As documented in webrtc:11908 this cleanup is fairly invasive and
when a part of a frequently executed code path, can be quite costly
in terms of performance overhead. This is currently the case with
synchronous calls between threads (Thread) as well with our proxy
api classes.
With this CL, all code in WebRTC should now either be using MessageHandlerAutoCleanup
or calling MessageHandler(false) explicitly.
Next steps will be to update external code to either depend on the
AutoCleanup variant, or call MessageHandler(false).
Changing the proxy classes to use TaskQueue set of concepts instead of
MessageHandler. This avoids the perf overhead related to the cleanup
above as well as incompatibility with the thread policy checks in
Thread that some current external users of the proxies would otherwise
run into (if we were to use Thread::Send() for synchronous call).
Following this we'll move the cleanup step into the AutoCleanup class
and an RTC_DCHECK that all calls to the MessageHandler are setting
the flag to false, before eventually removing the flag and make
MessageHandler pure virtual.
Bug: webrtc:11908
Change-Id: Idf4ff9bcc8438cb8c583777e282005e0bc511c8f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/183442
Reviewed-by: Artem Titov <titovartem@webrtc.org>
Commit-Queue: Tommi <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32049}
Send() was creating an instance of AutoThread for every call,
which is equivalent of instantiatiating a whole new instance of
Thread (AutoThread inherits from Thread) and not just ensuring that
a thread instance is registered for the current thread, as the
comments indicated.
Bug: webrtc:11908
Change-Id: I8bbb43ca83c30d9f5e1928205b3611271ecad053
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/183441
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Tommi <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32037}
Original patch contributed by andrey.semashev@gmail.com.
In PhysicalSocketServer::WaitEpoll(), the loop verifies that the
signalled dispatcher is in dispatchers_ set. It does so by looking up
the dispatcher pointer in the set. This is vulnerable to the ABA
problem because one dispatcher may be removed and destroyed and another
created and added with the same address before epoll reports an event
for the old dispatcher. The same issue exists for other Wait
implementations, if a dispatcher is removed and a new one added with
the same socket handle is the old.
This is avoided by using a 64-bit key for looking up the dispatcher
in the set. The key is set from a running counter which gets incremented
when a dispatcher is added to the set, so even if the same dispatcher
pointer is added, removed and added again, the key value will be
different.
This changes the storage of dispatchers_ from a set to a flat_hash_map,
which uses a bit more memory but has faster lookup (O(1) as opposed to
O(log n)).
Bug: webrtc:11124
Change-Id: I6d206e1a367b58ba971edca9b48af7664384b797
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/181027
Commit-Queue: Taylor <deadbeef@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#32019}
This test was added in
https://webrtc-review.googlesource.com/c/src/+/180620. On my machine it
passes about 98% of the time.
The test is meant to count the number of times the callback that the
operations chain is empty has been invoked, and does this by ensuring
the last operation to make the chain empty has completed before
expecting that the counter has increased.
The race happns when the operation has completed but the callback that
the chain is empty has not happened yet. This CL fixes that by using
EXPECT_EQ_WAIT instead.
TBR=hta@webrtc.org
Bug: chromium:1060083
Change-Id: I2ebfac3e635ef895d6602f7360e5ec6006fc1d0a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/182541
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31992}
This CL generates "negotiationneeded" events if negotiation is needed
when the Operations Chain becomes empty. This is only implemented in
Unified Plan to avoid Plan B regressions (the event is pretty useless
in Plan B as it fires repeatedly).
In order to implement the spec-compliant behavior of only firing the
event when the chain is empty, this CL introduces
PeerConnectionObserver::OnNegotiationNeededEvent() and
PeerConnectionInterface::ShouldFireNegotiationNeededEvent() to allow
validating the event before firing it. This is needed because the event
must not be fired until a task has been posted and subsequently chained
operations could invalidate it in the meantime.
Test coverage is added for both legacy and modern "negotiationneeded"
events.
Bug: chromium:1060083
Change-Id: I1dbaa8f6ddb1c6e7c8abd8da3b92efcb64060383
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180620
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31989}
Notably, this should detect whether an interface is "available" or not,
which should prevent the failure is with dual SIM card setups.
This is gated behind a field trial for now, to ensure this doesn't cause
any regressions due to false negatives (interfaces that are usable
but not listed as available by NWPathMonitor).
Bug: webrtc:10966
Change-Id: Ia3942c4c57b525d08d8b340e2325f3705cfd0304
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180923
Commit-Queue: Taylor <deadbeef@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Reviewed-by: Anders Carlsson <andersc@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31977}
It's now passed through CreatePeerConnectionFactory like all similar
injectable dependencies.
Also removing network_monitor.h from native_api/base, since clients
should all be moved to the new location now.
Bug: webrtc:9883
Change-Id: I3776262ce2a2a380d01c163f4d18a0dfad4b5b41
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/181401
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Commit-Queue: Taylor <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31931}
This patch adds an interface for os/firmware to set a network
preference NOT_PREFERRED / NEUTRAL that can be picked up by
an IceController and used when selection ice candidate pair.
The patch exposes this using an Android Intent based interface.
BUG: webrtc:11825
Change-Id: Ic12b6bf704fde7f9c912020dd7bc79ccae4613ab
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180883
Commit-Queue: Jonas Oreland <jonaso@webrtc.org>
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31877}
This is a reland of 003c9be817817ed0e3aef3f50c78ae5cb31bc0ff
Found some downstream code that relies on
NetworkMonitorFactory::SetFactory, so I'm adding those methods back
temporarily. BasicNetworkManager will fall back to the static factory
if the one passed into PeerConnectionFactory is null.
Original change's description:
> Pass NetworkMonitorFactory through PeerConnectionFactory.
>
> Previously the instance was set through a static method, which was
> really only done because it was difficult to add new
> PeerConnectionFactory construction arguments at the time.
>
> Now that we have PeerConnectionFactoryDependencies it's easy to clean
> this up.
>
> I'm doing this because I plan to add a NetworkMonitor implementation
> for iOS, and don't want to inherit this ugliness.
>
> Bug: webrtc:9883
> Change-Id: Id94dc061ab1c7186b81af8547393a6e336ff04c2
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180241
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
> Commit-Queue: Taylor <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#31815}
TBR=hta@webrtc.org, sakal@webrtc.org
Bug: webrtc:9883
Change-Id: I2e817c423f21936f87532a9694eb9a0a1b70c212
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180722
Reviewed-by: Taylor <deadbeef@webrtc.org>
Commit-Queue: Taylor <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31824}
This reverts commit 7ded73351870bfb45160fa6b9db71a94fe49397b.
Reason for revert: Found more code calling NetworkMonitorFactory::SetFactory...
Original change's description:
> Reland "Pass NetworkMonitorFactory through PeerConnectionFactory."
>
> This is a reland of 003c9be817817ed0e3aef3f50c78ae5cb31bc0ff
>
> Original change's description:
> > Pass NetworkMonitorFactory through PeerConnectionFactory.
> >
> > Previously the instance was set through a static method, which was
> > really only done because it was difficult to add new
> > PeerConnectionFactory construction arguments at the time.
> >
> > Now that we have PeerConnectionFactoryDependencies it's easy to clean
> > this up.
> >
> > I'm doing this because I plan to add a NetworkMonitor implementation
> > for iOS, and don't want to inherit this ugliness.
> >
> > Bug: webrtc:9883
> > Change-Id: Id94dc061ab1c7186b81af8547393a6e336ff04c2
> > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180241
> > Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> > Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
> > Commit-Queue: Taylor <deadbeef@webrtc.org>
> > Cr-Commit-Position: refs/heads/master@{#31815}
>
> TBR=hta@webrtc.org, sakal@webrtc.org
>
> Bug: webrtc:9883
> Change-Id: Ibf69a22e8f94226908636c7d50ff9eda65bd4129
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180720
> Reviewed-by: Taylor <deadbeef@webrtc.org>
> Commit-Queue: Taylor <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#31822}
TBR=deadbeef@webrtc.org,sakal@webrtc.org,hta@webrtc.org
Change-Id: Iae51b94072cec9abc021eed4e51d1fbeee998adc
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:9883
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180721
Reviewed-by: Taylor <deadbeef@webrtc.org>
Commit-Queue: Taylor <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31823}
This is a reland of 003c9be817817ed0e3aef3f50c78ae5cb31bc0ff
Original change's description:
> Pass NetworkMonitorFactory through PeerConnectionFactory.
>
> Previously the instance was set through a static method, which was
> really only done because it was difficult to add new
> PeerConnectionFactory construction arguments at the time.
>
> Now that we have PeerConnectionFactoryDependencies it's easy to clean
> this up.
>
> I'm doing this because I plan to add a NetworkMonitor implementation
> for iOS, and don't want to inherit this ugliness.
>
> Bug: webrtc:9883
> Change-Id: Id94dc061ab1c7186b81af8547393a6e336ff04c2
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180241
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
> Commit-Queue: Taylor <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#31815}
TBR=hta@webrtc.org, sakal@webrtc.org
Bug: webrtc:9883
Change-Id: Ibf69a22e8f94226908636c7d50ff9eda65bd4129
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180720
Reviewed-by: Taylor <deadbeef@webrtc.org>
Commit-Queue: Taylor <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31822}
This reverts commit 003c9be817817ed0e3aef3f50c78ae5cb31bc0ff.
Reason for revert: Breaks downstream build which is still using
SetFactory/ReleaseFactory. Probably will need to update this in lockstep.
Original change's description:
> Pass NetworkMonitorFactory through PeerConnectionFactory.
>
> Previously the instance was set through a static method, which was
> really only done because it was difficult to add new
> PeerConnectionFactory construction arguments at the time.
>
> Now that we have PeerConnectionFactoryDependencies it's easy to clean
> this up.
>
> I'm doing this because I plan to add a NetworkMonitor implementation
> for iOS, and don't want to inherit this ugliness.
>
> Bug: webrtc:9883
> Change-Id: Id94dc061ab1c7186b81af8547393a6e336ff04c2
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180241
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
> Commit-Queue: Taylor <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#31815}
TBR=deadbeef@webrtc.org,sakal@webrtc.org,hta@webrtc.org
Change-Id: I1f09df7be9c860017d515e5a87488340afa6eda6
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:9883
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180640
Reviewed-by: Taylor <deadbeef@webrtc.org>
Commit-Queue: Åsa Persson <asapersson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31818}
Previously the instance was set through a static method, which was
really only done because it was difficult to add new
PeerConnectionFactory construction arguments at the time.
Now that we have PeerConnectionFactoryDependencies it's easy to clean
this up.
I'm doing this because I plan to add a NetworkMonitor implementation
for iOS, and don't want to inherit this ugliness.
Bug: webrtc:9883
Change-Id: Id94dc061ab1c7186b81af8547393a6e336ff04c2
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/180241
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Commit-Queue: Taylor <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31815}
This name change communicates that the recursive critical section
should not be used for new code.
The relevant files are renamed rtc_base/critical_section* ->
rtc_base/deprecated/recursive_critical_section*
Bug: webrtc:11567
Change-Id: I73483a1c5e59c389407a981efbfc2cfe76ccdb43
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/179483
Commit-Queue: Markus Handell <handellm@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Tommi <tommi@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31754}
This change migrates a last stray consumer of GlobalLock
(SrtpSession) and removes all traces of GlobalLock/GlobalLockScope
from WebRTC.
Bug: webrtc:11567
Change-Id: I28059f2a10075815a4bdee8c357b9d3b6e50f18b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/179361
Commit-Queue: Tommi <tommi@webrtc.org>
Reviewed-by: Tommi <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31736}
GN recently added support for Apple frameworks to link, rather than
overloading the libs lists. This pulls .frameworks out of the libs
lists, so that GN can stop supporting .frameworks in libs in the
future.
Bug: chromium:1052560
Change-Id: I263230ddd3c468061584423bba9e1f887503bcaa
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/178601
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Sylvain Defresne <sdefresne@chromium.org>
Cr-Commit-Position: refs/heads/master@{#31632}
If num_wrap_ is zero this will result in a left shift of a
negative number, which is undefined behaviour.
Bug: webrtc:11742
Change-Id: I4a6ac448c5af82e15beb25ed55c16bab26e6ee5f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/178441
Commit-Queue: Dan Minor <dminor@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31608}
Policy will allow explicitly specify thread between which invokes are
allowed, or explicitly forbid any invokes.
Change-Id: I360e7cba3ce1c21abd5047c6f175d8c4e0e99c6f
Bug: webrtc:11728
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/177526
Reviewed-by: Tommi <tommi@webrtc.org>
Commit-Queue: Artem Titov <titovartem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31604}
When using a SequenceChecker, this adds a bit more information
about why the check failed.
Example (The "Expects" line is new):
# Fatal error in: foo.cc, line 380
# last system error: 0
# Check failed: (&thread_checker_)->IsCurrent()
# Expects: System queue: 0x7fff69541330, TaskQueue: 0x101804370 (not current), Thread: 0x10053cdc0
Bug: none
Change-Id: I3743e1d80f369f15219de5946e9e081f998b9b17
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/176569
Commit-Queue: Tommi <tommi@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31466}
Fairshare mutexes performed really badly during a Catalina
performance test. This change switches them to use
the _PTHREAD_MUTEX_POLICY_FIRSTFIT policy instead.
Bug: webrtc:11567, webrtc:11648
Change-Id: I2b8fbe3183beefc26f8d4ff3d63dc6958174605f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/176504
Reviewed-by: Tommi <tommi@webrtc.org>
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Commit-Queue: Tommi <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31456}
This change introduces a new non-reentrant mutex to WebRTC. It
enables eventual migration to Abseil's mutex.
The mutex types supportable by webrtc::Mutex are
- absl::Mutex
- CriticalSection (Windows only)
- pthread_mutex (POSIX only)
In addition to introducing the mutexes, the CL also changes
PacketBuffer to use the new mutex instead of rtc::CriticalSection.
The method of yielding from critical_section.cc was given a
mini-cleanup and YieldCurrentThread() was added to
rtc_base/synchronization/yield.h/cc.
Additionally, google_benchmark benchmarks for the mutexes were added
(test courtesy of danilchap@), and some results from a pthread/Abseil
shootout were added showing Abseil has the advantage in higher
contention.
Bug: webrtc:11567, webrtc:11634
Change-Id: Iaec324ccb32ec3851bf6db3fd290f5ea5dee4c81
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/176230
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Tommi <tommi@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31443}
Since we've passed IsConsistent(), if data_ is null, size_ must be
zero, so we might attempt to copy zero bytes from a nullptr. This does
not seem to cause problems in practice, but is still undefined
behaviour. This was caught on an UBsan test run in Firefox.
Bug: webrtc:11613
Change-Id: Iad795bf19ed69b56e066958a54a7e3a434b996cf
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/176280
Commit-Queue: Dan Minor <dminor@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#31386}
