NetEq currently only passes `jitterBufferDelay` to `getStats()`. We need its paired `jitterBufferEmittedCount` denominator stat for the calculations to be accurate.
Bug: webrtc:10192
Change-Id: I655aea629026ce9101409c2e0f18c2fa57a1c3ab
Reviewed-on: https://webrtc-review.googlesource.com/c/117320
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Commit-Queue: Chen Xing <chxg@google.com>
Cr-Commit-Position: refs/heads/master@{#26276}
Use size() accessor function. Also replace most nearby uses of _buffer
with data().
Bug: webrtc:9378
Change-Id: I1ac3459612f7c6151bd057d05448da1c4e1c6e3d
Reviewed-on: https://webrtc-review.googlesource.com/c/116783
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26273}
The new iceTransportState depends on the transports to signal when they have disconnected, this change ensures that they do so.
The logic is similar to what the old iceConnectionState did, but it uses the ice transports writable() flag instead of the one from the containing dtls transport.
Bug: webrtc:10199, webrtc:9308
Change-Id: I8a2a71a689b2a7027fe9117c79144811367d2165
Reviewed-on: https://webrtc-review.googlesource.com/c/117565
Commit-Queue: Jonas Olsson <jonasolsson@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26269}
It will be used to dump generated audio from TestAudioDeviceModule into
user defuned file in peer connection level test framework.
Bug: webrtc:10138
Change-Id: I6e3db36aaf1303ab148e8812937c4f9cd1b49315
Reviewed-on: https://webrtc-review.googlesource.com/c/117220
Commit-Queue: Artem Titov <titovartem@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26267}
Currently, CPU overuse settings for HW encoders are sometimes being used
even though the actual encoder is a SW encoder, e.g. in case of SW fallback
when the encoder is initialized. Polling is_hardware_accelerated after the
encoder has been created and initialized will improve choosing the correct
CPU overuse settings.
Bug: webrtc:10065
Change-Id: Ic6bd67630a040b5a121c13fa63dd074006973929
Reviewed-on: https://webrtc-review.googlesource.com/c/116688
Commit-Queue: Mirta Dvornicic <mirtad@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26266}
It is now possible to have dependencies with suffix _module without
clashing with the Android GN templates.
Bug: chromium:908819
Change-Id: I1b34aac60af93485ce23ebce295ab97c7c163d20
Reviewed-on: https://webrtc-review.googlesource.com/c/117161
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Oleh Prypin <oprypin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26265}
There are no plans to start using std::shared_ptr in WebRTC.
Bug: webrtc:10198
No-Try: True
Change-Id: I87a6c32b33b30d1b6b98eccda3400ce755a0ae95
Reviewed-on: https://webrtc-review.googlesource.com/c/117362
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26264}
The pitch gain type in ComputePitchGainThreshold() is wrong
(size_t instead of float).
The pitch period is an unsigned integer type, but it is safer to
switch to a signed type and add checks on the sign.
Bug: webrtc:9076
Change-Id: If69d182071edab9750a320f0fbfac24aa8052ee0
Reviewed-on: https://webrtc-review.googlesource.com/c/117302
Reviewed-by: Alex Loiko <aleloi@webrtc.org>
Commit-Queue: Alessio Bazzica <alessiob@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26259}
FFmpeg hasn't been rolled since [1] in order to avoid to break MSVC
trybots (//third_party/ffmpeg dropped MSVC support, in theory it is
possible to bring the support back but some work is needed every time
//third_party/ffmpeg gets updated).
Not rolling //third_party/ffmpeg is not enough to keep the Chromium
Roll working because -Wstring-plus-int becomes more chatty with clang 350768
and it has been suppressed in //third_party/ffmpeg/BUILD.gn [2].
Since WebRTC needs to update clang, //third_party/ffmpeg needs to be
updated. The only way to do it without fixing MSVC errors in
//third_party/ffmpeg is to enforce rtc_use_h264=False when MSVC is used.
PSA: https://groups.google.com/forum/#!topic/discuss-webrtc/cfkPPq5nvNE.
[1] - https://webrtc-review.googlesource.com/78402
[2] - https://chromium-review.googlesource.com/c/chromium/third_party/ffmpeg/+/1376376
Bug: webrtc:9213
Change-Id: I36bd7fb2db21012760e4ff7a791d81350e402ec0
Reviewed-on: https://webrtc-review.googlesource.com/c/116982
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Reviewed-by: Oleh Prypin <oprypin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26257}
A capturer with this flag was set in
WebRtcVideoChannelTest.PreviousAdaptationDoesNotApplyToScreenshare.
But the flag is used only by the VideoCapturerTrackSource class, which
isn't used in this test.
Bug: webrtc:6353
Change-Id: I58058c882c5a65b5cfa9921e302c422c8ccb20a9
Reviewed-on: https://webrtc-review.googlesource.com/c/117561
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Niels Moller <nisse@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26256}
This change adds a new way for test code to serialize the important information
from a VideoReceiveStream::Config so that it can be stored as configuration data
for WebRTC fuzzers. This code isn't included in the object itself as it is only
going to be used to generate new configurations for the fuzzer each time a new
error_correction or video format is added to WebRTC.
Bug: webrtc:10117
Change-Id: I9b6fb8e0345890ab16f6d319d91e4e316d1f2888
Reviewed-on: https://webrtc-review.googlesource.com/c/116920
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26255}
When using H264 encoder with profile level 3.1, the encoder may crash.
The reason is that we set the expected frame rate using kVTCompressionPropertyKey_ExpectedFrameRate
to the VideoToolBox. However, by iOS implementation, if our setting violates the sample rate limit
[1], the encoder will crash.
This CL fixes the bug by capping the expected frame rate with max allowed frame rate computed from the sample rate limit.
Change-Id: I090d7be8c20713c6a5a4ec80ed243c8fa7b4aa14
Bug: webrtc:10172
Reviewed-on: https://webrtc-review.googlesource.com/c/116056
Commit-Queue: Qiang Chen <qiangchen@chromium.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26254}
Bug: webrtc:10198
Change-Id: If510e6f508e34aaa36c9ccbbdc90dd33ad5fef10
Reviewed-on: https://webrtc-review.googlesource.com/c/116991
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26253}
This change moves the configuration parser that converts a JSON representation
of the VideoStreamReceiver::Config structure into a native object into the test
directory so that it can be shared with the new corpus_generator utility that is
being built. This rtc_source_set will have an additional utility function added
in a subsequent CL that will allow the generation of a VideoStreamSender::Config
from a given VideoStreamReceiver::Config and visa versa.
Bug: webrtc:10117
Change-Id: I3035826f799f8d1fcdeaa76997391f030c855a5c
Reviewed-on: https://webrtc-review.googlesource.com/c/116880
Reviewed-by: Sebastian Jansson <srte@webrtc.org>
Reviewed-by: Åsa Persson <asapersson@webrtc.org>
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26252}
EncodedImageIdInjector is responsible for injection of frame id into
encoded image before it will be sent to the transport layer. It will
help to track video frame from capturing on 1st peer side to rendering
on 2nd peer side and will make it possible to calculate video quality
stats between these frames.
This CL also introduces two different implementations for injector:
1. DefaultEncodedImageIdInjector will prepend all encoded images with
extra data and then will restore them on another side. This injector
can work even if peers are running on different devices.
2. SingleProcessEncodedImageIdInjector can work only when all peers
are running in the same process, but won't use any extra data
to propagate frame id between peers, so it won't affect any
transport level metrics and bitrate estimator.
This CL is first part of new video quality analyzer for end-2-end
peer connection level test framework.
Bug: webrtc:10138
Change-Id: I77defc8e8c95cb244a695a9732980a47bd7a2e9b
Reviewed-on: https://webrtc-review.googlesource.com/c/116682
Reviewed-by: Stefan Holmer <stefan@webrtc.org>
Reviewed-by: Peter Slatala <psla@webrtc.org>
Reviewed-by: Niels Moller <nisse@webrtc.org>
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Artem Titov <titovartem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26251}
I am planning to do a lot of work on adding additional fuzzing to WebRTC in Q1
of 2019. Given the limited number of available owners in this directory, and it
being non-prod I think this makes sense.
Bug: webrtc:10117
Change-Id: If2ad52fbce120c4d2bce51f4bfed99d83d78e6a5
Reviewed-on: https://webrtc-review.googlesource.com/c/117043
Reviewed-by: Alex Loiko <aleloi@webrtc.org>
Reviewed-by: Sam Zackrisson <saza@webrtc.org>
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Commit-Queue: Benjamin Wright <benwright@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26250}
This Config configuration will eventually replace the AudioProcessing::noise_suppression() interface.
This also introduces a proxy NoiseSuppression, returned by AudioProcessing::noise_suppression.
Without this proxy, ApplyConfig could overwrite NS settings for clients who currently use noise_suppression(). For example, the following code will not preserve the noise suppression level:
apm->noise_suppression()->set_level(NoiseSuppression::kHigh);
auto cfg = apm->GetConfig();
apm->ApplyConfig(cfg);
The NoiseSuppression instance returned by noise_suppression() has no way to update the config inside APM, so GetConfig() will return an out-of-date config which is then re-applied. This CL adds a proxy that makes this update, by forwarding Enable() and set_level() calls to ApplyConfig().
Drive-by change: AudioProcessing::Config substructs are reordered to mirror the capture processing pipeline.
Tested: Ran ToT and this CL builds of audioproc_f and verified identical settings/aecdumps.
Bug: webrtc:9947
Change-Id: I823eade894be115c254d656562564108b2b63b1f
Reviewed-on: https://webrtc-review.googlesource.com/c/116521
Reviewed-by: Alex Loiko <aleloi@webrtc.org>
Commit-Queue: Sam Zackrisson <saza@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26248}
Set render timestamp for all frames in the superframe.
Bug: chromium:912122
Change-Id: Ic9604620da9fb4176ad5c21b95df47fca8ddea31
Reviewed-on: https://webrtc-review.googlesource.com/c/116985
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26247}
Removes the deprecated video codec factories and the related flag and
helper classes.
Bug: webrtc:7925
Change-Id: I0a6d1666ece9ad074fefc79b626ba241765e1b98
Reviewed-on: https://webrtc-review.googlesource.com/c/113940
Commit-Queue: Anders Carlsson <andersc@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26245}
This is a space efficient way to store more records about decoded frames,
which is needed for long term references.
Bug: webrtc:9710
Change-Id: I051d59d34a966d48db011142466d9cd15304b7d9
Reviewed-on: https://webrtc-review.googlesource.com/c/116792
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Philip Eliasson <philipel@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26240}
Replaced with a combination of cricket::FakeFrameSource and
webrtc::test::FrameForwarder. This cl converts the first three
affected tests, the rest will follow.
Bug: webrtc:6353
Change-Id: I556f6b58f4ca81234ffae3dc6e1319f9c60a76ae
Reviewed-on: https://webrtc-review.googlesource.com/c/117260
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26239}
The change in
https://webrtc-review.googlesource.com/c/116683 made the fuzzer crash
at startup.
Bug: chromium:921050, webrtc:10185
Change-Id: Ie3eb26e12b4ae9b29c1c424af0d3eb287b5f1a73
Reviewed-on: https://webrtc-review.googlesource.com/c/117261
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26238}
Keep methods on subclasses where they are used: FifoBuffer and
MemoryStream. Also FileStream gets to keep SetPosition, because it's
used by a downstream subclass.
Bug: webrtc:6424
Change-Id: If2a00855aba7c2c9dc0909cda7c8a8ef00e0b9af
Reviewed-on: https://webrtc-review.googlesource.com/c/116487
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26237}
There is no need to explicitly do this, because when the GN
argument is_msan is True, the -fsanitize=memory is added by the
toolchain.
Bug: None
Change-Id: Id21dbd56df65636ca038e1abccaada0f44abfbb0
Reviewed-on: https://webrtc-review.googlesource.com/c/116992
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26236}
Some callback wrappers set the callback_ member at the start, but
most set it after calling any owned implementation of Start().
Setting it after the call means that the callback_ is not set up
for any callbacks that happen during the call.
This cl fixes that by setting the callback_ member before any
calls are made in Start().
Bug: chromium:916961
Change-Id: Id26f8cc98377ef217f928095834162f5526c1fdf
Reviewed-on: https://webrtc-review.googlesource.com/c/117040
Reviewed-by: Brave Yao <braveyao@webrtc.org>
Commit-Queue: Gary Kacmarcik <garykac@chromium.org>
Cr-Commit-Position: refs/heads/master@{#26231}
This is a following up cl to the fix of crbug.com/911110. On Windows,
if an App window is suspended, it will block some queries (which
causes Chromium freezing and is fixed in Chromium.) and won't be captured.
So there is no reason to list it in the window capture picker.
Notes: this cl can't fix the case that the select app window becomes
non-responsive just before capturing starts. Hope that an extreme corner
case that can be safely ingored.
Bug: chromium:911110
Change-Id: I0d14872ac699d559f40b3bff70f048efc67ca5d9
Reviewed-on: https://webrtc-review.googlesource.com/c/115441
Reviewed-by: Jamie Walch <jamiewalch@chromium.org>
Commit-Queue: Brave Yao <braveyao@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#26230}
