Optionally poison modules known to not work with pyfakefs

[gsnedders]

Is your feature request related to a problem? Please describe.
https://github.com/pytest-dev/pyfakefs/blob/main/docs/troubleshooting.rst#modules-not-working-with-pyfakefs documents a number of modules, most obviously multiprocessing and subprocess, as not playing nice with pyfakefs.

Describe the solution you'd like
I'd like to have a way for tests that call into functions in these modules to raise an exception (or at least raise a warning).

Describe alternatives you've considered
Trying to track down failures based on when pyfakefs does fail (often long after the call into problematic code in those modules).

[mrbean-bremen]

Thanks, this may be a good idea! I have to check how feasible this is (also performance-wise), but I can give it a go. I see that as an add-in feature, and probably configurable regarding the excluded modules.
Not sure when I will get to it though (first have to tackle a couple of bugs).

[davidlbaird]

I advise mocking subprocess directly through mock.patch.object since the called shell would not be able to interface with the fake filesystem.