uucore: build safe_traversal on AIX and Hurd by abendrothj · Pull Request #15030 · uutils/coreutils

abendrothj · 2026-10-02T08:08:35Z

Map the timespec time fields for AIX and Hurd inside the existing MetadataExt accessors, and gate safe_traversal and its users on Redox only, where nix has none of the *at calls.

With AIX on the safe_traversal path, chmod's AIX-only path-based symlink handling is no longer used and is removed: handle_symlink_during_traversal and its wrapper handle_symlink_during_recursion. If safe_traversal can't be brought to some other platform later, they are in this commit to restore from. The target-specific safe-traversal features of chmod, chown and install are also dropped, since perms already enables it.

Cross-checked with -Zbuild-std clippy for x86_64-unknown-hurd-gnu and i686-unknown-hurd-gnu; not run on Hurd. AIX (powerpc64-ibm-aix) doesn't build with libc 0.2.190 on main either (#15083); with libc pinned to 0.2.189 locally, uucore's safe_traversal builds. uu_du still fails to build on Hurd in fsext, as it does on main.

Closes #15028

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

codspeed · 2026-10-02T08:48:25Z

Merging this PR will degrade performance by 21.4%

⚡ 3 improved benchmarks
❌ 1 regressed benchmark
✅ 393 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	694.2 ms	-67.07%
⚡	Simulation	`thirteen_39_bit_primes`	9.2 s	8.7 s	+6.1%
⚡	Simulation	`du_wide_tree[(5000, 500)]`	15.9 ms	15.2 ms	+4.7%
⚡	Simulation	`tr_single_char_replace`	14.5 ms	13.9 ms	+4.36%
🆕	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 abendrothj:safe-traversal-aix-hurd (85d3e76) with main (968853a)²}

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. ↩
No successful run was found on main (ded92e1) 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 · 2026-10-02T08:56:16Z

GNU testsuite comparison:

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

xtqqczze · 2026-10-02T15:29:22Z


-[target.'cfg(all(unix, not(any(target_os = "aix", target_os = "hurd", target_os = "redox"))))'.dependencies]
+[target.'cfg(all(unix, not(target_os = "redox")))'.dependencies]
 uucore = { workspace = true, features = ["safe-traversal"] }


looks like the safe-traversal feature is unnecessary for chown

Right, perms already pulls it in. Dropped it from chown, and from chmod and install too, which had the same redundant section.

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

xtqqczze · 2026-10-02T21:19:55Z

Please note the removal of the fallback path handle_symlink_during_traversal in the description, in case we ever need it to implement safe_traversal for additional platforms.

abendrothj · 2026-10-02T21:41:24Z

Added to the description.

sylvestre · 2026-10-05T21:26:57Z


+    // aix and hurd only expose the times as `timespec` fields (`st_atim` and
+    // so on), whose types vary between them and with the pointer width.
+    #[cfg(any(target_os = "aix", target_os = "hurd"))]


almost the same function 6 times plus 6 not(any(aix, hurd)) gates on the existing ones. could you please add an aix/hurd branch inside the existing fns instead, like netbsd does?

Done: the six extra functions and the not(any(aix, hurd)) gates are gone; AIX and Hurd are a branch inside the existing MetadataExt accessors, as for NetBSD.

sylvestre · 2026-10-05T21:27:01Z

+    // aix and hurd only expose the times as `timespec` fields (`st_atim` and
+    // so on), whose types vary between them and with the pointer width.
+    #[cfg(any(target_os = "aix", target_os = "hurd"))]
+    #[allow(clippy::unnecessary_cast)]


the fields are all i32/i64 here, so .into() should work and we can drop the 6 #[allow(clippy::unnecessary_cast)], no?

The allows are gone, but I used as _ instead of .into(). On x86_64 Hurd the fields are already i64, so .into() gives six useless_conversion errors there (i686 Hurd was fine). as _ is clean under clippy -D warnings for x86_64 Hurd.

xtqqczze · 2026-10-05T21:39:01Z

sorry, compilation on AIX targets is blocked on #15083

The only thing keeping it off them was the stat time fields, which they expose as timespecs (st_atim and so on). Read those there, and gate safe_traversal and its users on Redox alone, so chmod, du, install and rm walk trees through descriptors there too. The path-based chmod helpers that only AIX still used are gone, as are the target-specific safe-traversal features of chmod, chown and install, which already get it through perms. Redox stays excluded: nix has none of the *at calls there.

Copilot

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

abendrothj · 2026-10-06T07:29:57Z

Yes, with libc 0.2.190 nix fails in statvfs.rs (fsid_t vs u64) before it reaches uucore, on main as well as here. With libc pinned to 0.2.189 locally, uucore's safe_traversal builds for AIX; I left the lockfile as it is on main. I'd keep AIX in, since it shares the code with Hurd and lets chmod drop its AIX-only path, so it builds once #15083 is sorted out. I can make this Hurd-only if you'd prefer.

sylvestre · 2026-10-07T06:58:15Z

Thanks for your PR

Copilot AI balanced review requested due to automatic review settings October 2, 2026 08:08

Copilot AI reviewed Oct 2, 2026

xtqqczze reviewed Oct 2, 2026

View reviewed changes

abendrothj force-pushed the safe-traversal-aix-hurd branch from 55e7c4c to 0df23c8 Compare October 2, 2026 20:12

Copilot AI balanced review requested due to automatic review settings October 2, 2026 20:12

Copilot AI reviewed Oct 2, 2026

sylvestre reviewed Oct 5, 2026

View reviewed changes

Copilot AI balanced review requested due to automatic review settings October 6, 2026 07:15

abendrothj force-pushed the safe-traversal-aix-hurd branch from 0df23c8 to 85d3e76 Compare October 6, 2026 07:15

Copilot AI reviewed Oct 6, 2026

sylvestre merged commit 4e2be1c into uutils:main Oct 7, 2026
191 of 192 checks passed

This was referenced Oct 7, 2026

chown, chgrp: use the descriptor walk for -R on all safe_traversal platforms #15128

Draft

chmod, chown: check --preserve-root on the directory actually descended into #14972

Open

mv: recreate special files when moving them across devices #13334

Open

Conversation

abendrothj commented Oct 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

codspeed Bot commented Oct 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Merging this PR will degrade performance by 21.4%

Performance Changes

Footnotes

Uh oh!

github-actions Bot commented Oct 2, 2026

Uh oh!

xtqqczze Oct 2, 2026

Choose a reason for hiding this comment

Uh oh!

abendrothj Oct 2, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

xtqqczze commented Oct 2, 2026

Uh oh!

abendrothj commented Oct 2, 2026

Uh oh!

sylvestre Oct 5, 2026

Choose a reason for hiding this comment

Uh oh!

abendrothj Oct 6, 2026

Choose a reason for hiding this comment

Uh oh!

sylvestre Oct 5, 2026

Choose a reason for hiding this comment

Uh oh!

abendrothj Oct 6, 2026

Choose a reason for hiding this comment

Uh oh!

xtqqczze commented Oct 5, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

abendrothj commented Oct 6, 2026

Uh oh!

Uh oh!

sylvestre commented Oct 7, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

abendrothj commented Oct 2, 2026 •

edited

Loading

codspeed Bot commented Oct 2, 2026 •

edited

Loading