-
Notifications
You must be signed in to change notification settings - Fork 5k
Occasional crashes in System.Net.Sockets.Tests on Alpine 3.15 #85666
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
Comments
Tagging subscribers to this area: @dotnet/ncl Issue DetailsWe see rare crashes of
|
I had a quick look in the crash dump and I could not find useful info (such as which test caused it). Some parts of the heap are inaccessible (dotnet-dump kept crashing) and there is no reference to a test state machine in the thread callstack :/ |
This comes from IMO it's very unlikely that we have a production bug here. If we don't want to bother to understand why is |
man page documents this for
it may be interesting to dump the related |
Similar crash on
|
Regardless, if dotnet dump crashes it would be good to have a big open on that. |
do you remember @rzikm @antonfirsov if the original issue was |
I believe it was ARM |
Not according to the yml files.
|
Ok, I'll check if this is still happening in other platforms. If it's not, I'll close this issue. |
I would be surprised if this is specific to 3.15. More interesting question IMHO is if any of the ARM(32) images would be able to reproduce it. This looks like rare race condition and product bug |
I can't repro this locally either with x64/ARM32/ARM64 |
Triage: 1 crash on different platform (Windows.Amd64) on 6/3. No crashes for this platform and arch in last 60 days, and can't repro this locally (on x64, ARM32/64 with both Windows/Linux). Closing this for now, and we can re-open again if we see any crash. Feel free to re-open if you don't agree with the decision. |
We have started seeing this again occasionally
This was on mcr.microsoft.com/dotnet-buildtools/prereqs:azurelinux-3.0-helix-amd64 it seems, to me that there is a race between SocketTestServerAPM.Dispose and (recursive) SocketTestServerAPM.OnAccept that runs on the background. On some platform, it leads to SocketError.InvalidArgument which is rethrown from the OnAccept method. the runtime/src/libraries/Common/tests/System/Net/Sockets/SocketTestServerAPM.cs Lines 32 to 39 in 8e5da14
runtime/src/libraries/Common/tests/System/Net/Sockets/SocketTestServerAPM.cs Lines 84 to 94 in 8e5da14
|
We see rare crashes of
System.Net.Sockets.Tests
on Alpine 3.15 because of aSocketException (22): Invalid argument
being thrown from a continuation callback invoked fromSocketAsyncEngine
/SocketAsyncContext
:https://helixre107v0xdcypoyl9e7f.blob.core.windows.net/dotnet-runtime-refs-heads-main-272f116de4bf4fea9f/System.Net.Sockets.Tests/1/console.48b03307.log?helixlogtype=result
The text was updated successfully, but these errors were encountered: