Emit future-incompatibility lint when calling/declaring functions with vectors that require missing target feature by veluca93 · Pull Request #127731 · rust-lang/rust

veluca93 · 2024-07-14T18:43:29Z

On some architectures, vector types may have a different ABI depending on whether the relevant target features are enabled. (The ABI when the feature is disabled is often not specified, but LLVM implements some de-facto ABI.)

As discussed in rust-lang/lang-team#235, this turns out to very easily lead to unsound code.

This commit makes it a post-monomorphization error to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. This ensures that these functions are always called with a consistent ABI.

See the nomination comment for more discussion.

r? RalfJung

Part of #116558

rustbot · 2024-07-14T18:43:36Z

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @RalfJung (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

rustbot · 2024-07-14T18:43:38Z

These commits modify the Cargo.lock file. Unintentional changes to Cargo.lock can be introduced when switching branches and rebasing PRs.

If this was unintentional then you should revert the changes before this PR is merged.
Otherwise, you can ignore this comment.

RalfJung

Please also add some tests so that we can see this check in action.

compiler/rustc_monomorphize/Cargo.toml

compiler/rustc_monomorphize/src/collector.rs

compiler/rustc_monomorphize/src/collector/abi_check.rs

RalfJung · 2024-08-01T09:10:25Z

Looks good for the initial draft, let's see what crater says. :)

@bors try

Emit error when calling/declaring functions with unavailable vectors. On some architectures, vector types may have a different ABI when relevant target features are enabled. As discussed in rust-lang/lang-team#235, this turns out to very easily lead to unsound code. This commit makes it an error to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. r? RalfJung

bors · 2024-08-01T09:11:39Z

⌛ Trying commit e8302b3 with merge 7587ff3...

bors · 2024-08-01T11:07:37Z

☀️ Try build successful - checks-actions
Build commit: 7587ff3 (7587ff3622fbec0abf6ac551eab5226f22f5d958)

RalfJung · 2024-08-01T11:16:20Z

@craterbot check

craterbot · 2024-08-01T11:16:28Z

👌 Experiment pr-127731 created and queued.
🤖 Automatically detected try build 7587ff3
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

bors · 2024-10-25T12:35:18Z

💔 Test failed - checks-actions

RalfJung · 2024-10-25T12:36:30Z

@bors retry error: linking with link.exe failed: exit code: 1104 on dist-x86_64-msvc

bors · 2024-10-25T15:17:52Z

⌛ Testing commit 5af56ca with merge 6faf0bd...

bors · 2024-10-25T17:54:57Z

☀️ Test successful - checks-actions
Approved by: RalfJung
Pushing 6faf0bd to master...

rust-timer · 2024-10-25T19:56:02Z

Finished benchmarking commit (6faf0bd): comparison URL.

Overall result: ❌ regressions - please read the text below

Our benchmarks found a performance regression caused by this PR.
This might be an actual regression, but it can also be just noise.

Next Steps:

If the regression was expected or you think it can be justified,
please write a comment with sufficient written justification, and add
@rustbot label: +perf-regression-triaged to it, to mark the regression as triaged.
If you think that you know of a way to resolve the regression, try to create
a new PR with a fix for the regression.
If you do not understand the regression or you think that it is just noise,
you can ask the @rust-lang/wg-compiler-performance working group for help (members of this group
were already notified of this PR).

@rustbot label: +perf-regression
cc @rust-lang/wg-compiler-performance

Instruction count

This is the most reliable metric that we have; it was used to determine the overall result at the top of this comment. However, even this metric can sometimes exhibit noise.

	mean	range	count
Regressions ❌ (primary)	5.0%	[0.3%, 16.9%]	77
Regressions ❌ (secondary)	4.7%	[0.1%, 29.4%]	30
Improvements ✅ (primary)	-0.1%	[-0.1%, -0.1%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	4.9%	[-0.1%, 16.9%]	78

Max RSS (memory usage)

Results (primary 4.1%, secondary 3.6%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	4.1%	[0.8%, 15.8%]	69
Regressions ❌ (secondary)	3.6%	[1.0%, 7.4%]	24
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	4.1%	[0.8%, 15.8%]	69

Cycles

Results (primary 11.4%, secondary 14.0%)

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	11.4%	[1.5%, 26.1%]	54
Regressions ❌ (secondary)	14.0%	[2.0%, 38.1%]	11
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	11.4%	[1.5%, 26.1%]	54

Binary size

This benchmark run did not return any relevant results for this metric.

Bootstrap: 790.301s -> 788.765s (-0.19%)
Artifact size: 333.71 MiB -> 333.61 MiB (-0.03%)

Kobzol · 2024-10-25T19:58:19Z

This looks pretty bad, even when ignoring the incr-unchanged results. I'd say that we revert it ASAP and then try to figure out how to improve perf.

RalfJung · 2024-10-25T20:40:11Z

Oh damn, we should have done a perf run... yeah, first order of business is a revert. @veluca93 can you prepare a revert?

lqd · 2024-10-25T20:48:31Z

I was preparing one and have opened #132152

Revert rust-lang#127731 "Emit error when calling/declaring functions with unavailable …" This reverts rust-lang#127731 due to the unexpected [perf regressions](rust-lang#127731 (comment)) and to give time to mitigate the regressions before re-landing it. r? `@RalfJung` cc `@veluca93`

On some architectures, vector types may have a different ABI depending on whether the relevant target features are enabled. (The ABI when the feature is disabled is often not specified, but LLVM implements some de-facto ABI.) As discussed in rust-lang/lang-team#235, this turns out to very easily lead to unsound code. This commit makes it a post-monomorphization future-incompat warning to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. This ensures that these functions are always called with a consistent ABI. See the [nomination comment](rust-lang#127731 (comment)) for more discussion. Part of rust-lang#116558

Emit warning when calling/declaring functions with unavailable vectors. On some architectures, vector types may have a different ABI depending on whether the relevant target features are enabled. (The ABI when the feature is disabled is often not specified, but LLVM implements some de-facto ABI.) As discussed in rust-lang/lang-team#235, this turns out to very easily lead to unsound code. This commit makes it a post-monomorphization future-incompat warning to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. This ensures that these functions are always called with a consistent ABI. See the [nomination comment](rust-lang#127731 (comment)) for more discussion. Part of rust-lang#116558 r? RalfJung

On some architectures, vector types may have a different ABI depending on whether the relevant target features are enabled. (The ABI when the feature is disabled is often not specified, but LLVM implements some de-facto ABI.) As discussed in rust-lang/lang-team#235, this turns out to very easily lead to unsound code. This commit makes it a post-monomorphization future-incompat warning to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. This ensures that these functions are always called with a consistent ABI. See the [nomination comment](rust-lang#127731 (comment)) for more discussion. Part of rust-lang#116558

Kobzol · 2024-10-29T19:37:06Z

Reverted in #132152.

@rustbot label: +perf-regression-triaged

rustbot assigned RalfJung Jul 14, 2024

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 Jul 14, 2024