Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Suboptimal Assembly Output from a Modulo on a Power of Two #129795

Closed
Altanis opened this issue Aug 30, 2024 · 9 comments
Closed

Suboptimal Assembly Output from a Modulo on a Power of Two #129795

Altanis opened this issue Aug 30, 2024 · 9 comments
Labels
A-LLVM Area: Code generation parts specific to LLVM. Both correctness bugs and optimization-related issues. C-optimization Category: An issue highlighting optimization opportunities or PRs implementing such E-needs-test Call for participation: An issue has been fixed and does not reproduce, but no test has been added. I-slow Issue: Problems and improvements with respect to performance of generated code. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue.

Comments

@Altanis
Copy link

Altanis commented Aug 30, 2024

I tried this code:

#[no_mangle]
pub fn index(key: usize, buckets: usize) -> usize {
    assert!(buckets.is_power_of_two());
    key % buckets
}

Since assert!(buckets.is_power_of_two()) is run (and .is_power_of_two() is an inbuilt function), the latter modulus operation should reduce to key & (buckets - 1).

Instead, the outputted assembly here shows that this reduction does not take place.

Meta

The compiler in the GodBolt link is version 1.80.0.

@Altanis Altanis added the C-bug Category: This is a bug. label Aug 30, 2024
@rustbot rustbot added the needs-triage This issue may need triage. Remove it if it has been sufficiently triaged. label Aug 30, 2024
@saethlin saethlin added A-LLVM Area: Code generation parts specific to LLVM. Both correctness bugs and optimization-related issues. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. C-optimization Category: An issue highlighting optimization opportunities or PRs implementing such and removed needs-triage This issue may need triage. Remove it if it has been sufficiently triaged. labels Aug 31, 2024
@scottmcm
Copy link
Member

scottmcm commented Sep 5, 2024

This is llvm/llvm-project#58996, and needs to be fixed in LLVM. Rust itself won't do this optimization.

@scottmcm scottmcm added I-slow Issue: Problems and improvements with respect to performance of generated code. and removed C-bug Category: This is a bug. labels Sep 5, 2024
dtcxzyw added a commit to llvm/llvm-project that referenced this issue Sep 10, 2024
This patch tries to infer is-power-of-2 from assumptions. I don't see
that this kind of assumption exists in my dataset.
Related issue: rust-lang/rust#129795

Close #58996.
@dtcxzyw
Copy link

dtcxzyw commented Sep 10, 2024

@scottmcm @dianqk @nikic This issue has been fixed by llvm/llvm-project#107745. Please add llvm-fixed-upstream.

@dianqk
Copy link
Member

dianqk commented Sep 10, 2024

You can also add it. :)
@rustbot label +llvm-fixed-upstream

@rustbot rustbot added the llvm-fixed-upstream Issue expected to be fixed by the next major LLVM upgrade, or backported fixes label Sep 10, 2024
@nikic
Copy link
Contributor

nikic commented Sep 10, 2024

@dtcxzyw Doesn't this issue need support for dominating condition rather than assumption?

@dtcxzyw
Copy link

dtcxzyw commented Sep 10, 2024

@dtcxzyw Doesn't this issue need support for dominating condition rather than assumption?

I am preparing a follow-up patch for this :)

@dianqk
Copy link
Member

dianqk commented Sep 10, 2024

@rustbot label -llvm-fixed-upstream

@rustbot rustbot removed the llvm-fixed-upstream Issue expected to be fixed by the next major LLVM upgrade, or backported fixes label Sep 10, 2024
dtcxzyw added a commit to llvm/llvm-project that referenced this issue Sep 13, 2024
@veera-sivarajan
Copy link
Contributor

@rustbot label +llvm-fixed-upstream

@rustbot rustbot added the llvm-fixed-upstream Issue expected to be fixed by the next major LLVM upgrade, or backported fixes label Nov 22, 2024
@nikic
Copy link
Contributor

nikic commented Feb 20, 2025

Fixed by LLVM 20 upgrade.

@nikic nikic added E-needs-test Call for participation: An issue has been fixed and does not reproduce, but no test has been added. and removed llvm-fixed-upstream Issue expected to be fixed by the next major LLVM upgrade, or backported fixes labels Feb 20, 2025
jieyouxu added a commit to jieyouxu/rust that referenced this issue Mar 16, 2025
Add codegen test for rust-lang#129795

Adds test for rust-lang#129795.

Min LLVM version is 20 because the optimization only happens since LLVM 20.
bors added a commit to rust-lang-ci/rust that referenced this issue Mar 16, 2025
Rollup of 16 pull requests

Successful merges:

 - rust-lang#133055 (Expand `CloneToUninit` documentation.)
 - rust-lang#137147 (Add exclude to config.toml)
 - rust-lang#137864 (Don't drop `Rvalue::WrapUnsafeBinder` during GVN)
 - rust-lang#137890 (doc: clarify that consume can be called after BufReader::peek)
 - rust-lang#137956 (Add RTN support to rustdoc)
 - rust-lang#137968 (Properly escape regexes in Python scripts)
 - rust-lang#138082 (Remove `#[cfg(not(test))]` gates in `core`)
 - rust-lang#138275 (expose `is_s390x_feature_detected!` from `std::arch`)
 - rust-lang#138303 (Fix Ptr inconsistency in {Rc,Arc})
 - rust-lang#138309 (Add missing doc for intrinsic (Fix PR135334))
 - rust-lang#138323 (Expand and organize `offset_of!` documentation.)
 - rust-lang#138329 (debug-assert that the size_hint is well-formed in `collect`)
 - rust-lang#138465 (linkchecker: bump html5ever)
 - rust-lang#138471 (Clean up some tests in tests/ui)
 - rust-lang#138472 (Add codegen test for rust-lang#129795)
 - rust-lang#138484 (Use lit span when suggesting suffix lit cast)

r? `@ghost`
`@rustbot` modify labels: rollup
rust-timer added a commit to rust-lang-ci/rust that referenced this issue Mar 16, 2025
Rollup merge of rust-lang#138472 - KonaeAkira:master, r=Mark-Simulacrum

Add codegen test for rust-lang#129795

Adds test for rust-lang#129795.

Min LLVM version is 20 because the optimization only happens since LLVM 20.
@scottmcm
Copy link
Member

With the test added in #138472 I think we're good here 🎉

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
A-LLVM Area: Code generation parts specific to LLVM. Both correctness bugs and optimization-related issues. C-optimization Category: An issue highlighting optimization opportunities or PRs implementing such E-needs-test Call for participation: An issue has been fixed and does not reproduce, but no test has been added. I-slow Issue: Problems and improvements with respect to performance of generated code. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue.
Projects
None yet
Development

No branches or pull requests

8 participants