[xtqqczze]

Error construction is infrequently executed compared with the normal success paths. Mark these functions with #[cold] so the compiler can treat them as unlikely code paths and keep cold code separate from hot code.

Use #[inline(never)] to prevent error-handling code from being inlined into their callers, avoiding unnecessary code-size growth and keeping the normal execution paths smaller.

We get a nice size reduction for release profile (33.7 KiB):

git checkout --quiet b198c70a2e8d2824605caf7a040d852debe1d566
cargo build --quiet --profile=release --target aarch64-apple-darwin && wc -c < target/aarch64-apple-darwin/release/coreutils
git checkout --quiet cold-error-size
cargo build --quiet --profile=release --target aarch64-apple-darwin && wc -c < target/aarch64-apple-darwin/release/coreutils

 9938064
 9903520

But but not for release-small:

git checkout --quiet b198c70a2e8d2824605caf7a040d852debe1d566
cargo build --quiet --profile=release-small --target aarch64-apple-darwin && wc -c < target/aarch64-apple-darwin/release-small/coreutils
git checkout --quiet cold-error-size
cargo build --quiet --profile=release-small --target aarch64-apple-darwin && wc -c < target/aarch64-apple-darwin/release-small/coreutils

 6057056
 6057088

Measurements taken from 3b62942 (head) vs b198c70 (base).

[codspeed]

Merging this PR will degrade performance by 10.03%

⚡ 5 improved benchmarks
❌ 2 regressed benchmarks
✅ 390 untouched benchmarks
🆕 2 new benchmarks
⏩ 54 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	three_39_bit_primes	228.6 ms	583.9 ms	-60.85%
❌	Simulation	five_38_bit_primes	1.8 s	1.8 s	-3.98%
⚡	Simulation	thirteen_39_bit_primes	9.2 s	8.6 s	+7.02%
⚡	Simulation	du_wide_tree[(5000, 500)]	15.9 ms	15.2 ms	+4.85%
⚡	Simulation	tr_single_char_replace	14.5 ms	13.9 ms	+4.26%
⚡	Simulation	unexpand_large_file[10]	286.2 ms	274.8 ms	+4.18%
⚡	Simulation	unexpand_many_lines[100000]	136.9 ms	131.4 ms	+4.17%
🆕	Memory	tr_delete_single_char	N/A	35.4 KB	N/A
🆕	Simulation	tr_delete_single_char	N/A	103.9 ms	N/A

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing xtqqczze:cold-error-size (1d5c49a) with main (968853a)²}

Footnotes

1. 54 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. ↩

2. No successful run was found on main (f29e308) during the generation of this report, so 968853a was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[github-actions]

GNU testsuite comparison:

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

[sylvestre]

gnu prints "Permission denied" with a capital P here. could you please keep the message and the io error separate?

[xtqqczze]

Right, this would be fixing a regression from #5004.

[sylvestre]

we drop e here, so any other permission-denied cause gets the same fixed text. please add a test in tests/by-util/test_cp.rs covering the new variant.