Re-enable extension of bool return values on x86_64 by saethlin · Pull Request #163917 · rust-lang/rust

saethlin · 2026-10-07T03:59:28Z

I think this fixes #163911

Though the test here is a bit hard to figure out, I'm not sure under what conditions expectation goes in the left or the right column. Do I decide that per-target or per-instance? Because x86_64 seems obligated to extend bool returns, so whether it is a "target that extends" or not depends on the type we're dealing with.

I also noticed that because of how the test is written, if codegen starts adding a zeroext on a return type where it is wrong to do so, the test will pass anyway.

rustbot · 2026-10-07T03:59:34Z

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

BoxyUwU · 2026-10-07T08:15:32Z

is this worth a stable/beta backport?

jieyouxu · 2026-10-07T09:09:24Z

is this worth a stable/beta backport?

Nominating just in case.
@rustbot label: +stable-nominated +beta-nominated

asquared31415

This PR should fix the issue, the emitted LLVM IR returns to what was emitted in 1.98. A small nit on the conditions to clean it up a bit though.

View changes since this review

rustbot · 2026-10-08T10:07:22Z

beta backport approved as per compiler team on Zulip. A backport PR will be authored by the release team at the end of the current development cycle. Backport labels are handled by them.

rustbot · 2026-10-08T10:07:35Z

stable backport approved as per compiler team on Zulip. A backport PR will be authored by the release team at the end of the current development cycle. Backport labels are handled by them.

apiraino · 2026-10-08T18:10:47Z

I have approved the backports without this PR being yet properly reviewed and merged (sorryu about that)

r? compiler

jieyouxu · 2026-10-09T01:07:35Z

A random reroll here is likely unhelpful, ABI handling experts are scarce. I'll ask in the private channel. I'll do a review pass as well.

jieyouxu

Change itself looks right to me, at least from what I can gather. Some non-blocking nits.
@rustbot author

View changes since this review

jieyouxu · 2026-10-09T01:34:57Z

 //                  ZERO/SIGN-EXTENDING TO 32 BITS            NON-EXTENDING
 //                  ==============================  =======================
-// x86_64:          void @c_arg_bool(i1 zeroext %_a)


I also noticed that because of how the test is written, if codegen starts adding a zeroext on a return type where it is wrong to do so, the test will pass anyway.

Yeah, that seems a bit awkward since it defeats the purpose of the return checks...

For the non-extending return check cases, can we instead match via patterns such as

// x86_64-linux: define {{(dso_local )?}}i1 @c_ret_bool()

That way if the return type gets a zeroext this check should fail. Using this pattern against the uefi one (where it does zeroext) indeed fails to match.

(This is not blocking; I think we can also do this as a follow-up, since this is pre-existing.)

Follow-up: #164016

Assuming we have the option to pass FileCheck arguments, you could also use something like --implicit-check-not=zeroext --implicit-check-not=signext.

Hm yeah. I'm not sure using the implicit flags is much better though, that also seems not very straightforward when looking at the actual diff right?

jieyouxu

Changes looks good to me, thanks. I think the test tightening is fine as a follow-up.

Might be worth waiting for another pair of eyes but don't feel too strongly about it. Feel free to r=me otherwise.

View changes since this review

nikic · 2026-10-09T06:57:51Z

+                    // Always extend bool returns, x86_64 psABI requires that on the ABI the top 7 bits are zero
+                    // https://github.com/llvm/llvm-project/issues/12579#issuecomment-2718032943


Suggested change

// Always extend bool returns, x86_64 psABI requires that on the ABI the top 7 bits are zero

// https://github.com/llvm/llvm-project/issues/12579#issuecomment-2718032943

// The x86_64 psABI requires bools to be zero-extended to 8-bits, with high bits unspecified.

// zeroext i1 has this behavior in the LLVM X86 backend.

I don't think the comment linked here is super relevant, it's talking about a mismatch in how arguments are handled, where Clang/LLVM apply too much extension.

I also think it's worth mentioning here that using zeroext here for partial extension is correct due to special handling in the X86 backend (which is noteworthy because the same wouldn't be true for the AArch64 backend).

I think it would also be good to make the bool case separate and call arg.extend_integer_width(8) for it. The bit width effectively gets ignored right now (as long as it's larger than the bit width we're starting with), but I think it's confusing to have the extend_integer_width_to(32) when it's not actually extending to 32 bits.

View changes since the review

I agree. The only hitch is that required updating the implementation of extend_integer_width_to.

RalfJung · 2026-10-09T09:17:17Z

r? @nikic

Add test coverage

294ff0f

rustbot assigned workingjubilee Oct 7, 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. labels Oct 7, 2026

rustatian mentioned this pull request Oct 7, 2026

fix: pin Rust 1.98.1 and align the build versions rapira-rs/rapira#202

Merged

6 tasks done

rustbot added beta-nominated Nominated for backporting to the compiler in the beta channel. stable-nominated Nominated for backporting to the compiler in the stable channel. labels Oct 7, 2026

jieyouxu added the O-x86_64 Target: x86-64 processors (like x86_64-*) (also known as amd64 and x64) label Oct 7, 2026

asquared31415 reviewed Oct 7, 2026

View reviewed changes

Comment thread compiler/rustc_target/src/callconv/x86_64.rs Outdated

saethlin force-pushed the bool-return-extend branch from 39167e7 to 61f40b0 Compare October 7, 2026 22:56

rustbot added the beta-accepted Accepted for backporting to the compiler in the beta channel. label Oct 8, 2026

rustbot added the stable-accepted Accepted for backporting to the compiler in the stable channel. label Oct 8, 2026

rustbot assigned JohnTitor and unassigned workingjubilee Oct 8, 2026

jieyouxu reviewed Oct 9, 2026

View reviewed changes

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 9, 2026

jieyouxu 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 Oct 9, 2026

jieyouxu assigned jieyouxu and unassigned JohnTitor Oct 9, 2026

jieyouxu mentioned this pull request Oct 9, 2026

fix(bootstrap/darwin): fix rpath for distributed LLD #163954

Merged

1 task

saethlin force-pushed the bool-return-extend branch from 61f40b0 to 3d42c2c Compare October 9, 2026 03:27

jieyouxu approved these changes Oct 9, 2026

View reviewed changes

jieyouxu mentioned this pull request Oct 9, 2026

Tighten (absence of) return attribute checks in some-abis-do-extend-params-to-32-bits.rs #164016

Open

nikic reviewed Oct 9, 2026

View reviewed changes

rustbot assigned nikic and unassigned jieyouxu Oct 9, 2026

Fix the bug by enabling extension for bool returns on x86_64

a0f1885

saethlin force-pushed the bool-return-extend branch from 3d42c2c to a0f1885 Compare October 10, 2026 00:56

		// Always extend bool returns, x86_64 psABI requires that on the ABI the top 7 bits are zero
		// https://github.com/llvm/llvm-project/issues/12579#issuecomment-2718032943

Conversation

saethlin commented Oct 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rustbot commented Oct 7, 2026

Uh oh!

BoxyUwU commented Oct 7, 2026

Uh oh!

jieyouxu commented Oct 7, 2026

Uh oh!

asquared31415 left a comment • 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!

Uh oh!

rustbot commented Oct 8, 2026

Uh oh!

rustbot commented Oct 8, 2026

Uh oh!

apiraino commented Oct 8, 2026

Uh oh!

jieyouxu commented Oct 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jieyouxu 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!

jieyouxu Oct 9, 2026 • 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!

jieyouxu Oct 9, 2026

Choose a reason for hiding this comment

Uh oh!

nikic Oct 9, 2026

Choose a reason for hiding this comment

Uh oh!

jieyouxu Oct 9, 2026

Choose a reason for hiding this comment

Uh oh!

jieyouxu left a comment • 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!

nikic Oct 9, 2026 • 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!

saethlin Oct 10, 2026

Choose a reason for hiding this comment

Uh oh!

RalfJung commented Oct 9, 2026

Uh oh!

This branch has not been deployed

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

10 participants

saethlin commented Oct 7, 2026 •

edited

Loading

asquared31415 left a comment •

edited

Loading

jieyouxu commented Oct 9, 2026 •

edited

Loading

jieyouxu left a comment •

edited by rustbot

Loading

jieyouxu Oct 9, 2026 •

edited

Loading

jieyouxu left a comment •

edited

Loading

nikic Oct 9, 2026 •

edited by rustbot

Loading