-
-
Notifications
You must be signed in to change notification settings - Fork 2.7k
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
Segmentation fault in versions >=3.1.1 #2875
Comments
Can you install pytest-faulthandler and maybe run with |
First of all, thank you for your fast response. I tried what you said. Finally, I discovered that just installing In addition, I have to say that I am using the module logging and the StreamHandler (might be it is useful to discover the problem). Thank you in advance. |
I started having a similar problem in multiple conda environments. Even empty tests seem to trigger the fault. It turned out that it was importing matplotlib for some reason and that has recently started causing segfaults with the anaconda5 serices. This was fixed for me by changing matplotlib. It was not clear to me why matplotlib was being imported.
|
I'm closing this for now, I think people have found suitable workarounds. In general, seg faults are not caused by pytest itself by some other module (with C extensions), so the hard part is diagnose which one. |
As suggested in pytest-dev/pytest#2875 the segmentation faults should not happen anymore when pytest-faulthandler is installed Travis-CI occasionally gives following error message: 1893 Segmentation fault (core dumped) PYTHONWARNINGS=all pytest --cov=
I am using pytest in a project running in a docker container. The docker image that I am using is the python:3.6-slip version from resin, resin/raspberrypi3-python:3.6-slim.
This is the pip3 list:
If I run pytest -v --duration=0, I get the
Segmentation fault (core dumped)
exception. But if I use the 3.1.0 version of pytest it works correctly. Moreover, if I run the docker container using the interactive mode (options -ti) it works fine even for versions >=3.1.1.The code I am testing uses asyncio, running in ARM architecture (raspberry pi 3).
Sorry for not reporting more information about the bug, but I cannot debug in interactive mode, so I am not sure about the best way to obtain more information about it.
The text was updated successfully, but these errors were encountered: