lib/mvfst
1
0
Fork 0
mirror of https://github.com/facebookincubator/mvfst.git synced 2025-08-08 09:42:06 +03:00
Code Activity

If there's lost data, we have data to write, take 2.

Browse Source 
Summary:
This is a bug that could prevent us from writing data if we ran out of connection flow control while we had lost data.

The last attempt missed a mistake in the scheduling of sequential priority streams.

Reviewed By: kvtsoy

Differential Revision: D33030784

fbshipit-source-id: e1b82234346a604875a9ffe9ab7bc5fb398450ed
This commit is contained in:
Matt Joras
2021-12-14 09:16:25 -08:00
committed by Facebook GitHub Bot
parent 8cde858b62
commit ca4705af0b
5 changed files with 107 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
quic/api/test/QuicTransportTest.cpp
View File

@@ -270,7 +270,17 @@ TEST_F(QuicTransportTest, NotAppLimitedWithLoss) {
auto stream = transport_->createBidirectionalStream().value();
auto lossStream = transport_->createBidirectionalStream().value();
conn.streamManager->addLoss(lossStream);
auto lossStreamState = conn.streamManager->findStream(lossStream);
ASSERT_TRUE(lossStreamState);
auto largeBuf = folly::IOBuf::createChain(conn.udpSendPacketLen * 20, 4096);
auto curBuf = largeBuf.get();
do {
curBuf->append(curBuf->capacity());
curBuf = curBuf->next();
} while (curBuf != largeBuf.get());
lossStreamState->lossBuffer.emplace_back(std::move(largeBuf), 31, false);
conn.streamManager->updateWritableStreams(*lossStreamState);
conn.streamManager->updateLossStreams(*lossStreamState);
transport_->writeChain(
stream, IOBuf::copyBuffer("An elephant sitting still"), false, nullptr);
EXPECT_CALL(*rawCongestionController, setAppLimited()).Times(0);