Reset() is called each time the encoder is reconfigured, but then it
happens the target bitrate isn't reset in encoder. So it might produce a
frame before next bitrate estimate is propagated to the metadata writer.
The incorrect zero bitrate would be treated as a paused encoder and would
cause metadata to be dropped.
Also, added unittest for that scenario at VideoStreamEncoder level.
Bug: webrtc:10460
Change-Id: I28024a527f1fb8474b172e2c5c2394fd38d69a07
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/140101
Reviewed-by: Erik Språng <sprang@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28159}
Raw RTP packetization is done using the existing RtpPacketizerGeneric
without adding the generic payload header. It is intended to be used
together with generic frame descriptor RTP header extension.
Bug: webrtc:10625
Change-Id: I2e3d0a766e4933ddc4ad4abc1449b9b91ba6cd35
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/138061
Commit-Queue: Mirta Dvornicic <mirtad@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Rasmus Brandt <brandtr@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28154}
SurfaceTextureHelper currently crashes if an OES texture is produced
before setTextureSize() has been called. This is annoying if the texture
size is not easily known beforehand. A real world example is MediaPlayer
that provides the video size with an asynchronous call to
setOnVideoSizeChangedListener(), but that might happen after the first
texture is produced on some devices.
This CL waits with delivering frames until the size has been sent,
rather than crashing.
Bug: webrtc:10709
Change-Id: I5d9ce542e0edaafe1153fd5fe7d64dba86d7e33c
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/140080
Reviewed-by: Sami Kalliomäki <sakal@webrtc.org>
Commit-Queue: Magnus Jedvert <magjed@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28151}
- This makes it consistent with ICE and MediaTransport ownership.
- Removes unnecessary datagram_transport() getter in DtlsTransportInternal
As a side effect this fixes bug in JsepTransportController, which moved datagram_transport to Dtls after creating it, then checked if (datagram_transport) to decide which RTP transport to create. As a result of this bug we were creating Sded instead of Unencrypted RTP with datagram transport.
Bug: webrtc:9719
Change-Id: Ic5b13a450ce6ac5b2a20d388657e3949aabef079
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139620
Commit-Queue: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Bjorn Mellem <mellem@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28146}
This documents in the API what is already true in the
implementation - that SessionDescription will eventually
delete MediaDescription objects passed to it.
The old API is preserved for backwards compatibility, but
marked as RTC_DEPRECATED.
Bug: webrtc:10701
Change-Id: I9a822b20cf3e58c5945fa51dbf6082960a332de8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139880
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28144}
This is an usual error while using neteq_quality_test. This tool
does not support wav files as input. Adding a validation.
Bug: webrtc:10690
Change-Id: I18ed308d2f688106728df5df25e0a58c7170f411
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139104
Commit-Queue: Pablo Barrera González <barrerap@webrtc.org>
Reviewed-by: Henrik Lundin <henrik.lundin@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28141}
The test RunPlayoutAndRecordingInFullDuplex makes the speakers play the
signal it simultaneously records from the microphone, which can cause
full howling.
The test itself measures buffer usage and does not depend on what signal
is played through the speakers. This change mutes the speakers to prevent
howling when running modules_unittests.
Bug: webrtc:10704
Change-Id: I6176adb2fb5ce0e4d86f6f447476be8a88c2f2cf
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139889
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Commit-Queue: Gustaf Ullberg <gustaf@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28140}
This change adds classes so that we later can plumb information about received packets to each audio and video frame. It's not wired up to do anything yet.
Bug: webrtc:10668
Change-Id: I962df493a76692f668314f78d6792d7636c5a31b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/138203
Commit-Queue: Chen Xing <chxg@google.com>
Reviewed-by: Minyue Li <minyue@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28138}
Removes CAPTURE_VIDEO_OUTPUT permission since it is not needed
when using MediaProjection API and not allowed in Android Q.
This is a reland of af4f1b41277ebdf0d7386cbd2903abc709cbc183
Original change's description:
> Target SDK level 27 in AppRTCMobile.
>
> Implements the dynamic permission model required by the newer SDK and
> changes the theme.
>
> Bug: webrtc:8803
> Change-Id: I3ea23a25b27f196fcffd018c7cdd2ff6255b62d9
> Reviewed-on: https://webrtc-review.googlesource.com/44400
> Commit-Queue: Sami Kalliomäki <sakal@webrtc.org>
> Reviewed-by: Anders Carlsson <andersc@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#21788}
Bug: webrtc:8803
Change-Id: Ie79d7f7b9e8fea255581d3b56772b5bb6fffff4c
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139881
Commit-Queue: Sami Kalliomäki <sakal@webrtc.org>
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28137}
Currently, if LNTF and NACK messages are both created, they will
be sent out in separate RTCP messages. This is wasteful.
This CL is the first of in a series of CLs that will ensure that
these feedback messages can be buffered together, without introducing
more of a delay than the CPU time required to process both messages.
Bug: webrtc:10336
Change-Id: I950324112ee346695a12a17d025483ea5e99c732
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139112
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Commit-Queue: Elad Alon <eladalon@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28136}
This change makes SwapQueue lockless in order to reduce lock contention
in the Audio Processing Module.
Bug: webrtc:10205
Change-Id: Idc3b2a85e959b467bc1653492e48eee42e236fa5
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/138901
Commit-Queue: Gustaf Ullberg <gustaf@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28135}
Instead of setting a runtime, allow neteq_quality_test to
consume a complete file using --runtime_ms -1
Bug: webrtc:10690
Change-Id: I90d35cf31996d9336fef817b9332a2cd1d04e77e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139101
Reviewed-by: Minyue Li <minyue@webrtc.org>
Commit-Queue: Pablo Barrera González <barrerap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28134}
H264 requires that the packetization modes are the same in order to
consider the code the same. This logic was added to VideoCodec::Matches
but was not reflected in IsSameCodec. This could manifest itself when a
remote description with an unsupported packetization mode is set.
Bug: webrtc:10693
Change-Id: Icda07f7d56a464895d2267a41cc0f2fd9d5f42ff
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/138983
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Amit Hilbuch <amithi@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28126}
This reverts commit 4436887ed2d3324279e0f2e091c9e9355392721a.
Reason for revert: The original revert was intended to be temporary.
Original change's description:
> Revert "Change default secure SCTP protocol to UDP/DTLS/SCTP"
>
> This reverts commit c3f4820e129d44471b366b8885a67b5392918d5a.
>
> Reason for revert: Will temporarily revert to fix an issue and reland afterwards.
>
> Original change's description:
> > Change default secure SCTP protocol to UDP/DTLS/SCTP
> >
> > The old value - DTLS/SCTP - is not standards conformant,
> > and the new value should be parsable since Chrome M61.
> >
> > Bug: webrtc:7706
> > Change-Id: I7468cc9597dec4ef4b102fccddc4e981fed7e8d8
> > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/136804
> > Reviewed-by: Steve Anton <steveanton@webrtc.org>
> > Commit-Queue: Harald Alvestrand <hta@webrtc.org>
> > Cr-Commit-Position: refs/heads/master@{#27940}
>
> TBR=steveanton@webrtc.org,mbonadei@webrtc.org,hbos@webrtc.org,hta@webrtc.org
>
> # Not skipping CQ checks because original CL landed > 1 day ago.
>
> Bug: webrtc:7706
> Change-Id: Ida8ae20767485c75edc44dff8a3fa1af2006f207
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139244
> Reviewed-by: Guido Urdaneta <guidou@webrtc.org>
> Reviewed-by: Henrik Boström <hbos@webrtc.org>
> Commit-Queue: Guido Urdaneta <guidou@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#28121}
TBR=steveanton@webrtc.org,mbonadei@webrtc.org,hbos@webrtc.org,hta@webrtc.org,guidou@webrtc.org
Change-Id: I381fa18b644874c20ddaa4cd13fec79a5fd9555a
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:7706
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139246
Reviewed-by: Guido Urdaneta <guidou@webrtc.org>
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Guido Urdaneta <guidou@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28122}
This reverts commit c3f4820e129d44471b366b8885a67b5392918d5a.
Reason for revert: Will temporarily revert to fix an issue and reland afterwards.
Original change's description:
> Change default secure SCTP protocol to UDP/DTLS/SCTP
>
> The old value - DTLS/SCTP - is not standards conformant,
> and the new value should be parsable since Chrome M61.
>
> Bug: webrtc:7706
> Change-Id: I7468cc9597dec4ef4b102fccddc4e981fed7e8d8
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/136804
> Reviewed-by: Steve Anton <steveanton@webrtc.org>
> Commit-Queue: Harald Alvestrand <hta@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#27940}
TBR=steveanton@webrtc.org,mbonadei@webrtc.org,hbos@webrtc.org,hta@webrtc.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: webrtc:7706
Change-Id: Ida8ae20767485c75edc44dff8a3fa1af2006f207
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139244
Reviewed-by: Guido Urdaneta <guidou@webrtc.org>
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Guido Urdaneta <guidou@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28121}
The datagram sink needs to know when datagrams are lost in addition to
when they are acked.
DatagramAck::receive_timestamp needs a default value so that
DatagramAck's default ctor is not implicitly deleted. Without a default
ctor, it's not possible to make this struct without specifying all its
fields, so users will still be broken when the interface adds a new
field.
Bug: webrtc:9719
Change-Id: I6688a938d68eea133f12b13a1228d4df4771d1b8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139480
Reviewed-by: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28117}
Since ScopedFieldTrials is not thread safe, if it goes out of
scope while other threads are still running and possibly querying
the field trials then it's possible to hit an MSAN failure.
Bug: webrtc:10694
Change-Id: I93c94f1008e4478d98ec1545bbc0a7536739e479
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139460
Reviewed-by: Qingsi Wang <qingsi@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28116}
Whenever a datagram is acked, the datagram transport will provide the
remote peer's receive timestamp in this field.
Bug: webrtc:9719
Change-Id: I516b9d602e62179a3deda001e0ee9b484aa20d37
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139440
Reviewed-by: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28114}
It does not make sense for DtlsTransport to own ICE, and this arrangement will
not work when negotiating datagram or DTLS transport. During negotiation, both
a DTLS transport and a datagram transport need to be ready to receive from the
same ICE transport, depending on which protocol is chosen by the answerer. Once
the answerer chooses a protocol, the transport that is not chosen must be
deleted, but ICE must be left intact for use by the remaining transport.
Bug: webrtc:9719
Change-Id: Ibab969b574c981e3834ced71f8ff88008cb26a6c
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/139340
Reviewed-by: Anton Sukhanov <sukhanov@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Commit-Queue: Bjorn Mellem <mellem@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#28113}
