gh-159606: Skip test_faulthandler signal tests on ASan

[vstinner]

The address sanitizer can catch signals such as SIGBUG and then hangs.

• Issue: test_faulthandler fails randomly on x86-64 MacOS Intel ASAN NoGIL 3.x #159606

[vstinner]

!buildbot x86-64 MacOS Intel ASAN NoGIL PR

[bedevere-bot]

🤖 New build scheduled with the buildbot fleet by @vstinner for commit 6f18149 🤖

Results will be shown at:

https://buildbot.python.org/all/#/grid?branch=refs%2Fpull%2F159607%2Fmerge

The command will test the builders whose names match following regular expression: x86-64 MacOS Intel ASAN NoGIL PR

The builders matched are:

• x86-64 MacOS Intel ASAN NoGIL PR

[vstinner]

I'm not sure why test_sigbus() hangs on "x86-64 MacOS Intel ASAN NoGIL 3.x". Since we already skip signal tests on TSan/UBSan/MSan, I think that it should be fine to also skip them on ASan.

cc @StanFromIreland

[vstinner]

Maybe the problem is on the macOS machine: see #159608 (comment). The bug was only seen on this machine for 2 days, whereas the code didn't change.

[StanFromIreland]

This does seem a little odd, but sure it won't hurt I suppose.

[itamaro]

the buildbot worker machine has been recycled and reprovisioned.

I think this PR is still worth merging, regardless of the machine!

[vstinner]

@StanFromIreland

This does seem a little odd, but sure it won't hurt I suppose.

@itamaro:

I think this PR is still worth merging, regardless of the machine!

test_faulthandler pass on Linux ASan. It also passed on this macOS ASan buildbot for many months, until only 2 days ago.

@itamaro rebooted the machine and test_faulthandler pass again. So I don't think that this change is needed. I close my PR.

If the situation changes, we can reconsider this change obviously.