mknod: fix SELinux cleanup when context setting fails

[ChiamakaUI]

Summary

Fixes SELinux cleanup logic in mknod when setting the security context fails.

Previously, uutils attempted to clean up a partially created device node using std::fs::remove_dir, which does not work for device nodes or FIFOs. This could leave behind a mislabeled node on SELinux-enforcing systems.

Changes

• Replaced std::fs::remove_dir with std::fs::remove_file for proper cleanup
• Ensure failed SELinux context application does not leave artifacts behind

Test / Reproduction Steps

mknod --context=invalid_context_t /tmp/testnode c 1 3
ls -Z /tmp/testnode

[github-actions]

GNU testsuite comparison:

Skipping an intermittent issue tests/shuf/shuf-reservoir (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/sort/sort-stale-thread-mem (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/tail/inotify-dir-recreate (passes in this run but fails in the 'main' branch)
Congrats! The gnu test tests/basenc/bounded-memory is now passing!

[ChiamakaUI]

It looks like CI failed in an unrelated dd test (test_bytes_oseek_bytes_oflag). My changes are limited to mknod SELinux cleanup logic and do not affect dd or shared code.

Happy to rebase or rerun CI if needed.

[ChrisDryden]

Can you make a regression test for this in test_mknod.rs?

[ChrisDryden]

test_selinux_invalid in tests/by-util/test_mkdir.rs is almost identical

You can use arg("p") for FIFO and at.file_exists() instead of at.dir_exists()

[github-actions]

GNU testsuite comparison:

Skipping an intermittent issue tests/shuf/shuf-reservoir (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/sort/sort-stale-thread-mem (passes in this run but fails in the 'main' branch)

[ChrisDryden]

can you run cargo fmt?

[ChiamakaUI]

can you run cargo fmt?

I ran cargo fmt and cargo fmt --check locally — both complete successfully with no changes.