`sort`: tests sometimes stack overflow on `Code Coverage (windows-latest, windows)`

[miDeb]

Example: https://github.com/uutils/coreutils/pull/2417/checks?check_run_id=2838351712 I really don't know why this happens. It seems to happen randomly on different sort tests, but I wasn't able to reproduce it in a windows vm. Interestingly the stack overflow always happens after the expected result is printed to stdout (i.e. stdout always contains the expected result, but stderr contains a message about the stack overflow) Any suggestions on how to debug this are appreciated!

[siebenHeaven]

Hello @miDeb, wanted to check what's going on here but the link points to a passing run instead of a failing one it seems? Could you please point me to where I can see failure logs?

[miDeb]

Thanks, I must have pasted the wrong link. Here's a failing run: https://github.com/uutils/coreutils/pull/2443/checks.

[siebenHeaven]

Thanks, taking a look. I'll try to see if I can reproduce locally and report back.

[siebenHeaven]

Could not reproduce the stack-overflow so far in a couple of runs, Hitting this for couple of sort tests though (with or without grcov related steps)

stdout =
 stderr = sort: Access is denied. (os error 5) at path "C:\\windows\\uutils_sortDg6kGW"
', tests\common\util.rs:160:13
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

    test_sort::test_buffer_sizes
    test_sort::test_compress_fail
    test_sort::test_merge_batch_size
    test_sort::test_merge_batches

(Note this is consistent locally for these tests)

[miDeb]

Thanks, I can actually reproduce that in a VM. I'll take a look (weird that it doesn't fail on CI though)

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.