tests: add test for priority inversion using overlapping mutexes

[haukepetersen]

rebased on #7445

This PR adds an additional test case for demonstrating priority inversion when using overlapping mutexes. As I am not sure (yet), how theoretical such a use case is in practice, I decided to put it in its own PR.

As pointed out in the comments for #7445, the solution for priority inheritance as proposed there does not work for this test case. So either I/we need to re-think #7445, or we decide that this overlapping case is not of our concern?!

[miri64]

Don't think I'm the right person to be assigned to this PR and with @bergzand being a maintainer now, I think it doesn't hurt if I revoke my assignment.

[smlng]

depending PR needs work too, postpone

[danpetry]

Hi @haukepetersen and @bergzand . Thanks for the nice work on exploring this edge case and potential solution. My initial feeling with this is that it's an edge case, and in fact the original priority inversion issue was largely theoretical - in other words, that we weren't having a problem with even the main issue, let alone edge cases. So I suggest that we just merge the main ones and pass on the edge case ones. In other words, close this issue and #7461 and save ourselves the trouble of working on it further. What do you think.
It will always be there in "closed" status in case we start having more of a problem, and people can search for closed issues on the subject, and subsequently re-open.

[bergzand]

@danpetry Sure, I agree with your reasoning. Feel free to close #7461

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you want me to ignore this issue, please mark it with the "State: don't stale" label. Thank you for your contributions.

[Teufelchen1]

@maribu given your solution in #17040, do you think this PR can be closed / reworked?

[maribu]

Yes, the upstream test does basically the same.