Enforce Sized return types on Fn* bounds
#83915
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
rust-highfive
added
the
S-waiting-on-review
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
I'm puzzled, I haven't been able to reproduce the problem locally. |
estebank
commented
Apr 6, 2021
| def_id: lang_items.sized_trait().unwrap(), | ||
| substs: self.tcx().mk_substs_trait(ty, &[]), | ||
| }; | ||
| if !matches!(ty.kind(), ty::Projection(_) | ty::Opaque(..)) { |
There was a problem hiding this comment.
For reviewer context, the filtering out of projections and opaque types is a hack and shouldn't be necessary long term if we find a better place to inject the Sized obligation.
If I don't keep this in, then we start failing stage 1 builds. This tells me there's a better place to inject this obligation, but I haven't been able to find it.
estebank
commented
Apr 6, 2021
| if [lang_items.fn_once_trait(), lang_items.fn_trait(), lang_items.fn_mut_trait()] | ||
| .contains(&Some(obligation.predicate.def_id())) | ||
| // Skip `MiscObligation`s for better output. | ||
| && matches!(obligation.cause.code, ObligationCauseCode::BindingObligation(_, _)) |
There was a problem hiding this comment.
You can look at the diff between the two commits to see what this check changes.
In a `fn() -> Out` bound, enforce `Out: Sized` to avoid unsoundness. Fix rust-lang#82633.
estebank
force-pushed
the
closure-return-must-be-sized
branch
from
093d25b to
670d20d
Compare
estebank
commented
Apr 6, 2021
| { | ||
| // Do not allow `foo::<fn() -> A>();` for `A: !Sized` (#82633) | ||
| let fn_sig = obligation.predicate.self_ty().skip_binder().fn_sig(self.tcx()); | ||
| let ty = self.infcx.replace_bound_vars_with_placeholders(fn_sig.output()); |
There was a problem hiding this comment.
Using replace_bound_vars_with_placeholders pacified CI, but I still don't understand why it wasn't triggering on my machine. Maybe the assert! wasn't being run locally?
|
Some discussion on Zulip: https://zulip-archive.rust-lang.org/144729wgtraits/32496PR83915.html |
|
Alternate branch: https://github.com/nikomatsakis/rust/tree/issue-82633-sized-return-types This fix is not perfect, but may be a better stepping stone. |
crlf0710
added
S-blocked
apiraino
added
the
T-compiler
|
☔ The latest upstream changes (presumably #94108) made this pull request unmergeable. Please resolve the merge conflicts. |
bors
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Sep 21, 2022
…zed, r=jackh726 a fn pointer doesn't implement `Fn`/`FnMut`/`FnOnce` if its return type isn't sized I stumbled upon rust-lang#83915 which hasn't received much attention recently, and I wanted to revive it since this is one existing soundness hole that seems pretty easy to fix. I'm not actually sure that the [alternative approach described here](rust-lang#83915 (comment)) is sufficient, given the `src/test/ui/function-pointer/unsized-ret.rs` example I provided below. Rebasing the branch mentioned in that comment and testing that UI test, it seems that we actually end up only observing that `str: !Sized` during monomorphization, whereupon we ICE. Even if we were to fix that ICE, ideally we'd be raising an error that a fn pointer is being used badly during _typecheck_ instead of monomorphization, hence adapting the original approach in rust-lang#83915. I am happy to close this if people would prefer we rebase the original PR and land that -- I am partly opening to be annoying and get people thinking about this unsoundness again ❤️ 😸 cc: `@estebank` and `@nikomatsakis` r? types Here's a link to the thread: https://rust-lang.zulipchat.com/#narrow/stream/144729-t-types/topic/PR.20.2383915/near/235421351 for more context.
|
I think this PR can now be closed in favor of #100096? 🤔 feel free to reopen if that isn't the case |
|
@lcnr I'm just happy it's finally handled |
RalfJung
pushed a commit
to RalfJung/rust-analyzer
that referenced
this pull request
Apr 20, 2024
…ckh726 a fn pointer doesn't implement `Fn`/`FnMut`/`FnOnce` if its return type isn't sized I stumbled upon #83915 which hasn't received much attention recently, and I wanted to revive it since this is one existing soundness hole that seems pretty easy to fix. I'm not actually sure that the [alternative approach described here](rust-lang/rust#83915 (comment)) is sufficient, given the `src/test/ui/function-pointer/unsized-ret.rs` example I provided below. Rebasing the branch mentioned in that comment and testing that UI test, it seems that we actually end up only observing that `str: !Sized` during monomorphization, whereupon we ICE. Even if we were to fix that ICE, ideally we'd be raising an error that a fn pointer is being used badly during _typecheck_ instead of monomorphization, hence adapting the original approach in #83915. I am happy to close this if people would prefer we rebase the original PR and land that -- I am partly opening to be annoying and get people thinking about this unsoundness again ❤️ 😸 cc: `@estebank` and `@nikomatsakis` r? types Here's a link to the thread: https://rust-lang.zulipchat.com/#narrow/stream/144729-t-types/topic/PR.20.2383915/near/235421351 for more context.
RalfJung
pushed a commit
to RalfJung/rust-analyzer
that referenced
this pull request
Apr 27, 2024
…ckh726 a fn pointer doesn't implement `Fn`/`FnMut`/`FnOnce` if its return type isn't sized I stumbled upon #83915 which hasn't received much attention recently, and I wanted to revive it since this is one existing soundness hole that seems pretty easy to fix. I'm not actually sure that the [alternative approach described here](rust-lang/rust#83915 (comment)) is sufficient, given the `src/test/ui/function-pointer/unsized-ret.rs` example I provided below. Rebasing the branch mentioned in that comment and testing that UI test, it seems that we actually end up only observing that `str: !Sized` during monomorphization, whereupon we ICE. Even if we were to fix that ICE, ideally we'd be raising an error that a fn pointer is being used badly during _typecheck_ instead of monomorphization, hence adapting the original approach in #83915. I am happy to close this if people would prefer we rebase the original PR and land that -- I am partly opening to be annoying and get people thinking about this unsoundness again ❤️ 😸 cc: `@estebank` and `@nikomatsakis` r? types Here's a link to the thread: https://rust-lang.zulipchat.com/#narrow/stream/144729-t-types/topic/PR.20.2383915/near/235421351 for more context.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In a
fn() -> Outbound, enforceOut: Sizedto avoid unsoundness.Fix #82633.
r? @pnkfelix cc @nikomatsakis