This CL can prepare downstream projects for being aware of
this new handover state.
This was extracted from change 299076.
Bug: webrtc:14997
Change-Id: I35bfbe040ffbaa5d7266eb67d58078b66083337a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/302980
Reviewed-by: Sergey Sukhanov <sergeysu@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39927}
This reverts commit a736f30a5fddfa9a6af02a0a916da09bcac49d0d.
Due to a downstream project not supporting this
new handover state, it fails. Handover state needs
to be submitted first, then downstream project needs
to be updated, and finally the code changes can be
submitted.
Bug: webrtc:14997
Change-Id: I8551e349408d9bf4eb593cb948279d659467fe20
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/302821
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Auto-Submit: Victor Boivie <boivie@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39923}
If configured, attempt to negotiate "zero checksum
acceptable" capability, which will make the outgoing
packets have a fixed checksum of zero. Received
packets will not be verified for a correct checksum
if it's zero.
Also includes some boilerplate:
- Setting capability in state cookie
- Adding capability to handover state
- Adding metric to know if the feature is used
This feature is not enabled by default, as it will be
first evaluated with an A/B experiment before making
it the default.
When the feature is enabled, lower CPU consumption for
both receiving and sending packets is expected. How
much depends on the architecture, as some architectures
have better support for generating CRC32 in hardware
than others.
Bug: webrtc:14997
Change-Id: If23f73e87220c7d42bd4f9a92772cda16bc18fcb
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/299076
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39920}
This reverts commit 45eae346930aedbf4624d054e0633c94b397b8ec.
It was found not to be the root cause of the performance
regression, so it's safe to reland.
Bug: webrtc:14997
Change-Id: I67c90752875bf4071cbdd5adfa462a37f4d4ceab
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/302162
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#39910}
This reverts commit bd46bb7660fbc9f31799196058f7a1957f50aa31.
Reason for revert: There is a slight performance degradation
pointing to this CL, so revert this to be able to confirm if
it is the culprit.
Original change's description:
> dcsctp: Support zero checksum packets
>
> If configured, the packet parser will allow packets with
> a set checksum of zero. In that case, the correct checksum
> will not even be calculated, avoiding a CPU intensive
> calculation.
>
> Also, if specified when building a packet, the checksum can
> be opted to be not calculated and written to the packet.
> This is to be used when draft-tuexen-tsvwg-sctp-zero-checksum
> has been negotiated, except for some packets during association
> establishment.
>
> This is mainly a preparation CL and follow-up CL will enable
> this feature.
>
> Low-Coverage-Reason: Affects debug logging code not run in tests
> Bug: webrtc:14997
> Change-Id: I3207ac3a626df039ee2990403c2edd6429f17617
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/298481
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Commit-Queue: Victor Boivie <boivie@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#39737}
Bug: webrtc:14997
Change-Id: Ie22267abb4bcd25d5af07875eb933c51ed5be853
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/301580
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39878}
This component is mostly "glue" and is heavily tested in the
socket tests, but not the ToString method, which results in
getting "low test coverage" warnings.
So for the sake of it, add a test that verifies that it works.
Bug: None
Change-Id: Id2b75e2798f334452be50631ef1ff15f53fe4157
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/300441
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39826}
If configured, the packet parser will allow packets with
a set checksum of zero. In that case, the correct checksum
will not even be calculated, avoiding a CPU intensive
calculation.
Also, if specified when building a packet, the checksum can
be opted to be not calculated and written to the packet.
This is to be used when draft-tuexen-tsvwg-sctp-zero-checksum
has been negotiated, except for some packets during association
establishment.
This is mainly a preparation CL and follow-up CL will enable
this feature.
Low-Coverage-Reason: Affects debug logging code not run in tests
Bug: webrtc:14997
Change-Id: I3207ac3a626df039ee2990403c2edd6429f17617
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/298481
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39737}
This will be included in INIT/INIT_ACK if the socket has
configured the "accept zero checksum" parameter, that will
be added in follow-up CLs.
Bug: webrtc:14997
Change-Id: I1a2823fbc77cfea8fe746b07c1c77593bc15efe9
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/298480
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39702}
The log_prefix frequently used in dcSCTP is intended to be used
to separate logs from different sockets within the same log output,
typically in unit tests. Every log entry always has the file and
line, so it's not important to add more information to the log prefix
that indicates _where_ it's logged. So those have been removed.
Also, since log_prefix is a string (typically 32 bytes) and it's
never changing during the lifetime of the socket, pass and store it
as a absl::string_view to save memory.
Bug: None
Change-Id: I10466710ca6c2badfcd3adc5630426a90ca74204
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/274704
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39571}
This CL migrates unit tests to the new TaskQueueBase interface.
Bug: chromium:1416199
Change-Id: Ic15c694b28eb67450ac99fdd56754de1246a4d95
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/295621
Commit-Queue: Markus Handell <handellm@webrtc.org>
Reviewed-by: Danil Chapovalov <danilchap@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#39434}
This tests all of the known integer unwrappers in WebRTC
* rtc::TimestampWrapAroundHandler
* webrtc::Unwrapper<T>
* webrtc::SequenceNumberUnwrapper<T>
* dcsctp::UnwrappedSequenceNumber
The conformance tests show differences between in behavior between the
unwrappers.
This change also adds const accessors for the internal values of dcsctp::UnwrappedSequenceNumber.
Bug: webrtc:13982
Change-Id: Ia1b483d7ef5ceb43fa4d93013a76cd0251a58c22
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/287620
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Evan Shrubsole <eshr@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#38900}
Most of the changes are trivial.
Bug: webrtc:14432
Change-Id: I0444527bf57c72c8d65f69754b4a4a1c1d7b2e92
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/275340
Reviewed-by: Erik Språng <sprang@webrtc.org>
Reviewed-by: Mirko Bonadei <mbonadei@webrtc.org>
Commit-Queue: Mirko Bonadei <mbonadei@webrtc.org>
Reviewed-by: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#38074}
This was a mistake from change 273800 in that it could try to send
packets if there wasn't a connection established - when tcb_ was
nullptr.
Bug: chromium:1360268
Change-Id: Idd4406071dbd8ac89303aef61840895505417569
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/274405
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Auto-Submit: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#38031}
The RTT will quickly be updated whenever a DATA chunk is acked, but
the initial value was set to zero. In unit tests, which often are
short and rarely increase the simulated time between a DATA is sent,
and acked, the smoothed RTT would usually stay at 0, which causes
e.g. the rate limiting of FORWARD not to work.
Bug: None
Change-Id: Ieb515fe875ce88d001777b00d6efd9762565a09d
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/273900
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#38013}
When re-receiving a stream reset request with the same request
sequence number, reply with the same result as previous time. In
case the original request was deferred, and "in progress" was
replied, it's important to not indicate that it was performed
successfully as that will make the sender believe it has completed
before it did.
Bug: webrtc:14277
Change-Id: I5c7082dc285180d62061d7ebebe05566e5c4195c
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/274080
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#38012}
When a stream reset response has been received, this may result
in unpausing the streams (either because it was successful or
because it failed - but streams will be unpaused). Directly after
receiving the response, send out any pending chunks that are
ready to be sent.
Before this CL, they would eventually be sent out, but that is
unnecessary and undeterministic.
Bug: webrtc:14277
Change-Id: Ic1ab38bc3cea96cfec7419e25001f12807523a3a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/273800
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#38009}
When a RECONFIG has been received with a last assigned TSN that is
not yet seen by the receiver, it will enter deferred reset mode
(https://www.rfc-editor.org/rfc/rfc6525#section-5.2.2, E2).
When more DATA is received, moving the cumulative acknowledgment point,
the request will finally be processed. But the last chunk that has the
same TSN as the last assigned TSN was before this CL not applied before
doing the reset - it was applied after.
This would result of a message getting lost or possibly getting
truncated or incorrectly merged with another.
Handling the message before resetting the stream is the simple
solution here.
Bug: webrtc:14277
Change-Id: Iea9fa227778077a9ff2f78bc77b5d93cc32b702b
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/273323
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37993}
To allow the transport to be able to know which ranges of
stream identifiers it can use, the negotiated incoming/inbound
and outgoing/outbound stream counts will be exposed. They are
added to Metrics, and guaranteed to be available from within
the OnConnected callback.
In this CL, dcSCTP will not validate that the client is sending
on a stream that is within the negotiated bounds. That will be
done as a follow-up CL.
Bug: webrtc:14277
Change-Id: Ic764e5f93f53d55633ee547df86246022f4097cf
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/272321
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37876}
To allow the transport to be able to know which ranges of
stream identifiers it can be use, the negotiated incoming/inbound
and outgoing/outbound stream counts will be exposed.
This is first added to handover state, with the actual implementation
to follow.
Bug: webrtc:14277
Change-Id: Idd821ecbd8fcb588c88d69f617889318b4b03d43
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/272320
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37863}
This adds the final piece, which makes the socket and the retransmission
queue generate the callbacks.
Bug: webrtc:5696
Change-Id: I1e28c98e9660bd018e817a3ba0fa6b03940fcd33
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/264125
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37455}
For all messages where the last fragment was _not_ put on the wire, the
send queue is responsible for generating lifecycle events, but once all
fragments have been put on the wire, it's the retransmission queue that
is responsible. It does that by marking the final fragment of a message
with the lifecycle identifier, and once that message has been fully
acked by the cumulative ack TSN, it's clear that the peer has fully
received all fragments of the message, as the last fragment was acked.
For abandoned messages - where FORWARD-TSNs are sent, those will be
replied to by a SACK as well, and then we report abandoned messages
separately, to later trigger `OnLifecycleMessageExpired`.
This CL adds support in OutstandingData, which doesn't generate the
callbacks itself, but just reports them in the AckInfo.
Bug: webrtc:5696
Change-Id: I64092f13bcfda685443b7df9967b04d54aedd36a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/264124
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37448}
The send queue is responsible for generating lifecycle events for all
messages that are still in the queue. Because, if they are still in the
queue, that means that the last fragment of the message hasn't been sent
yet (because then it would have been in the retransmission queue
instead). And if the last fragment hasn't been sent, the send queue is
responsible for generating the
`OnLifecycleMessageExpired(/*maybe_sent=*/false)` event.
Bug: webrtc:5696
Change-Id: Icd5956d6aa0f392cae54f2a05bd20728d9f7f0a6
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/264144
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37419}
Minor refactoring of the API, to put optional arguments last. Also
changed internal structures to reflect that order, for consistency.
Also reduced size of Item from 88 to 72 bytes, by packing fields better.
Bug: webrtc:5696
Change-Id: I1b9d50831a8e9a358224682d06a782a3269b8416
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/264123
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37403}
Let the OutgoingStream reference the parent instead of passing
references to individual items it needs, as follow-up CLs will add even
more items.
No functional change - pure refactoring.
Bug: webrtc:5696
Change-Id: I914e590c0d90e898d7d230a16170cf4faff2338c
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/264142
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37398}
This CL adds the API to enable message lifecycle events to be generated.
Those can in turn be used to generate metrics, e.g. latency metrics
tracking the time to send a message, the time until it's acknowledged,
and metrics tracking how often messages are expired.
This will be used to validate that message interleaving really improves
latency for high priority data channels.
The actual implementation of the API will be provided in follow-up CLs.
Bug: webrtc:5696
Change-Id: Ic06f8244d1c79a336975e35479130521dff17519
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/264141
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Reviewed-by: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37396}
There was also some refactoring to create the TCB at the same time,
to ensure the metric is always set.
Bug: webrtc:13052, webrtc:5696
Change-Id: I5557ad5f0fc4a0520de1eaaafa15459b3200c4f5
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/262259
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37388}
Before this CL, some components, e.g. the SendQueue, was first created
and then later restored from handover state, while some were created from
the handover state, as an optional parameter to their constructors.
This CL will make it consistent, by always creating the components in a
pristine state, and then modifying it when restoring them from handover
state. The name "RestoreFromState" was used to be consistent with SendQueue
and the socket.
This is just refactoring.
Bug: None
Change-Id: Ifad2d2e84a74a12a93abbfb0fe1027ebb9580e73
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/267006
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37384}
This adds support to enable message interleaving in the stream scheduler
from the socket, proxied by the send queue.
It also adds socket unit tests to ensure that prioritization and
interleaving works. Also, send queue test has been added to validate the
integration of the stream scheduler. But the actual scheduling parts of
it will be tested in the stream scheduler unit tests.
Bug: webrtc:5696
Change-Id: Ic7d3d2dc28405c77a107f0148f0096882961eec7
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/262248
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37355}
This mainly modifies the stream scheduler to add a weighted fair queuing
algorithm in addition to its round robin algorithm. The WFQ algorithm is
selected whenever interleaving is enabled, to ensure that the socket
stays backwards compatible in the normal (non-interleaved) scenario.
Adaptation to send queue and socket comes in a follow-up CL.
Bug: webrtc:5696
Change-Id: I8f0dbfa8c2f40f2e84cee536ea821e7ef4af6310
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/261947
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37330}
This is a revert of the revert of commit d729d12454906d924d5a142deb3432
which was reverted because it caused upstream test failures.
Contains fix in StreamScheduler::GetActiveStreamsForTesting.
This reverts commit 5df960d3073630c5619e00d79f89937bf6fabd69.
Bug: webrtc:5696
Change-Id: I89dada257a6fb1f149f50067ab66b17e24a7c01a
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/266368
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37299}
RFC 4960, 7.2.4. Fast Retransmit on Gap Reports
4) Restart the T3-rtx timer only if the last SACK acknowledged the
lowest outstanding TSN number sent to that address, or the
endpoint is retransmitting the first outstanding DATA chunk sent
to that address.
The second part of this sentence, "or the endpoint is retransmitting the
first outstanding DATA chunk sent to that address."
This means that on fast retransmit, and if the retransmitted chunks
weren't quickly acknowledged by the peer, the retransmission timer would
expire too quickly. With this CL, it will restart, as it should.
Bug: webrtc:12943
Change-Id: I7627253718530876acc516460562e66bfcc79533
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/265620
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37193}
This reverts commit d729d12454906d924d5a142deb3432e2d5fa97ae.
Reason for revert: Breaks downstream project.
Original change's description:
> dcsctp: Use stream scheduler in send queue
>
> Changing the currently embedded scheduler that was implemented using a
> revolving pointer, to the parameterized stream scheduler that is
> implemented using a "virtual finish time" approach.
>
> Also renamed StreamCallback to StreamProducer, per review comments.
>
> Bug: webrtc:5696
> Change-Id: I7719678776ddbe05b688ada1b52887e5ca2fb206
> Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/262160
> Reviewed-by: Harald Alvestrand <hta@webrtc.org>
> Commit-Queue: Victor Boivie <boivie@webrtc.org>
> Cr-Commit-Position: refs/heads/main@{#37170}
Bug: webrtc:5696
Change-Id: Iaf3608b52a31eb31b4ca604539edb2e8ca89399b
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/265389
Auto-Submit: Victor Boivie <boivie@webrtc.org>
Commit-Queue: Tomas Gunnarsson <tommi@webrtc.org>
Bot-Commit: rubber-stamper@appspot.gserviceaccount.com <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#37172}
Changing the currently embedded scheduler that was implemented using a
revolving pointer, to the parameterized stream scheduler that is
implemented using a "virtual finish time" approach.
Also renamed StreamCallback to StreamProducer, per review comments.
Bug: webrtc:5696
Change-Id: I7719678776ddbe05b688ada1b52887e5ca2fb206
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/262160
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37170}
This adds a stream scheduler using virtual finish time (as defined in
e.g. many Fair Queuing scheduler implementations), which indicates when
a stream's next sent packet is supposed to be sent.
In the initial version, this will be used to implement a round robin
scheduler, by emulating that a stream's virtual finish time - when
scheduled - is the "one more" than all existing virtual finish times.
That will make the scheduler simply iterate between the streams in
round robin order.
The stream scheduler component is tested in isolation, and follow-up
CLs will integrate it into the send queue.
Bug: webrtc:5696
Change-Id: Iaa2c204f9b9a00517f55355cb11cfd25bb415f9e
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/261946
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37157}
Make HasDataToSend not mutate any state, and let the Produce method do
all state mutation and possibly indicate if there is nothing that can be
sent. This is helpful preparation for extracting the scheduling part of
the send queue to a separate component.
Bug: webrtc:5696
Change-Id: I132779e77d3ce6a41e5fcf4432140d3728d03cdc
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/261945
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37141}
This is the receive-side part of supporting what is frequently called
"ndata", but actually RFC8260 - "User Message Interleaving".
This CL adds a new ReassemblyStreams implementation that can assemble
I-DATA chunks and process I-FORWARD-TSN for partial reliability.
Bug: webrtc:5696
Change-Id: I3cfbea62e7b6c02fbd3f51b43ba3fb7863cf0f88
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/218506
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Reviewed-by: Harald Alvestrand <hta@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37128}
This is a re-land of commit 3180a5ad0663900a39adf4b9974052c356c835fe.
This is an issue found in fuzzer, and doesn't really happen in WebRTC
as it never closes the connection and reconnects.
The issue is that the send queue outlives any connection since you're
allowed to send messages (well, enqueue them) before the association is
fully connected. So the send queue is always present but the TCB
(information about the connection) is torn down when the connection is
closed for example. And the TCB holds the Stream Reset handler, which is
responsible for e.g. keeping track of stream reset sequence numbers and
such - which is tied to the connection.
So to ensure that the Stream Reset Handler is in charge of deciding
if a stream reset is taking place, make sure that the send queue is in
a known good state when the Stream Reset handler is created.
Bug: webrtc:13994, chromium:1320194
Change-Id: Ib8254488523c7abb58057c602f76f411fce896fa
Reviewed-on: https://webrtc-review.googlesource.com/c/src/+/265000
Reviewed-by: Florent Castelli <orphis@webrtc.org>
Commit-Queue: Victor Boivie <boivie@webrtc.org>
Cr-Commit-Position: refs/heads/main@{#37115}
