Before this cl, ReadyToSend signaled false if sending a packet failed and transport->GetError() returns ECONN.
ECONN may be reported by the TCP connection (TcpConnection) if the remote closed the connection. TcpConnection will attempt to reconnect and should change the writable state if it fail.
Changing the state in the context of sending packets may cause recursive
calls and seems to cause problems with incorrect states.
It is simpler if RtpTransport::SendPacket ignore these failures and
upper layers treat these lost packets similar to if the packets had been
lost via UDP.
For safety, this change can be reverted by field trial WebRTC-SetReadyToSendFalseIfSendFail/Enabled/.
Bug: webrtc:361124449 b/359989715
Change-Id: I8e7016dfb4301862286215c4512aa8ac03a16685
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/360120
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Per Kjellander <perkj@webrtc.org>
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42868}
This fix an issue where the socket does not notify the port of sent packets after that the TCPConnection has opened a new socket. TcpConnection will open a new socket if the TCP connection has been closed.
Bug: webrtc:361124449 b/359989715
Change-Id: Id33c5fc5292ee7d1c2d1cad6c373e2d4355d4fe1
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/360260
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Commit-Queue: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42865}
and update some usage to use the "correct" stun attribute names
BUG=webrtc:42229250
Change-Id: If0c34d1d9b399766d7073661ea2a5515100256a5
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/359440
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Henrik Boström <hbos@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42810}
since we do not need two adapters with TLS and DTLS modes.
SSLAdapter is the TLS adapter,
SSLStreamAdapter is the DTLS adapter.
BUG=webrtc:353750117
Change-Id: I223917c71c88437339380e1f196dcf3c0e2021c8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/354940
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42713}
This is a reland of commit e13945bf0761d34b902ecbd4e1cc6deb1788a2c9
with additional backward compability defaulting to the new value.
Original change's description:
> Enable TLS Client Hello extension permutation by default
>
> similar to the previous change for DTLS. This affects native TURN/TLS
> connections which are already using this in Chromium.
>
> BUG=webrtc:422225803
>
> Change-Id: I605f106371f2dbe23b1ad5f8385e0e01abe7c48f
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/357903
> Commit-Queue: Philipp Hancke <phancke@meta.com>
> Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#42688}
Bug: webrtc:422225803
Change-Id: Ic194e4f763029e65c1a15a6bbaabcfbcd2866eac
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/358120
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42702}
similar to the previous change for DTLS. This affects native TURN/TLS
connections which are already using this in Chromium.
BUG=webrtc:422225803
Change-Id: I605f106371f2dbe23b1ad5f8385e0e01abe7c48f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/357903
Commit-Queue: Philipp Hancke <phancke@meta.com>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42688}
This can be tested by trying to connect to a TURN server that does not
listen on a specific TCP port.
BUG=None
Change-Id: I7029112afa4b1b4376220dfc2d613a30090e4f7d
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/354901
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Markus Handell <handellm@google.com>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42595}
Some dependencies still exist but are a bit more complex to remove.
This CL removes either unused or easily replaced with ToString()
instances of ostream usage. In one case, moving the operator<<
implementation to the one test file that requires it.
Bug: webrtc:8982
Change-Id: Ia5c840b12a42893494af401317a3daf2fe50ba9b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/356240
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42582}
The new constructor exposes an already existing constructor,
and is used to create a (test) UDPPort
with a socket...so that one does not (really) need a
socket factory.
Bug: b/339018639
Change-Id: Ib591fe6ae61519fe29cdea819192694448b071e0
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/356141
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Jonas Oreland <jonaso@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42571}
Using 1 as channel_id doesn't make it clear that the goal was to
provide an invalid channel.
Bug: webrtc:345518625
Change-Id: Ie64f25b9398eafd3d0a9c8bab106e5277adef7df
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/353984
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42494}
With the new version of UBsan, this test triggers:
../../p2p/base/turn_port.cc:1728:45: runtime error: left shift of negative value -1 because of:
message->AddAttribute(std::make_unique<StunUInt32Attribute>(
STUN_ATTR_CHANNEL_NUMBER, channel_id_ << 16));
Bug: b/344601229
Change-Id: I8e730e805eb6bb70f141fc1c178d48a59575b612
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/353900
Auto-Submit: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42444}
since it contains helpers mostly related to cryptographically secure random numbers and strings.
BUG=webrtc:339300437
Change-Id: I10db939534b25dc792ac1600a4721d1b84521880
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/352620
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42441}
to the only place where it is used.
BUG=webrtc:339300437
Change-Id: I0d4124b95d19a11578efaf7e8e0a1ff1d39eb59b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/351581
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42383}
since the tie breaker is owned by the allocator now.
BUG=webrtc:42224914
Change-Id: I76bd5ae714fb2a6df38e014991242f390ae87e6a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/351180
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42371}
with an intermediate step since Chromium depends on the openssl_stream_adapter.h which will move to the new target.
BUG=webrtc:339300437
Change-Id: Iea163e0a6e3923ce8a741a2e11e9a2a1e3f3e7a3
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/350887
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42362}
Introduces a struct CreatePortArgs that is similar to the existing
CreateRelayPortArgs and contains parameters that will be passed from
the constructors of descendents of the Port class to the Port.
This struct makes it easier to add new arguments to the port
constructor without changing all inheriting classes.
Rebase of https://webrtc-review.googlesource.com/c/src/+/341021
BUG=webrtc:14626
Change-Id: Id8e5c24a36149e1699e2b42c57e52002d27c86c8
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/345860
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42339}
This is a reland of commit 47bfe39ecfe45b2f94c616ace97949003d9e87b4
Original change's description:
> Split digest methods from ssl target into digest target
>
> in an attempt to break up the monolithic ssl target.
>
> BUG=None
>
> Change-Id: I38f5b3e2828742d5d918460db1af0a5797d6a5c2
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/349764
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
> Commit-Queue: Philipp Hancke <phancke@meta.com>
> Cr-Commit-Position: refs/heads/main@{#42249}
Bug: webrtc:339300437
Change-Id: I31bb79bbc6cc55a2634176f95ec67de195974e1b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/350260
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42304}
in an attempt to break up the monolithic ssl target.
BUG=None
Change-Id: I38f5b3e2828742d5d918460db1af0a5797d6a5c2
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/349764
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Cr-Commit-Position: refs/heads/main@{#42249}
In its stead, there's now a SetEventCallback() method.
Bug: webrtc:11943
Change-Id: If936d6e1e23e8a584f06feb123ecf2d450ea4145
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/319040
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42187}
Landing when last remaining usage in Chrome has been removed.
Bug: webrtc:11943
Change-Id: I62817e2cc0b67113126b82424b6f843c77e66f31
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/341001
Reviewed-by: Per Kjellander <perkj@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42157}
This avoids problems with the Chrome component build.
Bug: webrtc:11943
Change-Id: I120628ee7829aa0255e60e2f21ac0608374340b1
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/348723
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42146}
This is a step towards removing StreamInterface::SignalEvent.
Downstream dependency will need to be updated to call FireEvent()
before further changes can land in webrtc.
Bug: webrtc:11943
Change-Id: Ia7d3f1c43fda52b7cf5bfa082aef3f462553cd67
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/347884
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42143}
Earlier attempts have shown that this signal is multiply listened to.
Bug: webrtc:11943
Change-Id: I382df9a554925d214872d788c5d7a36f2f7c7b7e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/348661
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42139}
This reverts commit dc43cb24bd8ee85d6a8224c5928ceaf90de729b6.
Reason for revert: Converted the wrong signal, should have been GatheringState.
Original change's description:
> Convert P2PTransportChannel Candidate Pair Change to CallbackList
>
> Earlier attempts have shown that this signal is multiply listened to.
>
> Bug: webrtc:11943
> Change-Id: If9130a7f4c70714b5afda5aca0469b66c8e2612f
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/347981
> Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
> Commit-Queue: Harald Alvestrand <hta@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#42124}
Bug: webrtc:11943
Change-Id: I73d5d815ced8d7aef4df765c9cf54d7637c6769d
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/348220
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#42129}
Earlier attempts have shown that this signal is multiply listened to.
Bug: webrtc:11943
Change-Id: If9130a7f4c70714b5afda5aca0469b66c8e2612f
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/347981
Reviewed-by: Tomas Gunnarsson <tommi@webrtc.org>
Commit-Queue: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42124}
This removes many references to the unsupported ProxyInfo struct
but leaves temporary implementations for methods while downstream
code gets updated.
Bug: none
Change-Id: Iab4410b362a8296b2e00cf71080010e515f9f4ba
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/344660
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#42096}
While an empty realm attribute is technically allowed, it reduces
the amount of entropy that goes into the turn credentials hash.
This remains technically broken in the implementation as hash_ is
not recomputed when changing the realm from the initial empty string
value to the empty string. Before this change this lead to hash_ not
being set and the allocate request being treated as not having
enough details to authenticate, resulting in an endless loop of packets.
BUG=chromium:329978076
Change-Id: I3d1295f905a9fb58ca5bc6f82466896f79031865
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/344820
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Philipp Hancke <phancke@meta.com>
Reviewed-by: Christoffer Dewerin <jansson@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41996}
The parse error testcase creates a random byte string
and tries to parse it as a delta attribute expecting it to fail.
Ubsan detected that there was "unsafe" static_cast<>, where
a value from network is static_casted:ed into a enum.
That enum was then *checked* for validity, so I think it was
same before aswell.
This fix changes to do the check/convering as one step.
Bug: webrtc:15392
Change-Id: Ie2534deef8988bc3c3179e194155cfd48b0ee6e5
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/343980
Reviewed-by: Jonas Oreland <jonaso@webrtc.org>
Reviewed-by: Björn Terelius <terelius@webrtc.org>
Commit-Queue: Jonas Oreland <jonaso@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41942}
This restricts the interface such that only a single onclose handler
can be set and that only one OnClose() notification will be fired.
That behavior is the same as how the previous sigslot was being
used, but the difference is that, in addition to removing sigslot,
this pattern is now more explicitly checked in the design.
Bug: webrtc:11943
Change-Id: I469c8cab3d62544988c8145b326af60b06b76d8e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/343340
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41920}
Two new socket options are introduced OPT_SEND_ECN used for setting ECN bits. OPT_RECV_ECN used for reading the ECN bits.
If ECN bits are set on received IP packets, ECT(1) and CE is propagated via rtc::ReceivedPacket.
Bug: webrtc:15368
Change-Id: I3ac335007e2f7d30564569bbc80ce47fa541bef1
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/332380
Reviewed-by: Jonas Oreland <jonaso@google.com>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Per Kjellander <perkj@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41885}
This reverts commit ed8390d21a7b15091d01bc8e843193d0a6efd23a.
Reason for revert: Fix has landed in chrome, ready to reland.
Original change's description:
> Revert "Deprecate old constructors and set_type() in Candidate and Port"
>
> This reverts commit aaa6851d53741179a591d79fc82c4dd6651a7ba5.
>
> Reason for revert: breaks chromium webrtc import
>
> Original change's description:
> > Deprecate old constructors and set_type() in Candidate and Port
> >
> > * Deprecates constructors that use string based `type`
> > * Deprecates string based type functions in favor of enum based.
> > * Restrict possible values of Candidate::type. Ensure a valid value
> > is assigned at construction.
> > * Make Port constructors protected to limit their use to subclasses.
> > - The reason for this is to make sure that use of SharedSocket()
> > is controlled (it adds a bit of complexity).
> > * Simplify construction of Port (remove Construct() etc)
> >
> > Bug: webrtc:15846
> > Change-Id: If24ed674e175642efa49da37fd2bc847dd14f613
> > Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/339860
> > Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> > Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
> > Cr-Commit-Position: refs/heads/main@{#41865}
>
> Bug: webrtc:15846
> Change-Id: Ic8b7cba97f8fb207ef51a88900e704658ade28b7
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/342140
> Auto-Submit: Ilya Nikolaevskiy <ilnik@webrtc.org>
> Owners-Override: Ilya Nikolaevskiy <ilnik@webrtc.org>
> Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
> Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
> Cr-Commit-Position: refs/heads/main@{#41867}
Bug: webrtc:15846
Change-Id: I3d52643bbb537d1c072643528828d26eb18fea94
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/342200
Reviewed-by: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#41875}
This reverts commit aaa6851d53741179a591d79fc82c4dd6651a7ba5.
Reason for revert: breaks chromium webrtc import
Original change's description:
> Deprecate old constructors and set_type() in Candidate and Port
>
> * Deprecates constructors that use string based `type`
> * Deprecates string based type functions in favor of enum based.
> * Restrict possible values of Candidate::type. Ensure a valid value
> is assigned at construction.
> * Make Port constructors protected to limit their use to subclasses.
> - The reason for this is to make sure that use of SharedSocket()
> is controlled (it adds a bit of complexity).
> * Simplify construction of Port (remove Construct() etc)
>
> Bug: webrtc:15846
> Change-Id: If24ed674e175642efa49da37fd2bc847dd14f613
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/339860
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#41865}
Bug: webrtc:15846
Change-Id: Ic8b7cba97f8fb207ef51a88900e704658ade28b7
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/342140
Auto-Submit: Ilya Nikolaevskiy <ilnik@webrtc.org>
Owners-Override: Ilya Nikolaevskiy <ilnik@webrtc.org>
Commit-Queue: Ilya Nikolaevskiy <ilnik@webrtc.org>
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#41867}
