This makes the channel manager object into a factory, not a manager.
Bug: webrtc:13931
Change-Id: I59f7d818a739797a7c0a7a32e6583450834df122
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260467
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36718}
This calls out the fact that SetChannel() is only used on M-section activation; ClearChannel is called on deactivation, and we never change the channel while a transceiver is active.
Bug: webrtc:13931
Change-Id: I3a3bfeec7c1d27d98c3f94a9401bee2130754ed7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260461
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36709}
Fixing a race condition where session.sampleRate changes before AudioDeviceIOS::HandleValidRouteChange() finishes.
session.sampleRate is read into session_sample_rate at 576 and used at 623 to initialize the audio unit. However, in the call to SetupAudioBuffersForActiveAudioSession() the session.sampleRate is read again and may have changed, resulting in different sample rates used for the buffers and the audio unit. The consequence is a sample rate mismatch with either high pitched or low pitched audio.
The fix is to always use the buffer sample rate for the audio unit.
The DCHECK at 622 would save us in debug, but not in production, hence removed.
Change-Id: I562f1bf7f94d7447139ada2636b02ade7fcd6371
Bug: webrtc:14011
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260329
Reviewed-by: Henrik Andreasson <henrika@google.com>
Commit-Queue: Peter Hanspers <peterhanspers@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36708}
This CL replaces two booleans, that could never be active at the same
time (there is no such thing as an abandoned chunk that is scheduled
for retransmission), with a single enum.
Just for increased readability, and to understand that there is no such
thing as an abandoned chunk that will be retransmitted.
Bug: None
Change-Id: I1682c383aed692db07fd4ae1f84c0166db86c062
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/259864
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36707}
(Used to run gtests on iOS, but we don't want to depend on //base.)
Optimistically try to use the existing partial fork
Bug: webrtc:13402
Change-Id: I10528670862f2db67e77adb73b8a71b19642666d
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260328
Commit-Queue: Björn Terelius <terelius@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36703}
With MSVC, compiling with AVX2 support will also result in the
generation of BMI2 instructions. Some early Haswell CPUs reportedly support AVX2 but not BMI2. We have seen crashes (illegal instruction)
on these CPUs in our AVX2 code paths on MULX instructions (part of BMI2).
Including a check for BMI2 when checking for AVX2 support is
expected to solve the issue.
Please see the bug referenced below for more background on this issue.
Bug: chromium:1315519
Change-Id: I3a0a9838f1f632704ba505ecbb81a6f8b1889319
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260323
Commit-Queue: Gustaf Ullberg <gustaf@webrtc.org>
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Reviewed-by: Henrik Andreasson <henrika@google.com>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36701}
This makes it clearer which modules set the channel.
Also remove GetChannel() from PeerConnection public API
This was only used once, internally, and can better be inlined.
Part of reducing the exposure of Channel.
Bug: webrtc:13931
Change-Id: I5f44865230a0d8314d269c85afb91d4b503e8de0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260187
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36695}
This is achieved by wrapping a fake decoder inside the mock decoder, in
a sort of spy pattern.
This is preperation for moving the FrameBufferProxy tests into the main
VideoReceiveStream2 suite.
Bug: webrtc:14003
Change-Id: I7b9691cc5a1a8a3fadfb7aa6981752b647d5c73f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260113
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36691}
This is mainly an issue when sending items with partial reliability,
with high bandwidth on a link with packet loss.
In SCTP, when a fragment isn't included in the SACK a number of times,
it's scheduled to be retransmitted or abandoned, if it has been
retransmitted too many times already (depending configuration). Before
this CL, if a fragment was NACKed and scheduled for retransmission, but
couldn't be retransmitted immediately (due to congestion window not
allowing it), future received SACKs - that would still indicate that the
fragment hasn't been received yet - would still increment the
retransmission counter. Which wasn't fair, because this fragment hasn't
had a chance to be retransmitted yet.
With this CL, the fragment will only have its retransmission counter
increased when it is not already scheduled to be retransmitted, but
actually sent on the wire and considered in-flight again.
Bug: webrtc:12943
Change-Id: I2af08255650221c044cc14591a5835c885e94c58
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/259825
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36683}
Remove the `started_` member variable and some other minor updates to
follow conventions elsewhere in the code.
Bug: none
Change-Id: I4cbb914b39cb2e2787719b906ca937931dc3dad7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/258360
Reviewed-by: Markus Handell <handellm@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36677}
This changes most deletion paths of Connection objects to go through
the owner class of the Connection instances, Port.
In situations where Connection objects still need to be deleted
asynchronously, `async = true` can be passed to
`Port::DestroyConnection` and get the same behavior as
`Connection::Destroy` formerly gave.
The `Destroy()` method still exists for downstream compatibility, but
instead of deleting connection objects asynchronously, the deletion
now happens synchronously via the Port class.
Bug: webrtc:13892, webrtc:13865
Change-Id: I07edb7bb5e5d93b33542581b4b09def548de9e12
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/259826
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36676}
This is an implementation API, user classes should in principle
only use the channel_interface.h
Bug: webrtc:13931
Change-Id: I85c285217858dc087c90a50792e980f731f4439f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260185
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36674}
This will make it easier to extend testing, implement new features (e.g.
packet culling) and experiment with new variants.
Bug: webrtc:11340
Change-Id: I747f5f6cff61e11a420e43b06ffe0c4aba438c7b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260116
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36670}
Prior to this CL, calling RtpTransceiver::SetChannel() with null
arguments would cause the receiver's track to end. This is wrong,
because the channel can be nulled for other reasons than the transceiver
being stopped/removed - such as when the transceiver is rolled back but
still in use. Also, stopping a transceiver will end the track, so we
should simply ensure to always stop the transceiver when that is needed.
This CL makes sure that the transceiver is stopped or stopping in all
appropriate places, allowing us to remove the ability to end the source
for any other reason. A side-effect of this is that:
- The track never ends prematurely, fixing https://crbug.com/1315611.
- Removed transceivers are always stopped, fixing
https://crbug.com/webrtc/14005.
This CL fixes the issue of track being ended in the ontrack event when
running https://jsfiddle.net/henbos/nxebusjm/.
- We don't have WPT test coverage for this, so I'll add that separately.
With SetSourceEnded() removed, some stopping/stop in response to
rejecting locally SDP munged content had to be added in order not to
regress the existing test coverage for this:
*PeerConnectionInterfaceTest.RejectMediaContent/1
Bug: chromium:1315611, webrtc:14005.
Change-Id: I21f30a1259e51324066dc84f72a72485b9e0fadc
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/260180
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#36669}
