This reverts the async operation introduced here:
https://webrtc-review.googlesource.com/c/src/+/248170
The race that happened was that the "flush" operation in the dtor
of ChannelManager, could run _after_ PeerConnection::Close() which
is where the Call object gets deleted. Inside the dtor of Call, there
are DCHECKs that could hit when the pending deletions hadn't run.
In most cases the Invoke() that is used to delete the Call object
would run after the pending tasks, but there's still one code path
that I'm looking for that could trigger the deletion of a channel
after Call is destructed.
Bug: webrtc:11992, webrtc:13540, chromium:1291383
Change-Id: I160742907cc0c097a4b2bb1b7c3da03b4e8cd8d8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249780
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35822}
Thanks to machenbach@ for the heads up.
WebRTC's MB was generating a Swarming command that was using vpython
(defaulting on python2 on some platforms). This CL switches that to
vpython3 (fixing gtest-parallel-wrapper.py to be python3 compliant).
No-Presubmit: True
Bug: webrtc:13607
Change-Id: Icfa7d23b81e30cebfe8243d4ba65284955593465
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249604
Reviewed-by: Christoffer Jansson <jansson@google.com>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35821}
This reverts commit 9d230d54c7eef31ac1100f0aeef1374dd1ac62fa.
Reason for revert: Speculative revert to see if it's the cause of a few perf changes (some bad, some not so bad).
Bug: webrtc:13613
Original change's description:
> (Un/)Subscribe RtpVideoSender for feedback on the transport queue.
>
> * RtpVideoSender now registers/unregisters for feedback callback
> inside of SetActive(), which runs on the transport queue.
> * Transport feedback is given on the transport queue
> * Registration/unregistration for feedback is done on the same
> * Removed the last mutex from TransportFeedbackDemuxer.
>
> Ultimately, this work is related to moving state from the Call
> class, that's related to network configuration, but due to the code
> is currently written is attached to the worker thread, over to the
> Transport, where it's used (e.g. suspended_video_send_ssrcs_).
>
> Bug: webrtc:13517, webrtc:11993
> Change-Id: I057d0e2597e6cb746b335e0308599cd547350e56
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248165
> Reviewed-by: Erik Språng <sprang@webrtc.org>
> Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#35777}
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: webrtc:13517, webrtc:11993
Change-Id: I824623b3b1c14f0ca7049a2a0890c6d97b7fb608
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249600
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35815}
In SCTP, sending ACKs at the right time is presumably important in order
not to cause unnecessary retransmissions.
This CL sets the ACK timers to use high timer precision. This unblocks
experimentally lowering the precision of the default, "low", timers in
WebRTC.
// All bots are green, but mac_chromium_compile is randomly timing out
// independently of this CL and has been doing so for several days...
NOTRY=True
Bug: webrtc:13604
Change-Id: I1c81e3d0eeb477c3e00277d35649114c5edd249c
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249090
Reviewed-by: Victor Boivie <boivie@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35813}
Context: The timer precision of PostDelayedTask() is about to be lowered
to include up to 17 ms leeway. In order not to break use cases that
require high precision timers, PostDelayedHighPrecisionTask() will
continue to have the same precision that PostDelayedTask() has today.
webrtc::TaskQueueBase has an enum (kLow, kHigh) to decide which
precision to use when calling PostDelayedTaskWithPrecision().
See go/postdelayedtask-precision-in-webrtc for motivation and a table of
delayed task use cases in WebRTC that are "high" or "low" precision.
Most timers in DCSCTP are believed to only be needing low precision (see
table), but the delayed_ack_timer_ of DataTracker[1] is an example of a
use case that is likely to break if the timer precision is lowered (if
ACK is sent too late, retransmissions may occur). So this is considered
a high precision use case.
This CL makes it possible to specify the precision of dcsctp::Timer.
In a follow-up CL we will update delayed_ack_timer_ to kHigh precision.
[1] https://source.chromium.org/chromium/chromium/src/+/main:third_party/webrtc/net/dcsctp/rx/data_tracker.cc;l=340
Bug: webrtc:13604
Change-Id: I8eec5ce37044096978b5dd1985fbb00bc0d8fb7e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249081
Reviewed-by: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35809}
This is a test-only class that today has high precision for its frames.
We intend to make the default task queue precision lower, so high
precision use cases have to opt-in to kHigh if they want to continue to
use high precision.
Bug: webrtc:13604
Change-Id: I25babec2a64d91d45548ad017200b806a60efe11
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249362
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35808}
This std::map was used to look up audio streams from ssrcs when
creating/destroying and/or modifying streams. Those operations aren't
frequent enough to justify having a separate lookup map. Removing
the variable, simplifies the thread ownership work a bit.
Bug: webrtc:11993
Change-Id: I94f9f2f56c138051a8b9c5f6a6d7cae3a4e78b48
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249091
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35806}
This makes fuzzer test cases fail if there's an assert failure in
the helper functions called by the test.
Bug: None
Change-Id: Ic187d72b8d4e016659a68a7bdcaadb78ab2aab05
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/246400
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35804}
This is a delegate that is used by video_receive_stream2 to handle frame
buffer tasks like threading, and stats. This will be used in a follow up
to use FrameBuffer3 as a strategy selected by field trial.
Unit-tests will be used in follow-up CLs containing Frame Buffer 3, and
are expected to work with both Frame buffer proxy versions.
Change-Id: I524279343d60a348d044d9085d618f12d7bf3a23
Bug: webrtc:13343
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/241605
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35803}
This is a safe cleanup change since top-level const applied to
parameters in function declarations (that are not also
definitions) are ignored by the compiler. Hence, such changes do
not change the type of the declared functions and are simply
no-ops.
Bug: webrtc:13610
Change-Id: Ibafb92c45119a6d8bdb6f9109aa8dad6385163a9
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249086
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Ali Tofigh <alito@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35802}
Just applied a short sed script. See bug description for
the motiviation for this change.
This is the command that was used to generate the changes:
$ find . -type f \( -iname '*.cc' -o -iname '*.h' \) -print0 | \
xargs -0 sed -i -e 's/(const override)/(const, override)/'
Bug: webrtc:13090
Change-Id: Iec7d280f9d55263a972dbb3bd644ebfcd2eb38cf
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249088
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Ali Tofigh <alito@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35801}
This CL adds a SequenceChecker, receive_11993_checker_, specifically for
variables that need to move to the network thread. Once migrated,
the checker will be replaced with a check for the network thread.
In the meantime, the checker will match with one of worker [x]or
network threads.
As a first step, this checker is used to isolate access to
`receive_rtp_config_` which is used from object factory paths (Create/
Destroy routines) as well as paths that handle network packets.
Bug: webrtc:11993
Change-Id: Ia58423583cf99492018f218eb1640535e3919193
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249080
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35800}
Move deletion of channel objects over to the RtpTransceiver instead
of having it done by SdpOfferAnswer.
The deletion is now also done via PostTask rather than Invoke.
Bug: webrtc:11992, webrtc:13540
Change-Id: I5aff14956d5e572ca8816bbfef8739bb609b4484
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248170
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35798}
Some of the state that's managed in VideoStreamEncoder, is updated
and accessed on the encoder queue, but a method that's used for testing
only (GetAdaptationResources()), represents a race between PostTask
operations that update state and checking for said state on the worker
thread.
This CL removes Wait() operations related to adaptation resources from
the common path and puts one in the test path instead.
Bug: webrtc:13612
Change-Id: Ie3e018e815e24951bc0634ed70de17eaf336a508
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249220
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35797}
This adds a field trial to change the pacing rate to pace
at a rate relative to max(bwe ,lower link capacity)
Bug: none
Change-Id: Ibe9ef3e08eb422e9abff6488780e82188958eeeb
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248865
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35795}
Avoid implicit conversion from rtc::scoped_refptr to raw pointer, when
calling the MethodCall and ConstMethodCall constructors. Fixed by
defining a private accessor T* c() which always returns a raw pointer,
with different implementation for BEGIN_PROXY_MAP and
BEGIN_OWNED_PROXY_MAP.
Avoid using the same name for macro argument and function argument
within the macro definition, like `c` in:
```
#define PRIMARY_PROXY_MAP_BOILERPLATE(c) \
protected: \
c##ProxyWithInternal(rtc::Thread* primary_thread, INTERNAL_CLASS* c) \
: primary_thread_(primary_thread), c_(c) {} \
```
Fixed by using `class_name` as the macro argument throughout.
Bug: webrtc:13464
Change-Id: I9faf095b004afd5225f5ec09fb782ef4b5c9dc5d
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/249089
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35794}
This unblocks lowering the precision of low precision tasks which are
the default.
Bug: webrtc:13604
Change-Id: Icd663cbbf5b0bf87ac83a4a0abd58699e6e27e8f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248862
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35782}
Use cases of TaskQueue or TaskQueueBase that are considered high
precision are updated to make use of PostDelayedHighPrecisionTask
(see go/postdelayedtask-precision-in-webrtc) instead of PostDelayedTask.
The cases here are the ones covered by that document, plus some
testing-only uses. The FrameBuffer2 and DataTracker use cases will
be covered by separate CLs because FrameBuffer2 uses
RepeatingTaskHandle and DataTracker uses dcsctp::Timer.
This protects these use cases against regressions when PostDelayedTask
gets its precision lowered.
This CL also adds TaskQueue::PostDelayedHighPrecisionTask which calls
TaskQueueBase::PostDelayedHighPrecisionTask (same pattern as for
PostDelayedTask).
Bug: webrtc:13604
Change-Id: I7dcab59cbe4d274d27b734ceb4fc06daa12ffd0f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248864
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35781}
Intended to be compatible with absl::StrSplit, but without the binary
cost of that dependency.
Bug: webrtc:13579
Change-Id: I167726903d74b8d5f299886cfb3e5d60610ddb93
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/247185
Reviewed-by: Ali Tofigh <alito@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35780}
* RtpVideoSender now registers/unregisters for feedback callback
inside of SetActive(), which runs on the transport queue.
* Transport feedback is given on the transport queue
* Registration/unregistration for feedback is done on the same
* Removed the last mutex from TransportFeedbackDemuxer.
Ultimately, this work is related to moving state from the Call
class, that's related to network configuration, but due to the code
is currently written is attached to the worker thread, over to the
Transport, where it's used (e.g. suspended_video_send_ssrcs_).
Bug: webrtc:13517, webrtc:11993
Change-Id: I057d0e2597e6cb746b335e0308599cd547350e56
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248165
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35777}
See go/postdelayedtask-precision-in-webrtc for context of which use
cases are considered "high" or "low". Most use cases are "low" which
is the default, but this CL allows opting in to "high".
Will be used by FrameBuffer2.
Bug: webrtc:13604
Change-Id: Iebf6eea44779873e78746da749a39e1101b92819
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248861
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35776}
Before, this call was being made from the SendPacket path of the
pacer. The transport will post a task to the transport queue regardless
so this change moves the lock inside of the demuxer away from the
pacer and over to the transport queue.
Moving forward, the calls to register/unregister with the feedback
demuxer, will occur on the transport queue as well and we can change
the transport OnTransportFeedback() implementation to forward the
calls to the demuxer on the transport queue as well. That will bring
all calls into the same execution context and we won't need a lock.
Bug: webrtc:13517, webrtc:11993
Change-Id: If714ca6d2b164a1a2b6bcb8c99787372064a31a0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/248164
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35775}
Move StopPermanentlyAndGetRtpStates closer to being the last step of
the destruction process.
Bug: webrtc:11993
Change-Id: I83d86c505b05f5c10d0ce802494baba9aa645027
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/239182
Auto-Submit: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#35774}
