Skip to content

Don't FCW assoc consts in patterns #140467

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

Merged
merged 1 commit into from
Apr 30, 2025
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 10 additions & 9 deletions compiler/rustc_middle/src/mir/interpret/queries.rs
Original file line number Diff line number Diff line change
Expand Up @@ -115,15 +115,16 @@ impl<'tcx> TyCtxt<'tcx> {
// @lcnr believes that successfully evaluating even though there are
// used generic parameters is a bug of evaluation, so checking for it
// here does feel somewhat sensible.
if !self.features().generic_const_exprs() && ct.args.has_non_region_param() {
let def_kind = self.def_kind(instance.def_id());
assert!(
matches!(
def_kind,
DefKind::InlineConst | DefKind::AnonConst | DefKind::AssocConst
),
"{cid:?} is {def_kind:?}",
);
if !self.features().generic_const_exprs()
&& ct.args.has_non_region_param()
// We only FCW for anon consts as repeat expr counts with anon consts are the only place
// that we have a back compat hack for. We don't need to check this is a const argument
// as only anon consts as const args should get evaluated "for the type system".
//
// If we don't *only* FCW anon consts we can wind up incorrectly FCW'ing uses of assoc
// consts in pattern positions. #140447
&& self.def_kind(instance.def_id()) == DefKind::AnonConst
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think there should be a comment explaining why we check the def_kind here.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

doh.... yeah, why would I put it only in the test 😆

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

updated. let me know if the comment doesn't make sense to you and it needs elaborating somewhere :)

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I barely know what this entire function does, let alone what the FCW is for, so I'm afraid I am mostly lost here, but this way at least there are some breadcrumbs to follow for someone who stumbles upon this curious extra check in the future. :)

{
let mir_body = self.mir_for_ctfe(instance.def_id());
if mir_body.is_polymorphic {
let Some(local_def_id) = ct.def.as_local() else { return };
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,23 @@
//@ check-pass

// Previously the `CONST_EVALUATABLE_UNCHECKED` FCW would fire on const evaluation of
// associated consts. This is unnecessary as the FCW only needs to apply for repeat expr
// counts which are anon consts with generic parameters provided. #140447

pub struct Foo<const N: usize>;

impl<const N: usize> Foo<N> {
const UNUSED_PARAM: usize = {
let _: [(); N];
3
};

pub fn bar() {
match 1 {
Self::UNUSED_PARAM => (),
_ => (),
}
}
}

fn main() {}
Loading