-
Notifications
You must be signed in to change notification settings - Fork 939
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
[Question] How can we debug gevent related segfaults? #1751
Comments
My first step in debugging would be to run with |
In addition, I would also try running with |
After a long time debugging I have managed to boil it down to a combination of a few tests running in a specific order. I am running with both I also once only got a non-segfault error at the same location which may be helpful:
I am at a loss as to how to continue debugging this. I can get in pdb in the test a few calls before the segfault. But not even sure what to check to figure out why it happens. |
This turns out to be a bug in web3 and its use of a C data structure. The same problem can happen with threads. |
Thanks a lot for all the help @jamadden. I guess we can now close this issue as there is nothing actionable on gevent's side. |
Description:
For a very long time in our project we had this constraint on gevent in requirements:
All worked well. But we decided to try and upgrade to the latest gevent since the one we use is quite old and also remove the greenlet constraint since well from my understanding these 2 go together.
But it seems that our test suite in the CI is now segfaulting at the same place iff the entire directory of tests is ran. If tests are ran individually no segfault happens. That makes it hard to debug since that directory contains test that take a long time.
The traceback is:
What I've run:
As described above we ran a particular test directory of our test suite.
Here is the run: https://github.com/rotki/rotki/pull/2132/checks?check_run_id=1731324419
And here is the PR: rotki/rotki#2132
Naturally I don't expect you guys to debug it for me. Just would like some hints as to where to look or some gotchas I may be missing for finding the root cause of this.
Keep in mind everything works fine with
gevent==1.5a2
andgreenlet==0.4.16
for some reason. I have since tried various other newer combinations and they all fail with the same traceback.Thanks for your time! And if this is not the right place to ask please point me to the right one.
The text was updated successfully, but these errors were encountered: