mknod: match GNU's error for a bad MAJOR/MINOR device number

[arbelonson-source]

What

MAJOR and MINOR validate their values with clap's own
value_parser!(u32), so an unparseable or overflowing value produces
clap's own generic wording instead of GNU's:

$ mknod f c 4294967296 1

# GNU
mknod: invalid major device number '4294967296'

# uutils, before this PR
error: invalid value '4294967296' for '[major]': 4294967296 is not in 0..=4294967295

Fix

Resolve the value by hand: parse as u32, reject anything that
doesn't fit (unparseable or overflowing) with GNU's wording,
distinguishing which of the two operands (major/minor) it was for --
same wording shape as ptx --gap-size/--width (#14315), no separate
"value too large" case needed since GNU's own message doesn't
distinguish overflow from any other bad value here either.

A negative value passed as its own positional argument (e.g. mknod f c -5 5) is deliberately left alone, still reported as clap's own
"unexpected argument" rejection: GNU's own getopt also rejects it,
but as an unrecognized option (invalid option -- '5') rather
than an invalid device number -- a different and much deeper mismatch
(clap's vs. getopt's own unknown-argument handling in general, not
anything specific to this option) that's out of scope for this fix.

Testing

• cargo test --features mknod -p uu_mknod / full tests/by-util/test_mknod.rs suite: 14 passed, 0 failed.
• Fixed 2 existing tests (3 assertions) that had encoded clap's wrong wording as the expected result, one with a comment explicitly noting it ("clap generated message, thats fine").
• Manually diffed 7 cases (invalid, overflow, valid) on both major and minor, plus the negative-value positional-argument case, against GNU mknod 9.11 under LC_ALL=C (via sudo, since device-node creation requires it).
• cargo clippy -p uu_mknod --all-targets -- -D warnings and cargo fmt --check: clean.

---

This PR was written with AI assistance (Claude Opus 5, via Claude Code). I've tested the changes but please review the code carefully.

[oech3]

clap's error message is better at here because it provides what you should do too.
@arbelonson-source Please stop removing clap provided error.

[sylvestre]

i think our error msg is better, so, let's keep it.

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/rm/isatty (fails in this run but passes in the 'main' branch)
Skip an intermittent issue tests/tail/retry (fails in this run but passes in the 'main' branch)
Skipping an intermittent issue tests/cut/bounded-memory (passes in this run but fails in the 'main' branch)
Skipping an intermittent issue tests/date/date-locale-hour (passes in this run but fails in the 'main' branch)
Note: The gnu test tests/rm/many-dir-entries-vs-OOM is now being skipped but was previously passing.
Skip an intermittent issue tests/pr/bounded-memory (was skipped on 'main', now failing)