callconv: fix mips64 aggregate argument passing for C FFI #148294

chenx97 · 2025-10-30T13:10:56Z

MIPS64 needs to put a padding argument before an aggregate argument when
this argument is in an odd-number position, starting from 0, and has an
alignment of 16 bytes or higher, e.g.
void foo(int a, max_align_t b); is the same as
void foo(int a, long _padding, max_align_t b);

rustbot · 2025-10-30T13:11:01Z

r? @petrochenkov

rustbot has assigned @petrochenkov.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

chenx97 · 2025-11-04T10:07:49Z

r? @workingjubilee

rustbot · 2025-11-04T10:07:54Z

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

chenx97 · 2025-11-13T02:12:27Z

r? @petrochenkov

petrochenkov · 2025-11-13T12:14:34Z

r? @workingjubilee
I've stolen one PR from workingjubilee's queue instead.

rustbot · 2025-11-13T12:14:38Z

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

workingjubilee · 2025-11-13T19:53:20Z

Fair trade.

workingjubilee · 2025-11-13T19:55:20Z

@chenx97 re: your PR message: a is an alignment, and a is (8 < a) && (a <= 16)? Since 8 is not less than a if a == 8, then doesn't that mean a == 16? Or do you mean let 8 | 16 = a;?

workingjubilee · 2025-11-13T20:07:59Z

compiler/rustc_target/src/callconv/mips64.rs

+    let align = arg.layout.align.abi.max(dl.i64_align).min(dl.i128_align);
+    let pad_i32 = !offset.is_aligned(align);


So we take the max of align_of::<ArgType>() and align_of::<i64>(), guaranteeing it is at least align_of::<i64>(), and then we take the minimum of align_of::<i128>() and align_of::<ArgType>(), guaranteeing it is no more than align_of::<i128>()...

So, is there a reason this isn't just Ord::clamp(arg.layout.align.abi, dl.i64_align, dl.128_align)?

And then we ask if the offset is aligned to this? ... but we're changing the alignment so that u8, u16, etc. don't align...? You can easily convince me this PR is correct for MIPS, but I'm not sure how much this makes sense with respect to the PR description.

So, is there a reason this isn't just Ord::clamp(arg.layout.align.abi, dl.i64_align, dl.128_align)?

There is no reason other than that this line comes from existing code in mips.rs:28. I'll update this patch accordingly.

but we're changing the alignment so that u8, u16, etc. don't align...?

The offset starts from 0 and only increments in an aligned manner, and u8, u16 and u32 are always as aligned as u64: *offset = offset.align_to(align) + size.align_to(align);

You can easily convince me this PR is correct for MIPS, but I'm not sure how much this makes sense with respect to the PR description.

In that case, I need help with improving my description and commit message.

Well, usually when I note a value is in a range, even though it's a commit message or PR, I deliberately use Rust's range syntax and say it's in the range instead of < or <=, because I make mistakes like the one you made if I don't.

Similar to mips, mips64 also adds a padding register when an aggregate argument is not passed with an aligned offset,

I do not think "add a padding register when an aggregate argument is not passed with an aligned offset" is quite right. That makes it sound a bit like when you pass a 1-byte struct, it uses 9 bytes, that byte plus an entire register! It's more like, "pads aggregates up to the 64-bit register size", right?

Hopefully, the new description can clean up some confusion.

I'm not sure I was accurate in my supposition of how it would work, here, but yes, this is much better, thank you.

bors · 2025-11-20T17:51:55Z

☔ The latest upstream changes (presumably #149132) made this pull request unmergeable. Please resolve the merge conflicts.

MIPS64 needs to put a padding argument before an aggregate argument when this argument is in an odd-number position, starting from 0, and has an alignment of 16 bytes or higher, e.g. `void foo(int a, max_align_t b);` is the same as `void foo(int a, long _padding, max_align_t b);` This fix uses an i32 padding, but it should work just fine because i32 is aligned like i64 for arguments.

rustbot · 2025-11-24T03:12:32Z

This PR was rebased onto a different main commit. Here's a range-diff highlighting what actually changed.

Rebasing is a normal part of keeping PRs up to date, so no action is needed—this note is just to help reviewers.

chenx97 · 2025-11-24T03:13:07Z

@workingjubilee would you mind giving it another review?

workingjubilee · 2025-12-10T03:17:41Z

Sorry about that, taking another look.

workingjubilee · 2025-12-10T03:27:00Z

Huh, I see.

What an odd little quirk in a calling convention.

@bors r+ rollup

bors · 2025-12-10T03:27:03Z

📌 Commit d60e7cf has been approved by workingjubilee

It is now in the queue for this repository.

…rgs, r=workingjubilee callconv: fix mips64 aggregate argument passing for C FFI MIPS64 needs to put a padding argument before an aggregate argument when this argument is in an odd-number position, starting from 0, and has an alignment of 16 bytes or higher, e.g. `void foo(int a, max_align_t b);` is the same as `void foo(int a, long _padding, max_align_t b);`

Rollup of 12 pull requests Successful merges: - #147602 (Deduplicate higher-ranked lifetime capture errors in impl Trait) - #147725 (Remove -Zoom=panic) - #148294 (callconv: fix mips64 aggregate argument passing for C FFI) - #148491 ( Correctly provide suggestions when encountering `async fn` with a `dyn Trait` return type) - #149417 (tidy: Detect outdated workspaces in workspace list) - #149458 (Run clippy on cg_gcc in CI) - #149679 (Restrict spe_acc to PowerPC SPE targets) - #149781 (Don't suggest wrapping attr in unsafe if it may come from proc macro) - #149795 (Use `let`...`else` instead of `match foo { ... _ => return };` and `if let ... else return` in std) - #149816 (Make typo in field and name suggestions verbose) - #149824 (Add a regression test for issue 145748) - #149826 (compiletest: tidy up `adb_path`/`adb_test_dir` handling) r? `@ghost` `@rustbot` modify labels: rollup

Rollup of 10 pull requests Successful merges: - #147725 (Remove -Zoom=panic) - #148294 (callconv: fix mips64 aggregate argument passing for C FFI) - #148491 ( Correctly provide suggestions when encountering `async fn` with a `dyn Trait` return type) - #149458 (Run clippy on cg_gcc in CI) - #149679 (Restrict spe_acc to PowerPC SPE targets) - #149781 (Don't suggest wrapping attr in unsafe if it may come from proc macro) - #149795 (Use `let`...`else` instead of `match foo { ... _ => return };` and `if let ... else return` in std) - #149816 (Make typo in field and name suggestions verbose) - #149824 (Add a regression test for issue 145748) - #149826 (compiletest: tidy up `adb_path`/`adb_test_dir` handling) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of #148294 - chenx97:mips64-padding-aggregate-args, r=workingjubilee callconv: fix mips64 aggregate argument passing for C FFI MIPS64 needs to put a padding argument before an aggregate argument when this argument is in an odd-number position, starting from 0, and has an alignment of 16 bytes or higher, e.g. `void foo(int a, max_align_t b);` is the same as `void foo(int a, long _padding, max_align_t b);`

Rollup of 10 pull requests Successful merges: - rust-lang/rust#147725 (Remove -Zoom=panic) - rust-lang/rust#148294 (callconv: fix mips64 aggregate argument passing for C FFI) - rust-lang/rust#148491 ( Correctly provide suggestions when encountering `async fn` with a `dyn Trait` return type) - rust-lang/rust#149458 (Run clippy on cg_gcc in CI) - rust-lang/rust#149679 (Restrict spe_acc to PowerPC SPE targets) - rust-lang/rust#149781 (Don't suggest wrapping attr in unsafe if it may come from proc macro) - rust-lang/rust#149795 (Use `let`...`else` instead of `match foo { ... _ => return };` and `if let ... else return` in std) - rust-lang/rust#149816 (Make typo in field and name suggestions verbose) - rust-lang/rust#149824 (Add a regression test for issue 145748) - rust-lang/rust#149826 (compiletest: tidy up `adb_path`/`adb_test_dir` handling) r? `@ghost` `@rustbot` modify labels: rollup

rustbot assigned petrochenkov Oct 30, 2025

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. labels Oct 30, 2025

chenx97 marked this pull request as draft October 31, 2025 11:47

rustbot 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 Oct 31, 2025

This comment was marked as resolved.

Sign in to view

chenx97 force-pushed the mips64-padding-aggregate-args branch from fa68eff to 2b8161f Compare November 2, 2025 16:34

chenx97 marked this pull request as ready for review November 2, 2025 16:35

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels Nov 2, 2025

rustbot assigned workingjubilee and unassigned petrochenkov Nov 4, 2025

chenx97 force-pushed the mips64-padding-aggregate-args branch from 2b8161f to 4c08196 Compare November 7, 2025 09:33

This comment has been minimized.

Sign in to view

rustbot assigned petrochenkov and unassigned workingjubilee Nov 13, 2025

rustbot assigned workingjubilee and unassigned petrochenkov Nov 13, 2025

workingjubilee reviewed Nov 13, 2025

View reviewed changes

chenx97 force-pushed the mips64-padding-aggregate-args branch from 4c08196 to e03382d Compare November 14, 2025 06:13

This comment has been minimized.

Sign in to view

chenx97 force-pushed the mips64-padding-aggregate-args branch from e03382d to a23c4cc Compare November 14, 2025 06:20

chenx97 changed the title ~~callconv: adapt mips padding logic to mips64~~ callconv: fix mips64 aggregate argument passing for C FFI Nov 14, 2025

chenx97 requested a review from workingjubilee November 17, 2025 06:08

chenx97 force-pushed the mips64-padding-aggregate-args branch from a23c4cc to d60e7cf Compare November 24, 2025 03:12

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 Dec 10, 2025

workingjubilee closed this Dec 10, 2025

workingjubilee reopened this Dec 10, 2025

rustbot added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Dec 10, 2025

workingjubilee removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Dec 10, 2025

Zalathar mentioned this pull request Dec 10, 2025

Rollup of 13 pull requests #149828

Closed

Zalathar mentioned this pull request Dec 10, 2025

Rollup of 12 pull requests #149830

Closed

matthiaskrgr mentioned this pull request Dec 10, 2025

Rollup of 10 pull requests #149836

Merged

bors merged commit 397de87 into rust-lang:main Dec 10, 2025
22 checks passed

rustbot added this to the 1.94.0 milestone Dec 10, 2025

chenx97 deleted the mips64-padding-aggregate-args branch December 11, 2025 05:40

github-actions bot mentioned this pull request Dec 12, 2025

Toolchain upgrade to nightly-2025-12-11 flux-rs/flux#1345

Draft

		let align = arg.layout.align.abi.max(dl.i64_align).min(dl.i128_align);
		let pad_i32 = !offset.is_aligned(align);

Uh oh!

callconv: fix mips64 aggregate argument passing for C FFI #148294

callconv: fix mips64 aggregate argument passing for C FFI #148294

Conversation

chenx97 commented Oct 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Oct 30, 2025

Uh oh!

This comment was marked as resolved.

chenx97 commented Nov 4, 2025

Uh oh!

rustbot commented Nov 4, 2025

Uh oh!

This comment has been minimized.

chenx97 commented Nov 13, 2025

Uh oh!

petrochenkov commented Nov 13, 2025

Uh oh!

rustbot commented Nov 13, 2025

Uh oh!

workingjubilee commented Nov 13, 2025

Uh oh!

workingjubilee commented Nov 13, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

workingjubilee Nov 13, 2025

Choose a reason for hiding this comment

Uh oh!

chenx97 Nov 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

workingjubilee Nov 14, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

chenx97 Nov 17, 2025

Choose a reason for hiding this comment

Uh oh!

workingjubilee Dec 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

This comment has been minimized.

bors commented Nov 20, 2025

Uh oh!

rustbot commented Nov 24, 2025

Uh oh!

chenx97 commented Nov 24, 2025

Uh oh!

workingjubilee commented Dec 10, 2025

Uh oh!

workingjubilee commented Dec 10, 2025

Uh oh!

bors commented Dec 10, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

chenx97 commented Oct 30, 2025 •

edited

Loading

workingjubilee commented Nov 13, 2025 •

edited

Loading

chenx97 Nov 14, 2025 •

edited

Loading

workingjubilee Nov 14, 2025 •

edited

Loading

workingjubilee Dec 10, 2025 •

edited

Loading