Add repr(align(2)) to RcInner and ArcInner#146886

taiki-e · 2025-09-22T13:07:07Z

Rc currently assumes that RcInner has at least 2-byte alignment, but on AVR, usize has 1-byte alignment (this is because the AVR has 1-byte register sizes, so having 2-byte alignment is generally useless), breaking this assumption.

rust/library/alloc/src/rc.rs

Lines 3005 to 3008 in 9f32ccf

    
           // `Weak::new` sets this to `usize::MAX` so that it doesn’t need 
        
           // to allocate space on the heap. That's not a value a real pointer 
        
           // will ever have because RcInner has alignment at least 2. 
        
           ptr: NonNull<RcInner<T>>,

This PR adds repr(align(2)) to force RcInner to always have at least 2-byte alignment.

Note that ArcInner doesn't need repr(align(2)) because atomic types have the alignment same as its size. This PR adds a comment about this.

rustbot · 2025-09-22T13:07:12Z

r? @Mark-Simulacrum

rustbot has assigned @Mark-Simulacrum.
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

hkBst · 2025-09-22T16:46:12Z

library/alloc/src/sync.rs

It may not be necessary, but why not write it in code as well?

This is my thought as well. Even though it has no effect, it documents the fact that the type expects that alignment anyway.

Mark-Simulacrum · 2025-09-28T17:46:39Z

r=me with ArcInner also similarly updated (per comments above).

If you don't have the bandwidth to do that, happy to merge without it, but otherwise would be good to have it in a single commit/PR IMO.

taiki-e · 2025-09-30T13:43:39Z

Added repr(align(2)) also to ArcInner, and updated the comment to reflect it:

Unlike RcInner, repr(align(2)) is not strictly required because atomic types have the alignment same as its size, but we use it for consistency and clarity.

@rustbot ready

Mark-Simulacrum · 2025-09-30T13:45:22Z

@bors r+ rollup

bors · 2025-09-30T13:45:25Z

📌 Commit 9e79fac has been approved by Mark-Simulacrum

It is now in the queue for this repository.

Rollup of 6 pull requests Successful merges: - #142506 (Add `Path::has_trailing_sep` and related methods) - #146886 (Add repr(align(2)) to RcInner and ArcInner) - #147166 (several small `proc_macro` cleanups) - #147172 (bootstrap: build bootstrap docs with in-tree rustdoc) - #147181 (cg_llvm: Replace enum `MetadataType` with a list of `MetadataKindId` constants) - #147187 (remove unnecessary test directives) r? `@ghost` `@rustbot` modify labels: rollup

Rollup merge of #146886 - taiki-e:rc-inner-align, r=Mark-Simulacrum Add repr(align(2)) to RcInner and ArcInner `Rc` currently assumes that `RcInner` has at least 2-byte alignment, but on AVR, `usize` has 1-byte alignment (this is because the AVR has 1-byte register sizes, so having 2-byte alignment is generally useless), breaking this assumption. https://github.com/rust-lang/rust/blob/9f32ccf35fb877270bc44a86a126440f04d676d0/library/alloc/src/rc.rs#L3005-L3008 This PR adds `repr(align(2))` to force `RcInner` to always have at least 2-byte alignment. Note that `ArcInner` doesn't need `repr(align(2))` because atomic types have the alignment same as its size. This PR adds a comment about this.

…ulacrum Add repr(align(2)) to RcInner and ArcInner `Rc` currently assumes that `RcInner` has at least 2-byte alignment, but on AVR, `usize` has 1-byte alignment (this is because the AVR has 1-byte register sizes, so having 2-byte alignment is generally useless), breaking this assumption. https://github.com/rust-lang/rust/blob/9f32ccf35fb877270bc44a86a126440f04d676d0/library/alloc/src/rc.rs#L3005-L3008 This PR adds `repr(align(2))` to force `RcInner` to always have at least 2-byte alignment. Note that `ArcInner` doesn't need `repr(align(2))` because atomic types have the alignment same as its size. This PR adds a comment about this.

rustbot assigned Mark-Simulacrum Sep 22, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-libs Relevant to the library team, which will review and decide on the PR/issue. labels Sep 22, 2025

hkBst reviewed Sep 22, 2025

View reviewed changes

Mark-Simulacrum 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 Sep 28, 2025

Add repr(align(2)) to RcInner and ArcInner

9e79fac

taiki-e force-pushed the rc-inner-align branch from 71f9091 to 9e79fac Compare September 30, 2025 13:39

taiki-e changed the title ~~Add repr(align(2)) to RcInner~~ Add repr(align(2)) to ArcInner Sep 30, 2025

taiki-e changed the title ~~Add repr(align(2)) to ArcInner~~ Add repr(align(2)) to RcInner and ArcInner Sep 30, 2025

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 Sep 30, 2025

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 Sep 30, 2025

matthiaskrgr mentioned this pull request Sep 30, 2025

Rollup of 6 pull requests #147197

Merged

bors merged commit f7c7e34 into rust-lang:master Sep 30, 2025
10 checks passed

rustbot added this to the 1.92.0 milestone Sep 30, 2025

taiki-e deleted the rc-inner-align branch September 30, 2025 23:48

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Add repr(align(2)) to RcInner and ArcInner#146886