Normalize opaques with late-bound vars again
We have a hack in the compiler where if an opaque has escaping late-bound vars, we skip revealing it even though we *could* reveal it from a technical perspective. First of all, this is weird, since we really should be revealing all opaques in `Reveal::All` mode. Second of all, it causes subtle bugs (linked below).
I attempted to fix this in #100980, which was unfortunately reverted due to perf regressions on codebases that used really deeply nested futures in some interesting ways. The worst of which was #103423, which caused the project to hang on build. Another one was #104842, which was just a slow-down, but not a hang. I took some time afterwards to investigate how to rework `normalize_erasing_regions` to take advantage of better caching, but that effort kinda fizzled out (#104133).
However, recently, I was made aware of more bugs whose root cause is not revealing opaques during codegen. That made me want to fix this again -- in the process, interestingly, I took the the minimized example from https://github.com/rust-lang/rust/issues/103423#issuecomment-1292947043, and it doesn't seem to hang any more...
Thinking about this harder, there have been some changes to the way we lower and typecheck async futures that may have reduced the pathologically large number of outlives obligations (see description of #103423) that we were encountering when normalizing opaques with bound vars the last time around:
* #104321 (lower `async { .. }` directly as a generator that implements `Future`, removing the `from_generator` shim)
* #104833 (removing an `identity_future` fn that was wrapping desugared future generators)
... so given that I can see:
* No significant regression on rust perf bot (https://github.com/rust-lang/rust/pull/107620#issuecomment-1600070317)
* No timeouts in crater run I did (https://github.com/rust-lang/rust/pull/107620#issuecomment-1605428952, rechecked failing crates in https://github.com/rust-lang/rust/pull/107620#issuecomment-1605973434)
... and given that this PR:
* Fixes#104601
* Fixes#107557
* Fixes#109464
* Allows us to remove a `DefiningAnchor::Bubble` from codegen (75a8f681837c70051e0200a14f58ae07dbe58e66)
I'm inclined to give this another shot at landing this. Best case, it just works -- worst case, we get more examples to study how we need to improve the compiler to make this work.
r? types
Make associated type bounds in supertrait position implied
`trait A: B<Assoc: C> {}` should be able to imply both `Self: B` and `<Self as B>::Assoc: C`. Adjust the way that we collect implied predicates to do so.
Fixes#112573Fixes#112568
Fix runnable detection for `#[tokio::test]`
fix#15141
It is hacky, and it wouldn't work for e.g. this case:
```Rust
use ::core::prelude;
#[prelude::v1::test]
fn foo() {
}
```
But it works for the tokio case. We should use the name resolution here somehow, and after that we should probably also get rid of the ast based `test_related_attribute` function.
Refactor metadata emission to avoid visiting HIR
This PR refactors metadata emission to be based on tables and iteration over definitions.
In a first part, this PR moves information from the `EntryKind` enum to tables, until removing the `EntryKind` enum.
In a second part, the iteration scheme is refactored to avoid fetching HIR unless strictly necessary.
r? `@ghost`
internal: add `library` fixture meta
Currently, there is no way to specify `CrateOrigin` of a file fixture ([this] might be a bug?). This PR adds `library` meta to explicitly specify the fixture to be `CrateOrigin::Library` and also makes sure crates that belong to a library source root are set `CrateOrigin::Library`.
(`library` isn't really the best name. It essentially means that the crate is outside workspace but `non_workspace_member` feels a bit too long. Suggestions for the better name would be appreciated)
Additionally:
- documents the fixture meta syntax as thoroughly as possible
- refactors relevant code
[this]: 4b06d3c595/crates/base-db/src/fixture.rs (L450)
Rollup of 6 pull requests
Successful merges:
- #111571 (Implement proposed API for `proc_macro_span`)
- #112236 (Simplify computation of killed borrows)
- #112867 (More `ImplSource` nits)
- #113019 (add note for non-exhaustive matches with guards)
- #113094 (Fix invalid HTML DIV tag used in HEAD)
- #113111 (add myself to review for t-types stuff)
r? `@ghost`
`@rustbot` modify labels: rollup
add note for non-exhaustive matches with guards
Associated issue: #92197
When a match statement includes guards on every match arm (and is therefore necessarily non-exhaustive), add a note to the error E0004 diagnostic noting this.
Fix `self` and `super` path resolution in block modules
This PR fixes `self` and `super` path resolution with block modules involved.
Previously, we were just going up the module tree count-of-`super` times without considering block modules in the way, and then if we ended up in a block `DefMap`, we adjust "to the containing crate-rooted module". While this seems to work in most real-world cases, we failed to resolve them within peculiar module structures.
`self` and `super` should actually be resolved to the nearest non-block module, and the paths don't necessarily resolve to a crate-rooted module. This PR makes sure every `self` and `super` segment in paths are resolved to a non-block module.
internal: support `#[rustc_coinductive]`
rust-lang/rust#100386 changed the trait solver so that `Sized` is treated as coinductive trait, just like auto traits. This is now controlled by the perma-unstable `#[rustc_coinductive]` attribute (rust-lang/rust#108033), which this PR adds support for.
In practice, I don't think this matters much if at all. Currently we don't give chalk enough information so chalk cannot precisely (dis)prove `Sized` bounds.
editor/code: Use `@tsconfig/strictest` to define type checking rules
Motivation
-----------
This change aims to make it easier to manage tsconfig by [``@tsconfig/strictest`](https://www.npmjs.com/package/@tsconfig/strictest)` and intend to leave to create "ideal" rules about TypeScript's type checking.
Implementation
---------------
This change removes some duplicated rules defined in ``@tsconfig/strictest`` and add disabing some strict rules that fails with the current codebase.
Validate fluent variable references in tests
Closes#101109
Under `cfg(test)`, the `fluent_messages` macro will emit a list of variables referenced by each message and its attributes. The derive attribute will now emit a `#[test]` that checks that each referenced variable exists in the structure it's applied to.
mir opt + codegen: handle subtyping
fixes#107205
the same issue was caused in multiple places:
- mir opts: both copy and destination propagation
- codegen: assigning operands to locals (which also propagates values)
I changed codegen to always update the type in the operands used for locals which should guard against any new occurrences of this bug going forward. I don't know how to make mir optimizations more resilient here. Hopefully the added tests will be enough to detect any trivially wrong optimizations going forward.
Motivation
-----------
This change aims to make it easier to manage tsconfig.
This intend to leave to create "ideal" rules about TypeScript's type checking.
Implementation
---------------
This change removes some duplicated rules defined in `@tsconfig/strictest` and add disabing some strict rules
that fails with the current codebase.
