fix(quic): batch openBidiStream under one lock hold for multiplexing

Diagnosis from yet another qlog round: streams/packet still ~1 even
with the prepareRequest/awaitResponse split. Root cause: openBidiStream
is suspend due to lock.withLock, and each call releases the lock between
iterations. The send loop is queued on the lock; it grabs it the moment
we release, drains the one stream of data we just enqueued, and the
next prepareRequest call has to re-acquire after the send loop releases.
Net: one stream per drain per packet, same useless coalescing as before.

Fix is structural:
  - QuicConnection.openBidiStreamLocked() — public, lock-not-acquired
    version of openBidiStream. Caller MUST hold conn.lock.
  - GetClient.prepareRequests(authority, paths) — batch API that
    holds conn.lock once, opens + enqueues all N streams in a single
    critical section, releases. Send loop can't interject; when it
    next drains it sees ALL N streams' data ready and packs them
    into coalesced packets.
  - Http3GetClient + HqInteropGetClient: implement prepareRequests
    using openBidiStreamLocked under conn.lock.withLock { ... }.
  - InteropClient's chunked-multiplex loop: uses prepareRequests
    (batch) instead of N x prepareRequest.

Single-stream paths still use prepareRequest / get(); behavior unchanged.

The fundamental architectural improvement (per-stream / per-level lock
split, or actor-model dispatch) is a follow-up; this commit gets us
the throughput we need from the existing single-mutex shape by holding
the lock for the full chunk's worth of work.

https://claude.ai/code/session_01HcvfQq1ttPV9PkRoJb4nyT
This commit is contained in:
Claude
2026-05-07 02:06:58 +00:00
parent 7ed3d55b31
commit 57ba23519d
4 changed files with 80 additions and 20 deletions
@@ -706,22 +706,34 @@ class QuicConnection(
* check capacity proactively if the caller wants to back-pressure rather
* than throw.
*/
suspend fun openBidiStream(): QuicStream =
lock.withLock {
if (nextLocalBidiIndex >= peerMaxStreamsBidi) {
throw QuicStreamLimitException(
"peer-granted bidi stream cap reached " +
"(used=$nextLocalBidiIndex limit=$peerMaxStreamsBidi)",
)
}
val id = StreamId.build(StreamId.Kind.CLIENT_BIDI, nextLocalBidiIndex++)
val stream = QuicStream(id, QuicStream.Direction.BIDIRECTIONAL)
stream.sendCredit = peerTransportParameters?.initialMaxStreamDataBidiRemote ?: config.initialMaxStreamDataBidiRemote
stream.receiveLimit = config.initialMaxStreamDataBidiLocal
streams[id] = stream
streamsList += stream
stream
suspend fun openBidiStream(): QuicStream = lock.withLock { openBidiStreamLocked() }
/**
* The connection-lock-holding part of [openBidiStream]. Public so
* batched-multiplex callers (interop multiplexing testcase, MoQ
* audio rooms emitting many group streams in quick succession) can
* acquire [lock] ONCE and bracket multiple stream opens — preventing
* the send loop from interjecting between opens and draining one
* stream's data per packet. Caller MUST hold [lock].
*
* The non-batched paths use [openBidiStream] which holds the lock
* for one call only.
*/
fun openBidiStreamLocked(): QuicStream {
if (nextLocalBidiIndex >= peerMaxStreamsBidi) {
throw QuicStreamLimitException(
"peer-granted bidi stream cap reached " +
"(used=$nextLocalBidiIndex limit=$peerMaxStreamsBidi)",
)
}
val id = StreamId.build(StreamId.Kind.CLIENT_BIDI, nextLocalBidiIndex++)
val stream = QuicStream(id, QuicStream.Direction.BIDIRECTIONAL)
stream.sendCredit = peerTransportParameters?.initialMaxStreamDataBidiRemote ?: config.initialMaxStreamDataBidiRemote
stream.receiveLimit = config.initialMaxStreamDataBidiLocal
streams[id] = stream
streamsList += stream
return stream
}
/**
* Allocate a new client-initiated unidirectional (write-only) stream.