Lisandro Dalcin

Tentative fix in branch [`bugfix/mpich-3.0`](https://github.com/mpi4py/mpi4py/tree/bugfix/mpich-3.0) @Xingyu-Romantic Could you please try my changes? For your convenience, the quickest way would be ```console pip install --no-cache-dir https://github.com/mpi4py/mpi4py/archive/bugfix/mpich-3.0.tar.gz ```

@Xingyu-Romantic Any update from your side?

@hppritcha Sorry for the late reply, I've been offline for a few days. I need a couple extra days to take a look about what is going on. It is...

@hppritcha I pushed a quick fix to mpi4py/mpi4py@master. Can you try restarting the workflow here?

Fantastic! Now we have phishing in GitHub comments? :facepalm:

Here you have a reproducer that at least show different behavior between hydra and gforker, and their interaction with the child process in case of SIGNIT vs no signal. ```bash...

Indeed, the following patch makes hydra and gforker behave identically with my reproducer. I think it would be a good idea to push these changes to make the two PMs...

At this point I believe these weird 221 return codes I'm getting in GitHub Actions are something specific to such virtualized environment.

Just in case this is useful to others, I paste below a self-contained hunk of monkeypatching that allowed me to get reproducible (same sha256 hash) sdist tarballs. This hunk of...

> Better to use uid = gid = 0, and set uname/gname to empty string. I've updated the code snippet as per your recommendation. Thanks.