |
| bool raw_payload) { |
| payload_type_map_.emplace( |
| video_codec.plType, |
| raw_payload ? std::make_unique<VideoRtpDepacketizerRaw>() |
| : CreateVideoRtpDepacketizer(video_codec.codecType)); |
| pt_codec_params_.emplace(video_codec.plType, codec_params); |
| } |
| |
| absl::optional<Syncable::Info> RtpVideoStreamReceiver::GetSyncInfo() const { |
| Syncable::Info info; |
| if (rtp_rtcp_->RemoteNTP(&info.capture_time_ntp_secs, |
| &info.capture_time_ntp_frac, nullptr, nullptr, |
| &info.capture_time_source_clock) != 0) { |
| return absl::nullopt; |
| } |
| { |
| rtc::CritScope lock(&sync_info_lock_); |
| if (!last_received_rtp_timestamp_ || !last_received_rtp_system_time_ms_) { |
| return absl::nullopt; |
| } |
| info.latest_received_capture_timestamp = *last_received_rtp_timestamp_; |
| info.latest_receive_time_ms = *last_received_rtp_system_time_ms_; |
| } |
| |
| // Leaves info.current_delay_ms uninitialized. |
| return info; |
| } |
| |
| void RtpVideoStreamReceiver::OnReceivedPayloadData( |
| rtc::CopyOnWriteBuffer codec_payload, |
| const RtpPacketReceived& rtp_packet, |
| const RTPVideoHeader& video) { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| auto packet = std::make_unique<video_coding::PacketBuffer::Packet>( |
| rtp_packet, video, ntp_estimator_.Estimate(rtp_packet.Timestamp()), |
| clock_->TimeInMilliseconds()); |
| |
| // Try to extrapolate absolute capture time if it is missing. |
| // TODO(bugs.webrtc.org/10739): Add support for estimated capture clock |
| // offset. |
| packet->packet_info.set_absolute_capture_time( |
| absolute_capture_time_receiver_.OnReceivePacket( |
| AbsoluteCaptureTimeReceiver::GetSource(packet->packet_info.ssrc(), |
| packet->packet_info.csrcs()), |
| packet->packet_info.rtp_timestamp(), |
| // Assume frequency is the same one for all video frames. |
| kVideoPayloadTypeFrequency, |
| packet->packet_info.absolute_capture_time())); |
| |
| RTPVideoHeader& video_header = packet->video_header; |
| video_header.rotation = kVideoRotation_0; |
| video_header.content_type = VideoContentType::UNSPECIFIED; |
| video_header.video_timing.flags = VideoSendTiming::kInvalid; |
| video_header.is_last_packet_in_frame |= rtp_packet.Marker(); |
| video_header.frame_marking.temporal_id = kNoTemporalIdx; |
| |
| if (const auto* vp9_header = |
| absl::get_if<RTPVideoHeaderVP9>(&video_header.video_type_header)) { |
| video_header.is_last_packet_in_frame |= vp9_header->end_of_frame; |
| video_header.is_first_packet_in_frame |= vp9_header->beginning_of_frame; |
| } |
| |
| rtp_packet.GetExtension<VideoOrientation>(&video_header.rotation); |
| rtp_packet.GetExtension<VideoContentTypeExtension>( |
| &video_header.content_type); |
| rtp_packet.GetExtension<VideoTimingExtension>(&video_header.video_timing); |
| rtp_packet.GetExtension<PlayoutDelayLimits>(&video_header.playout_delay); |
| rtp_packet.GetExtension<FrameMarkingExtension>(&video_header.frame_marking); |
| |
| RtpGenericFrameDescriptor& generic_descriptor = |
| packet->generic_descriptor.emplace(); |
| if (rtp_packet.GetExtension<RtpGenericFrameDescriptorExtension01>( |
| &generic_descriptor)) { |
| if (rtp_packet.HasExtension<RtpGenericFrameDescriptorExtension00>()) { |
| RTC_LOG(LS_WARNING) << "RTP packet had two different GFD versions."; |
| return; |
| } |
| generic_descriptor.SetByteRepresentation( |
| rtp_packet.GetRawExtension<RtpGenericFrameDescriptorExtension01>()); |
| } else if ((rtp_packet.GetExtension<RtpGenericFrameDescriptorExtension00>( |
| &generic_descriptor))) { |
| generic_descriptor.SetByteRepresentation( |
| rtp_packet.GetRawExtension<RtpGenericFrameDescriptorExtension00>()); |
| } else { |
| packet->generic_descriptor = absl::nullopt; |
| } |
| if (packet->generic_descriptor != absl::nullopt) { |
| video_header.is_first_packet_in_frame = |
| packet->generic_descriptor->FirstPacketInSubFrame(); |
| video_header.is_last_packet_in_frame = |
| packet->generic_descriptor->LastPacketInSubFrame(); |
| |
| if (packet->generic_descriptor->FirstPacketInSubFrame()) { |
| video_header.frame_type = |
| packet->generic_descriptor->FrameDependenciesDiffs().empty() |
| ? VideoFrameType::kVideoFrameKey |
| : VideoFrameType::kVideoFrameDelta; |
| |
| auto& descriptor = video_header.generic.emplace(); |
| int64_t frame_id = |
| frame_id_unwrapper_.Unwrap(packet->generic_descriptor->FrameId()); |
| descriptor.frame_id = frame_id; |
| descriptor.spatial_index = packet->generic_descriptor->SpatialLayer(); |
| descriptor.temporal_index = packet->generic_descriptor->TemporalLayer(); |
| descriptor.discardable = |
| packet->generic_descriptor->Discardable().value_or(false); |
| for (uint16_t fdiff : |
| packet->generic_descriptor->FrameDependenciesDiffs()) { |
| descriptor.dependencies.push_back(frame_id - fdiff); |
| } |
| } |
| |
| video_header.width = packet->generic_descriptor->Width(); |
| video_header.height = packet->generic_descriptor->Height(); |
| } |
| |
| // Color space should only be transmitted in the last packet of a frame, |
| // therefore, neglect it otherwise so that last_color_space_ is not reset by |
| // mistake. |
| if (video_header.is_last_packet_in_frame) { |
| video_header.color_space = rtp_packet.GetExtension<ColorSpaceExtension>(); |
| if (video_header.color_space || |
| video_header.frame_type == VideoFrameType::kVideoFrameKey) { |
| // Store color space since it's only transmitted when changed or for key |
| // frames. Color space will be cleared if a key frame is transmitted |
| // without color space information. |
| last_color_space_ = video_header.color_space; |
| } else if (last_color_space_) { |
| video_header.color_space = last_color_space_; |
| } |
| } |
| |
| if (loss_notification_controller_) { |
| if (rtp_packet.recovered()) { |
| // TODO(bugs.webrtc.org/10336): Implement support for reordering. |
| RTC_LOG(LS_INFO) |
| << "LossNotificationController does not support reordering."; |
| } else if (!packet->generic_descriptor) { |
| RTC_LOG(LS_WARNING) << "LossNotificationController requires generic " |
| "frame descriptor, but it is missing."; |
| } else { |
| if (video_header.is_first_packet_in_frame) { |
| RTC_DCHECK(video_header.generic); |
| LossNotificationController::FrameDetails frame; |
| frame.is_keyframe = |
| video_header.frame_type == VideoFrameType::kVideoFrameKey; |
| frame.frame_id = video_header.generic->frame_id; |
| frame.frame_dependencies = video_header.generic->dependencies; |
| loss_notification_controller_->OnReceivedPacket( |
| rtp_packet.SequenceNumber(), &frame); |
| } else { |
| loss_notification_controller_->OnReceivedPacket( |
| rtp_packet.SequenceNumber(), nullptr); |
| } |
| } |
| } |
| |
| if (nack_module_) { |
| const bool is_keyframe = |
| video_header.is_first_packet_in_frame && |
| video_header.frame_type == VideoFrameType::kVideoFrameKey; |
| |
| packet->times_nacked = nack_module_->OnReceivedPacket( |
| rtp_packet.SequenceNumber(), is_keyframe, rtp_packet.recovered()); |
| } else { |
| packet->times_nacked = -1; |
| } |
| |
| if (codec_payload.size() == 0) { |
| NotifyReceiverOfEmptyPacket(packet->seq_num); |
| rtcp_feedback_buffer_.SendBufferedRtcpFeedback(); |
| return; |
| } |
| |
| if (packet->codec() == kVideoCodecH264) { |
| // Only when we start to receive packets will we know what payload type |
| // that will be used. When we know the payload type insert the correct |
| // sps/pps into the tracker. |
| if (packet->payload_type != last_payload_type_) { |
| last_payload_type_ = packet->payload_type; |
| InsertSpsPpsIntoTracker(packet->payload_type); |
| } |
| |
| video_coding::H264SpsPpsTracker::FixedBitstream fixed = |
| tracker_.CopyAndFixBitstream( |
| rtc::MakeArrayView(codec_payload.cdata(), codec_payload.size()), |
| &packet->video_header); |
| |
| switch (fixed.action) { |
| case video_coding::H264SpsPpsTracker::kRequestKeyframe: |
| rtcp_feedback_buffer_.RequestKeyFrame(); |
| rtcp_feedback_buffer_.SendBufferedRtcpFeedback(); |
| ABSL_FALLTHROUGH_INTENDED; |
| case video_coding::H264SpsPpsTracker::kDrop: |
| return; |
| case video_coding::H264SpsPpsTracker::kInsert: |
| packet->video_payload = std::move(fixed.bitstream); |
| break; |
| } |
| |
| } else { |
| packet->video_payload = std::move(codec_payload); |
| } |
| |
| rtcp_feedback_buffer_.SendBufferedRtcpFeedback(); |
| frame_counter_.Add(packet->timestamp); |
| OnInsertedPacket(packet_buffer_.InsertPacket(std::move(packet))); |
| } |
| |
| void RtpVideoStreamReceiver::OnRecoveredPacket(const uint8_t* rtp_packet, |
| size_t rtp_packet_length) { |
| RtpPacketReceived packet; |
| if (!packet.Parse(rtp_packet, rtp_packet_length)) |
| return; |
| if (packet.PayloadType() == config_.rtp.red_payload_type) { |
| RTC_LOG(LS_WARNING) << "Discarding recovered packet with RED encapsulation"; |
| return; |
| } |
| |
| packet.IdentifyExtensions(rtp_header_extensions_); |
| packet.set_payload_type_frequency(kVideoPayloadTypeFrequency); |
| // TODO(nisse): UlpfecReceiverImpl::ProcessReceivedFec passes both |
| // original (decapsulated) media packets and recovered packets to |
| // this callback. We need a way to distinguish, for setting |
| // packet.recovered() correctly. Ideally, move RED decapsulation out |
| // of the Ulpfec implementation. |
| |
| ReceivePacket(packet); |
| } |
| |
| // This method handles both regular RTP packets and packets recovered |
| // via FlexFEC. |
| void RtpVideoStreamReceiver::OnRtpPacket(const RtpPacketReceived& packet) { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| |
| if (!receiving_) { |
| return; |
| } |
| |
| if (!packet.recovered()) { |
| // TODO(nisse): Exclude out-of-order packets? |
| int64_t now_ms = clock_->TimeInMilliseconds(); |
| { |
| rtc::CritScope cs(&sync_info_lock_); |
| last_received_rtp_timestamp_ = packet.Timestamp(); |
| last_received_rtp_system_time_ms_ = now_ms; |
| } |
| // Periodically log the RTP header of incoming packets. |
| if (now_ms - last_packet_log_ms_ > kPacketLogIntervalMs) { |
| rtc::StringBuilder ss; |
| ss << "Packet received on SSRC: " << packet.Ssrc() |
| << " with payload type: " << static_cast<int>(packet.PayloadType()) |
| << ", timestamp: " << packet.Timestamp() |
| << ", sequence number: " << packet.SequenceNumber() |
| << ", arrival time: " << packet.arrival_time_ms(); |
| int32_t time_offset; |
| if (packet.GetExtension<TransmissionOffset>(&time_offset)) { |
| ss << ", toffset: " << time_offset; |
| } |
| uint32_t send_time; |
| if (packet.GetExtension<AbsoluteSendTime>(&send_time)) { |
| ss << ", abs send time: " << send_time; |
| } |
| RTC_LOG(LS_INFO) << ss.str(); |
| last_packet_log_ms_ = now_ms; |
| } |
| } |
| |
| ReceivePacket(packet); |
| |
| // Update receive statistics after ReceivePacket. |
| // Receive statistics will be reset if the payload type changes (make sure |
| // that the first packet is included in the stats). |
| if (!packet.recovered()) { |
| rtp_receive_statistics_->OnRtpPacket(packet); |
| } |
| |
| for (RtpPacketSinkInterface* secondary_sink : secondary_sinks_) { |
| secondary_sink->OnRtpPacket(packet); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::RequestKeyFrame() { |
| // TODO(bugs.webrtc.org/10336): Allow the sender to ignore key frame requests |
| // issued by anything other than the LossNotificationController if it (the |
| // sender) is relying on LNTF alone. |
| if (keyframe_request_sender_) { |
| keyframe_request_sender_->RequestKeyFrame(); |
| } else { |
| rtp_rtcp_->SendPictureLossIndication(); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::SendLossNotification( |
| uint16_t last_decoded_seq_num, |
| uint16_t last_received_seq_num, |
| bool decodability_flag, |
| bool buffering_allowed) { |
| RTC_DCHECK(config_.rtp.lntf.enabled); |
| rtp_rtcp_->SendLossNotification(last_decoded_seq_num, last_received_seq_num, |
| decodability_flag, buffering_allowed); |
| } |
| |
| bool RtpVideoStreamReceiver::IsUlpfecEnabled() const { |
| return config_.rtp.ulpfec_payload_type != -1; |
| } |
| |
| bool RtpVideoStreamReceiver::IsRetransmissionsEnabled() const { |
| return config_.rtp.nack.rtp_history_ms > 0; |
| } |
| |
| void RtpVideoStreamReceiver::RequestPacketRetransmit( |
| const std::vector<uint16_t>& sequence_numbers) { |
| rtp_rtcp_->SendNack(sequence_numbers); |
| } |
| |
| bool RtpVideoStreamReceiver::IsDecryptable() const { |
| return frames_decryptable_.load(); |
| } |
| |
| void RtpVideoStreamReceiver::OnInsertedPacket( |
| video_coding::PacketBuffer::InsertResult result) { |
| for (std::unique_ptr<video_coding::RtpFrameObject>& frame : result.frames) { |
| OnAssembledFrame(std::move(frame)); |
| } |
| if (result.buffer_cleared) { |
| RequestKeyFrame(); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::OnAssembledFrame( |
| std::unique_ptr<video_coding::RtpFrameObject> frame) { |
| RTC_DCHECK_RUN_ON(&network_tc_); |
| RTC_DCHECK(frame); |
| |
| const absl::optional<RTPVideoHeader::GenericDescriptorInfo>& descriptor = |
| frame->GetRtpVideoHeader().generic; |
| |
| if (loss_notification_controller_ && descriptor) { |
| loss_notification_controller_->OnAssembledFrame( |
| frame->first_seq_num(), descriptor->frame_id, descriptor->discardable, |
| descriptor->dependencies); |
| } |
| |
| // If frames arrive before a key frame, they would not be decodable. |
| // In that case, request a key frame ASAP. |
| if (!has_received_frame_) { |
| if (frame->FrameType() != VideoFrameType::kVideoFrameKey) { |
| // |loss_notification_controller_|, if present, would have already |
| // requested a key frame when the first packet for the non-key frame |
| // had arrived, so no need to replicate the request. |
| if (!loss_notification_controller_) { |
| RequestKeyFrame(); |
| } |
| } |
| has_received_frame_ = true; |
| } |
| |
| rtc::CritScope lock(&reference_finder_lock_); |
| // Reset |reference_finder_| if |frame| is new and the codec have changed. |
| if (current_codec_) { |
| bool frame_is_newer = |
| AheadOf(frame->Timestamp(), last_assembled_frame_rtp_timestamp_); |
| |
| if (frame->codec_type() != current_codec_) { |
| if (frame_is_newer) { |
| // When we reset the |reference_finder_| we don't want new picture ids |
| // to overlap with old picture ids. To ensure that doesn't happen we |
| // start from the |last_completed_picture_id_| and add an offset in case |
| // of reordering. |
| reference_finder_ = |
| std::make_unique<video_coding::RtpFrameReferenceFinder>( |
| this, last_completed_picture_id_ + |
| std::numeric_limits<uint16_t>::max()); |
| current_codec_ = frame->codec_type(); |
| } else { |
| // Old frame from before the codec switch, discard it. |
| return; |
| } |
| } |
| |
| if (frame_is_newer) { |
| last_assembled_frame_rtp_timestamp_ = frame->Timestamp(); |
| } |
| } else { |
| current_codec_ = frame->codec_type(); |
| last_assembled_frame_rtp_timestamp_ = frame->Timestamp(); |
| } |
| |
| if (buffered_frame_decryptor_ == nullptr) { |
| reference_finder_->ManageFrame(std::move(frame)); |
| } else { |
| buffered_frame_decryptor_->ManageEncryptedFrame(std::move(frame)); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::OnCompleteFrame( |
| std::unique_ptr<video_coding::EncodedFrame> frame) { |
| { |
| rtc::CritScope lock(&last_seq_num_cs_); |
| video_coding::RtpFrameObject* rtp_frame = |
| static_cast<video_coding::RtpFrameObject*>(frame.get()); |
| last_seq_num_for_pic_id_[rtp_frame->id.picture_id] = |
| rtp_frame->last_seq_num(); |
| } |
| last_completed_picture_id_ = |
| std::max(last_completed_picture_id_, frame->id.picture_id); |
| complete_frame_callback_->OnCompleteFrame(std::move(frame)); |
| } |
| |
| void RtpVideoStreamReceiver::OnDecryptedFrame( |
| std::unique_ptr<video_coding::RtpFrameObject> frame) { |
| rtc::CritScope lock(&reference_finder_lock_); |
| reference_finder_->ManageFrame(std::move(frame)); |
| } |
| |
| void RtpVideoStreamReceiver::OnDecryptionStatusChange( |
| FrameDecryptorInterface::Status status) { |
| frames_decryptable_.store( |
| (status == FrameDecryptorInterface::Status::kOk) || |
| (status == FrameDecryptorInterface::Status::kRecoverable)); |
| } |
| |
| void RtpVideoStreamReceiver::SetFrameDecryptor( |
| rtc::scoped_refptr<FrameDecryptorInterface> frame_decryptor) { |
| RTC_DCHECK_RUN_ON(&network_tc_); |
| if (buffered_frame_decryptor_ == nullptr) { |
| buffered_frame_decryptor_ = |
| std::make_unique<BufferedFrameDecryptor>(this, this); |
| } |
| buffered_frame_decryptor_->SetFrameDecryptor(std::move(frame_decryptor)); |
| } |
| |
| void RtpVideoStreamReceiver::UpdateRtt(int64_t max_rtt_ms) { |
| if (nack_module_) |
| nack_module_->UpdateRtt(max_rtt_ms); |
| } |
| |
| absl::optional<int64_t> RtpVideoStreamReceiver::LastReceivedPacketMs() const { |
| return packet_buffer_.LastReceivedPacketMs(); |
| } |
| |
| absl::optional<int64_t> RtpVideoStreamReceiver::LastReceivedKeyframePacketMs() |
| const { |
| return packet_buffer_.LastReceivedKeyframePacketMs(); |
| } |
| |
| void RtpVideoStreamReceiver::AddSecondarySink(RtpPacketSinkInterface* sink) { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| RTC_DCHECK(!absl::c_linear_search(secondary_sinks_, sink)); |
| secondary_sinks_.push_back(sink); |
| } |
| |
| void RtpVideoStreamReceiver::RemoveSecondarySink( |
| const RtpPacketSinkInterface* sink) { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| auto it = absl::c_find(secondary_sinks_, sink); |
| if (it == secondary_sinks_.end()) { |
| // We might be rolling-back a call whose setup failed mid-way. In such a |
| // case, it's simpler to remove "everything" rather than remember what |
| // has already been added. |
| RTC_LOG(LS_WARNING) << "Removal of unknown sink."; |
| return; |
| } |
| secondary_sinks_.erase(it); |
| } |
| |
| void RtpVideoStreamReceiver::ReceivePacket(const RtpPacketReceived& packet) { |
| if (packet.payload_size() == 0) { |
| // Padding or keep-alive packet. |
| // TODO(nisse): Could drop empty packets earlier, but need to figure out how |
| // they should be counted in stats. |
| NotifyReceiverOfEmptyPacket(packet.SequenceNumber()); |
| return; |
| } |
| if (packet.PayloadType() == config_.rtp.red_payload_type) { |
| ParseAndHandleEncapsulatingHeader(packet); |
| return; |
| } |
| |
| const auto type_it = payload_type_map_.find(packet.PayloadType()); |
| if (type_it == payload_type_map_.end()) { |
| return; |
| } |
| absl::optional<VideoRtpDepacketizer::ParsedRtpPayload> parsed_payload = |
| type_it->second->Parse(packet.PayloadBuffer()); |
| if (parsed_payload == absl::nullopt) { |
| RTC_LOG(LS_WARNING) << "Failed parsing payload."; |
| return; |
| } |
| |
| OnReceivedPayloadData(std::move(parsed_payload->video_payload), packet, |
| parsed_payload->video_header); |
| } |
| |
| void RtpVideoStreamReceiver::ParseAndHandleEncapsulatingHeader( |
| const RtpPacketReceived& packet) { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| if (packet.PayloadType() == config_.rtp.red_payload_type && |
| packet.payload_size() > 0) { |
| if (packet.payload()[0] == config_.rtp.ulpfec_payload_type) { |
| // Notify video_receiver about received FEC packets to avoid NACKing these |
| // packets. |
| NotifyReceiverOfEmptyPacket(packet.SequenceNumber()); |
| } |
| if (!ulpfec_receiver_->AddReceivedRedPacket( |
| packet, config_.rtp.ulpfec_payload_type)) { |
| return; |
| } |
| ulpfec_receiver_->ProcessReceivedFec(); |
| } |
| } |
| |
| // In the case of a video stream without picture ids and no rtx the |
| // RtpFrameReferenceFinder will need to know about padding to |
| // correctly calculate frame references. |
| void RtpVideoStreamReceiver::NotifyReceiverOfEmptyPacket(uint16_t seq_num) { |
| { |
| rtc::CritScope lock(&reference_finder_lock_); |
| reference_finder_->PaddingReceived(seq_num); |
| } |
| OnInsertedPacket(packet_buffer_.InsertPadding(seq_num)); |
| if (nack_module_) { |
| nack_module_->OnReceivedPacket(seq_num, /* is_keyframe = */ false, |
| /* is _recovered = */ false); |
| } |
| if (loss_notification_controller_) { |
| // TODO(bugs.webrtc.org/10336): Handle empty packets. |
| RTC_LOG(LS_WARNING) |
| << "LossNotificationController does not expect empty packets."; |
| } |
| } |
| |
| bool RtpVideoStreamReceiver::DeliverRtcp(const uint8_t* rtcp_packet, |
| size_t rtcp_packet_length) { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| |
| if (!receiving_) { |
| return false; |
| } |
| |
| rtp_rtcp_->IncomingRtcpPacket(rtcp_packet, rtcp_packet_length); |
| |
| int64_t rtt = 0; |
| rtp_rtcp_->RTT(config_.rtp.remote_ssrc, &rtt, nullptr, nullptr, nullptr); |
| if (rtt == 0) { |
| // Waiting for valid rtt. |
| return true; |
| } |
| uint32_t ntp_secs = 0; |
| uint32_t ntp_frac = 0; |
| uint32_t rtp_timestamp = 0; |
| uint32_t recieved_ntp_secs = 0; |
| uint32_t recieved_ntp_frac = 0; |
| if (rtp_rtcp_->RemoteNTP(&ntp_secs, &ntp_frac, &recieved_ntp_secs, |
| &recieved_ntp_frac, &rtp_timestamp) != 0) { |
| // Waiting for RTCP. |
| return true; |
| } |
| NtpTime recieved_ntp(recieved_ntp_secs, recieved_ntp_frac); |
| int64_t time_since_recieved = |
| clock_->CurrentNtpInMilliseconds() - recieved_ntp.ToMs(); |
| // Don't use old SRs to estimate time. |
| if (time_since_recieved <= 1) { |
| ntp_estimator_.UpdateRtcpTimestamp(rtt, ntp_secs, ntp_frac, rtp_timestamp); |
| } |
| |
| return true; |
| } |
| |
| void RtpVideoStreamReceiver::FrameContinuous(int64_t picture_id) { |
| if (!nack_module_) |
| return; |
| |
| int seq_num = -1; |
| { |
| rtc::CritScope lock(&last_seq_num_cs_); |
| auto seq_num_it = last_seq_num_for_pic_id_.find(picture_id); |
| if (seq_num_it != last_seq_num_for_pic_id_.end()) |
| seq_num = seq_num_it->second; |
| } |
| if (seq_num != -1) |
| nack_module_->ClearUpTo(seq_num); |
| } |
| |
| void RtpVideoStreamReceiver::FrameDecoded(int64_t picture_id) { |
| int seq_num = -1; |
| { |
| rtc::CritScope lock(&last_seq_num_cs_); |
| auto seq_num_it = last_seq_num_for_pic_id_.find(picture_id); |
| if (seq_num_it != last_seq_num_for_pic_id_.end()) { |
| seq_num = seq_num_it->second; |
| last_seq_num_for_pic_id_.erase(last_seq_num_for_pic_id_.begin(), |
| ++seq_num_it); |
| } |
| } |
| if (seq_num != -1) { |
| packet_buffer_.ClearTo(seq_num); |
| rtc::CritScope lock(&reference_finder_lock_); |
| reference_finder_->ClearTo(seq_num); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::SignalNetworkState(NetworkState state) { |
| rtp_rtcp_->SetRTCPStatus(state == kNetworkUp ? config_.rtp.rtcp_mode |
| : RtcpMode::kOff); |
| } |
| |
| void RtpVideoStreamReceiver::StartReceive() { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| receiving_ = true; |
| } |
| |
| void RtpVideoStreamReceiver::StopReceive() { |
| RTC_DCHECK_RUN_ON(&worker_task_checker_); |
| receiving_ = false; |
| } |
| |
| void RtpVideoStreamReceiver::UpdateHistograms() { |
| FecPacketCounter counter = ulpfec_receiver_->GetPacketCounter(); |
| if (counter.first_packet_time_ms == -1) |
| return; |
| |
| int64_t elapsed_sec = |
| (clock_->TimeInMilliseconds() - counter.first_packet_time_ms) / 1000; |
| if (elapsed_sec < metrics::kMinRunTimeInSeconds) |
| return; |
| |
| if (counter.num_packets > 0) { |
| RTC_HISTOGRAM_PERCENTAGE( |
| "WebRTC.Video.ReceivedFecPacketsInPercent", |
| static_cast<int>(counter.num_fec_packets * 100 / counter.num_packets)); |
| } |
| if (counter.num_fec_packets > 0) { |
| RTC_HISTOGRAM_PERCENTAGE("WebRTC.Video.RecoveredMediaPacketsInPercentOfFec", |
| static_cast<int>(counter.num_recovered_packets * |
| 100 / counter.num_fec_packets)); |
| } |
| if (config_.rtp.ulpfec_payload_type != -1) { |
| RTC_HISTOGRAM_COUNTS_10000( |
| "WebRTC.Video.FecBitrateReceivedInKbps", |
| static_cast<int>(counter.num_bytes * 8 / elapsed_sec / 1000)); |
| } |
| } |
| |
| void RtpVideoStreamReceiver::InsertSpsPpsIntoTracker(uint8_t payload_type) { |
| auto codec_params_it = pt_codec_params_.find(payload_type); |
| if (codec_params_it == pt_codec_params_.end()) |
| return; |
| |
| RTC_LOG(LS_INFO) << "Found out of band supplied codec parameters for" |
| " payload type: " |
| << static_cast<int>(payload_type); |
| |
| H264SpropParameterSets sprop_decoder; |
| auto sprop_base64_it = |
| codec_params_it->second.find(cricket::kH264FmtpSpropParameterSets); |
| |
| if (sprop_base64_it == codec_params_it->second.end()) |
| return; |
| |
| if (!sprop_decoder.DecodeSprop(sprop_base64_it->second.c_str())) |
| return; |
| |
| tracker_.InsertSpsPpsNalus(sprop_decoder.sps_nalu(), |
| sprop_decoder.pps_nalu()); |
| } |
| |
| } // namespace webrtc |