Make no_mangle on foreign items explicit instead of implicit by jdonszelmann · Pull Request #144678 · rust-lang/rust

jdonszelmann · 2025-07-30T11:24:44Z

for a followup PR I'm working on I need some foreign items to mangle. I could add a new attribute: no_no_mangle or something silly like that but by explicitly putting no_mangle in the codegen fn attrs of foreign items we can default it to no_mangle and then easily remove it when we don't want it.

I guess you'd know about this r? @bjorn3. Shouldn't be too hard to review :)

Builds on #144655 which should merge first.

rustbot · 2025-07-30T11:24:50Z

Some changes occurred in compiler/rustc_codegen_ssa

cc @WaffleLapkin

bjorn3 · 2025-07-30T11:34:44Z

Can you please squash the last two commits?

jdonszelmann · 2025-07-30T13:53:56Z

Can do

bjorn3 · 2025-07-31T12:41:35Z

    ///
+    /// Note that this returns true for foreign items.
+    /// However, in some places that care about `contains_extern_indicator`, foreign items
+    /// (in an `extern` block) should explicitly be ignored.


Is it possible to make this not return true for foreign items? Foreign items don't need to be exported and the point of this method is to check if an item needs to be exported.

it's annoying since we don't have a did or TyCtxt here. We could, but that'd involve adding another flag in the codegen fn attributes. I decided against that but it's perfectly possible.

bjorn3 · 2025-08-01T08:52:04Z

@bors r+

bors · 2025-08-01T08:52:07Z

📌 Commit 9003a3b has been approved by bjorn3

It is now in the queue for this repository.

…jorn3 Make no_mangle on foreign items explicit instead of implicit for a followup PR I'm working on I need some foreign items to mangle. I could add a new attribute: `no_no_mangle` or something silly like that but by explicitly putting `no_mangle` in the codegen fn attrs of foreign items we can default it to `no_mangle` and then easily remove it when we don't want it. I guess you'd know about this r? `@bjorn3.` Shouldn't be too hard to review :) Builds on rust-lang#144655 which should merge first.

Rollup of 13 pull requests Successful merges: - #143857 (Port #[macro_export] to the new attribute parsing infrastructure) - #144070 (Implement `hash_map` macro ) - #144322 (Add lint against dangling pointers from local variables) - #144667 (`AlignmentEnum` should just be `repr(usize)` now) - #144678 (Make no_mangle on foreign items explicit instead of implicit) - #144790 (Multiple bounds checking elision failures) - #144794 (Port `#[coroutine]` to the new attribute system) - #144805 (compiletest: Preliminary cleanup of `ProcRes` printing/unwinding) - #144808 (`Interner` arg to `EarlyBinder` does not affect auto traits) - #144816 (Update E0562 to account for the new impl trait positions) - #144822 (Return a struct with named fields from `hash_owner_nodes`) - #144824 (Updated test links in compiler) - #144829 (Use full flag name in strip command for Darwin) r? `@ghost` `@rustbot` modify labels: rollup

Zalathar · 2025-08-03T09:46:57Z

Trying to diagnose rollup failure #144846 (comment)

@bors try jobs=test-various

rust-bors · 2025-08-03T09:47:02Z

⌛ Trying commit 9003a3b with merge 50d1271…

To cancel the try build, run the command @bors try cancel.

Make no_mangle on foreign items explicit instead of implicit try-job: test-various

rust-bors · 2025-08-03T10:38:50Z

💔 Test failed (CI). Failed jobs:

bors build finished (web logs, extended logs)
try - test-various (web logs, extended logs)

Zalathar · 2025-08-03T10:38:56Z

@bors r-

jdonszelmann · 2025-08-12T22:21:21Z

sigh, ye that's significant I'm afraid :/

bjorn3 · 2025-08-13T14:36:04Z

    }
+
+    // Foreign items by default use no mangling for their symbol name.
+    if tcx.is_foreign_item(did) {


I suspect the extra query call here is the cause for the perf regression.

Do we think #145429 will resolve that? I think it moved the call around maybe but didn't eliminate it? (I couldn't quite tell at a quick skim). Or do we need to wait for some future date where there's additional validation added at an earlier stage (attribute parsing)?

#145429 doesn't move this call around, only the one in rustc_symbol_mangling which already existed. It also adds an extra call for #[link_name] that can be removed once extra validation is done during attribute parsing. I don't think that extra call in that PR would have much effect. Only a small fraction of all functions for which codegen_fn_attrs is called has a #[link_name] attribute.

tl;dr: #145429 has no effect on the regression in this PR.

Mhm, I doubt it too. I just haven't quite worked out what to do about this since this really is the behavior I want but it's a perf problem it seems. We could revert but the alternative will be much less pretty. It'd involve having an extra flag in codegen fn attrs to explicitly not no-mangle foreign items.

…ts, r=jdonszelmann Couple of codegen_fn_attrs improvements As noted in rust-lang#144678 (comment) here is no need to keep link_name and export_name separate, which the third commit fixes by merging them. The second commit removes some dead code and the first commit merges two ifs with equivalent conditions. The last commit is an unrelated change which removes an unused `feature(autodiff)`.

Rollup merge of #145429 - bjorn3:codegen_fn_attrs_improvements, r=jdonszelmann Couple of codegen_fn_attrs improvements As noted in #144678 (comment) here is no need to keep link_name and export_name separate, which the third commit fixes by merging them. The second commit removes some dead code and the first commit merges two ifs with equivalent conditions. The last commit is an unrelated change which removes an unused `feature(autodiff)`.

…nszelmann Couple of codegen_fn_attrs improvements As noted in rust-lang/rust#144678 (comment) here is no need to keep link_name and export_name separate, which the third commit fixes by merging them. The second commit removes some dead code and the first commit merges two ifs with equivalent conditions. The last commit is an unrelated change which removes an unused `feature(autodiff)`.

Make no_mangle on foreign items explicit instead of implicit for a followup PR I'm working on I need some foreign items to mangle. I could add a new attribute: `no_no_mangle` or something silly like that but by explicitly putting `no_mangle` in the codegen fn attrs of foreign items we can default it to `no_mangle` and then easily remove it when we don't want it. I guess you'd know about this r? `@bjorn3.` Shouldn't be too hard to review :) Builds on rust-lang#144655 which should merge first.

…nszelmann Couple of codegen_fn_attrs improvements As noted in rust-lang/rust#144678 (comment) here is no need to keep link_name and export_name separate, which the third commit fixes by merging them. The second commit removes some dead code and the first commit merges two ifs with equivalent conditions. The last commit is an unrelated change which removes an unused `feature(autodiff)`.

jdonszelmann · 2025-08-25T16:54:56Z

ok, based on discussion above I wanted to go ahead with a revert, but when doing that I finally had an idea how to actually fix it so let's try that quickly in #145855

This commit fixes an accidental regression from rust-lang#144678 where wasm targets would now accidentally use the wrong import module map for a symbol causing a symbol to skip mangling. This can result in compilation failures when symbols are used in cross-crate situations. Closes rust-lang#148347

This commit fixes an accidental regression from rust-lang#144678 where wasm targets would now accidentally use the wrong import module map for a symbol causing a symbol to skip mangling. This can result in compilation failures when symbols are used in cross-crate situations. Closes 148347

…=jackh726 Fix `wasm_import_module` attribute cross-crate This commit fixes an accidental regression from rust-lang#144678 where wasm targets would now accidentally use the wrong import module map for a symbol causing a symbol to skip mangling. This can result in compilation failures when symbols are used in cross-crate situations. Closes rust-lang#148347

Rollup merge of #148363 - alexcrichton:fix-wasm-link-name, r=jackh726 Fix `wasm_import_module` attribute cross-crate This commit fixes an accidental regression from #144678 where wasm targets would now accidentally use the wrong import module map for a symbol causing a symbol to skip mangling. This can result in compilation failures when symbols are used in cross-crate situations. Closes #148347

@petrochenkov

…est, r=petrochenkov Add test for parallel compiler reproducible build Adds a test for reproducible binaries when using parallel compiler. Confirmed with bisection and manual testing that rust-lang#144722 fixed rust-lang#140413, added test fails before commit (075ce31) from rust-lang#144722. ```rust ******************************************************************************** Regression in nightly-2025-08-14 ******************************************************************************** fetching https://static.rust-lang.org/dist/2025-08-13/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-13: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 357.18 KB/s converted 2025-08-13 to 8e62bfd fetching https://static.rust-lang.org/dist/2025-08-14/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-14: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 123.86 KB/s converted 2025-08-14 to 3672a55 looking for regression commit between 2025-08-13 and 2025-08-14 fetching (via remote github) commits from max(8e62bfd, 2025-08-11) to 3672a55 ending github query because we found starting sha: 8e62bfd get_commits_between returning commits, len: 7 commit[0] 2025-08-12: Auto merge of rust-lang#144678 - jdonszelmann:no-mangle-extern, r=bjorn3 commit[1] 2025-08-12: Auto merge of rust-lang#145295 - Kobzol:unify-stages, r=jieyouxu commit[2] 2025-08-13: Auto merge of rust-lang#145093 - nikic:dead-on-return, r=nnethercote commit[3] 2025-08-13: Auto merge of rust-lang#145334 - Kobzol:rollup-fs5a133, r=Kobzol commit[4] 2025-08-13: Auto merge of rust-lang#144722 - ywxt:parallel-reproducibile, r=SparrowLii commit[5] 2025-08-13: Auto merge of rust-lang#145298 - nikic:llvm21-rc3, r=cuviper commit[6] 2025-08-13: Auto merge of rust-lang#145366 - GuillaumeGomez:rollup-v0a6v3u, r=GuillaumeGomez ``` Fixes rust-lang#140413. r? @petrochenkov

@petrochenkov

…est, r=petrochenkov Add test for parallel compiler reproducible build Adds a test for reproducible binaries when using parallel compiler. Confirmed with bisection and manual testing that rust-lang#144722 fixed rust-lang#140413, added test fails before commit (075ce31) from rust-lang#144722. ```rust ******************************************************************************** Regression in nightly-2025-08-14 ******************************************************************************** fetching https://static.rust-lang.org/dist/2025-08-13/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-13: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 357.18 KB/s converted 2025-08-13 to 8e62bfd fetching https://static.rust-lang.org/dist/2025-08-14/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-14: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 123.86 KB/s converted 2025-08-14 to 3672a55 looking for regression commit between 2025-08-13 and 2025-08-14 fetching (via remote github) commits from max(8e62bfd, 2025-08-11) to 3672a55 ending github query because we found starting sha: 8e62bfd get_commits_between returning commits, len: 7 commit[0] 2025-08-12: Auto merge of rust-lang#144678 - jdonszelmann:no-mangle-extern, r=bjorn3 commit[1] 2025-08-12: Auto merge of rust-lang#145295 - Kobzol:unify-stages, r=jieyouxu commit[2] 2025-08-13: Auto merge of rust-lang#145093 - nikic:dead-on-return, r=nnethercote commit[3] 2025-08-13: Auto merge of rust-lang#145334 - Kobzol:rollup-fs5a133, r=Kobzol commit[4] 2025-08-13: Auto merge of rust-lang#144722 - ywxt:parallel-reproducibile, r=SparrowLii commit[5] 2025-08-13: Auto merge of rust-lang#145298 - nikic:llvm21-rc3, r=cuviper commit[6] 2025-08-13: Auto merge of rust-lang#145366 - GuillaumeGomez:rollup-v0a6v3u, r=GuillaumeGomez ``` Fixes rust-lang#140413. r? @petrochenkov

@petrochenkov

…est, r=petrochenkov Add test for parallel compiler reproducible build Adds a test for reproducible binaries when using parallel compiler. Confirmed with bisection and manual testing that rust-lang#144722 fixed rust-lang#140413, added test fails before commit (075ce31) from rust-lang#144722. ```rust ******************************************************************************** Regression in nightly-2025-08-14 ******************************************************************************** fetching https://static.rust-lang.org/dist/2025-08-13/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-13: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 357.18 KB/s converted 2025-08-13 to 8e62bfd fetching https://static.rust-lang.org/dist/2025-08-14/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-14: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 123.86 KB/s converted 2025-08-14 to 3672a55 looking for regression commit between 2025-08-13 and 2025-08-14 fetching (via remote github) commits from max(8e62bfd, 2025-08-11) to 3672a55 ending github query because we found starting sha: 8e62bfd get_commits_between returning commits, len: 7 commit[0] 2025-08-12: Auto merge of rust-lang#144678 - jdonszelmann:no-mangle-extern, r=bjorn3 commit[1] 2025-08-12: Auto merge of rust-lang#145295 - Kobzol:unify-stages, r=jieyouxu commit[2] 2025-08-13: Auto merge of rust-lang#145093 - nikic:dead-on-return, r=nnethercote commit[3] 2025-08-13: Auto merge of rust-lang#145334 - Kobzol:rollup-fs5a133, r=Kobzol commit[4] 2025-08-13: Auto merge of rust-lang#144722 - ywxt:parallel-reproducibile, r=SparrowLii commit[5] 2025-08-13: Auto merge of rust-lang#145298 - nikic:llvm21-rc3, r=cuviper commit[6] 2025-08-13: Auto merge of rust-lang#145366 - GuillaumeGomez:rollup-v0a6v3u, r=GuillaumeGomez ``` Fixes rust-lang#140413. r? @petrochenkov

@petrochenkov

…est, r=petrochenkov Add test for parallel compiler reproducible build Adds a test for reproducible binaries when using parallel compiler. Confirmed with bisection and manual testing that rust-lang#144722 fixed rust-lang#140413, added test fails before commit (075ce31) from rust-lang#144722. ```rust ******************************************************************************** Regression in nightly-2025-08-14 ******************************************************************************** fetching https://static.rust-lang.org/dist/2025-08-13/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-13: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 357.18 KB/s converted 2025-08-13 to 8e62bfd fetching https://static.rust-lang.org/dist/2025-08-14/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-14: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 123.86 KB/s converted 2025-08-14 to 3672a55 looking for regression commit between 2025-08-13 and 2025-08-14 fetching (via remote github) commits from max(8e62bfd, 2025-08-11) to 3672a55 ending github query because we found starting sha: 8e62bfd get_commits_between returning commits, len: 7 commit[0] 2025-08-12: Auto merge of rust-lang#144678 - jdonszelmann:no-mangle-extern, r=bjorn3 commit[1] 2025-08-12: Auto merge of rust-lang#145295 - Kobzol:unify-stages, r=jieyouxu commit[2] 2025-08-13: Auto merge of rust-lang#145093 - nikic:dead-on-return, r=nnethercote commit[3] 2025-08-13: Auto merge of rust-lang#145334 - Kobzol:rollup-fs5a133, r=Kobzol commit[4] 2025-08-13: Auto merge of rust-lang#144722 - ywxt:parallel-reproducibile, r=SparrowLii commit[5] 2025-08-13: Auto merge of rust-lang#145298 - nikic:llvm21-rc3, r=cuviper commit[6] 2025-08-13: Auto merge of rust-lang#145366 - GuillaumeGomez:rollup-v0a6v3u, r=GuillaumeGomez ``` Fixes rust-lang#140413. r? @petrochenkov

@petrochenkov

…est, r=petrochenkov Add test for parallel compiler reproducible build Adds a test for reproducible binaries when using parallel compiler. Confirmed with bisection and manual testing that rust-lang#144722 fixed rust-lang#140413, added test fails before commit (075ce31) from rust-lang#144722. ```rust ******************************************************************************** Regression in nightly-2025-08-14 ******************************************************************************** fetching https://static.rust-lang.org/dist/2025-08-13/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-13: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 357.18 KB/s converted 2025-08-13 to 8e62bfd fetching https://static.rust-lang.org/dist/2025-08-14/channel-rust-nightly-git-commit-hash.txt nightly manifest 2025-08-14: 40 B / 40 B [============================================================================================================================================================================] 100.00 % 123.86 KB/s converted 2025-08-14 to 3672a55 looking for regression commit between 2025-08-13 and 2025-08-14 fetching (via remote github) commits from max(8e62bfd, 2025-08-11) to 3672a55 ending github query because we found starting sha: 8e62bfd get_commits_between returning commits, len: 7 commit[0] 2025-08-12: Auto merge of rust-lang#144678 - jdonszelmann:no-mangle-extern, r=bjorn3 commit[1] 2025-08-12: Auto merge of rust-lang#145295 - Kobzol:unify-stages, r=jieyouxu commit[2] 2025-08-13: Auto merge of rust-lang#145093 - nikic:dead-on-return, r=nnethercote commit[3] 2025-08-13: Auto merge of rust-lang#145334 - Kobzol:rollup-fs5a133, r=Kobzol commit[4] 2025-08-13: Auto merge of rust-lang#144722 - ywxt:parallel-reproducibile, r=SparrowLii commit[5] 2025-08-13: Auto merge of rust-lang#145298 - nikic:llvm21-rc3, r=cuviper commit[6] 2025-08-13: Auto merge of rust-lang#145366 - GuillaumeGomez:rollup-v0a6v3u, r=GuillaumeGomez ``` Fixes rust-lang#140413. r? @petrochenkov

rustbot assigned bjorn3 Jul 30, 2025

rustbot added A-attributes Area: Attributes (`#[…]`, `#![…]`) 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. labels Jul 30, 2025

bjorn3 reviewed Jul 30, 2025

View reviewed changes

Comment thread loge Outdated

This comment has been minimized.

Sign in to view

jdonszelmann force-pushed the no-mangle-extern branch 2 times, most recently from 1790424 to 5240989 Compare July 30, 2025 21:37

bjorn3 reviewed Jul 31, 2025

View reviewed changes

Comment thread compiler/rustc_symbol_mangling/src/lib.rs Outdated

jdonszelmann force-pushed the no-mangle-extern branch from 5240989 to e75b93d Compare July 31, 2025 12:06

bjorn3 reviewed Jul 31, 2025

View reviewed changes

Comment thread compiler/rustc_passes/src/reachable.rs Outdated

jdonszelmann force-pushed the no-mangle-extern branch from e75b93d to 9003a3b Compare July 31, 2025 12:40

bjorn3 reviewed Jul 31, 2025

View reviewed changes

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Aug 1, 2025

Zalathar mentioned this pull request Aug 3, 2025

Rollup of 13 pull requests #144846

Closed

rust-bors Bot added a commit that referenced this pull request Aug 3, 2025

Auto merge of #144678 - jdonszelmann:no-mangle-extern, r=<try>

50d1271

Make no_mangle on foreign items explicit instead of implicit try-job: test-various

This comment has been minimized.

Sign in to view

bors removed the S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. label Aug 3, 2025

rustbot added the perf-regression Performance regression. label Aug 12, 2025

bjorn3 reviewed Aug 13, 2025

View reviewed changes

bjorn3 mentioned this pull request Aug 15, 2025

Couple of codegen_fn_attrs improvements #145429

Merged

Mark-Simulacrum added the perf-regression-triaged The performance regression has been triaged. label Aug 19, 2025

alexcrichton mentioned this pull request Nov 1, 2025

Fix wasm_import_module attribute cross-crate #148363

Merged

marc0olo mentioned this pull request Apr 16, 2026

ic-rusqlite 0.4.3 incompatible with Rust 1.91.0: WASM import module mismatch for time wasm-forge/ic-rusqlite#2

Open

Uh oh!

Uh oh!

Conversation

jdonszelmann commented Jul 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Jul 30, 2025

Uh oh!

Uh oh!

bjorn3 commented Jul 30, 2025

Uh oh!

This comment has been minimized.

jdonszelmann commented Jul 30, 2025

Uh oh!

Uh oh!

Uh oh!

bjorn3 Jul 31, 2025

Choose a reason for hiding this comment

Uh oh!

jdonszelmann Jul 31, 2025

Choose a reason for hiding this comment

Uh oh!

bjorn3 commented Aug 1, 2025

Uh oh!

bors commented Aug 1, 2025

Uh oh!

Zalathar commented Aug 3, 2025

Uh oh!

rust-bors Bot commented Aug 3, 2025

Uh oh!

This comment has been minimized.

rust-bors Bot commented Aug 3, 2025

Uh oh!

Zalathar commented Aug 3, 2025

Uh oh!

jdonszelmann commented Aug 12, 2025

Uh oh!

bjorn3 Aug 13, 2025

Choose a reason for hiding this comment

Uh oh!

Mark-Simulacrum Aug 19, 2025

Choose a reason for hiding this comment

Uh oh!

bjorn3 Aug 19, 2025

Choose a reason for hiding this comment

Uh oh!

jdonszelmann Aug 19, 2025

Choose a reason for hiding this comment

Uh oh!

jdonszelmann commented Aug 25, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

11 participants

jdonszelmann commented Jul 30, 2025 •

edited

Loading