So far the code assumed that there is only one implementation of AudioSourceInterface: LocalAudioSource.
That is not true. This change allows custom implementations to still set options (such as audio network adaptation) on the source.
Long term solution should include refactoring options so that they are passed to peer connection or call object, and not be defined on audio source.
Bug: webrtc:9719
Change-Id: Ic3b92219502bc73a964adbbb9c5cd7156aa382e1
Reviewed-on: https://webrtc-review.googlesource.com/c/110681
Commit-Queue: Peter Slatala <psla@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25626}
This is a follow up of https://webrtc-review.googlesource.com/c/src/+/43201.
Issue 43201 didn't do the job properly.
1. The audio rtcp report interval is not properly hooked up.
2. We don't need to propagate audio rtcp interval into video send stream or vice versa.
3. We don't need to propagate rtcp report interval to any receiving streams.
Bug: webrtc:8789
Change-Id: I1f637d6e5173608564ef0702d7eda6fc93b3200f
Reviewed-on: https://webrtc-review.googlesource.com/c/110105
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Magnus Flodman <mflodman@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Commit-Queue: Jiawei Ou <ouj@fb.com>
Cr-Commit-Position: refs/heads/master@{#25610}
It's not currently used or maintained, so it shouldn't be a part of out API.
Bug: webrtc:9824
Change-Id: Ic44c5ea3a9eab8fb75e87a5005cbf6cdd4b1d4ad
Reviewed-on: https://webrtc-review.googlesource.com/c/107645
Commit-Queue: Jonas Olsson <jonasolsson@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Minyue Li <minyue@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25593}
This changes PeerConnection to allow sending and receiving data channel
messages over the media transport. If |use_media_transport_for_data_channels|
is set, PeerConnection will use a DCT_MEDIA_TRANSPORT mode for data
channels.
DCT_MEDIA_TRANSPORT acts exactly like DCT_SCTP within the data channel
and peer connection layers. On the transport layer, it uses the media
transport instead of SCTP. It appears as an RTP data channel in SDP
(just as media over media-transport appears as RTP in SDP).
Bug: webrtc:9719
Change-Id: I6a90142bd3f43668479c825ed02689dcd0d58b78
Reviewed-on: https://webrtc-review.googlesource.com/c/109740
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25575}
Compared the original CL: https://webrtc-review.googlesource.com/c/src/+/94782
This new CL added backward compatible functions to WebRtcMediaEngineFactory so that internal projects will not be broken.
Because of that, now we can revert all the changes to SDK and PeerConnection and do it in following CLs. This makes this CL cleaner.
One temporary disadvantage of this is the media engine now need to take a dependency onto builtin video bitrate factory, but practically it just moved code around and should not result in a large binary size change. We can remove this dependency later if needed.
Bug: webrtc:9513
Change-Id: I38708762ff365e4ca05974b99fac71edc739a756
Reviewed-on: https://webrtc-review.googlesource.com/c/109040
Commit-Queue: Jiawei Ou <ouj@fb.com>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25574}
If rate controller is trusted, we disable the frame dropper in the
media optimization module.
This is a re-land of
https://webrtc-review.googlesource.com/c/src/+/105020
Bug: webrtc:9890
Change-Id: I418e47a43a1a98cb2fd5295c03360b954f2288f2
Reviewed-on: https://webrtc-review.googlesource.com/c/109141
Commit-Queue: Erik Språng <sprang@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25570}
This reverts commit 61c6e5643e.
Reason for revert: downstream projects prepared for this change
Original change's description:
> Revert "Isolating APM API build target: making :api an actual target."
>
> This reverts commit a7f77a7c05.
>
> Reason for revert: breaking downstream
>
> Original change's description:
> > Isolating APM API build target: making :api an actual target.
> >
> > This CL is part of a refactoring work to unblock other CLs
> > that would generate a circular dependency when including
> > modules/audio_processing. It will also allow to easily move
> > the APM interface part under //api.
> >
> > More in detail, this change moves the APM interface files from
> > the build target modules/audio_processing to
> > modules/audio_processing:api. It also adds :api as dependency
> > where needed.
> >
> > Bug: webrtc:9535
> > Change-Id: I72829e22d08ba4d75985f0421e6e8bf0216ebecd
> > Reviewed-on: https://webrtc-review.googlesource.com/c/109501
> > Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
> > Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
> > Reviewed-by: Niels Moller <nisse@webrtc.org>
> > Reviewed-by: Sam Zackrisson <saza@webrtc.org>
> > Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
> > Cr-Commit-Position: refs/heads/master@{#25539}
>
> TBR=saza@webrtc.org,alessiob@webrtc.org,kwiberg@webrtc.org,nisse@webrtc.org,kthelgason@webrtc.org
>
> Change-Id: I974c6237311e7c06970aa62e5f6940f3aa80113d
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: webrtc:9535
> Reviewed-on: https://webrtc-review.googlesource.com/c/109820
> Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
> Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#25540}
TBR=saza@webrtc.org,alessiob@webrtc.org,kwiberg@webrtc.org,nisse@webrtc.org,kthelgason@webrtc.org
Change-Id: Ic8ed4cc3baf43d639ce13cae256c007728c3ad92
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:9535
Reviewed-on: https://webrtc-review.googlesource.com/c/109884
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25547}
This removes the constructor that takes all members as parameters.
A main goal with this struct is to make it easy to add or remove fields
without having to refactor implementations and usage all over the palce,
this breaks down if we need to update the arguments in this constructor
every time.
Also cleans up a few commented-out lines that were forgotten in an
earliers cl.
Bug: webrtc:9722
Change-Id: I22db35d399469bdb3bfd8bdc531b24ead9fe1c3e
Reviewed-on: https://webrtc-review.googlesource.com/c/109570
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Commit-Queue: Erik Språng <sprang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25542}
This reverts commit a7f77a7c05.
Reason for revert: breaking downstream
Original change's description:
> Isolating APM API build target: making :api an actual target.
>
> This CL is part of a refactoring work to unblock other CLs
> that would generate a circular dependency when including
> modules/audio_processing. It will also allow to easily move
> the APM interface part under //api.
>
> More in detail, this change moves the APM interface files from
> the build target modules/audio_processing to
> modules/audio_processing:api. It also adds :api as dependency
> where needed.
>
> Bug: webrtc:9535
> Change-Id: I72829e22d08ba4d75985f0421e6e8bf0216ebecd
> Reviewed-on: https://webrtc-review.googlesource.com/c/109501
> Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
> Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
> Reviewed-by: Niels Moller <nisse@webrtc.org>
> Reviewed-by: Sam Zackrisson <saza@webrtc.org>
> Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#25539}
TBR=saza@webrtc.org,alessiob@webrtc.org,kwiberg@webrtc.org,nisse@webrtc.org,kthelgason@webrtc.org
Change-Id: I974c6237311e7c06970aa62e5f6940f3aa80113d
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:9535
Reviewed-on: https://webrtc-review.googlesource.com/c/109820
Reviewed-by: Alessio Bazzica <alessiob@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25540}
This CL is part of a refactoring work to unblock other CLs
that would generate a circular dependency when including
modules/audio_processing. It will also allow to easily move
the APM interface part under //api.
More in detail, this change moves the APM interface files from
the build target modules/audio_processing to
modules/audio_processing:api. It also adds :api as dependency
where needed.
Bug: webrtc:9535
Change-Id: I72829e22d08ba4d75985f0421e6e8bf0216ebecd
Reviewed-on: https://webrtc-review.googlesource.com/c/109501
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25539}
Adds a field |use_media_transport_for_data_channels| to RTCConfiguration.
PeerConnection requires a media transport factory to be set if this bit
is set. As with |use_media_transport|, the value may not be modified
after setting the local or remote description.
If either |use_media_transport| or |use_media_transport_for_data_channel| is
set, PeerConnection uses its media transport factory when creating a JSEP
transport controller.
PeerConnection stops unconditionally using media transport in
CreateVoiceChannel, as it may be present only for use in data channels. It uses
the media transport if it is present and |use_media_transport| is set.
Bug: webrtc:9719
Change-Id: I59d4ce8f7531fd19d9c17eefe033f063f663ebcc
Reviewed-on: https://webrtc-review.googlesource.com/c/109041
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Peter Slatala <psla@webrtc.org>
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25507}
The SetChannelParameters function was used when WebRTC supported decoding
with errors, which we no longer do.
This cleanup CL is related to the work tracked by 9946.
Bug: webrtc:9946
Change-Id: Id2d5ed23031388f890c42651bfbe5f79eda701e5
Reviewed-on: https://webrtc-review.googlesource.com/c/108861
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Commit-Queue: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25505}
This enables PeerConnection tests to use LoopbackMediaTransport to test
data-channel-over-media-transport code.
Also changes LoopbackMediaTransport to invoke callbacks asynchronously.
This is more accurate, as these callbacks are triggered by network
events. The caller should not block while the callback executes.
Since LoopbackMediaTransport is used for testing, it provides a
FlushAsyncInvokes() method which may be used to ensure that callbacks
occur deterministically (eg. before checking that data has been
received).
Bug: webrtc:9719
Change-Id: Ib8ea9aebf4a0ad3d5934a6fe4ab33432c68523fd
Tbr: stefan@webrtc.org
Reviewed-on: https://webrtc-review.googlesource.com/c/109060
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25489}
IceConnected state (transport state) now includes the state of the
MediaTransport.
This is a first change of two. Second change will add state change
signals to the PeerConnectionInterface informing separately about
ice+media transport vs ice+dtls.
Bug: webrtc:9719
Change-Id: I5731530073e8f26dfc8b188778d268b815da7052
Reviewed-on: https://webrtc-review.googlesource.com/c/108901
Commit-Queue: Peter Slatala <psla@webrtc.org>
Reviewed-by: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25473}
This makes it safer to reason about the common case where send
time information is available. We don't have to either assume that
it's available, or check it everywhere the PacketResult struct is used.
To achieve this, a new field is added to TransportPacketsFeedback
and a new interface is introduced to clearly separate which field is
used. A possible followup would be to introduce a separate struct.
That would complicate the signature of ProcessTransportFeedback.
Bug: webrtc:9934
Change-Id: I2b319e4df2b557fbd4de66b812744bca7d91ca15
Reviewed-on: https://webrtc-review.googlesource.com/c/107080
Commit-Queue: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25465}
Additionally, AudioProcessing::GetStatistics(bool) is made pure
virtual and the default implementation in AudioProcessing is removed.
Deprecation PSA:
https://groups.google.com/forum/#!msg/discuss-webrtc/NgqEPvkNuDE/7HtwnMmADgAJ
Bug: webrtc:9947, webrtc:8572
Change-Id: I123402bf7d6c49f3613154c469b818109d8fad43
Reviewed-on: https://webrtc-review.googlesource.com/c/108783
Commit-Queue: Sam Zackrisson <saza@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Ivo Creusen <ivoc@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25463}
This change deletes the default implementations of state and data
channel methods (SetMediaTransportStateCallback, SendData, CloseChannel,
and SetDataSink). It adds stub implementations to LoopbackMediaTransport
and FakeMediaTransport.
Bug: webrtc:9719
Change-Id: I49b7780c055b552330546b460c2e79ce8df81833
Reviewed-on: https://webrtc-review.googlesource.com/c/108940
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25457}
Adds the types and methods required for sending and receiving data
channel messages over the media transport. These are:
- A DataMessageType to distinguish between text, binary, and control
messages
- A parameters struct for sending data messages, which specifies the
channel id, type, and ordering/reliability parameters
- A sink for data-channel related callbacks (receive data, begin
closing procedure, and end closing procedure)
- A method to set the sink for data channels
- Methods to open, close, and send on data channels
These methods, combined with the state sink, allow PeerConnection to
implement the DataChannelProviderInterface using MediaTransport as the
underlying transport.
Change-Id: Iccb2ba374594762a5b4f995564e2a1ff7d8805f5
Bug: webrtc:9719
Reviewed-on: https://webrtc-review.googlesource.com/c/108541
Reviewed-by: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Peter Slatala <psla@webrtc.org>
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25454}
This allows to use secure, end to end communication if SDES cryptos are
passed. MediaTransport can use a derived key to secure its own
communication.
Bug: webrtc:9719
Change-Id: If1a20b136b3b4af0cb24f10b52fc5ce1eb31daa2
Reviewed-on: https://webrtc-review.googlesource.com/c/108504
Commit-Queue: Peter Slatala <psla@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Qingsi Wang <qingsi@webrtc.org>
Reviewed-by: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Bjorn Mellem <mellem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25452}
This CL consistently use:
* relative paths for WebRTC dependent targets (test_support)
* absolute paths for shared dependent targets (abseil)
This is a necessary (but insufficient) step to build WebRTC tests
from Chromium tree (rtc_include_tests=true), since test/ doesn't
sit anymore in the top level directory.
We also make sure that target declarations and uses are
consistent in regard to build_with_chromium flag.
Bug: webrtc:9943
Bug: webrtc:9855
Change-Id: I21dea98894df2fd4bfe2fd7ee7b71ba971e0ab5b
Reviewed-on: https://webrtc-review.googlesource.com/c/108720
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Commit-Queue: Yves Gerey <yvesg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25445}
This CL utilizes the existing, but unused, ability to set
different histogram thresholds for early and late delay
estimation. It does so by tuning the parameters for these.
On top of that, some corrections are added to correctly
handle resets and the use of the hysteresis thresholds.
Bug: webrtc:19886,chromium:896334
Change-Id: I950ac107c124541af8f02b4403f477dda71cc1a1
Reviewed-on: https://webrtc-review.googlesource.com/c/106706
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Reviewed-by: Gustaf Ullberg <gustaf@webrtc.org>
Commit-Queue: Per Åhgren <peah@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25443}
Clients of media_transport_interface need the ability to monitor BWE
estimates, and this change adds a TargetBitrate observer to the media
transport interface.
Bug: webrtc:9719
Change-Id: I90ebbf684c6f269e0c3cd58428010cfa511cc970
Reviewed-on: https://webrtc-review.googlesource.com/c/108106
Reviewed-by: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Commit-Queue: Peter Slatala <psla@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25415}
The struct is more generic and easier to extend than parameters to the
Factory. In addition, the list of parameters to the factory might grow,
making invocations awkward if not difficult to read.
Bug: webrtc:9719
Change-Id: I4b98e26f1f4c0d5ea840f9c28e7ed7abee072b74
Reviewed-on: https://webrtc-review.googlesource.com/c/107984
Commit-Queue: Peter Slatala <psla@webrtc.org>
Reviewed-by: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25413}
There was a suggestion in a previous CL to add an end to end test case to
prevent future regressions. I have enabled this by adding two fakes that
perform fake encryption and enabling an end to end test with VP8 and the
GenericDescriptor.
Bug: webrtc:9927
Change-Id: Icf96eeed541ada1e0579eb81b6f87a46d1c43d96
Reviewed-on: https://webrtc-review.googlesource.com/c/108020
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25399}
After the problem fixed by [1], it seems wiser to only check that
`include_rules` contains all the top level directories and delegate the
action of adding a rule for .cc file to the first user of an header
from a new root level directory.
This should make the presubmit check more actionable (in [1] for example
the solution was to *not* add "+ios" but to consider ios/ as a non
WebRTC directory).
[1] - https://webrtc-review.googlesource.com/c/107707
Bug: webrtc:9887, webrtc:9924
Change-Id: Ic85e2153a2b83a4874c8faec3c5d1a8c61fe6faf
Reviewed-on: https://webrtc-review.googlesource.com/c/107731
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25386}
This will allow downstream projects to use it to construct their own
injected codecs without pulling in dependencies on software codecs that
they don't need.
TBR=shampson@webrtc.org
Bug: webrtc:7925
Change-Id: Ie5c246bbf8e2ef1b27562887f717af9e719a1edf
Reviewed-on: https://webrtc-review.googlesource.com/c/107698
Commit-Queue: Jonathan Yu <yujo@chromium.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Anders Carlsson <andersc@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25379}
With the expanding use cases for webrtc::CryptoOptions it makes more sense for
it to be be available per peer connection instead of only as a factory option.
To support backwards compatability for now this code will support the factory
method of setting crypto options by default. However it will completely
overwrite these settings if an RTCConfiguration.crypto_options is provided.
Got LGTM offline from Sami, adding him to TBR if he has any further comments.
TBR=sakal@webrtc.org
Bug: webrtc:9891
Change-Id: I86914cab69284ad82afd7285fd84ec5f4f2c4986
Reviewed-on: https://webrtc-review.googlesource.com/c/107029
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#25375}
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}
