[eduardomourar]

Centralizes the ~460 #[cfg_attr(wasi_runner, ignore = "...")] annotations behind #[uutests::wasi_ignore(ReasonKey)], a proc-macro attribute in the new tests/uutests_procs crate.

Context: PR 14884 and PR 13687 flagged wasi_runner as a cfg duplicated across the suite as free-text strings. The tests always compile for the host, because WASI cannot spawn the
child process they launch. So target_os cannot replace the cfg. wasi_runner is the only signal that a run drives a wasm32-wasip1/wasm32-wasip2 binary through wasmtime. This change keeps the cfg and removes the string duplication. Every reason is now defined once, alphabetically, in tests/uutests_procs/src/lib.rs: 40 keys, down from 57 after merging duplicates. Several keys previously had no explanation.

#[uutests::wasi_ignore(Reason)] also applies to a mod { ... } block. It recurses into nested modules and attaches the reason to every test inside. tee's linux_only module and truncate's fifo module use this.

Not touched: structural gates that are not ignore-reasons (feat_diagnostics module gating, #[cfg(wasi_runner)] branches that select alternate passing assertions).

Ignore gates converted to compile-time gates

The PR also changes some annotations between the two WASI skip mechanisms. #[cfg_attr(wasi_runner, ignore = "...")] still compiles the test under the WASI runner and skips it at runtime. #[cfg(not(wasi_runner))] drops the test from the build under the WASI runner. The second form fits a test that is genuinely platform-specific under WASI, where no WASI API could make it pass. This change converts some annotations from the first form to the second, and the reverse in a few places. It affects cp, mv, truncate and shuf. The gate change rides along with the reason-string centralization rather than a separate PR.

[github-actions]

GNU testsuite comparison:

GNU test failed: tests/id/setgid. tests/id/setgid is passing on 'main'. Maybe you have to rebase?

[codspeed]

Merging this PR will not alter performance

1 warning, 1 note

⚠️ 124 benchmarks are dominated by system calls

System calls cannot be simulated consistently and are left out of the measure. The reported value understates the real cost. Switch to the Walltime instrument to measure them.

Learn more about system calls

See affected benchmarks

ℹ️ 1 benchmark has no flamegraph

The flamegraph wasn't generated: most of the time was spent in runtime internals, which would leave nothing useful to analyze.

See affected benchmarks

✅ 341 untouched benchmarks
⏩ 130 skipped benchmarks¹

---

_{Comparing eduardomourar:refactor/wasi-ignore-reasons (8a50254) with main (aea67c0)²}

Footnotes

1. 130 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 (90b48bf) during the generation of this report, so aea67c0 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[eduardomourar]

Maybe we can remove this whole file and add those as comments in tests/uutests_procs/src/lib.rs