I found one additional way a crash could occur: "OnRtpTransportChanged"
being called instead of "OnDtlsTransportChanged", due to a mixup of m=
section types. I could reproduce this by:
1. Applying description with RTP data channel m= section.
2. Applying description with both a rejected RTP data channel m=
section and rejected SCTP m= section.
This is a very strange scenario, but maybe there are other ways to
reproduce that I haven't thought of.
The solution is to combine "OnRtpTransportChanged" and
"OnDtlsTransportChanged", and not do anything with the content type.
This simplifies the code a bit as well.
Bug: chromium:827917
Change-Id: If6818ea0c41573255831534060b30c76a6544e04
Reviewed-on: https://webrtc-review.googlesource.com/70360
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22893}
This is a reland of fc43d11717e16dd427ac84fee614e5511e43cefd
Original change's description:
> Add thread checker to PortAllocator and its subclasses and fix a bug
> causing memory contention by threads.
>
> PortAllocator and its subclasses assume all of their methods except the
> constructor must be called on the same thread (the network thread in
> practice). This CL adds a thread checker to PortAllocator and its
> subclasses for thread safety, and fixes bugs of invoking some of their
> methods in PeerConnection on the signaling thread.
>
> Bug: webrtc:9112
> Change-Id: I33ba9bae72ec09a45ec70435962f3f25cd31583c
> Reviewed-on: https://webrtc-review.googlesource.com/66945
> Commit-Queue: Qingsi Wang <qingsi@google.com>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22814}
Bug: webrtc:9112
Change-Id: I5c7377f05c0daccbe469e2fdbdfacabc5c222f4c
Reviewed-on: https://webrtc-review.googlesource.com/69422
Commit-Queue: Qingsi Wang <qingsi@google.com>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22889}
RFC4566 says to use this if a session has no meaningful name. But we
were rejecting it due to another rule that says "whitespace MUST NOT be
used on either side of the = sign".
Bug: chromium:590625
Change-Id: I5d632f2cb371060adee794febe19bdfe76cb20ed
Reviewed-on: https://webrtc-review.googlesource.com/70262
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22887}
BundleFilter is replaced by RtpDemuxer in RtpTransport for payload
type-based demuxing. RtpTransport will support MID-based demuxing later.
Each BaseChannel has its own RTP demuxing criteria and when connecting
to the RtpTransport, BaseChannel will register itself as a demuxer sink.
The inheritance model is changed. New inheritance chain:
DtlsSrtpTransport->SrtpTransport->RtpTranpsort
The JsepTransport2 is renamed to JsepTransport.
NOTE:
When RTCP packets are received, Call::DeliverRtcp will be called for
multiple times (webrtc:9035) which is an existing issue. With this CL,
it will become more of a problem and should be fixed.
Bug: webrtc:8587
Change-Id: Ibd880e7b744bd912336a691309950bc18e42cf62
Reviewed-on: https://webrtc-review.googlesource.com/65786
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Benjamin Wright <benwright@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22867}
This adds confusion to the native API and is only needed for
Chromium UMA metrics, so the appropriate metrics have been moved
upstream and kDefault option removed.
Bug: chromium:811683
Change-Id: I666d7f7793765b8d6edcd99416c8b6c957766f00
Reviewed-on: https://webrtc-review.googlesource.com/59261
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22864}
This CL fixes issues when applying a description with an empty BUNDLE
group (previously it would fail, after recent refactoring it started
crashing).
This CL also will cause an empty BUNDLE group to be generated when it
should be. Namely, when responding to an offer that had a BUNDLE group,
rejecting everything in it.
Bug: chromium:831996
Change-Id: I4e705a328daef4e81f8f1ace6aa73ddfa13c0107
Reviewed-on: https://webrtc-review.googlesource.com/69720
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22844}
This was previously not working because the answerer wasn't generating
ICE credentials when it should have been.
This was fixed inadvertently by:
https://webrtc-review.googlesource.com/c/src/+/46380
But we should really also have a PeerConnection-level regression test
for this.
Bug: webrtc:6023
Change-Id: I3da900edcc8db8034ed61a7bb981d9c0e616254e
Reviewed-on: https://webrtc-review.googlesource.com/69403
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22832}
This was working before, but somewhat by accident (because an error
wasn't being surfaced).
This CL also starts surfacing that error, from
JsepTransportController::AddRemoteCandidates to PeerConnection.
Bug: None
Change-Id: Ib48c9c00ea2a5baa5f7e3210c5dc7a339498b2d0
Reviewed-on: https://webrtc-review.googlesource.com/69015
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22830}
This changes the PeerConnection when in Unified Plan mode to reject
SDP applied with SetLocalDescription or SetRemoteDescription if the
SDP has multiple "Plan B tracks" (a=ssrc lines) in a media section.
The error is to inform developers that the given SDP will not be
interpreted as they might expect.
Bug: None
Change-Id: I7a0e11282fbf63dac06038cd22a66683517a87d0
Reviewed-on: https://webrtc-review.googlesource.com/68764
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22829}
This reverts commit fc43d11717e16dd427ac84fee614e5511e43cefd.
Reason for revert: Crashes downstream tests
Original change's description:
> Add thread checker to PortAllocator and its subclasses and fix a bug
> causing memory contention by threads.
>
> PortAllocator and its subclasses assume all of their methods except the
> constructor must be called on the same thread (the network thread in
> practice). This CL adds a thread checker to PortAllocator and its
> subclasses for thread safety, and fixes bugs of invoking some of their
> methods in PeerConnection on the signaling thread.
>
> Bug: webrtc:9112
> Change-Id: I33ba9bae72ec09a45ec70435962f3f25cd31583c
> Reviewed-on: https://webrtc-review.googlesource.com/66945
> Commit-Queue: Qingsi Wang <qingsi@google.com>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22814}
TBR=deadbeef@webrtc.org,pthatcher@google.com,pthatcher@webrtc.org,qingsi@google.com,honghaiz@webrtc.org
Change-Id: I2db6561d5d6366d38caa58c3e719d0d48eda70c2
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:9112
Reviewed-on: https://webrtc-review.googlesource.com/69200
Reviewed-by: Patrik Höglund <phoglund@webrtc.org>
Commit-Queue: Patrik Höglund <phoglund@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22818}
causing memory contention by threads.
PortAllocator and its subclasses assume all of their methods except the
constructor must be called on the same thread (the network thread in
practice). This CL adds a thread checker to PortAllocator and its
subclasses for thread safety, and fixes bugs of invoking some of their
methods in PeerConnection on the signaling thread.
Bug: webrtc:9112
Change-Id: I33ba9bae72ec09a45ec70435962f3f25cd31583c
Reviewed-on: https://webrtc-review.googlesource.com/66945
Commit-Queue: Qingsi Wang <qingsi@google.com>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22814}
Reject the local/remote description trying to change the pre-negotiated
BUNDLE tag.
Reject an answer containing a BUNDLE group that's not a subset of the offered group.
Reject an offer/answer with a BUNDLE group containing a MID that no m= section has.
Reject an answer removes an m= section from an established BUNDLE group without
rejecting it.
Bug: chromium:827917
Change-Id: If334eefb00b1c1c1e24f9afba0cb00b5867f5590
Reviewed-on: https://webrtc-review.googlesource.com/67190
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22813}
This argument was previously used to implement track muting
(black frames) in the video engine, but that now happens in
the VideoTrack/VideoBroadcaster upstream.
Bug: webrtc:6983
Change-Id: Ib721b297d9fbe55b641c56690dbbd37a52edbb2f
Reviewed-on: https://webrtc-review.googlesource.com/67341
Commit-Queue: Niels Moller <nisse@webrtc.org>
Reviewed-by: Magnus Jedvert <magjed@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22785}
A downstream bug ocurred because of a lack of symmetry when adding and
removing a remote sender in Plan B that specifies SSRCs, but doesn't
specify stream IDs. The issue when the first remote description is
applied "default" for the stream ID on the remote sender, but the
second time it's applied the current remote sender's "default" stream
ID does not match the new remote description's empty stream ID. This
was incorrectly interpreted as a new remote sender (which removed/added
the sender).
Bug: webrtc:7933
Change-Id: I87191b9e887b3450ef15111b5e867023c723a86e
Reviewed-on: https://webrtc-review.googlesource.com/67191
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Noah Richards <noahric@chromium.org>
Commit-Queue: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22760}
This profile will now not be used unless the application explicitly
sets the flag in CryptoOptions to true. As a result, an 80-bit
authentication tag will be used instead of a 32-bit one. See bug for
more details.
Bug: webrtc:7670
Change-Id: I7c0a118fd7b1e7aac23b9eb8717099f055de0441
Reviewed-on: https://webrtc-review.googlesource.com/66600
Reviewed-by: Benjamin Wright <benwright@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22757}
This CL removes PushCaptureData(), which is unused.
The reason I'm removing it is since this method is cauing chromium-style
violations for all files that includes
modules/audio_device/include/audio_device_defines.h, and it's annoying
to suppress it everywhere.
Bug: webrtc:8659
Change-Id: I9133d05259075d8e8ec89b764be934f37b5fa77e
Reviewed-on: https://webrtc-review.googlesource.com/66404
Reviewed-by: Henrik Andreassson <henrika@webrtc.org>
Commit-Queue: Magnus Jedvert <magjed@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22717}
Currently in the SDP we require an a=ssrc line in the m= section in
order for a StreamParams object to be created with that
MediaContentDescription. This change creates a StreamParams object
without ssrcs in the case that a=msid lines are signaled, but ssrcs
are not. When the remote description is set, this allows us to store
the "unsignaled" StreamParams object in the media channel to later
be used when the first packet is received and we create the
receive stream.
Bug: webrtc:7932, webrtc:7933
Change-Id: Ib6734abeee62b8ed688a8208722c402134c074ef
Reviewed-on: https://webrtc-review.googlesource.com/63141
Commit-Queue: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22712}
In previous implementation, the SctpTransport always assumes the
DtlsTransport underneath is non-null, which is not true after switching
to new JsepTransportController model.
This CL adds nullptr when connecting/disconnecting the SctpTransport with
the DtlsTransport.
The "channel" related methods and variables are also renamed.
Bug: chromium:827917, chromium:828220
Change-Id: I95aa2900d23b0885f45500e2c53def771abdccad
Reviewed-on: https://webrtc-review.googlesource.com/66160
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22700}
This is a reland of f351c3408a0c7f695447a2a9f4e6a1719a0d6a26
Reland history:
The original CL broke tests in chromium which were manually tested in
the first reland. Another small fix was added to the reland to fix a
downstream bug, which caused separate tests to fail in chromium.
These were not caught because the chromium trybot was down. These
are temporarily disabled in chrome to allow this change to roll in.
Original change's description:
> Reland "Adds support for multiple or no media stream ids."
>
> This is a reland of 1550292efe680ac79a18004705c908b1cdca54cb
>
> Original change's description:
> > Adds support for multiple or no media stream ids.
> >
> > With Unified Plan SDP semantics, this adds support for specifying
> > either no media stream ids or multiple media stream ids for a
> > transceiver/sender/receiver. This includes serializing/deserializing
> > SDPs with multiple a=msid lines in a m section, or an "a=msid:-
> > <appdata>" line to indicate the no stream case. Note that this does
> > not synchronize between multiple streams, this is still just supported
> > based upon the first media stream id.
> >
> > Bug: webrtc:7932, webrtc:7933
> > Change-Id: Ib7433929af7b2925abe2824b485b360cec12f275
> > Reviewed-on: https://webrtc-review.googlesource.com/61341
> > Commit-Queue: Seth Hampson <shampson@webrtc.org>
> > Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> > Reviewed-by: Steve Anton <steveanton@webrtc.org>
> > Cr-Commit-Position: refs/heads/master@{#22611}
>
> Bug: webrtc:7932, webrtc:7933
> Change-Id: Ica272ac18088103e65cccf6b96a6d3ecccb178ed
> Reviewed-on: https://webrtc-review.googlesource.com/65560
> Commit-Queue: Seth Hampson <shampson@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Reviewed-by: Steve Anton <steveanton@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22687}
TBR=deadbeef@webrtc.org
Bug: webrtc:7932, webrtc:7933
Change-Id: Ideb30219b2f952dd51428cd4e8bd43ef49df5b17
Reviewed-on: https://webrtc-review.googlesource.com/66280
Commit-Queue: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22699}
The RTCStatsCollector was only iterating through RtpTransceivers
in order to find the active transports for which to generate stats.
But for data channel only connections, there were no
RtpTransceivers so no transports were being identified.
This CL changes the stats collector to include the transport names
of the SCTP and RTP data channel if active.
Bug: chromium:826972
Change-Id: I762b253b3bbf0f0d7861bc281b8908decbb9b0d9
Reviewed-on: https://webrtc-review.googlesource.com/65788
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22697}
The TransportController will be replaced by the JsepTransportController
and JsepTransport will be replace be JsepTransport2.
The JsepTransportController will take the entire SessionDescription
and handle the RtcpMux, Sdes and BUNDLE internally.
The ownership model is also changed. The P2P layer transports are not
ref-counted and will be owned by the JsepTransport2.
In ORTC aspect, RtpTransportAdapter is now a wrapper over RtpTransport
or SrtpTransport and it implements the public and internal interface
by calling the transport underneath.
Bug: webrtc:8587
Change-Id: Ia7fa61288a566f211f8560072ea0eecaf19e48df
Reviewed-on: https://webrtc-review.googlesource.com/59586
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22693}
This reverts commit f351c3408a0c7f695447a2a9f4e6a1719a0d6a26.
Reason for revert: Breaks chromium import
https://ci.chromium.org/p/chromium/builders/luci.chromium.try/linux_chromium_rel_ng/58012
Failin tests:
WebRtcRtpBrowserTest.TrackAddedToSecondStream
WebRtcRtpBrowserTest.TrackSwitchingStream
Original change's description:
> Reland "Adds support for multiple or no media stream ids."
>
> This is a reland of 1550292efe680ac79a18004705c908b1cdca54cb
>
> Original change's description:
> > Adds support for multiple or no media stream ids.
> >
> > With Unified Plan SDP semantics, this adds support for specifying
> > either no media stream ids or multiple media stream ids for a
> > transceiver/sender/receiver. This includes serializing/deserializing
> > SDPs with multiple a=msid lines in a m section, or an "a=msid:-
> > <appdata>" line to indicate the no stream case. Note that this does
> > not synchronize between multiple streams, this is still just supported
> > based upon the first media stream id.
> >
> > Bug: webrtc:7932, webrtc:7933
> > Change-Id: Ib7433929af7b2925abe2824b485b360cec12f275
> > Reviewed-on: https://webrtc-review.googlesource.com/61341
> > Commit-Queue: Seth Hampson <shampson@webrtc.org>
> > Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> > Reviewed-by: Steve Anton <steveanton@webrtc.org>
> > Cr-Commit-Position: refs/heads/master@{#22611}
>
> Bug: webrtc:7932, webrtc:7933
> Change-Id: Ica272ac18088103e65cccf6b96a6d3ecccb178ed
> Reviewed-on: https://webrtc-review.googlesource.com/65560
> Commit-Queue: Seth Hampson <shampson@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Reviewed-by: Steve Anton <steveanton@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22687}
TBR=steveanton@webrtc.org,deadbeef@webrtc.org,shampson@webrtc.org
Change-Id: I1835419f963762bc308a91d81c423d8e7bf65026
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:7932, webrtc:7933
Reviewed-on: https://webrtc-review.googlesource.com/65700
Reviewed-by: Tomas Gunnarsson <tommi@chromium.org>
Commit-Queue: Tomas Gunnarsson <tommi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#22690}
This is a reland of 1550292efe680ac79a18004705c908b1cdca54cb
Original change's description:
> Adds support for multiple or no media stream ids.
>
> With Unified Plan SDP semantics, this adds support for specifying
> either no media stream ids or multiple media stream ids for a
> transceiver/sender/receiver. This includes serializing/deserializing
> SDPs with multiple a=msid lines in a m section, or an "a=msid:-
> <appdata>" line to indicate the no stream case. Note that this does
> not synchronize between multiple streams, this is still just supported
> based upon the first media stream id.
>
> Bug: webrtc:7932, webrtc:7933
> Change-Id: Ib7433929af7b2925abe2824b485b360cec12f275
> Reviewed-on: https://webrtc-review.googlesource.com/61341
> Commit-Queue: Seth Hampson <shampson@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Reviewed-by: Steve Anton <steveanton@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22611}
Bug: webrtc:7932, webrtc:7933
Change-Id: Ica272ac18088103e65cccf6b96a6d3ecccb178ed
Reviewed-on: https://webrtc-review.googlesource.com/65560
Commit-Queue: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22687}
IceConfig contains a set of parameters that affect the behavior of ICE.
Inconsistent or conflicting parameters lead to erroneous or
unpredicatble behavior in the network stack. Sanity checks are now added
to validate IceConfig.
TBR=magjed@webrtc.org
Bug: webrtc:8993
Change-Id: I708bc3f1ef970872754a82a47a509bda15061ca6
Reviewed-on: https://webrtc-review.googlesource.com/60847
Commit-Queue: Qingsi Wang <qingsi@google.com>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Peter Thatcher <pthatcher@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22664}
This reverts commit 97d5e5b32c77bf550f1d788454f2db10ac9fbb1c.
Reason for revert: <INSERT REASONING HERE>
Original change's description:
> Revert "Replace BundleFilter with RtpDemuxer in RtpTransport."
>
> This reverts commit ea8b62a3e74fe91cd6bf66304839cd5677880a4e.
>
> Reason for revert: Broke chromium tests.
> Original change's description:
> > Replace BundleFilter with RtpDemuxer in RtpTransport.
> >
> > BundleFilter is replaced by RtpDemuxer in RtpTransport for payload
> > type-based demuxing. RtpTransport will support MID-based demuxing later.
> >
> > Each BaseChannel has its own RTP demuxing criteria and when connecting
> > to the RtpTransport, BaseChannel will register itself as a demuxer sink.
> >
> > The inheritance model is changed. New inheritance chain:
> > DtlsSrtpTransport->SrtpTransport->RtpTranpsort
> >
> > NOTE:
> > When RTCP packets are received, Call::DeliverRtcp will be called for
> > multiple times (webrtc:9035) which is an existing issue. With this CL,
> > it will become more of a problem and should be fixed.
> >
> > Bug: webrtc:8587
> > Change-Id: I1d8a00443bd4bcbacc56e5e19b7294205cdc38f0
> > Reviewed-on: https://webrtc-review.googlesource.com/61360
> > Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
> > Reviewed-by: Steve Anton <steveanton@webrtc.org>
> > Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> > Cr-Commit-Position: refs/heads/master@{#22613}
>
> TBR=steveanton@webrtc.org,deadbeef@webrtc.org,zhihuang@webrtc.org
>
> Change-Id: If245da9d1ce970ac8dab7f45015e9b268a5dbcbd
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Bug: webrtc:8587
> Reviewed-on: https://webrtc-review.googlesource.com/64860
> Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
> Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22614}
TBR=steveanton@webrtc.org,deadbeef@webrtc.org,zhihuang@webrtc.org
Change-Id: I3c272588ab4388ecadc4edc6786d5195c701855f
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:8587
Reviewed-on: https://webrtc-review.googlesource.com/64862
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22615}
This reverts commit ea8b62a3e74fe91cd6bf66304839cd5677880a4e.
Reason for revert: Broke chromium tests.
Original change's description:
> Replace BundleFilter with RtpDemuxer in RtpTransport.
>
> BundleFilter is replaced by RtpDemuxer in RtpTransport for payload
> type-based demuxing. RtpTransport will support MID-based demuxing later.
>
> Each BaseChannel has its own RTP demuxing criteria and when connecting
> to the RtpTransport, BaseChannel will register itself as a demuxer sink.
>
> The inheritance model is changed. New inheritance chain:
> DtlsSrtpTransport->SrtpTransport->RtpTranpsort
>
> NOTE:
> When RTCP packets are received, Call::DeliverRtcp will be called for
> multiple times (webrtc:9035) which is an existing issue. With this CL,
> it will become more of a problem and should be fixed.
>
> Bug: webrtc:8587
> Change-Id: I1d8a00443bd4bcbacc56e5e19b7294205cdc38f0
> Reviewed-on: https://webrtc-review.googlesource.com/61360
> Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
> Reviewed-by: Steve Anton <steveanton@webrtc.org>
> Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
> Cr-Commit-Position: refs/heads/master@{#22613}
TBR=steveanton@webrtc.org,deadbeef@webrtc.org,zhihuang@webrtc.org
Change-Id: If245da9d1ce970ac8dab7f45015e9b268a5dbcbd
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: webrtc:8587
Reviewed-on: https://webrtc-review.googlesource.com/64860
Reviewed-by: Zhi Huang <zhihuang@webrtc.org>
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22614}
BundleFilter is replaced by RtpDemuxer in RtpTransport for payload
type-based demuxing. RtpTransport will support MID-based demuxing later.
Each BaseChannel has its own RTP demuxing criteria and when connecting
to the RtpTransport, BaseChannel will register itself as a demuxer sink.
The inheritance model is changed. New inheritance chain:
DtlsSrtpTransport->SrtpTransport->RtpTranpsort
NOTE:
When RTCP packets are received, Call::DeliverRtcp will be called for
multiple times (webrtc:9035) which is an existing issue. With this CL,
it will become more of a problem and should be fixed.
Bug: webrtc:8587
Change-Id: I1d8a00443bd4bcbacc56e5e19b7294205cdc38f0
Reviewed-on: https://webrtc-review.googlesource.com/61360
Commit-Queue: Zhi Huang <zhihuang@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22613}
With Unified Plan SDP semantics, this adds support for specifying
either no media stream ids or multiple media stream ids for a
transceiver/sender/receiver. This includes serializing/deserializing
SDPs with multiple a=msid lines in a m section, or an "a=msid:-
<appdata>" line to indicate the no stream case. Note that this does
not synchronize between multiple streams, this is still just supported
based upon the first media stream id.
Bug: webrtc:7932, webrtc:7933
Change-Id: Ib7433929af7b2925abe2824b485b360cec12f275
Reviewed-on: https://webrtc-review.googlesource.com/61341
Commit-Queue: Seth Hampson <shampson@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Steve Anton <steveanton@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22611}
This exposes the stats selection algorithm[1] on the PeerConnection.
Per-spec, there are four flavors of getStats():
1. RTCPeerConnection.getStats().
2. RTCPeerConnection.getStats(MediaStreamTrack selector).
3. RTCRtpSender.getStats().
4. RTCRtpReceiver.getStats().
1) is the parameterless getStats() which is already shipped.
2) is the same as 3) and 4) except the track is used to look up the
corresponding sender/receiver to use as the selector.
3) and 4) perform stats collection with a filter, which is implemented
in RTCStatsCollector.GetStatsReport(selector).
For technical reasons, it is easier to place GetStats() on the
PeerConnection where the RTCStatsCollector lives than to place it on the
sender/receiver. Passing the selector as an argument or as a "this"
makes little difference other than style. Wiring Chrome up such that the
JavaScript APIs is like the spec is trivial after GetStats() is added to
PeerConnectionInterface.
This CL also adds comments documenting our intent to deprecate and
remove the legacy GetStats() APIs some time in the future.
[1] https://w3c.github.io/webrtc-pc/#dfn-stats-selection-algorithm
Bug: chromium:680172
Change-Id: I09316ba6f20b25d4f9c11785d0a1a1262d6062a1
Reviewed-on: https://webrtc-review.googlesource.com/62900
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22602}
This flag (added to CryptoOptions) will allow applications to opt-in to
use of this suite, before it's disabled by default later. See bug for
more details.
TBR=magjed@webrtc.org
Bug: webrtc:7670
Change-Id: I800bedd4b26d807b6b7ac66b505d419c3323e454
Reviewed-on: https://webrtc-review.googlesource.com/64390
Commit-Queue: Taylor Brandstetter <deadbeef@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22586}
The convention is reinforced so that setting a rtc::Optional IceConfig
parameter to null restores the default value. Helper getters are added
to IceConfig to provide either user-defined value or the default.
Shared constants and config defaults used in p2p are moved to
p2pconstants.h/cc for future management with sanity checks.
Bug: webrtc:8993
Change-Id: I976cf1eef5a654b8911f449248bb2f3086279db8
Reviewed-on: https://webrtc-review.googlesource.com/61149
Commit-Queue: Qingsi Wang <qingsi@google.com>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22575}
We want to ensure that encoders and decoders actually get IDs with the
desired properties.
Bug: webrtc:8941
Change-Id: Ie64b67c2e9cb67171725d27f92e954afd1b77834
Reviewed-on: https://webrtc-review.googlesource.com/63300
Commit-Queue: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Oskar Sundbom <ossu@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22547}
This implements the stats selection algorithm[1] in RTCStatsCollector by
obtaining the selector's inbound-rtp/outbound-rtp stats and performing
the stats traversal algorithm (TakeReferencedStats)[2] on a copy of the
cached report with the rtps as starting point.
Changes:
- RTCStatsCollector.GetStatsReport() with selector arguments added.
- RequestInfo added, "callbacks_" is replaced by "requests_".
- RTCStatsReport.Copy() added.
- New test for sender selector and receiver selector,
RTCStatsCollectorTest.GetStatsWithSelector.
[1] https://w3c.github.io/webrtc-pc/#dfn-stats-selection-algorithm
[2] https://cs.chromium.org/chromium/src/third_party/webrtc/pc/rtcstatstraversal.h
Bug: chromium:680172
Change-Id: I9eff00738a1f24c94c9c8ecd13c1304452e962cf
Reviewed-on: https://webrtc-review.googlesource.com/62141
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Henrik Boström <hbos@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22499}
The eventual goal is to allow PlatformThread to use
SequencedTaskChecker, but getting to that point will require
some more detangling.
Here are (roughly) the steps taken in this CL:
* Make constructormagic a separate target.
* Move atomicops and arraysize to separate targets
* Move platform_thread_types to a separate target
* Move criticalsection to a separate target
* Move thread_checker to separate target
* Make sequenced_task_checker not depend on base_approved
* Move ptr_util to a separate target
* Move scoped_ptr to ptr_util
* Make rtc_task_queue_api not depend on base_approved
* Make sequenced_task_checker depend on rtc_task_queue_api
* Move rtc::Event to its own target
* Move basictypes.h to constructormagic
* Move format_macros and stringize_macros into constructormagic
* Rename constructormagic target to... macromagic
* Move stringencode to stringutils
* New target for safe_conversions
* Move timeutils to a new target.
* Move logging to a new target.
* Move platform_thread to a new target.
* Make refcount a new target (refcount, refcountedobject, refcounter).
* Remove rtc_base_approved from deps of TQ
* Remove a circular dependency between event tracer and platform thread.
Further steps will probably be to factor TaskQueue::Current() to not
be a part of the TaskQueue class itself and have it declared+implemented
in a target that's lower level than TQ itself. SequencedTaskChecker can
then depend on that target and avoid the TQ dependency. Once we're there,
PlatformThread will be able to depend on SequencedTaskChecker.
Attempted but eventually removed from this CL:
* Make TQ a part of rtc_base_approved
* Remove direct dependencies on sequenced_task_checker.
* Profit.
A few include-what-you-use updates along the way.
Fix a few targets that were depending on rtc_task_queue_api
Change-Id: Iee79aa2e81d978444c51b3005db9df7dc12d92a9
Bug: webrtc:8957
Reviewed-on: https://webrtc-review.googlesource.com/58480
Commit-Queue: Tommi <tommi@webrtc.org>
Reviewed-by: Karl Wiberg <kwiberg@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22487}
Downstreams have been updated, and this now updates all uses of label()
to id() within WebRTC code. This change also makes id() pure virtual and
removes label().
Bug: webrtc:8977
Change-Id: Ib045ea4fabba6f14447c64875c7aeba87dc2be24
Reviewed-on: https://webrtc-review.googlesource.com/60382
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Reviewed-by: Taylor Brandstetter <deadbeef@webrtc.org>
Commit-Queue: Seth Hampson <shampson@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#22431}
