Skip to content

Conversation

@blueyed
Copy link
Contributor

@blueyed blueyed commented Apr 4, 2020

No description provided.

@blueyed blueyed force-pushed the cleanup-test_tmpdir branch from 89b5bbe to 731af26 Compare April 4, 2020 11:27
@bluetech
Copy link
Member

bluetech commented Apr 4, 2020

The test_timeout faulthandler test is very flaky recently...

@blueyed
Copy link
Contributor Author

blueyed commented Apr 4, 2020

Yeah.
Maybe since 2cc3227?

@bluetech
Copy link
Member

bluetech commented Apr 4, 2020

Maybe...

Should we increase the extended sleep added in 81a9df6 some more?

@blueyed
Copy link
Contributor Author

blueyed commented Apr 4, 2020

It's crashing - not being slow (AFAICT).

@blueyed
Copy link
Contributor Author

blueyed commented Apr 4, 2020

@bluetech
An alternative would be to pin twisted for the oldattrs job (blueyed@66d28b9#diff-b91f3d5bd63fcd17221b267e851608e8R45) - I guess the current combination triggers some bug somewhere.

@blueyed blueyed merged commit e01dcbf into pytest-dev:master Apr 4, 2020
@blueyed blueyed deleted the cleanup-test_tmpdir branch April 4, 2020 12:25
@bluetech
Copy link
Member

bluetech commented Apr 4, 2020

Hmm right, it seems like the timeout (very confusing name btw, it doesn't timeout anything...) works, but when it's running something in coverage crashes:

File "c:\hostedtoolcache\windows\python\3.8.2\x64\lib\concurrent\futures\_base.py"Windows fatal exception: access violation

I'm not sure about 2cc3227 - why would twisted start to trigger it?

Curiously the help for the faulthandler_timeout option says "Not available on Windows.", but seems like support was added in Python 3.7.

My inclination is to allow it to fail on Windows...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants