Avoid weak-linking `gettid` on musl by windsornguyen · Pull Request #160114 · rust-lang/rust

windsornguyen · 2026-07-29T02:08:21Z

Mitigates #154439.

Brief

current_os_id weak-links gettid so older glibc can fall back to a raw syscall.

Details

During fat LTO, LLVM can merge the weak declaration with a strong declaration from another crate and keep the weak linkage. A static linker then does not extract gettid.o from libc.a, leaving the strong call without a definition.

@bjorn3 traced the underlying compiler problem to fat LTO using LLVM’s legacy ModuleLinker. Changing that compiler-level behavior is intentionally out of scope here, so that issue should remain open.

Note that Rust-supported musl versions provide gettid, so musl does not need that probe.

This PR calls gettid directly on musl. It follows the musl-specific branch suggested by @tgross35 in #154439.

r? @tgross35

`current_os_id` uses an extern-weak `gettid` declaration so older glibc can fall back to a raw syscall. Rust-supported musl versions provide `gettid`, so musl does not need this probe. During fat LTO, LLVM can merge the weak declaration with a strong reference from another crate and keep the weak linkage. The static linker then does not extract `gettid.o` from `libc.a`, leaving the strong call without a definition. Call `gettid` directly on musl and add a regression test for the returned thread ID. This removes std's contribution to the collision; the underlying LTO linkage bug remains.

rustbot · 2026-07-29T02:08:26Z

tgross35 is currently at their maximum review capacity.
They may take a while to respond.

rustbot · 2026-07-29T02:08:27Z

Thanks for the pull request, and welcome! The Rust Project is excited to review your changes, and you should hear from @tgross35 (or someone else) some time within the next two weeks.

Please see the contribution instructions for more information. Namely, in order to ensure the minimum review times lag, PR authors and assigned reviewers should ensure that the review label (S-waiting-on-review and S-waiting-on-author) stays updated, invoking these commands when appropriate:

@rustbot author: the review is finished, PR author should check the comments and take action accordingly
@rustbot review: the author is ready for a review, this PR will be queued again in the reviewer's queue

windsornguyen · 2026-07-29T02:09:07Z

@bjorn3, since you identified the ModuleLinker root cause in #154439, would you be willing to give the LTO scope here a second look?

tgross35 · 2026-08-04T09:17:43Z

@bors try jobs=test-various

Avoid weak-linking `gettid` on musl try-job: test-various

tgross35

Two small requests than r=me if the try job passes.

No need for the Test section in the PR description, we can pretty much see all of that from the test itself. (Please handwrite things like commit messages and PR descriptions)

View changes since this review

rust-bors · 2026-08-04T09:31:27Z

💔 Test for a5ba248 failed: CI. Failed job:

try - test-various (web logs, enhanced plaintext logs)

tgross35 · 2026-08-04T09:37:44Z

Looks like a download failure, @bors retry

rust-bors · 2026-08-04T09:37:46Z

❗ You can only retry pull requests that are approved and have a previously failed auto build.

tgross35 · 2026-08-04T09:37:56Z

@bors try jobs=test-various

Avoid weak-linking `gettid` on musl try-job: test-various

rust-bors · 2026-08-04T09:45:38Z

💔 Test for de974aa failed: CI. Failed job:

try - test-various (web logs, enhanced plaintext logs)

The regression test also works on glibc. Change its target gate from musl to Linux so both libc environments exercise it. Signed-off-by: windsornguyen <win@dedaluslabs.ai>

Comments outside cfg_select arms describe multiple targets. Move the musl-specific explanation into its arm so it cannot be read as applying to the neighboring Linux/Android path. Signed-off-by: windsornguyen <win@dedaluslabs.ai>

tgross35 · 2026-08-04T10:00:35Z

If this fails again, will hold off for a while longer

@bors try jobs=test-various

Please squash as well

Avoid weak-linking `gettid` on musl try-job: test-various

rust-bors · 2026-08-04T10:08:53Z

💔 Test for 1868143 failed: CI. Failed job:

try - test-various (web logs, enhanced plaintext logs)

rust-log-analyzer · 2026-08-04T11:38:58Z

A job failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

#15 134.2 gcc-9.4.0.tar.xz: OK
#15 134.2 mv sources/gcc-9.4.0.tar.xz.tmp/gcc-9.4.0.tar.xz sources/gcc-9.4.0.tar.xz
#15 134.2 rm -rf sources/gcc-9.4.0.tar.xz.tmp
#15 134.2 failed: Connection timed out.
#15 134.2 Connecting to musl.libc.org (musl.libc.org)|2001:19f0:4009:4061:5400:ff:fe11:6da2|:443... failed: Network is unreachable.
#15 134.2 make: *** [Makefile:87: sources/musl-1.2.5.tar.gz] Error 4
#15 ERROR: process "/bin/sh -c bash musl-toolchain.sh x86_64 && rm -rf build" did not complete successfully: exit code: 1
------
 > [ 9/14] RUN bash musl-toolchain.sh x86_64 && rm -rf build:
134.2 2026-08-04 10:05:52 (33.0 MB/s) - 'gcc-9.4.0.tar.xz' saved [72411232/72411232]
134.2 
134.2 cd sources/gcc-9.4.0.tar.xz.tmp && touch gcc-9.4.0.tar.xz
134.2 cd sources/gcc-9.4.0.tar.xz.tmp && sha1sum -c /build/musl-cross-make/hashes/gcc-9.4.0.tar.xz.sha1
134.2 gcc-9.4.0.tar.xz: OK
134.2 mv sources/gcc-9.4.0.tar.xz.tmp/gcc-9.4.0.tar.xz sources/gcc-9.4.0.tar.xz
134.2 rm -rf sources/gcc-9.4.0.tar.xz.tmp
134.2 failed: Connection timed out.
134.2 Connecting to musl.libc.org (musl.libc.org)|2001:19f0:4009:4061:5400:ff:fe11:6da2|:443... failed: Network is unreachable.
134.2 make: *** [Makefile:87: sources/musl-1.2.5.tar.gz] Error 4
------
Dockerfile:40
--------------------
  38 |     COPY scripts/musl-cve-2026-6042.diff /build/
  39 |     COPY scripts/musl-cve-2026-40200.diff /build/

rust-log-analyzer · 2026-08-04T11:44:19Z

The job x86_64-gnu-next-trait-solver-polonius failed! Check out the build log: (web) (plain enhanced) (plain)

Click to see the possible cause of the failure (guessed by this bot)

test [ui] tests/ui/lint/must_not_suspend/allocator.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/boxed.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/feature-gate-must_not_suspend.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/dedup.rs ... ok
test [ui] tests/ui/linking/lto-weak-gettid-issue-154439.rs ... FAILED
test [ui] tests/ui/lint/must_not_suspend/gated.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/generic.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/mutex.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/handled.rs ... ok
test [ui] tests/ui/lint/must_not_suspend/other_items.rs ... ok
---
test [ui] tests/ui/zero-sized/zero-sized-btreemap-insert.rs ... ok

failures:

---- [ui] tests/ui/linking/lto-weak-gettid-issue-154439.rs stdout ----

error: test did not exit with success! code=None so test would pass with `run-crash`
status: signal: 11 (SIGSEGV) (core dumped)
command: cd "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/linking/lto-weak-gettid-issue-154439" && RUSTC="/checkout/obj/build/x86_64-unknown-linux-gnu/stage1/bin/rustc" RUST_TEST_THREADS="4" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/linking/lto-weak-gettid-issue-154439/a"
stdout: none
stderr: none

---- [ui] tests/ui/linking/lto-weak-gettid-issue-154439.rs stdout end ----

failures:
    [ui] tests/ui/linking/lto-weak-gettid-issue-154439.rs

test result: FAILED. 21648 passed; 1 failed; 238 ignored; 0 measured; 0 filtered out; finished in 608.55s

Some tests failed in compiletest suite=ui mode=ui host=x86_64-unknown-linux-gnu target=x86_64-unknown-linux-gnu
Bootstrap failed while executing `test tests/ui --compare-mode polonius --stage 1`

Important

For more information how to resolve CI failures of this job, visit this link.

windsornguyen · 2026-08-04T17:58:56Z

It looks like the glibc path still weak-links gettid. Investigating.

tgross35 · 2026-08-05T00:45:56Z

Huh - I wasn't able to reproduce the failure locally with glibc before. Wonder why it shows up here

rustbot assigned tgross35 Jul 29, 2026

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Jul 29, 2026

windsornguyen mentioned this pull request Jul 29, 2026

LTO turns weak linkage into a miscompile (identified via gettid segfault on x86_64-unknown-linux-musl) #154439

Open

This comment has been minimized.

Sign in to view

rust-bors Bot pushed a commit that referenced this pull request Aug 4, 2026

Auto merge of #160114 - windsornguyen:fix/musl-lto-weak-gettid, r=<try>

a5ba248

Avoid weak-linking `gettid` on musl try-job: test-various

tgross35 reviewed Aug 4, 2026

View reviewed changes

Comment thread tests/ui/linking/lto-weak-gettid-issue-154439.rs Outdated

Comment thread library/std/src/sys/thread/unix.rs

rust-bors Bot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Aug 4, 2026

This comment has been minimized.

Sign in to view

rust-bors Bot pushed a commit that referenced this pull request Aug 4, 2026

Auto merge of #160114 - windsornguyen:fix/musl-lto-weak-gettid, r=<try>

de974aa

Avoid weak-linking `gettid` on musl try-job: test-various

windsornguyen added 2 commits August 4, 2026 02:46

tests: run weak gettid LTO test on Linux

f3cdd0f

The regression test also works on glibc. Change its target gate from musl to Linux so both libc environments exercise it. Signed-off-by: windsornguyen <win@dedaluslabs.ai>

std: scope the musl gettid comment to its cfg arm

74fb8fb

Comments outside cfg_select arms describe multiple targets. Move the musl-specific explanation into its arm so it cannot be read as applying to the neighboring Linux/Android path. Signed-off-by: windsornguyen <win@dedaluslabs.ai>

This comment has been minimized.

Sign in to view

rust-bors Bot pushed a commit that referenced this pull request Aug 4, 2026

Auto merge of #160114 - windsornguyen:fix/musl-lto-weak-gettid, r=<try>

1868143

Avoid weak-linking `gettid` on musl try-job: test-various

This comment has been minimized.

Sign in to view

Uh oh!

Uh oh!

Conversation

windsornguyen commented Jul 29, 2026 • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Brief

Details

Uh oh!

rustbot commented Jul 29, 2026

Uh oh!

rustbot commented Jul 29, 2026

Uh oh!

windsornguyen commented Jul 29, 2026

Uh oh!

tgross35 commented Aug 4, 2026

Uh oh!

This comment has been minimized.

tgross35 left a comment • edited by rustbot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

rust-bors Bot commented Aug 4, 2026

Uh oh!

tgross35 commented Aug 4, 2026

Uh oh!

rust-bors Bot commented Aug 4, 2026

Uh oh!

tgross35 commented Aug 4, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Aug 4, 2026

Uh oh!

tgross35 commented Aug 4, 2026

Uh oh!

This comment has been minimized.

rust-bors Bot commented Aug 4, 2026

Uh oh!

This comment has been minimized.

This comment has been minimized.

rust-log-analyzer commented Aug 4, 2026

Uh oh!

rust-log-analyzer commented Aug 4, 2026

Uh oh!

windsornguyen commented Aug 4, 2026

Uh oh!

tgross35 commented Aug 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

windsornguyen commented Jul 29, 2026 •

edited by rustbot

Loading

tgross35 left a comment •

edited by rustbot

Loading