-
Notifications
You must be signed in to change notification settings - Fork 12.9k
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
Pretty-print always-const trait predicates correctly #119476
Conversation
--> $DIR/call-generic-method-nonconst.rs:23:34 | ||
| | ||
LL | pub const EQ: bool = equals_self(&S); | ||
| ----------- ^^ the trait `~const Foo` is not implemented for `S` | ||
| ----------- ^^ the trait `const Foo` is not implemented for `S` |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This regression is a bit unfortunate. Of course, from the perspective of the lowered code, it makes perfect sense that we register always-const trait predicate obligations in const item bodies, so technically speaking it's correct.
Still, it's bad that we're now basically saying S: const Trait
is unsatisfied // required by bound ~const Trait
in equals_self
.
This comment has been minimized.
This comment has been minimized.
2d550a4
to
19e4eb2
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
r=me unless you want a specific review
@@ -20,7 +20,8 @@ const fn equals_self<T: ~const Foo>(t: &T) -> bool { | |||
// Calling `equals_self` with something that has a non-const impl should throw an error, despite | |||
// it not using the impl. | |||
|
|||
// FIXME(effects): We should print `S: ~const Foo` here instead. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why? I believe that S: const Foo
makes sense here, since this is in an always-const context...
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It does make sense to me as I've written above, I'm just not super keen on the
diagnostic note required by this bound, *points at ~const
*.
I can imagine ppl will open gh issues for that. Anyways.
= note: expected constant `false` | ||
found constant `true` |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We may want to intercept the TypeError
here and special-case it for host params (as a follow-up)...
19e4eb2
to
01ac44a
Compare
I've removed that FIXME. @bors r=compiler-errors rollup |
🌲 The tree is currently closed for pull requests below priority 100. This pull request will be tested once the tree is reopened. |
…, r=compiler-errors Pretty-print always-const trait predicates correctly Follow-up to rust-lang#119099. r? fee1-dead
…, r=compiler-errors Pretty-print always-const trait predicates correctly Follow-up to rust-lang#119099. r? fee1-dead
Rollup of 21 pull requests Successful merges: - rust-lang#119086 (Query panic!() to useful diagnostic) - rust-lang#119239 (Remove unnecessary arm in `check_expr_yield`) - rust-lang#119298 (suppress change-tracker warnings in CI containers) - rust-lang#119319 (Document that File does not buffer reads/writes) - rust-lang#119434 (rc: Take *const T in is_dangling) - rust-lang#119444 (Rename `TyCtxt::is_closure` to `TyCtxt::is_closure_or_coroutine`) - rust-lang#119474 (Update tracking issue of naked_functions) - rust-lang#119476 (Pretty-print always-const trait predicates correctly) - rust-lang#119477 (rustdoc ui: adjust tooltip z-index to be above sidebar) - rust-lang#119479 (Remove two unused feature gates from rustc_query_impl) - rust-lang#119487 (Minor improvements in comment on `freshen.rs`) - rust-lang#119492 (Update books) - rust-lang#119494 (Deny defaults for higher-ranked generic parameters) - rust-lang#119498 (Update deadlinks of `strict_provenance` lints) - rust-lang#119505 (Don't synthesize host effect params for trait associated functions marked const) - rust-lang#119510 (Report I/O errors from rmeta encoding with emit_fatal) - rust-lang#119512 (Mark myself as back from leave) - rust-lang#119514 (coverage: Avoid a query stability hazard in `function_coverage_map`) - rust-lang#119523 (llvm: Allow `noundef` in codegen tests) - rust-lang#119534 (Update `thread_local` examples to use `local_key_cell_methods`) - rust-lang#119544 (Fix: Properly set vendor in i686-win7-windows-msvc target) r? `@ghost` `@rustbot` modify labels: rollup
Rollup merge of rust-lang#119476 - fmease:pp-always-const-trait-preds, r=compiler-errors Pretty-print always-const trait predicates correctly Follow-up to rust-lang#119099. r? fee1-dead
Follow-up to #119099.
r? fee1-dead