本文基于 WebRTC 中的示例应用 peerconnection_client 分析 WebRTC Audio 接收和发送的关键过程。首先是发送的过程，然后是接收的过程。

创建 webrtc::AudioState

应用程序择机初始化 PeerConnectionFactory：

#0  Init () at webrtc/src/pc/channel_manager.cc:121
#1  Initialize () at webrtc/src/pc/peer_connection_factory.cc:139
#6  webrtc::CreateModularPeerConnectionFactory(webrtc::PeerConnectionFactoryDependencies) () at webrtc/src/pc/peer_connection_factory.cc:55
#7  webrtc::CreatePeerConnectionFactory(rtc::Thread*, rtc::Thread*, rtc::Thread*, rtc::scoped_refptr<webrtc::AudioDeviceModule>, rtc::scoped_refptr<webrtc::AudioEncoderFactory>, rtc::scoped_refptr<webrtc::AudioDecoderFactory>, std::__1::unique_ptr<webrtc::VideoEncoderFactory, std::__1::default_delete<webrtc::VideoEncoderFactory> >, std::__1::unique_ptr<webrtc::VideoDecoderFactory, std::__1::default_delete<webrtc::VideoDecoderFactory> >, rtc::scoped_refptr<webrtc::AudioMixer>, rtc::scoped_refptr<webrtc::AudioProcessing>) () at webrtc/src/api/create_peerconnection_factory.cc:65
#8  InitializePeerConnection () at webrtc/src/examples/peerconnection/client/conductor.cc:132
#9  ConnectToPeer () at webrtc/src/examples/peerconnection/client/conductor.cc:422
#10 OnRowActivated () at webrtc/src/examples/peerconnection/client/linux/main_wnd.cc:433
#11 (anonymous namespace)::OnRowActivatedCallback(_GtkTreeView*, _GtkTreePath*, _GtkTreeViewColumn*, void*) () at webrtc/src/examples/peerconnection/client/linux/main_wnd.cc:70

由 Conductor::InitializePeerConnection() 的代码可知，PeerConnectionFactory 竟然是随同 peer connection一起创建的。

在 webrtc/src/pc/channel_manager.cc 文件里定义的 ChannelManager::Init() 函数中会在另一个线程中起一个 task，调用 media_engine_->Init()，完成媒体引擎的初始化：

bool ChannelManager::Init() {RTC_DCHECK(!initialized_);if (initialized_) {return false;}RTC_DCHECK(network_thread_);RTC_DCHECK(worker_thread_);if (!network_thread_->IsCurrent()) {// Do not allow invoking calls to other threads on the network thread.network_thread_->Invoke<void>(RTC_FROM_HERE, [&] { network_thread_->DisallowBlockingCalls(); });}if (media_engine_) {initialized_ = worker_thread_->Invoke<bool>(RTC_FROM_HERE, [&] { return media_engine_->Init(); });RTC_DCHECK(initialized_);} else {initialized_ = true;}return initialized_;
}

媒体引擎初始化过程中，将会创建 AudioState：

#0  webrtc::AudioState::Create(webrtc::AudioState::Config const&) () at webrtc/src/audio/audio_state.cc:188
#1  Init () at webrtc/src/media/engine/webrtc_voice_engine.cc:260
#2  cricket::CompositeMediaEngine::Init() () at webrtc/src/media/base/media_engine.cc:155
#3  cricket::ChannelManager::Init()::$_3::operator()() const () at webrtc/src/pc/channel_manager.cc:135

AudioState 伴随着 PeerConnectionFactory 、ChannelManager 和 MediaEngine 的创建及初始化一起创建。

创建 WebRTC Call

应用程序根据需要创建 peer connection：

#0  CreatePeerConnection () at webrtc/src/pc/peer_connection_factory.cc:240
#1  webrtc::PeerConnectionFactory::CreatePeerConnection(webrtc::PeerConnectionInterface::RTCConfiguration const&, std::__1::unique_ptr<cricket::PortAllocator, std::__1::default_delete<cricket::PortAllocator> >, std::__1::unique_ptr<rtc::RTCCertificateGeneratorInterface, std::__1::default_delete<rtc::RTCCertificateGeneratorInterface> >, webrtc::PeerConnectionObserver*) () at webrtc/src/pc/peer_connection_factory.cc:233
#7  CreatePeerConnection () at webrtc/src/examples/peerconnection/client/conductor.cc:184
#8  InitializePeerConnection () at webrtc/src/examples/peerconnection/client/conductor.cc:148
#9  ConnectToPeer () at webrtc/src/examples/peerconnection/client/conductor.cc:422
#10 OnRowActivated () at webrtc/src/examples/peerconnection/client/linux/main_wnd.cc:433
#11 (anonymous namespace)::OnRowActivatedCallback(_GtkTreeView*, _GtkTreePath*, _GtkTreeViewColumn*, void*) () at webrtc/src/examples/peerconnection/client/linux/main_wnd.cc:70

PeerConnectionFactory::CreatePeerConnection() 在创建 connection 时，会在另一个线程中起一个task 来创建 Call：

rtc::scoped_refptr<PeerConnectionInterface>
PeerConnectionFactory::CreatePeerConnection(const PeerConnectionInterface::RTCConfiguration& configuration,PeerConnectionDependencies dependencies) {RTC_DCHECK(signaling_thread_->IsCurrent());// Set internal defaults if optional dependencies are not set.if (!dependencies.cert_generator) {dependencies.cert_generator =absl::make_unique<rtc::RTCCertificateGenerator>(signaling_thread_,network_thread_);}if (!dependencies.allocator) {network_thread_->Invoke<void>(RTC_FROM_HERE, [this, &configuration,&dependencies]() {dependencies.allocator = absl::make_unique<cricket::BasicPortAllocator>(default_network_manager_.get(), default_socket_factory_.get(),configuration.turn_customizer);});}// TODO(zstein): Once chromium injects its own AsyncResolverFactory, set// |dependencies.async_resolver_factory| to a new// |rtc::BasicAsyncResolverFactory| if no factory is provided.network_thread_->Invoke<void>(RTC_FROM_HERE,rtc::Bind(&cricket::PortAllocator::SetNetworkIgnoreMask,dependencies.allocator.get(), options_.network_ignore_mask));std::unique_ptr<RtcEventLog> event_log =worker_thread_->Invoke<std::unique_ptr<RtcEventLog>>(RTC_FROM_HERE,rtc::Bind(&PeerConnectionFactory::CreateRtcEventLog_w, this));std::unique_ptr<Call> call = worker_thread_->Invoke<std::unique_ptr<Call>>(RTC_FROM_HERE,rtc::Bind(&PeerConnectionFactory::CreateCall_w, this, event_log.get()));rtc::scoped_refptr<PeerConnection> pc(new rtc::RefCountedObject<PeerConnection>(this, std::move(event_log),std::move(call)));ActionsBeforeInitializeForTesting(pc);if (!pc->Initialize(configuration, std::move(dependencies))) {return nullptr;}return PeerConnectionProxy::Create(signaling_thread(), pc);
}

创建 WebRTC Call 的过程如下：

#0  webrtc::Call::Create(webrtc::CallConfig const&) () at webrtc/src/call/call.cc:424
#1  webrtc::CallFactory::CreateCall(webrtc::CallConfig const&) () at webrtc/src/call/call_factory.cc:84
#2  CreateCall_w () at webrtc/src/pc/peer_connection_factory.cc:364

不难看出，在 WebRTC 中，Call 是 per peer connection 的。

为 WebRTC Call 注入的 AudioState 来自于全局的 MediaEngine 的 VoiceEngine。AudioState 是全局的，而 Call 则是 connection 局部的。

创建 WebRtcAudioReceiveStream

WebRTC 应用需要起一个专门的专门的连接，用于接收媒体协商信息。在收到媒体协商信息之后，则将媒体协商信息进行层层传递及处理：

#0  cricket::BaseChannel::SetRemoteContent(cricket::MediaContentDescription const*, webrtc::SdpType, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >*) () at webrtc/src/pc/channel.cc:299
#1  PushdownMediaDescription () at webrtc/src/pc/peer_connection.cc:5700
#2  UpdateSessionState () at webrtc/src/pc/peer_connection.cc:5668
#3  ApplyRemoteDescription () at webrtc/src/pc/peer_connection.cc:2668
#4  SetRemoteDescription () at webrtc/src/pc/peer_connection.cc:2562
#5  webrtc::PeerConnection::SetRemoteDescription(webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*) () at webrtc/src/pc/peer_connection.cc:2506#6  void webrtc::ReturnType<void>::Invoke<webrtc::PeerConnectionInterface, void (webrtc::PeerConnectionInterface::*)(webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*), webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*>(webrtc::PeerConnectionInterface*, void (webrtc::PeerConnectionInterface::*)(webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*), webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*) () at webrtc/src/api/proxy.h:131
#7  webrtc::MethodCall2<webrtc::PeerConnectionInterface, void, webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*>::OnMessage(rtc::Message*) () at webrtc/src/api/proxy.h:252
#8  webrtc::internal::SynchronousMethodCall::Invoke(rtc::Location const&, rtc::Thread*) () at webrtc/src/api/proxy.cc:24
#9  webrtc::MethodCall2<webrtc::PeerConnectionInterface, void, webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*>::Marshal(rtc::Location const&, rtc::Thread*) () at webrtc/src/api/proxy.h:246
#10 webrtc::PeerConnectionProxyWithInternal<webrtc::PeerConnectionInterface>::SetRemoteDescription(webrtc::SetSessionDescriptionObserver*, webrtc::SessionDescriptionInterface*) () at webrtc/src/api/peer_connection_proxy.h:101#11 OnMessageFromPeer () at webrtc/src/examples/peerconnection/client/conductor.cc:351
#12 PeerConnectionClient::OnMessageFromPeer(int, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&) ()at webrtc/src/examples/peerconnection/client/peer_connection_client.cc:250
#13 OnHangingGetRead () at webrtc/src/examples/peerconnection/client/peer_connection_client.cc:403
#18 rtc::SocketDispatcher::OnEvent(unsigned int, int) () at webrtc/src/rtc_base/physical_socket_server.cc:790
#19 rtc::ProcessEvents(rtc::Dispatcher*, bool, bool, bool) () at webrtc/src/rtc_base/physical_socket_server.cc:1379
#20 WaitEpoll () at webrtc/src/rtc_base/physical_socket_server.cc:1620
#21 rtc::PhysicalSocketServer::Wait(int, bool) () at webrtc/src/rtc_base/physical_socket_server.cc:1328
#22 CustomSocketServer::Wait(int, bool) () at webrtc/src/examples/peerconnection/client/linux/main.cc:56
#23 Get () at webrtc/src/rtc_base/message_queue.cc:329
#24 ProcessMessages () at webrtc/src/rtc_base/thread.cc:525
#25 rtc::Thread::Run() () at webrtc/src/rtc_base/thread.cc:351
#26 main () at webrtc/src/examples/peerconnection/client/linux/main.cc:111

在 webrtc/src/pc/channel.cc 文件里定义的 BaseChannel::SetRemoteContent() 函数中将收到的媒体协商信息，抛给另一个线程进行处理：

bool BaseChannel::SetRemoteContent(const MediaContentDescription* content,SdpType type,std::string* error_desc) {TRACE_EVENT0("webrtc", "BaseChannel::SetRemoteContent");return InvokeOnWorker<bool>(RTC_FROM_HERE,Bind(&BaseChannel::SetRemoteContent_w, this, content, type, error_desc));
}

WebRtcAudioReceiveStream 最终由 webrtc/src/media/engine/webrtc_voice_engine.cc 文件中定义的 WebRtcVoiceMediaChannel::AddRecvStream() 创建：

#0  AddRecvStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:1854
#1  AddRecvStream_w () at webrtc/src/pc/channel.cc:599
#2  UpdateRemoteStreams_w () at webrtc/src/pc/channel.cc:714
#3  SetRemoteContent_w () at webrtc/src/pc/channel.cc:951

创建 WebRtcAudioReceiveStream 时，也会一并创建 Call 的 AudioReceiveStream：

#0  CreateAudioReceiveStream () at webrtc/src/call/call.cc:779
#1  RecreateAudioReceiveStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:1224
#2  WebRtcAudioReceiveStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:1090
#3  AddRecvStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:1889
#4  AddRecvStream_w () at webrtc/src/pc/channel.cc:599
#5  UpdateRemoteStreams_w () at webrtc/src/pc/channel.cc:714

WebRtcAudioReceiveStream 创建完成后，随即将其加进 mixer，作为 mixer 的 audio source 之一：

#0  AddSource () at webrtc/src/modules/audio_mixer/audio_mixer_impl.cc:160
#1  AddReceivingStream () at webrtc/src/audio/audio_state.cc:60
#2  Start () at webrtc/src/audio/audio_receive_stream.cc:161
#3  SetPlayout () at webrtc/src/media/engine/webrtc_voice_engine.cc:1173
#4  AddRecvStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:1899
#5  AddRecvStream_w () at webrtc/src/pc/channel.cc:599

WebRTC 中音频接收处理的关键流程

1. 从网络收到 UDP 包

#0  OnPacketReceived () at webrtc/src/pc/channel.cc:507
#1  cricket::BaseChannel::OnRtpPacket(webrtc::RtpPacketReceived const&) () at webrtc/src/pc/channel.cc:468
#2  webrtc::RtpDemuxer::OnRtpPacket(webrtc::RtpPacketReceived const&) () at webrtc/src/call/rtp_demuxer.cc:177
#3  DemuxPacket () at webrtc/src/pc/rtp_transport.cc:194
#4  OnRtpPacketReceived () at webrtc/src/pc/srtp_transport.cc:230
#5  OnReadPacket () at webrtc/src/pc/rtp_transport.cc:268
#10 OnReadPacket () at webrtc/src/p2p/base/dtls_transport.cc:600
#15 OnReadPacket () at webrtc/src/p2p/base/p2p_transport_channel.cc:2499
#20 OnReadPacket () at webrtc/src/p2p/base/connection.cc:415
#21 OnReadPacket () at webrtc/src/p2p/base/stun_port.cc:407
#22 cricket::UDPPort::HandleIncomingPacket(rtc::AsyncPacketSocket*, char const*, unsigned long, rtc::SocketAddress const&, long) () at webrtc/src/p2p/base/stun_port.cc:348
#23 OnReadPacket () at webrtc/src/p2p/client/basic_port_allocator.cc:1673
#28 OnReadEvent () at webrtc/src/rtc_base/async_udp_socket.cc:132
#33 rtc::SocketDispatcher::OnEvent(unsigned int, int) () at webrtc/src/rtc_base/physical_socket_server.cc:790
#34 rtc::ProcessEvents(rtc::Dispatcher*, bool, bool, bool) () at webrtc/src/rtc_base/physical_socket_server.cc:1379
#35 WaitEpoll () at webrtc/src/rtc_base/physical_socket_server.cc:1620
#36 rtc::PhysicalSocketServer::Wait(int, bool) () at webrtc/src/rtc_base/physical_socket_server.cc:1328

在 webrtc/src/rtc_base/physical_socket_server.cc 文件里 PhysicalSocketServer
类的 Wait() 函数中，通过 epoll 机制等待网络数据包的到来。当数据包到来时，经过层层处理及传递，一直被传到 webrtc/src/pc/channel.cc 文件里 BaseChannel 类的 OnPacketReceived() 函数中，该函数又将数据包抛进另一个线程进行处理：

void BaseChannel::OnPacketReceived(bool rtcp,const rtc::CopyOnWriteBuffer& packet,int64_t packet_time_us) {if (!has_received_packet_ && !rtcp) {has_received_packet_ = true;signaling_thread()->Post(RTC_FROM_HERE, this, MSG_FIRSTPACKETRECEIVED);}if (!srtp_active() && srtp_required_) {// Our session description indicates that SRTP is required, but we got a// packet before our SRTP filter is active. This means either that// a) we got SRTP packets before we received the SDES keys, in which case//    we can't decrypt it anyway, or// b) we got SRTP packets before DTLS completed on both the RTP and RTCP//    transports, so we haven't yet extracted keys, even if DTLS did//    complete on the transport that the packets are being sent on. It's//    really good practice to wait for both RTP and RTCP to be good to go//    before sending  media, to prevent weird failure modes, so it's fine//    for us to just eat packets here. This is all sidestepped if RTCP mux//    is used anyway.RTC_LOG(LS_WARNING)<< "Can't process incoming "<< RtpPacketTypeToString(rtcp ? RtpPacketType::kRtcp: RtpPacketType::kRtp)<< " packet when SRTP is inactive and crypto is required";return;}invoker_.AsyncInvoke<void>(RTC_FROM_HERE, worker_thread_,Bind(&BaseChannel::ProcessPacket, this, rtcp, packet, packet_time_us));
}

2. 媒体引擎对收到的音频包的处理

#0  InsertPacketInternal () at webrtc/src/modules/audio_coding/neteq/neteq_impl.cc:467
#1  webrtc::NetEqImpl::InsertPacket(webrtc::RTPHeader const&, rtc::ArrayView<unsigned char const, -4711l>, unsigned int) () at webrtc/src/modules/audio_coding/neteq/neteq_impl.cc:153
#2  InsertPacket () at webrtc/src/modules/audio_coding/acm2/acm_receiver.cc:117
#3  IncomingPacket () at webrtc/src/modules/audio_coding/acm2/audio_coding_module.cc:667
#4  OnReceivedPayloadData () at webrtc/src/audio/channel_receive.cc:283
#5  ReceivePacket () at webrtc/src/audio/channel_receive.cc:669
#6  webrtc::voe::(anonymous namespace)::ChannelReceive::OnRtpPacket(webrtc::RtpPacketReceived const&) () at webrtc/src/audio/channel_receive.cc:622
#7  webrtc::RtpDemuxer::OnRtpPacket(webrtc::RtpPacketReceived const&) () at webrtc/src/call/rtp_demuxer.cc:177
#8  webrtc::RtpStreamReceiverController::OnRtpPacket(webrtc::RtpPacketReceived const&) () at webrtc/src/call/rtp_stream_receiver_controller.cc:54
#9  DeliverRtp () at webrtc/src/call/call.cc:1423
#10 DeliverPacket () at webrtc/src/call/call.cc:1461
#11 OnPacketReceived () at webrtc/src/media/engine/webrtc_voice_engine.cc:2070
#12 ProcessPacket () at webrtc/src/pc/channel.cc:540

webrtc/src/pc/channel.cc 文件里 BaseChannel 类的 ProcessPacket() 函数将收到的音频包送进媒体引擎进行处理，这一过程包括，根据 RTP 包的 ssrc 派发进不同的 channel，ACM receiver 的处理，一直到最终插入 NetEq 的缓冲区。在 NetEq 中将会完成数据包的重排序，网络对抗，音频的解码等处理操作。

音频数据的解码及播放

AudioDevice 组件被初始化时，即会启动一个播放线程，如 (webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc)：

AudioDeviceGeneric::InitStatus AudioDeviceLinuxPulse::Init() {RTC_DCHECK(thread_checker_.IsCurrent());if (_initialized) {return InitStatus::OK;}// Initialize PulseAudioif (InitPulseAudio() < 0) {RTC_LOG(LS_ERROR) << "failed to initialize PulseAudio";if (TerminatePulseAudio() < 0) {RTC_LOG(LS_ERROR) << "failed to terminate PulseAudio";}return InitStatus::OTHER_ERROR;}#if defined(WEBRTC_USE_X11)// Get X display handle for typing detection_XDisplay = XOpenDisplay(NULL);if (!_XDisplay) {RTC_LOG(LS_WARNING)<< "failed to open X display, typing detection will not work";}
#endif// RECORDING_ptrThreadRec.reset(new rtc::PlatformThread(RecThreadFunc, this,"webrtc_audio_module_rec_thread",rtc::kRealtimePriority));_ptrThreadRec->Start();// PLAYOUT_ptrThreadPlay.reset(new rtc::PlatformThread(PlayThreadFunc, this, "webrtc_audio_module_play_thread",rtc::kRealtimePriority));_ptrThreadPlay->Start();_initialized = true;return InitStatus::OK;
}

这个线程不断地从媒体引擎中拿数据进行播放，这个过程如下：

#0  GetNextAudioInterleaved () at webrtc/src/modules/audio_coding/neteq/sync_buffer.cc:85
#1  GetAudioInternal () at webrtc/src/modules/audio_coding/neteq/neteq_impl.cc:897
#2  GetAudio () at webrtc/src/modules/audio_coding/neteq/neteq_impl.cc:215
#3  GetAudio () at webrtc/src/modules/audio_coding/acm2/acm_receiver.cc:134
#4  PlayoutData10Ms () at webrtc/src/modules/audio_coding/acm2/audio_coding_module.cc:704
#5  GetAudioFrameWithInfo () at webrtc/src/audio/channel_receive.cc:332
#6  webrtc::internal::AudioReceiveStream::GetAudioFrameWithInfo(int, webrtc::AudioFrame*) () at webrtc/src/audio/audio_receive_stream.cc:276
#7  GetAudioFromSources () at webrtc/src/modules/audio_mixer/audio_mixer_impl.cc:186
#8  Mix () at webrtc/src/modules/audio_mixer/audio_mixer_impl.cc:130
#9  NeedMorePlayData () at webrtc/src/audio/audio_transport_impl.cc:193
#10 RequestPlayoutData () at webrtc/src/modules/audio_device/audio_device_buffer.cc:301
#11 PlayThreadProcess () at webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc:2121
#12 webrtc::AudioDeviceLinuxPulse::PlayThreadFunc(void*) () at webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc:1984

WebRTC 中对于音频，是即解码即播放的，播放和解码在同一个线程中完成，此外从解码到播放，还将完成回声消除，混音等处理。

上面创建 WebRtcAudioReceiveStream，并接收到音频数据包，是这里能够从 mixer 拿到数据并播放的基础。

webrtc::AudioSendStream 的创建

webrtc::AudioSendStream 的创建由应用程序发起：

#0  cricket::BaseChannel::SetLocalContent(cricket::MediaContentDescription const*, webrtc::SdpType, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >*) () at webrtc/src/pc/channel.cc:290
#1  PushdownMediaDescription () at webrtc/src/pc/peer_connection.cc:5699
#2  UpdateSessionState () at webrtc/src/pc/peer_connection.cc:5668
#3  ApplyLocalDescription () at webrtc/src/pc/peer_connection.cc:2356
#4  SetLocalDescription () at webrtc/src/pc/peer_connection.cc:2187
#10 Conductor::OnSuccess(webrtc::SessionDescriptionInterface*) () at webrtc/src/examples/peerconnection/client/conductor.cc:544
#11 OnMessage () at webrtc/src/pc/webrtc_session_description_factory.cc:299
#12 Dispatch () at webrtc/src/rtc_base/message_queue.cc:513
#13 ProcessMessages () at webrtc/src/rtc_base/thread.cc:527
#14 rtc::Thread::Run() () at webrtc/src/rtc_base/thread.cc:351
#15 main () at webrtc/src/examples/peerconnection/client/linux/main.cc:111

webrtc/src/pc/channel.cc 文件里的 BaseChannel::SetLocalContent() 将 MediaContentDescription 抛进 worker_thread 中进一步处理：

bool BaseChannel::SetLocalContent(const MediaContentDescription* content,SdpType type,std::string* error_desc) {TRACE_EVENT0("webrtc", "BaseChannel::SetLocalContent");return InvokeOnWorker<bool>(RTC_FROM_HERE,Bind(&BaseChannel::SetLocalContent_w, this, content, type, error_desc));
}

webrtc::AudioSendStream 最终在 Call 中创建：

#0  CreateAudioSendStream () at webrtc/src/call/call.cc:707
#1  WebRtcAudioSendStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:735
#2  AddSendStream () at webrtc/src/media/engine/webrtc_voice_engine.cc:1803
#3  UpdateLocalStreams_w () at webrtc/src/pc/channel.cc:671
#4  SetLocalContent_w () at webrtc/src/pc/channel.cc:906

音频数据的发送

如前面看到的，AudioDevice 组件被初始化时，在启动播放线程的同时，还会启动一个录制线程。录制线程捕获录制的音频数据，一路传递进行处理：

#0  ProcessAndEncodeAudio () at webrtc/src/audio/channel_send.cc:1101
#1  SendAudioData () at webrtc/src/audio/audio_send_stream.cc:365
#2  RecordedDataIsAvailable () at webrtc/src/audio/audio_transport_impl.cc:164
#3  DeliverRecordedData () at webrtc/src/modules/audio_device/audio_device_buffer.cc:269
#4  webrtc::AudioDeviceLinuxPulse::ProcessRecordedData(signed char*, unsigned int, unsigned int) ()at webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc:1971
#5  webrtc::AudioDeviceLinuxPulse::ReadRecordedData(void const*, unsigned long) ()at webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc:1918
#6  RecThreadProcess () at webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc:2229
#7  webrtc::AudioDeviceLinuxPulse::RecThreadFunc(void*) () at webrtc/src/modules/audio_device/linux/audio_device_pulse_linux.cc:1990

AudioDevice 将录制的音频数据一直传递到 webrtc/src/audio/channel_send.cc 文件里的 ChannelSend::ProcessAndEncodeAudio()：

void ChannelSend::ProcessAndEncodeAudio(std::unique_ptr<AudioFrame> audio_frame) {RTC_DCHECK_RUNS_SERIALIZED(&audio_thread_race_checker_);struct ProcessAndEncodeAudio {void operator()() {RTC_DCHECK_RUN_ON(&channel->encoder_queue_);if (!channel->encoder_queue_is_active_) {return;}channel->ProcessAndEncodeAudioOnTaskQueue(audio_frame.get());}std::unique_ptr<AudioFrame> audio_frame;ChannelSend* const channel;};// Profile time between when the audio frame is added to the task queue and// when the task is actually executed.audio_frame->UpdateProfileTimeStamp();encoder_queue_.PostTask(ProcessAndEncodeAudio{std::move(audio_frame), this});
}

在该函数中，录制获得的 AudioFrame 被抛进编码的线程中的 task 进行编码，封装为 RTP 包，并送进 PacedSender，PacedSender 将 RTP 包放进 queue 中：

#0  InsertPacket () at webrtc/src/modules/pacing/paced_sender.cc:200
#1  non-virtual thunk to webrtc::PacedSender::InsertPacket(webrtc::RtpPacketSender::Priority, unsigned int, unsigned short, long, unsigned long, bool) ()
#2  webrtc::voe::(anonymous namespace)::RtpPacketSenderProxy::InsertPacket(webrtc::RtpPacketSender::Priority, unsigned int, unsigned short, long, unsigned long, bool) () at webrtc/src/audio/channel_send.cc:391
#3  SendToNetwork () at webrtc/src/modules/rtp_rtcp/source/rtp_sender.cc:963
#4  webrtc::RTPSenderAudio::LogAndSendToNetwork(std::__1::unique_ptr<webrtc::RtpPacketToSend, std::__1::default_delete<webrtc::RtpPacketToSend> >, webrtc::StorageType) () at webrtc/src/modules/rtp_rtcp/source/rtp_sender_audio.cc:363
#5  SendAudio () at webrtc/src/modules/rtp_rtcp/source/rtp_sender_audio.cc:260
#6  SendRtpAudio () at webrtc/src/audio/channel_send.cc:568
#7  SendData () at webrtc/src/audio/channel_send.cc:497
#8  Encode () at webrtc/src/modules/audio_coding/acm2/audio_coding_module.cc:385
#9  webrtc::(anonymous namespace)::AudioCodingModuleImpl::Add10MsData(webrtc::AudioFrame const&) ()at webrtc/src/modules/audio_coding/acm2/audio_coding_module.cc:430
#10 ProcessAndEncodeAudioOnTaskQueue () at webrtc/src/audio/channel_send.cc:1152

PacedSender 中的另一个线程从 queue 中拿到 RTP 包并发送：

#0  SendPacket () at webrtc/src/pc/channel.cc:397
#1  cricket::BaseChannel::SendPacket(rtc::CopyOnWriteBuffer*, rtc::PacketOptions const&) () at webrtc/src/pc/channel.cc:328
#2  cricket::MediaChannel::DoSendPacket(rtc::CopyOnWriteBuffer*, bool, rtc::PacketOptions const&) () at webrtc/src/media/base/media_channel.h:328
#3  cricket::MediaChannel::SendPacket(rtc::CopyOnWriteBuffer*, rtc::PacketOptions const&) () at webrtc/src/media/base/media_channel.h:249
#4  cricket::WebRtcVideoChannel::SendRtp(unsigned char const*, unsigned long, webrtc::PacketOptions const&) () at webrtc/src/media/engine/webrtc_video_engine.cc:1690
#5  SendPacketToNetwork () at webrtc/src/modules/rtp_rtcp/source/rtp_sender.cc:550
#6  PrepareAndSendPacket () at webrtc/src/modules/rtp_rtcp/source/rtp_sender.cc:791
#7  webrtc::RTPSender::TimeToSendPacket(unsigned int, unsigned short, long, bool, webrtc::PacedPacketInfo const&) () at webrtc/src/modules/rtp_rtcp/source/rtp_sender.cc:604
#8  webrtc::ModuleRtpRtcpImpl::TimeToSendPacket(unsigned int, unsigned short, long, bool, webrtc::PacedPacketInfo const&) () at webrtc/src/modules/rtp_rtcp/source/rtp_rtcp_impl.cc:415
#9  webrtc::PacketRouter::TimeToSendPacket(unsigned int, unsigned short, long, bool, webrtc::PacedPacketInfo const&) ()at webrtc/src/modules/pacing/packet_router.cc:123
#10 Process () at webrtc/src/modules/pacing/paced_sender.cc:390

RTP 包被 PacedSender 一直递到 webrtc/src/pc/channel.cc 文件里定义的 BaseChannel::SendPacket()。BaseChannel::SendPacket() 将包抛进网络发送线程中发送：

bool BaseChannel::SendPacket(bool rtcp,rtc::CopyOnWriteBuffer* packet,const rtc::PacketOptions& options) {// Until all the code is migrated to use RtpPacketType instead of bool.RtpPacketType packet_type = rtcp ? RtpPacketType::kRtcp : RtpPacketType::kRtp;// SendPacket gets called from MediaEngine, on a pacer or an encoder thread.// If the thread is not our network thread, we will post to our network// so that the real work happens on our network. This avoids us having to// synchronize access to all the pieces of the send path, including// SRTP and the inner workings of the transport channels.// The only downside is that we can't return a proper failure code if// needed. Since UDP is unreliable anyway, this should be a non-issue.if (!network_thread_->IsCurrent()) {// Avoid a copy by transferring the ownership of the packet data.int message_id = rtcp ? MSG_SEND_RTCP_PACKET : MSG_SEND_RTP_PACKET;SendPacketMessageData* data = new SendPacketMessageData;data->packet = std::move(*packet);data->options = options;network_thread_->Post(RTC_FROM_HERE, this, message_id, data);return true;}

网络发送线程将数据包通过系统 socket 发送到网络：

#0  rtc::PhysicalSocket::DoSendTo(int, char const*, int, int, sockaddr const*, unsigned int) () at webrtc/src/rtc_base/physical_socket_server.cc:479
#1  SendTo () at webrtc/src/rtc_base/physical_socket_server.cc:344
#2  rtc::AsyncUDPSocket::SendTo(void const*, unsigned long, rtc::SocketAddress const&, rtc::PacketOptions const&) () at webrtc/src/rtc_base/async_udp_socket.cc:84
#3  SendTo () at webrtc/src/p2p/base/stun_port.cc:301
#4  Send () at webrtc/src/p2p/base/connection.cc:1162
#5  SendPacket () at webrtc/src/p2p/base/p2p_transport_channel.cc:1473
#6  SendPacket () at webrtc/src/p2p/base/dtls_transport.cc:409
#7  SendPacket () at webrtc/src/pc/rtp_transport.cc:147
#8  SendRtpPacket () at webrtc/src/pc/srtp_transport.cc:173
#9  SendPacket () at webrtc/src/pc/channel.cc:457
#10 OnMessage () at webrtc/src/pc/channel.cc:757

webrtc/src/rtc_base/physical_socket_server.cc 文件里定义的 PhysicalSocket::DoSendTo() 函数将数据包通过系统的 socket 发送到网络上：

int PhysicalSocket::DoSendTo(SOCKET socket,const char* buf,int len,int flags,const struct sockaddr* dest_addr,socklen_t addrlen) {return ::sendto(socket, buf, len, flags, dest_addr, addrlen);
}

WebRTC Audio 接收和发送的关键过程相关推荐

WebRTC音频系统音频发送和接收
文章目录 3.1音频数据流发送流程 3.2 发送中的编码.RTP打包 3.3 AudioSendStream类关系 3.4`webrtc::AudioSendStream` 创建和初始化 3.5 创建 ...
STM32F103RBT6 串口1正常接收，发送过程也很正常，但TXD引脚没有波形
STM32F103RBT6 串口1正常接收,发送过程也很正常,但TXD引脚没有波形,这个程序前几天还是正常工作,百思不得其解.后来找来开发板串口通讯例程对比发现GPIO初始化缺少一句,GPIO_Ini ...
webrtc audio
音频的基本概念采样频率单位时间内对模拟信号的采样次数.采样频率越高,声音的还原就越真实越自然,当然数据量就越大.采样率根据使用类型不同大概有以下几种: 8khz:电话等使用,对于记录人声已经足够使 ...
来自网页的消息服务器繁处理忙,EventSource 对象用于接收服务器发送事件通知,是网页自动获取来自服务器的更新...
//--------------------------------客户端代码----------------------------- if(typeof(EventSource) !== &quo ...
java使用Socket类接收和发送数据
java使用Socket类接收和发送数据网络应用分为客户端和服务端两部分,而Socket类是负责处理客户端通信的Java类.通过这个类可以连接到指定IP或域名的服务器上,并且可以和服务器互相发送和接 ...
CMM关键过程域剖析：需求管理
需求管理是CMM二级中列出的第一个关键域,这是因为它实际上是二级引入到开发过程中的所有管理原则的先决条件.只有在开发的目标被清楚明白地表述和理解的情况下,软件开发才能以一种有计划的有序的方式进行.实际 ...
TCP/IP传输层协议实现 - TCP接收窗口/发送窗口/通告窗口(lwip)
1.tcp通告窗口/接收窗口/发送窗口接收端有一个接收窗口大小,接收端只能接收这么多数据,接收窗口的数据需要被上层接收后才释放更大接收空间,才可以接收更多数据:接收窗口之前的数据已经被接收,再次接收 ...
domino服务器打开邮件,Domino邮件服务器配置(接收、发送).doc
Domino邮件服务器配置(接收.发送).doc Domino邮件服务器配置-多台 (mail+smtp) 用户使用可使用Pop3 客户端(如:Foxmail,Outlook 等)接收邮件,也可以使用 ...
c++ grpc 实现一个传图服务(异步方式，流式接收与发送)
~!转载请注明出处异步传输官方示例只给了普通Unary元对象的传输,没有流式传输示例,经过摸索调试,实现了grpc的异步流式传输(目前只是单向流,服务端推流至客户端,或者客户端上送流至服务端). 1 ...

WebRTC Audio 接收和发送的关键过程