Fix bool calling convention for aarch64, etc.

[workingjubilee]

View all comments

We have been making a broad assumption about the way bool should be handled that may be correct for Rust calling Rust but is wildly incorrect for cross-language FFI calls. The ABI of bool is generally poorly-specified in any C psABI, and often deliberately allows for deviation after the first 8 bits.

This has surfaced as a bug that can corrupt Rust programs on aarch64, but it is a potential latent bug when it comes to every ABI. The simplest solution is to simply stop applying the special case when generating bool's ArgAttributes. This yields control to the later ABI implementations, which may then sign or zero extend the bool (or not) according to their needs.

Fixes #159244

[rustbot]

r? @mejrs

rustbot has assigned @mejrs.
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

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: compiler
• compiler expanded to 75 candidates
• Random selection from 20 candidates

[workingjubilee]

Might as well look at the baseline regression-or-not that we might be facing.

@bors try @rust-timer queue

[rust-bors]

☀️ Try build successful (CI)
Build commit: 47fb3c6 (47fb3c6d799d920ef079445db3076cc4e029e601)
Base parent: d0babd8 (d0babd8b6b05ef9bb65d42f928cef4129d64cf65)

[rust-bors]

☀️ Test successful - CI
Approved by: davidtwco
Duration: 3h 29m 21s
Pushing ad0c9dc to main...

[github-actions]

What is this?

This is an experimental post-merge analysis report that shows differences in test outcomes between the merged PR and its parent PR.

Comparing dfbea5b (parent) -> ad0c9dc (this PR)

Test differences

Show 2 test diffs

Stage 1

• [assembly] tests/assembly-llvm/aarch64/mask-bool-ffi-return.rs: [missing] -> pass (J1)

Stage 2

• [assembly] tests/assembly-llvm/aarch64/mask-bool-ffi-return.rs: [missing] -> pass (J0)

Job group index

• J0: aarch64-apple, aarch64-apple-macos-26, aarch64-gnu, aarch64-gnu-llvm-21-1, aarch64-msvc-1, arm-android, armhf-gnu, dist-i586-gnu-i586-i686-musl, i686-gnu-1, i686-gnu-nopt-1, i686-msvc-1, test-various, x86_64-gnu, x86_64-gnu-llvm-21-2, x86_64-gnu-llvm-22-2, x86_64-gnu-nopt, x86_64-gnu-stable, x86_64-mingw-1, x86_64-msvc-1
• J1: x86_64-gnu-llvm-21-3, x86_64-gnu-llvm-22-3

Test dashboard

Run

cargo run --manifest-path src/ci/citool/Cargo.toml -- \
    test-dashboard ad0c9dce27a22416b65946bc0010edaf22ac6c83 --output-dir test-dashboard

And then open test-dashboard/index.html in your browser to see an overview of all executed tests.

Job duration changes

1. x86_64-gnu-stable: 1h 28m -> 2h 20m (+59.2%)
2. x86_64-gnu-llvm-21-3: 1h 10m -> 1h 46m (+51.4%)
3. x86_64-gnu-gcc-core-tests: 14m 43s -> 7m 38s (-48.1%)
4. i686-gnu-1: 2h 29m -> 1h 23m (-44.4%)
5. test-various: 2h 15m -> 1h 30m (-33.7%)
6. dist-i586-gnu-i586-i686-musl: 1h 13m -> 1h 38m (+33.3%)
7. dist-x86_64-solaris: 1h 42m -> 1h 11m (-29.7%)
8. dist-android: 22m 59s -> 29m 48s (+29.7%)
9. i686-gnu-nopt-1: 2h 17m -> 1h 37m (-29.5%)
10. aarch64-apple: 2h 59m -> 2h 7m (-28.9%)

How to interpret the job duration changes?

Job durations can vary a lot, based on the actual runner instance
that executed the job, system noise, invalidated caches, etc. The table above is provided
mostly for t-infra members, for simpler debugging of potential CI slow-downs.

[rust-timer]

Finished benchmarking commit (ad0c9dc): comparison URL.

Overall result: ❌ regressions - no action needed

@rustbot label: -perf-regression

Instruction count

Our most reliable metric. Used to determine the overall result above. However, even this metric can be noisy.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	0.1%	[0.1%, 0.1%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Max RSS (memory usage)

Results (secondary 2.0%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	2.0%	[2.0%, 2.0%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-	-	0

Cycles

Results (primary 4.1%, secondary -2.6%)

A less reliable metric. May be of interest, but not used to determine the overall result above.

	mean	range	count
Regressions ❌ (primary)	4.1%	[4.1%, 4.1%]	1
Regressions ❌ (secondary)	4.2%	[4.2%, 4.2%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-4.9%	[-6.0%, -2.8%]	3
All ❌✅ (primary)	4.1%	[4.1%, 4.1%]	1

Binary size

This perf run didn't have relevant results for this metric.

Bootstrap: 486.74s -> 495.576s (1.82%)
Artifact size: 387.74 MiB -> 389.78 MiB (0.53%)