webrtc_m130/pc/video_rtp_receiver.h

184 lines
6.7 KiB
C
Raw Normal View History

/*
* Copyright 2019 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.
*/
#ifndef PC_VIDEO_RTP_RECEIVER_H_
#define PC_VIDEO_RTP_RECEIVER_H_
#include <stdint.h>
#include <string>
#include <vector>
#include "absl/types/optional.h"
#include "api/crypto/frame_decryptor_interface.h"
#include "api/dtls_transport_interface.h"
#include "api/frame_transformer_interface.h"
#include "api/media_stream_interface.h"
#include "api/media_types.h"
#include "api/rtp_parameters.h"
#include "api/rtp_receiver_interface.h"
#include "api/scoped_refptr.h"
#include "api/sequence_checker.h"
#include "api/transport/rtp/rtp_source.h"
#include "api/video/video_frame.h"
#include "api/video/video_sink_interface.h"
#include "api/video/video_source_interface.h"
#include "media/base/media_channel.h"
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
#include "pc/jitter_buffer_delay.h"
#include "pc/media_stream_track_proxy.h"
#include "pc/rtp_receiver.h"
#include "pc/video_rtp_track_source.h"
#include "pc/video_track.h"
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
#include "rtc_base/system/no_unique_address.h"
#include "rtc_base/thread.h"
#include "rtc_base/thread_annotations.h"
namespace webrtc {
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
class VideoRtpReceiver : public RtpReceiverInternal {
public:
// An SSRC of 0 will create a receiver that will match the first SSRC it
// sees. Must be called on signaling thread.
VideoRtpReceiver(rtc::Thread* worker_thread,
std::string receiver_id,
std::vector<std::string> streams_ids);
// TODO(hbos): Remove this when streams() is removed.
// https://crbug.com/webrtc/9480
VideoRtpReceiver(
rtc::Thread* worker_thread,
const std::string& receiver_id,
const std::vector<rtc::scoped_refptr<MediaStreamInterface>>& streams);
virtual ~VideoRtpReceiver();
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
rtc::scoped_refptr<VideoTrackInterface> video_track() const { return track_; }
// RtpReceiverInterface implementation
rtc::scoped_refptr<MediaStreamTrackInterface> track() const override {
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
return track_;
}
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
rtc::scoped_refptr<DtlsTransportInterface> dtls_transport() const override;
std::vector<std::string> stream_ids() const override;
std::vector<rtc::scoped_refptr<MediaStreamInterface>> streams()
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
const override;
cricket::MediaType media_type() const override {
return cricket::MEDIA_TYPE_VIDEO;
}
std::string id() const override { return id_; }
RtpParameters GetParameters() const override;
void SetFrameDecryptor(
rtc::scoped_refptr<FrameDecryptorInterface> frame_decryptor) override;
rtc::scoped_refptr<FrameDecryptorInterface> GetFrameDecryptor()
const override;
void SetDepacketizerToDecoderFrameTransformer(
rtc::scoped_refptr<FrameTransformerInterface> frame_transformer) override;
// RtpReceiverInternal implementation.
void Stop() override;
void SetupMediaChannel(uint32_t ssrc) override;
void SetupUnsignaledMediaChannel() override;
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
uint32_t ssrc() const override;
void NotifyFirstPacketReceived() override;
void set_stream_ids(std::vector<std::string> stream_ids) override;
void set_transport(
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
rtc::scoped_refptr<DtlsTransportInterface> dtls_transport) override;
void SetStreams(const std::vector<rtc::scoped_refptr<MediaStreamInterface>>&
streams) override;
void SetObserver(RtpReceiverObserverInterface* observer) override;
void SetJitterBufferMinimumDelay(
absl::optional<double> delay_seconds) override;
void SetMediaChannel(cricket::MediaChannel* media_channel) override;
int AttachmentId() const override { return attachment_id_; }
std::vector<RtpSource> GetSources() const override;
// Combines SetMediaChannel, SetupMediaChannel and
// SetupUnsignaledMediaChannel.
void SetupMediaChannel(absl::optional<uint32_t> ssrc,
cricket::MediaChannel* media_channel);
private:
void RestartMediaChannel(absl::optional<uint32_t> ssrc)
RTC_RUN_ON(&signaling_thread_checker_);
void RestartMediaChannel_w(absl::optional<uint32_t> ssrc,
MediaSourceInterface::SourceState state)
RTC_RUN_ON(worker_thread_);
void SetSink(rtc::VideoSinkInterface<VideoFrame>* sink)
RTC_RUN_ON(worker_thread_);
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
void SetMediaChannel_w(cricket::MediaChannel* media_channel)
RTC_RUN_ON(worker_thread_);
// VideoRtpTrackSource::Callback
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
void OnGenerateKeyFrame();
void OnEncodedSinkEnabled(bool enable);
void SetEncodedSinkEnabled(bool enable) RTC_RUN_ON(worker_thread_);
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
class SourceCallback : public VideoRtpTrackSource::Callback {
public:
explicit SourceCallback(VideoRtpReceiver* receiver) : receiver_(receiver) {}
~SourceCallback() override = default;
private:
void OnGenerateKeyFrame() override { receiver_->OnGenerateKeyFrame(); }
void OnEncodedSinkEnabled(bool enable) override {
receiver_->OnEncodedSinkEnabled(enable);
}
VideoRtpReceiver* const receiver_;
} source_callback_{this};
RTC_NO_UNIQUE_ADDRESS SequenceChecker signaling_thread_checker_;
rtc::Thread* const worker_thread_;
const std::string id_;
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
cricket::VideoMediaChannel* media_channel_ RTC_GUARDED_BY(worker_thread_) =
nullptr;
absl::optional<uint32_t> ssrc_ RTC_GUARDED_BY(worker_thread_);
Reland "Use backticks not vertical bars to denote variables in comments for /pc" Original change's description: > Revert "Use backticks not vertical bars to denote variables in comments for /pc" > > This reverts commit 37ee0f5e594dd772ec6d620b5e5ea8a751b684f0. > > Reason for revert: Revert in order to be able to revert https://webrtc-review.googlesource.com/c/src/+/225642 > > Original change's description: > > Use backticks not vertical bars to denote variables in comments for /pc > > > > Bug: webrtc:12338 > > Change-Id: I88cf10afa5fc810b95d2a585ab2e895dcc163b63 > > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/226953 > > Reviewed-by: Harald Alvestrand <hta@webrtc.org> > > Commit-Queue: Artem Titov <titovartem@webrtc.org> > > Cr-Commit-Position: refs/heads/master@{#34575} > > TBR=hta@webrtc.org,titovartem@webrtc.org,webrtc-scoped@luci-project-accounts.iam.gserviceaccount.com > > Change-Id: I5eddd3a14e1f664bf831e5c294fbc4de5f6a88af > No-Presubmit: true > No-Tree-Checks: true > No-Try: true > Bug: webrtc:12338 > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/227082 > Reviewed-by: Björn Terelius <terelius@webrtc.org> > Commit-Queue: Björn Terelius <terelius@webrtc.org> > Cr-Commit-Position: refs/heads/master@{#34577} Bug: webrtc:12338 Change-Id: I96bd229b73613c162d11d75fa4f5934e1b4295c7 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/227087 Commit-Queue: Artem Titov <titovartem@webrtc.org> Reviewed-by: Harald Alvestrand <hta@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34611}
2021-07-30 22:30:23 +02:00
// `source_` is held here to be able to change the state of the source when
// the VideoRtpReceiver is stopped.
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
const rtc::scoped_refptr<VideoRtpTrackSource> source_;
const rtc::scoped_refptr<VideoTrackProxyWithInternal<VideoTrack>> track_;
std::vector<rtc::scoped_refptr<MediaStreamInterface>> streams_
RTC_GUARDED_BY(&signaling_thread_checker_);
RtpReceiverObserverInterface* observer_
RTC_GUARDED_BY(&signaling_thread_checker_) = nullptr;
bool received_first_packet_ RTC_GUARDED_BY(&signaling_thread_checker_) =
false;
const int attachment_id_;
rtc::scoped_refptr<FrameDecryptorInterface> frame_decryptor_
RTC_GUARDED_BY(worker_thread_);
rtc::scoped_refptr<DtlsTransportInterface> dtls_transport_
RTC_GUARDED_BY(&signaling_thread_checker_);
rtc::scoped_refptr<FrameTransformerInterface> frame_transformer_
RTC_GUARDED_BY(worker_thread_);
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
// Stores the minimum jitter buffer delay. Handles caching cases
Reland "Use backticks not vertical bars to denote variables in comments for /pc" Original change's description: > Revert "Use backticks not vertical bars to denote variables in comments for /pc" > > This reverts commit 37ee0f5e594dd772ec6d620b5e5ea8a751b684f0. > > Reason for revert: Revert in order to be able to revert https://webrtc-review.googlesource.com/c/src/+/225642 > > Original change's description: > > Use backticks not vertical bars to denote variables in comments for /pc > > > > Bug: webrtc:12338 > > Change-Id: I88cf10afa5fc810b95d2a585ab2e895dcc163b63 > > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/226953 > > Reviewed-by: Harald Alvestrand <hta@webrtc.org> > > Commit-Queue: Artem Titov <titovartem@webrtc.org> > > Cr-Commit-Position: refs/heads/master@{#34575} > > TBR=hta@webrtc.org,titovartem@webrtc.org,webrtc-scoped@luci-project-accounts.iam.gserviceaccount.com > > Change-Id: I5eddd3a14e1f664bf831e5c294fbc4de5f6a88af > No-Presubmit: true > No-Tree-Checks: true > No-Try: true > Bug: webrtc:12338 > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/227082 > Reviewed-by: Björn Terelius <terelius@webrtc.org> > Commit-Queue: Björn Terelius <terelius@webrtc.org> > Cr-Commit-Position: refs/heads/master@{#34577} Bug: webrtc:12338 Change-Id: I96bd229b73613c162d11d75fa4f5934e1b4295c7 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/227087 Commit-Queue: Artem Titov <titovartem@webrtc.org> Reviewed-by: Harald Alvestrand <hta@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34611}
2021-07-30 22:30:23 +02:00
// if `SetJitterBufferMinimumDelay` is called before start.
VideoRtpReceiver & AudioRtpReceiver threading fixes. For implementations where the signaling and worker threads are not the same thread, this significantly cuts down on Thread::Invoke()s that would block the signaling thread while waiting for the worker thread. For Audio and Video Rtp receivers, the following methods now do not block the signaling thread: * GetParameters * SetJitterBufferMinimumDelay * GetSources * SetFrameDecryptor / GetFrameDecryptor * SetDepacketizerToDecoderFrameTransformer Importantly this change also makes the track() accessor accessible directly from the application thread (bypassing the proxy) since for receiver objects, the track object is const. Other changes: * Remove RefCountedObject inheritance, use make_ref_counted instead. * Every member variable in the rtp receiver classes is now RTC_GUARDED * Stop() now fully clears up worker thread state, and Stop() is consistently called before destruction. This means that there's one thread hop instead of at least 4 before (sometimes more), per receiver. * OnChanged triggered volume for audio tracks is done asynchronously. * Deleted most of the JitterBufferDelay implementation. Turns out that it was largely unnecessary overhead and complexity. It seems that these two classes are copy/pasted to a large extent so further refactoring would be good in the future, as to not have to fix each issue twice. Bug: chromium:1184611 Change-Id: I1ba5c3abbd1b0571f7d12850d64004fd2d83e5e2 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218605 Commit-Queue: Tommi <tommi@webrtc.org> Reviewed-by: Markus Handell <handellm@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34022}
2021-05-17 14:50:10 +02:00
JitterBufferDelay delay_ RTC_GUARDED_BY(worker_thread_);
Reland "Use backticks not vertical bars to denote variables in comments for /pc" Original change's description: > Revert "Use backticks not vertical bars to denote variables in comments for /pc" > > This reverts commit 37ee0f5e594dd772ec6d620b5e5ea8a751b684f0. > > Reason for revert: Revert in order to be able to revert https://webrtc-review.googlesource.com/c/src/+/225642 > > Original change's description: > > Use backticks not vertical bars to denote variables in comments for /pc > > > > Bug: webrtc:12338 > > Change-Id: I88cf10afa5fc810b95d2a585ab2e895dcc163b63 > > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/226953 > > Reviewed-by: Harald Alvestrand <hta@webrtc.org> > > Commit-Queue: Artem Titov <titovartem@webrtc.org> > > Cr-Commit-Position: refs/heads/master@{#34575} > > TBR=hta@webrtc.org,titovartem@webrtc.org,webrtc-scoped@luci-project-accounts.iam.gserviceaccount.com > > Change-Id: I5eddd3a14e1f664bf831e5c294fbc4de5f6a88af > No-Presubmit: true > No-Tree-Checks: true > No-Try: true > Bug: webrtc:12338 > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/227082 > Reviewed-by: Björn Terelius <terelius@webrtc.org> > Commit-Queue: Björn Terelius <terelius@webrtc.org> > Cr-Commit-Position: refs/heads/master@{#34577} Bug: webrtc:12338 Change-Id: I96bd229b73613c162d11d75fa4f5934e1b4295c7 Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/227087 Commit-Queue: Artem Titov <titovartem@webrtc.org> Reviewed-by: Harald Alvestrand <hta@webrtc.org> Cr-Commit-Position: refs/heads/master@{#34611}
2021-07-30 22:30:23 +02:00
// Records if we should generate a keyframe when `media_channel_` gets set up
// or switched.
bool saved_generate_keyframe_ RTC_GUARDED_BY(worker_thread_) = false;
bool saved_encoded_sink_enabled_ RTC_GUARDED_BY(worker_thread_) = false;
};
} // namespace webrtc
#endif // PC_VIDEO_RTP_RECEIVER_H_