This reverts commit 9a0662ac7e.
Reason for revert: breaks some av sync perf tests
Original change's description:
> Use only first payload timestamp for RTCP SR generation for audio
>
> Since now RTP rate is set correctly for audio, there's no need to
> use the very last data packet rtp/capture timestamps for generating
> RTCP SR packets.
>
> Using only one (first) packet timestamp eliminates the jitter between
> rtp and capture timestamps for audio. This jitter comes from the fact
> that capture timestamp for audio is unknown and we generate bogus
> timestamp at arbitrary, non-constant offset from the real capture time.
>
> Bug: webrtc:9905
> Change-Id: I855556184cfe994be39ab7780836a050f5a38c35
> Reviewed-on: https://webrtc-review.googlesource.com/c/108580
> Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
> Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
> Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#25430}
TBR=danilchap@webrtc.org,ilnik@webrtc.org,ossu@webrtc.org
Change-Id: I208a659379b1075258ee94613e42afd9aebe4754
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:9905
Reviewed-on: https://webrtc-review.googlesource.com/c/108623
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25435}
Since now RTP rate is set correctly for audio, there's no need to
use the very last data packet rtp/capture timestamps for generating
RTCP SR packets.
Using only one (first) packet timestamp eliminates the jitter between
rtp and capture timestamps for audio. This jitter comes from the fact
that capture timestamp for audio is unknown and we generate bogus
timestamp at arbitrary, non-constant offset from the real capture time.
Bug: webrtc:9905
Change-Id: I855556184cfe994be39ab7780836a050f5a38c35
Reviewed-on: https://webrtc-review.googlesource.com/c/108580
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25430}
This change corrects a potential race condition when updating a FrameEncryptor
for the audio send channel. If a FrameEncryptor is set on an active audio
stream it is possible for the current FrameEncryptor attached to the audio channel to be deallocated due to
the FrameEncryptors reference count reaching zero before the new FrameEncryptor is set on the
channel.
To address this issue the ChannelSend is now holds a scoped_reftptr<FrameEncryptor>
to only allow deallocation when it is actually set on the encoder queue.
ChannelSend is unique in this respect as the Audio Receiver a long with the
Video Sender and Video Receiver streams all recreate themselves when they have
a configuration change. ChannelSend instead reconfigures itself using the
existing channel object.
Added Seth as TBR as this only introduces mocks.
TBR=shampson@webrtc.org
Bug: webrtc:9907
Change-Id: Ibf391dc9cecdbed1874e0252ff5c2cb92a5c64f4
Reviewed-on: https://webrtc-review.googlesource.com/c/107664
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Fredrik Solenberg <solenberg@webrtc.org>
Reviewed-by: Qingsi Wang <qingsi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25374}
This change adds a new subcategory to the public native webrtc::CryptoOptions
structure: webrtc::CryptoOptions::Frame.
This new structure has a single off by default property:
crypto_options.frame.require_frame_encryption.
This new flag if set prevents RtpSenders from sending outgoing payloads unless
a frame_encryptor_ is attached and prevents RtpReceivers from receiving
incoming payloads unless a frame_decryptor_ is attached.
This option is important to enforce no unencrypted data can ever leave the
device or be received.
I have also attached bindings for Java and Objective-C.
I have implemented this functionality for E2EE audio but not E2EE video
since the changes are still in review.
Bug: webrtc:9681
Change-Id: Ie184711190e0cdf5ac781f69e9489ceec904736f
Reviewed-on: https://webrtc-review.googlesource.com/c/105540
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25238}
This change integrates the FrameDecryptorInterface and the FrameEncryptorInterface into
the audio media path. If a FrameEncryptorInterface is set on an outgoing audio RTPSender
then each outgoing audio payload will first pass through the provided FrameEncryptor which
will have a chance to modify the payload contents for the purposes of encryption.
If a FrameDecryptorInterface is set on an incoming audio RtpReceiver then each incoming
audio payload will first pass through the provided FrameDecryptor which have a chance to
modify the payload contents for the purpose of decryption.
While AEAD is supported by the FrameDecryptor/FrameEncryptor interfaces this CL does not
use it and so it is left as null.
Bug: webrtc:9681
Change-Id: Ic383a9dce280528739f9d271357c2220e0a0dccf
Reviewed-on: https://webrtc-review.googlesource.com/c/101702
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Fredrik Solenberg <solenberg@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Emad Omara <emadomara@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25001}
