Detecting CMOR seg faults

[paolap]

Occasionally CMOR will produce a set fault (see their GitHub for examples of when this happens). As the segfault signal can't be detected in a try/except or other usual ways as it happens in a different package (i.e. CMOR) nothing get passed to the the working process and from multiprocessing Pool and this just goes on indefinitely until the PBS jobs reaches walltime.
I'm trying to move to concurrent to test these approaches:
https://stackoverflow.com/questions/70356876/python-multiprocessing-skip-child-segfault
so far no luck, but I haven't try yet to introduce a timeout