Summary:
We currently pause all producers if the sum of the egress buffers of all transactions exceeds the write buffer limit. This turns out to be deterimental to prioritization.
Now, we pass the underlying transport pause state or connection flow control state back to the handlers. The previous diff in this stack introduces a per-stream buffer limit (64kb default). To limit total session buffer size, limit the number of concurrent streams or lower the per-stream limit.
Reviewed By: lnicco
Differential Revision: D17097138
fbshipit-source-id: 9025c5be8b318963311c3aaad9ee9a03c0e2265e
Summary:
I think not doing this, and throw an error at the point of output to
files is kinda surprising to users
Reviewed By: sharma95
Differential Revision: D18016183
fbshipit-source-id: b16881b365ea82b75842f89dc364054804808116
Summary:
Mainly based on Junqi's changes in D15230415
Added the start of several error checking. Section 19.15
of d-23
(https://tools.ietf.org/html/draft-ietf-quic-transport-23#section-19.15)
specifies 2 cases where PROTOCOL_VIOLATION should be thrown during processing:
* retire_prior_to is larger than sequence number
* new_connection_id is sent by peer using 0-len conn id (will handle later)
I don't think it was specified what happens if we get a NEW_CONN_ID frame
after the the endpoint already has enough of the peer's conn ids - just ignore
the frame.
Reviewed By: sharma95
Differential Revision: D17578167
fbshipit-source-id: 6e410b6110970a6e52970576ac3ff2b1c7d5c06a
Summary:
Prior to this we had an incorrect check for something being a stream frame. It technically should have worked for minimally encoded frame types, but would not work for a spec-incompliant frame type.
Instead of using a mask this makes it so each stream frame type is explicitly enumerated. This isn't as clean looking but is easier to validate as correct.
This also revealed some places where we were not correctly plumbing through the version for unit tests, so fixing those as a driveby.
Reviewed By: yangchi
Differential Revision: D16676631
fbshipit-source-id: 835dcc6e1f431bbe3fa4a5c6b8e616863c126155
Summary:
Previously we tried to put this API in connectionCallback, but that has
some lifecycle problem. So we decided to make this API independent of mvfst's
state. Application (e.g. HTTP/3, QPACK) will have to either make the 2 functions
stateless or maintain the lifecycle of any captured objects.
Reviewed By: afrind, lnicco
Differential Revision: D16074471
fbshipit-source-id: b0a3baa6870bee56372797754972ab5c2c1f2232
Summary:
This diff adds a DebugState in the QuicConnectionState to track the
reason we schedule transport WriteLooper to run, the reason we end up not
writing and the number of times such write happens consecutively. And when it
reaches a predefined limit, we trigger a callback on a loop detector interface.
Reviewed By: kvtsoy
Differential Revision: D15433881
fbshipit-source-id: d903342c00bc4dccf8d7320726368d23295bec66
Summary:
Previously updateAppLimited was not accounting for peer created streams.
This moves app limited accounting into the stream manager
This also makes a subtle change to the applimited callback. We assume that
we start off as not app limited now, so if we create a stream we will not call
the congestion controller app limited callback.
Reviewed By: mjoras
Differential Revision: D15166114
fbshipit-source-id: 1a8d573533861f53bb1bd9fdc605dfefe68902dc
Summary: The previous diff removed automatically sending a rst stream when receiving a reset stream. This adds a helper to do that from the application layer.
Reviewed By: mjoras, pkir
Differential Revision: D15208997
fbshipit-source-id: b8d3898e85537e97df61cc66854299698a48c825
Summary:
For the sender, upon egress or ingress skip we must check registered
delivery callbacks and invoke `onCancel()` for each offset below the skipped
offset.
Reviewed By: mjoras
Differential Revision: D15113113
fbshipit-source-id: 221badce8e92abde48e8ab2c8f3c1da9e506f54e
