TST: periodic segfault in test suite on windows-latest

[brendan-ward]

This may be a non-issue, but tagging so we can keep an eye on it:
https://github.com/geopandas/pyogrio/actions/runs/3970511768/jobs/6807152242

Has happened 2x intermittently in the last 2 days.

[jorisvandenbossche]

So it's not a specific test that crashes (they all passed), but it segfaults at exit (clean-up of the python process). Did you see this happen before as well, or only the last two days? Because then wondering if it is related to #209

[brendan-ward]

I don't recall seeing it previously, but because it is intermittent and resolves when I run the failed job again on CI, it is hard to relate it back to a specific changeset. I also don't have a Windows machine to test on.

The only thing I can think of from #209 that would contribute is if we are calling OGR_F_Destroy on a NULL pointer, but that is specifically checked in all cases.

Just something for us to keep an eye on.

[brendan-ward]

We're still getting this; but it goes away on rerun in CI:
failed: https://github.com/geopandas/pyogrio/actions/runs/4021285808/jobs/6910581866
rerun passed: https://github.com/geopandas/pyogrio/actions/runs/4021285808/jobs/6910649547

[jorisvandenbossche]

This is happening regularly now on multiple of the Windows wheel test builds, not just the Python 3.8 one ..