Make sure that outer opaques capture inner opaques's lifetimes even with precise capturing syntax #131789
Conversation
This file contains 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
rustbot
added
S-waiting-on-review
fmease
approved these changes
Oct 19, 2024
There was a problem hiding this comment.
Makes sense to me. r=me with nitpicks addressed at your leisure unless you want sb. else to review it.
Yesterday I finally spent quite some time carefully reading up on the way we deal with opaque types, namely AST lowering, the whole parameter duplication thing, variances of opaques and member constraints.
| _ => false, | ||
| }) | ||
| { | ||
| for (ident, id, _) in self.resolver.extra_lifetime_params(opaque_ty_node_id) { |
There was a problem hiding this comment.
Right so IIUC, we can no longer take the extra lifetime params because if we did, they would no longer be available when lowering any nested opaques (here: impl Sized) after having lowered any overarching opaques (here: impl IntoIterator<Item = impl Sized>).
There was a problem hiding this comment.
Yep, we need to be able to access these extra lifetimes both when walking the bounds of the parent opaque, and also lowering the child opaque itself. I have no idea why we were using take before, it's not like we really enforced it.
…ith precise capturing syntax
compiler-errors
force-pushed
the
capture-more
branch
from
77cc69c
to
70746d0
Compare
|
@bors r=fmease |
bors
added
S-waiting-on-bors
bors
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Oct 19, 2024
…iaskrgr Rollup of 12 pull requests Successful merges: - rust-lang#116863 (warn less about non-exhaustive in ffi) - rust-lang#127675 (Remove invalid help diagnostics for const pointer) - rust-lang#131772 (Remove `const_refs_to_static` TODO in proc_macro) - rust-lang#131789 (Make sure that outer opaques capture inner opaques's lifetimes even with precise capturing syntax) - rust-lang#131795 (Stop inverting expectation in normalization errors) - rust-lang#131920 (Add codegen test for branchy bool match) - rust-lang#131921 (replace STATX_ALL with (STATX_BASIC_STATS | STATX_BTIME) as former is deprecated) - rust-lang#131925 (Warn on redundant `--cfg` directive when revisions are used) - rust-lang#131931 (Remove unnecessary constness from `lower_generic_args_of_path`) - rust-lang#131932 (use tracked_path in rustc_fluent_macro) - rust-lang#131936 (feat(rustdoc-json-types): introduce rustc-hash feature) - rust-lang#131939 (Get rid of `OnlySelfBounds`) Failed merges: - rust-lang#131181 (Compiletest: Custom differ) r? `@ghost` `@rustbot` modify labels: rollup
rust-timer
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Oct 20, 2024
Rollup merge of rust-lang#131789 - compiler-errors:capture-more, r=fmease Make sure that outer opaques capture inner opaques's lifetimes even with precise capturing syntax When lowering an opaque, we must capture and duplicate all of the lifetimes in the opaque's bounds to correctly lower the opaque's bounds. We do this *even if* the lifetime is not captured according to the `+ use<>` precise capturing bound; in that case, we will later reject that captured lifetime. For example, Given an opaque like `impl Sized + 'a + use<>`, we will still duplicate `'a` but later error that it is not mentioned in the `use<>` bound. The current heuristic was not properly handling cases like: ``` //@ edition: 2024 fn foo<'a>() -> impl Trait<Assoc = impl Trait2> + use<> {} ``` Which forces the outer `impl Trait` to capture `'a` since `impl Trait2` *implicitly* captures `'a` due to the new lifetime capture rules for edition 2024. We were only capturing lifetimes syntactically mentioned in the bounds. (Note that this still is an error; we just need to capture `'a` so it is handled later in the compiler correctly -- hence the ICE in rust-lang#131769 where a late-bound lifetime was being referenced outside of its binder). This PR reworks the way we collect lifetimes to capture and duplicate in AST lowering to fix this. Fixes rust-lang#131769
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.
When lowering an opaque, we must capture and duplicate all of the lifetimes in the opaque's bounds to correctly lower the opaque's bounds. We do this even if the lifetime is not captured according to the
+ use<>precise capturing bound; in that case, we will later reject that captured lifetime. For example, Given an opaque likeimpl Sized + 'a + use<>, we will still duplicate'abut later error that it is not mentioned in theuse<>bound.The current heuristic was not properly handling cases like:
Which forces the outer
impl Traitto capture'asinceimpl Trait2implicitly captures'adue to the new lifetime capture rules for edition 2024. We were only capturing lifetimes syntactically mentioned in the bounds. (Note that this still is an error; we just need to capture'aso it is handled later in the compiler correctly -- hence the ICE in #131769 where a late-bound lifetime was being referenced outside of its binder).This PR reworks the way we collect lifetimes to capture and duplicate in AST lowering to fix this.
Fixes #131769