cp: keep permissions

[mtvb]

delete_dest_if_needed_and_allowed in cp.rs decided whether "cp -f" needs to recreate an existing destination by checking fs::metadata(dest)?.permissions().readonly().

Its only a mode-bit check, and ignores that a privileged process (root) can write to a file regardless of its mode bits, so it always treated a chmod 000 destination as needs recreating.

Fixes #14175

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/tail/inotify-dir-recreate (fails in this run but passes in the 'main' branch)
Skipping an intermittent issue tests/tail/tail-n0f (passes in this run but fails in the 'main' branch)
Congrats! The gnu test tests/csplit/csplit-heap is now passing!
Congrats! The gnu test tests/seq/seq-epipe is now passing!

[sylvestre]

GNU test failed: tests/cp/special-f. tests/cp/special-f is passing on 'main'. Maybe you have to rebase?

you probably regressed this test

[mtvb]

Handling for non-regular destinations was missing.
Added them, see new commit.

[codspeed-hq]

Merging this PR will not alter performance

✅ 11 untouched benchmarks
⏩ 400 skipped benchmarks¹

---

_{Comparing mtvb:cp_permissions_fix (4d6e997) with main (0d8310c)}

Footnotes

1. 400 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[oech3]

Please use rustix directly. We will remove it soon.

[mtvb]

Moved over to rustix.

test_cp_f_i_verbose_non_writeable_destination_y has been changed to check for root first, as a privileged process can write to a 000-mode file regardless.

Moved this testcase to rustix also.