chown, chgrp: report a verbose stdout write error instead of panicking

[leeewee]

Fixes #10572

The verbose output in the shared ChownExecutor (uucore/src/lib/features/perms.rs) was printed with println!, which aborts (exit 134) when stdout can't be written — e.g. output redirected to a full device. Because the site is shared, both chown and chgrp abort, so this also covers the chgrp case of the same bug:

$ chown -v $(whoami) missing > /dev/full        # and: chgrp -v <group> missing > /dev/full
thread 'main' panicked at library/std/src/io/stdio.rs:...:
failed printing to stdout: No space left on device
Aborted (core dumped)          # exit 134

Verbose lines now go through a helper that propagates the write failure: on error it reports write error: <reason> to stderr and returns a non-zero code, matching GNU (exit 1). The three stdout println! sites (the obtain_meta → None branch in traverse, and both print_verbose_ownership_retained_as lines) use it, and print_verbose_ownership_retained_as now returns its exit code to its callers.

After:

$ chgrp -v <group> missing > /dev/full
chgrp: cannot dereference 'missing': No such file or directory
chgrp: write error: No space left on device          # exit 1

This matches GNU (exit 1, write error: No space left on device). Regression tests are added for both chgrp and chown (/dev/full, Linux-gated), each exercising the trigger with only stdout on the full sink.

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/date/resolution (fails in this run but passes in the 'main' branch)
Skipping an intermittent issue tests/pr/bounded-memory (passes in this run but fails in the 'main' branch)

[xtqqczze]

Please use the translate! macro, and preferably a PermsError enum, like #13397.

[leeewee]

Updated per review:

• Switched to the thiserror + translate! pattern — added a PermsError::Write(io::Error) enum reusing the existing common-write-error locale key (mirrors checksum/mod.rs), so write_verbose_line no longer hand-rolls the message.
• While here, aligned the reporting with GNU's close_stdout: the write error is now detected once at a single final stdout().flush() in exec() and reported a single time (previously it printed once per failed verbose line). The two main.rs files use bin!(…, no_flush) so the util owns that flush and there's no extra Error flushing stdout: line. Verified byte-for-byte against GNU on /dev/full (exit 1, one write error: line).

[codspeed-hq]

Merging this PR will improve performance by 7.71%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
✅ 338 untouched benchmarks
⏩ 46 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	complex_relative_date	148 µs	137.4 µs	+7.71%

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing leeewee:perms-fix-verbose-write-error (d2d44ea) with main (2772896)}

Footnotes

1. 46 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. ↩