-
Notifications
You must be signed in to change notification settings - Fork 4.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
undefined behavior in quic integration test #34353
Labels
Comments
zuercher
added
area/quic
area/test flakes
and removed
triage
Issue requires triage
labels
May 28, 2024
cc @danzh2010 |
UBSAN alerts in Envoy::Http::Http1::ServerConnectionImpl, instead of QUIC objects. I'm not an expert in that code stack though. |
@alyssawilk I don't think this is related to HTTP/3. The call stack has |
Sure but AFIK it's only showing up in an H3 test, so it may be a lifetime bug in that particular test |
This was referenced Jun 11, 2024
alyssawilk
pushed a commit
that referenced
this issue
Jun 14, 2024
…NoBPF (#34697) Commit Message: switch to use autonomous upstream in testMultipleQuicConnections(). The ASAN complains about the multiple resets of fake_upstream_connection_ during the tests which races with the FakeUpstream's network connection's I/O events in the upstream thread. This test helper function doesn't need to use controlled test upstream, so using autonomous upstream should avoid the race. This potentially fix #34353 which I wasn't able to reproduce locally under ASAN. This test has to run with --jobs=1 if we set --runs_under_test to be larger than 1. 1000 runs already took more than 1h to finish. Risk Level: low, test onlhy Testing: existing test passes under ASAN Docs Changes: N/A Release Notes: N/A Platform Specific Features: N/A Fixes #34353 Signed-off-by: Dan Zhang <[email protected]> Co-authored-by: Dan Zhang <[email protected]>
Nealsoni00
pushed a commit
to Nealsoni00/envoy
that referenced
this issue
Jun 18, 2024
…NoBPF (envoyproxy#34697) Commit Message: switch to use autonomous upstream in testMultipleQuicConnections(). The ASAN complains about the multiple resets of fake_upstream_connection_ during the tests which races with the FakeUpstream's network connection's I/O events in the upstream thread. This test helper function doesn't need to use controlled test upstream, so using autonomous upstream should avoid the race. This potentially fix envoyproxy#34353 which I wasn't able to reproduce locally under ASAN. This test has to run with --jobs=1 if we set --runs_under_test to be larger than 1. 1000 runs already took more than 1h to finish. Risk Level: low, test onlhy Testing: existing test passes under ASAN Docs Changes: N/A Release Notes: N/A Platform Specific Features: N/A Fixes envoyproxy#34353 Signed-off-by: Dan Zhang <[email protected]> Co-authored-by: Dan Zhang <[email protected]> Signed-off-by: Neal Soni <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Log https://dev.azure.com/cncf/envoy/_build/results?buildId=171258&view=logs&jobId=1439b9f7-a348-5b50-b5fe-ea612ea91241&j=1439b9f7-a348-5b50-b5fe-ea612ea91241&t=1002ac43-da84-5fae-70b2-98833b702d09&s=4f7d954b-a765-565f-91a2-c04870dab43f
The text was updated successfully, but these errors were encountered: