The existing methods SetEncrypedHeaderExtensionIds in SrtpTransport and SrtpSession
are removed because those methods could be confusing. When these methods are called
the head extension IDs are not actually updated and the user need to call SetRtpParams
again to make that happen. The existing setter just caches the new IDs.
To make it less confusing, the SetEncryptedHeaderExtensionIds is removed and the new
extension IDs will be set immediately when setting the crypto params.
For SDES, the crypto params and the header extension IDs will be set at the same time.
For DTLS, the new header extensions are cached in BaseChannel and will be set when
the DTLS handshake is completed.
Another major change is that when doing DTLS-SRTP, the encrypted header extension
IDs will be updated only when they are changed.
Bug: webrtc:7013
Change-Id: Ib70d4797456ae5ecb61b3dfff15c7e3e7ede89bd
Reviewed-on: https://webrtc-review.googlesource.com/15860
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20639}
PeerConnectionIntegrationTest.AddMediaToConnectedBundleDoesNotRestartIce
--> Fixed by an earlier CL (https://webrtc-review.googlesource.com/c/src/+/16261)
so re-enabled.
PeerConnectionIntegrationTest/PeerConnectionIntegrationIceStatesTest.VerifyIceStates
--> An existing bug causes this to by flaky when using a fake clock.
Fake clock removed with a TODO to change it back when the bug is fixed.
PeerConnectionIntegrationTest.TrackStatsUpdatedCorrectlyWhenUnsignaledSsrcChanges
--> The heuristic that >25% concealed audio samples is abnormal is
unfortunately not reliable enough on certain slow trybots. Bump the
threshold to 50% in hopes that is enough.
Bug: webrtc:8496
Change-Id: I17cfdf956a8a72ac399212c3c7efcdd2236be00d
Reviewed-on: https://webrtc-review.googlesource.com/20963
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Ivo Creusen <ivoc@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20638}
Bug: None
Change-Id: Ifd07c10dc1d3655e0138900c9a9897810cec3d54
Reviewed-on: https://webrtc-review.googlesource.com/18080
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20593}
This literally copies & pastes the code from WebRtcSession into
PeerConnection as private methods. The only other changes were to
inline the WebRtcSession construction/initialization/destruction
into PeerConnection and fix issues using rtc::Bind on the
reference-counted PeerConnection.
Bug: webrtc:8323
Change-Id: Ib3f071ac10d18566a21a3b04813b1d4ec691ef3c
Reviewed-on: https://webrtc-review.googlesource.com/15160
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20574}
This commit prepares WebRtcSession so that it can be cleanly
copy & pasted into PeerConnection in the next commit. To accomplish
this, the following was done:
1. Added a pointer to the owning PeerConnection to WebRtcSession.
2. Replace WebRtcSession state enum with signaling state.
3. All signals/observers only observed by PeerConnection were
replaced with direct calls to PeerConnection methods.
4. All duplicated fields were moved to PeerConnection.
5. The remaining tests that still use WebRtcSession for mocks were
updated to minimize dependence on WebRtcSession construction.
Bug: webrtc:8323
Change-Id: Ifc1a4ee819dcc9beca5363291012f7d5563ff7b1
Reviewed-on: https://webrtc-review.googlesource.com/9020
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20573}
EXPECT_EQ(0u, caller()->ice_connection_state_history().size());
Was failing since it cleared the connection state history before
the caller had finished transitioning to Completed, so the initial
transitions would be mistaken for later transitions.
Bug: None
Change-Id: I7043638f077ac5dcaeeca0d3ea6accc93c920364
Reviewed-on: https://webrtc-review.googlesource.com/16261
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Commit-Queue: Peter Thatcher <pthatcher@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20570}
We need to get rid of the ones that don't take audio codec factory
arguments in order to eliminate the dependency on audio codec
implementations.
BUG=webrtc:8396
Change-Id: Id0c1c3b70c2b3479da81ba1056cc69e857e454bd
Reviewed-on: https://webrtc-review.googlesource.com/12281
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20555}
Follow-up on https://webrtc-review.googlesource.com/c/src/+/17784.
Adds a new PC integration test using the newly added StartRecording API.
Bug: webrtc:7313
Change-Id: Ibd59910ca5d8f8ac96cfb891f41039759a18b6f6
Reviewed-on: https://webrtc-review.googlesource.com/17940
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20549}
Second attempt to land https://webrtc-review.googlesource.com/c/src/+/16180
Now removes voice_engine dependency from peerconnection and fixes a minor
const issue in NullAudioPoller.
TBR=solenberg
Bug: webrtc:7313
Change-Id: Ibfddbdc76118581e4a4dc64575203f84c1659e5c
Reviewed-on: https://webrtc-review.googlesource.com/17784
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20526}
This reverts commit 90bace095806a635411edd40fb8490a144e59e63.
Reason for revert: The original problem of this CL has been fixed in https://webrtc-review.googlesource.com/17540 but sounds like it is also adding voice_engine as a dependency of pc:peerconnection. We should investigate this because probably we can avoid it.
Original change's description:
> Add SetAudioPlayout and SetAudioRecording methods to the PeerConnection API
>
> (this CL is based on the work by Taylor and Steve in https://webrtc-review.googlesource.com/c/src/+/10201)
>
> This SetAudioPlayout method lets applications disable audio playout while
> still processing incoming audio data and generating statistics on the
> received audio.
>
> This may be useful if the application wants to set up media flows as
> soon as possible, but isn't ready to play audio yet. Currently, native
> applications don't have any API point to control this, unless they
> completely implement their own AudioDeviceModule.
>
> The SetAudioRecording works in a similar fashion but for the recorded
> audio. One difference is that calling SetAudioRecording(false) does not
> keep any audio processing alive.
>
> TBR=solenberg
>
> Bug: webrtc:7313
> Change-Id: I0aa075f6bfef9818f1080f85a8ff7842fb0750aa
> Reviewed-on: https://webrtc-review.googlesource.com/16180
> Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
> Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
> Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#20499}
TBR=solenberg@webrtc.org,henrika@webrtc.org,kwiberg@webrtc.org
Change-Id: I8431227e21dbffcfed3dd0e6bd7ce26c4ce09394
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:7313
Reviewed-on: https://webrtc-review.googlesource.com/17701
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20512}
(this CL is based on the work by Taylor and Steve in https://webrtc-review.googlesource.com/c/src/+/10201)
This SetAudioPlayout method lets applications disable audio playout while
still processing incoming audio data and generating statistics on the
received audio.
This may be useful if the application wants to set up media flows as
soon as possible, but isn't ready to play audio yet. Currently, native
applications don't have any API point to control this, unless they
completely implement their own AudioDeviceModule.
The SetAudioRecording works in a similar fashion but for the recorded
audio. One difference is that calling SetAudioRecording(false) does not
keep any audio processing alive.
TBR=solenberg
Bug: webrtc:7313
Change-Id: I0aa075f6bfef9818f1080f85a8ff7842fb0750aa
Reviewed-on: https://webrtc-review.googlesource.com/16180
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20499}
Enable cpplint check in the PRESUBMIT for pc/ and fix all existing
warnings.
Bug: webrtc:5583
Change-Id: If39994692ab6f6f3c83c74f23850f02fdfe810e8
Reviewed-on: https://webrtc-review.googlesource.com/16540
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20482}
The DataChannel OPEN message is sent in-band in this test, and waiting
for the signaling state to change is not sufficient to guarantee that the
callee received this message
Bug: webrtc:8443
Change-Id: I76fa6348b6f8e1e70fb41a4e644aee805b2ef4de
Reviewed-on: https://webrtc-review.googlesource.com/15060
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20415}
WebRtcSession is being merged into PeerConnection, and to make the
code review easier this is the first step towards achieving that.
Bug: webrtc:8323
Change-Id: I33778e46f20cb14089dff4328947868e207476bd
Reviewed-on: https://webrtc-review.googlesource.com/8760
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20413}
Originally, the idea was to implement QUIC data channels as a
PeerConnection API. Now, the effort has shifted to implementing it as a
part of ORTC which will live in Chromium. Since this code has not been
maintained and is not currently being used, remove it to reduce
maintenance overhead while a copy will be retained in the Git history.
Bug: webrtc:8385
Change-Id: I2719c007a0de0118b67d41a425f900b66c52f65a
Reviewed-on: https://webrtc-review.googlesource.com/14100
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20412}
In SrtpTransport::SetRtcpParams, send_rtcp_session_ should really call
SetSend rather than SetRecv.
Modified the LOG message in SrtpTransport::SetRtpParams.
Bug: webrtc:8436
Change-Id: Iccbfbc5ef2d4f4ebd5f876c3f6dcc81671fdc631
Reviewed-on: https://webrtc-review.googlesource.com/14562
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20403}
This is a reland of b49b66109ea8a0a33a3192ebccf91366af2e49ae.
Original change's description:
> Rewrite WebRtcSession BUNDLE tests as PeerConnection tests
>
> Bug: webrtc:8222
> Change-Id: Id47e4544dc073564ad7e63d02865ca80dd5a85ff
> Reviewed-on: https://webrtc-review.googlesource.com/8280
> Commit-Queue: Steve Anton <steveanton@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#20365}
Bug: webrtc:8222
Change-Id: If3dcd8090875c641881e2b9e92fc1db387ba1de5
Reviewed-on: https://webrtc-review.googlesource.com/14400
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20397}
This is a reland of 3df5dcac9b339ba4d3f4969602f094c2c8035b51
Original change's description:
> Rewrite WebRtcSession media tests as PeerConnection tests
>
> Bug: webrtc:8222
> Change-Id: I782a3227e30de70eb8f6c26a48723cb3510a84ad
> Reviewed-on: https://webrtc-review.googlesource.com/6640
> Commit-Queue: Steve Anton <steveanton@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#20364}
Bug: webrtc:8222
Change-Id: I0a5398170d469eb9223bc781bfb417a85a72a2d2
Reviewed-on: https://webrtc-review.googlesource.com/14380
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20377}
AddRef() now returns void, and Release() returns an enum
RefCountReleaseStatus, to indicate whether or not this Release
call implied deletion.
Bug: webrtc:8270
Change-Id: If2fb77f26118b61751b51c856af187c72112c630
Reviewed-on: https://webrtc-review.googlesource.com/3320
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20366}
This reverts commit 54d1da13a584680ae80a1f229291e5bb7e76e6e1.
Reason for revert: Breaking tests
Original change's description:
> BWE allocation strategy allows controlling of bitrate allocation with WEBRTC external logic.
>
> This CL implements the main logic and IOS appRTC integration.
>
> Unit tests and Android appRTC will be in separate CL.
>
> Bug: webrtc:8243
> Change-Id: If8e5195294046a47316e9fade1b0dfec211155e1
> Reviewed-on: https://webrtc-review.googlesource.com/4860
> Commit-Queue: Alex Narest <alexnarest@webrtc.org>
> Reviewed-by: Niels Moller <nisse@webrtc.org>
> Reviewed-by: Stefan Holmer <stefan@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#20329}
TBR=deadbeef@webrtc.org,nisse@webrtc.org,stefan@webrtc.org,alexnarest@webrtc.org
Change-Id: I5be1da78f360f72be66f9d56dd6b88c1cc13e963
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:8243
Reviewed-on: https://webrtc-review.googlesource.com/12560
Reviewed-by: Lu Liu <lliuu@webrtc.org>
Commit-Queue: Lu Liu <lliuu@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20330}
This CL implements the main logic and IOS appRTC integration.
Unit tests and Android appRTC will be in separate CL.
Bug: webrtc:8243
Change-Id: If8e5195294046a47316e9fade1b0dfec211155e1
Reviewed-on: https://webrtc-review.googlesource.com/4860
Commit-Queue: Alex Narest <alexnarest@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20329}
They're about to be removed.
BUG=webrtc:8396
Change-Id: Ie9a45f4c0dccb4414d2a2f939aa5f142edc6e4b6
Reviewed-on: https://webrtc-review.googlesource.com/12280
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20328}
This test was disabled for being flaky on the bots.
Try reenabling it to see if something has changed.
Bug: webrtc:7247
Change-Id: I65ce2cf6ce7a3761247369255d9ba106aa3e53f9
Reviewed-on: https://webrtc-review.googlesource.com/3262
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Commit-Queue: Kári Helgason <kthelgason@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20320}
By having a unique_ptr own the callback data instead of a raw pointer,
the compiler helps us ensure that it's destroyed exactly once,
and never used after being destroyed.
(This made the callback object move-only, so I had to add support
for move-only callbacks to rtc::Thread::Invoke().)
BUG=webrtc:8111
Change-Id: Ia0804e4662e63e91e5cee18ecc3f38d2cfe8a26b
Reviewed-on: https://webrtc-review.googlesource.com/10812
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20317}
Unit test now checks that ADM:Init() works before any test runs.
It means that all tests will be skipped on bots that lack Pulse
support which is as how it worked before this CL as well. But then,
we detected the lack of support by checking that the audio layer had
changed from Pulse to Alsa.
As a consequence, I also decided to inject fake/mock ADMs in more
unit tests. One was actually already injected for other reasons
(see https://codereview.webrtc.org/2997383002/) but it had accidentally
been "reverted" later in combination with other changes.
To summarize: before this change we had a set of unit tests where real
audio was tested but it was not the intention of the test or required.
In addition, some Linux bots (VM:s) did not support PulseAudio and on
them the tests relied on a fallback mechanism to ALSA to work, i.e.,
we had a rather complex dependency on hardware. This dependency has now
been removed and it should result in more stable tests.
Bug: webrtc:7306, webrtc:7806
Change-Id: Ia0485658c04a4ef3b3f2dc0d557d73738067304b
Reviewed-on: https://webrtc-review.googlesource.com/8640
Commit-Queue: Henrik Andreassson <henrika@webrtc.org>
Reviewed-by: Fredrik Solenberg <solenberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20307}
This reverts commit a8264dbdd97f5e125d45fd0e84356f2e1f747df1.
Reason for revert: Reverting to unblock rolls into Chromium.
See failure here:
https://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_rel_ng/builds/565449
Fails: external/wpt/webrtc/RTCPeerConnection-setRemoteDescription-offer.html
I'm guessing these lines from the output are relevant:
12:15:32.525 11839 [1:19:1015/121532.495175:16438293900:ERROR:webrtcsession.cc(350)] Failed to set remote offer sdp: The order of m-lines in subsequent offer doesn't match order from previous offer/answer.
12:15:32.525 11839 [1:20:1015/121532.497199:16438296127:WARNING:delay_based_bwe.cc(326)] BWE Setting start bitrate to: 300000
12:15:32.525 11839 [1:1:1015/121532.498272:16438296963:ERROR:webrtcsdp.cc(359)] Failed to parse: "Invalid SDP". Reason: Expect line: v=
12:15:32.525 11839 [1:1:1015/121532.498364:16438297040:ERROR:rtc_peer_connection_handler.cc(2183)] Failed to create native session description. Type: offer SDP: Invalid SDP
12:15:32.525 11839 [1:1:1015/121532.498432:16438297104:ERROR:rtc_peer_connection_handler.cc(1458)] Failed to parse SessionDescription. Invalid SDP Expect line: v=
Original change's description:
> Reject the subsequent offer with fewer m= sections.
>
> If the subsequent offer contains fewer m= sections than the existing
> description, it would be rejected.
>
> The helper method MediaSectionsInSameOrder is modified and it will
> compare the number of m= sections before matching the media type.
>
> Bug: chromium:773620
> Change-Id: Ic8999445f4bc023da1d85a65659583db1687ec37
> Reviewed-on: https://webrtc-review.googlesource.com/9621
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#20298}
TBR=deadbeef@webrtc.org,zhihuang@webrtc.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: chromium:773620
Change-Id: I4a3ff7a42abb95144615b1dd37fb21585ee07b5d
Reviewed-on: https://webrtc-review.googlesource.com/10920
Reviewed-by: Tommi <tommi@webrtc.org>
Commit-Queue: Tommi <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20300}
If the subsequent offer contains fewer m= sections than the existing
description, it would be rejected.
The helper method MediaSectionsInSameOrder is modified and it will
compare the number of m= sections before matching the media type.
Bug: chromium:773620
Change-Id: Ic8999445f4bc023da1d85a65659583db1687ec37
Reviewed-on: https://webrtc-review.googlesource.com/9621
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20298}
This will allow Chrome to provide a RtcEventLogOutput object that reports the log back to Chrome, allowing Chrome to manage the log by itself - write it to a file, upload it to a server, etc.
Bug: webrtc:8111
Change-Id: I6a2a6945fc8586ef10e0fb9c56eaa8fda00dfc98
Reviewed-on: https://webrtc-review.googlesource.com/8081
Commit-Queue: Elad Alon <eladalon@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20295}
This will cause the application to be aborted before it encounters
something worse like a heap overflow, in case any bug in this code
exists or is introduced in the future.
TBR=zhihuang@webrtc.org
Bug: chromium:773620
Change-Id: Idd4e31aa63a3f673eefd3e8cb2ae3f4a5092ca4e
Reviewed-on: https://webrtc-review.googlesource.com/9040
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20293}
This is the same thing we're doing for usrsctp. Before this CL, the
first SrtpSession to call SetKey would initialize libsrtp, and
ChannelManager's destructor would deinitialize it. This works for an
application that only uses one instance of ChannelManager simultaneously
(or one instance of PeerConnectionFactory), but not one that uses
multiple.
Now, libsrtp is effectively reference-counted, with the first
SrtpSession to take a reference initializing it, and the last to remove
its reference deinitializing it.
This issue was discovered recently due to a change that resulted in
using srtp_update. Without using that method, the issue went unnoticed;
maybe srtp_protect/srtp_unprotect don't require initialization?
Bug: webrtc:8388
Change-Id: If1329360f0b469e454810e62e9b5acfbd4cba100
Reviewed-on: https://webrtc-review.googlesource.com/9000
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#20262}
