Implement `advance_by` via `try_fold` for `Sized` iterators
When `try_fold` is overriden, it is usually easier for compilers to optimize.
Example difference: https://iter.godbolt.org/z/z8cEfnKro
Use the fn_span when emitting function calls for better debug info.
This especially improves the developer experience for long chains of function calls that span multiple lines, which is common with builder patterns, chains of iterator/future combinators, etc.
try-job: armhf-gnu
try-job: test-various
try-job: x86_64-msvc-1
try-job: arm-android
r? `@jieyouxu`
Update IDEs to use rustfmt 2024, fix Zed settings
Update IDEs to use rustfmt 2024, fix Zed settings
- Update IDE `rust-analyzer` settings to use 2024 rather than 2021.
- Fix Zed settings by removing `${workspaceFolder}/` from paths.
Fast path for processing some obligations in the new solver
Fast path applies to:
- Dyn compatibility predicates
- Region and type outlives predicates
- Trivially sized predicates
Emit a warning if the doctest `main` function will not be run
Fixes#140310.
I think we could try to go much further like adding a "link" (ie UI annotations) on the `main` function in the doctest. However that will require some more computation, not sure if it's worth it or not. Can still be done in a follow-up if we want it.
For now, this PR does two things:
1. Pass the `DiagCtxt` to the doctest parser to emit the warning.
2. Correctly generate the `Span` to where the doctest is starting (I hope the way I did it isn't too bad either...).
cc `@fmease`
r? `@notriddle`
Match on lang item kind instead of using an if/else chain
Similar to how the new solver does this. Just noticed while I was adding a new entry to the chain 😆
[win][ci] Update LLVM toolchain used to build LLVM to 20
While trying to get the aarch64-msvc build working correctly (#140136), I needed to update the version of LLVM used to build LLVM in Windows CI runners to 20 (as this has improved support for Arm64 and Arm64EC on Windows).
This catches Windows up to Linux which was updated to 20 by #137189
try-job: `x86_64-apple-*`
try-job: `aarch64-apple`
try-job: `x86_64-msvc-*`
try-job: `i686-msvc-*`
try-job: `x86_64-mingw-*`
update llvm-tools logic for `dist` and `install` steps
First commit aligns `build_steps::compile` and `build_steps::dist` logics for copying llvm-tools, and the second commit adds the correct `should_run` condition for `LlvmTools` step as the previous one was clearly incorrect.
Fixes#140913
Merge mir query analysis invocations
r? `@ghost`
same thing as https://github.com/rust-lang/rust/pull/140854 just a different set of queries
Doing this in general has some bad cache coherence issues because the query caches are laid out in Vec<QueryResult> lists per query where each index refers to a DefId in the same order as we're iterating. Iterating two or more lists at the same time does have cache issues, so I want to poke a bit at it to see if we can't merge just a few of them at a time.
Invoke a query only when it doesn't return immediately anyway
This should cause less query key caching and less dep graph data, hopefully resulting in some perf improvements
Remove manual WF hack
We do not need this hack anymore since we fixed the candidate selection problems with `Sized` bounds. We prefer built-in sized bounds now since #138176, which fixes the only regression this hack was intended to fix.
While this theoretically is broken for some code, for example, when there a param-env bound that shadows an impl or built-in trait, we don't see it in practice and IMO it's not worth the burden of having to maintain this wart in `compare_method_predicate_entailment`.
The code that regresses is, for example:
```rust
trait Bar<'a> {}
trait Foo<'a, T> {
fn method(&self)
where
Self: Bar<'a>;
}
struct W<'a, T>(&'a T)
where
Self: Bar<'a>;
impl<'a, 'b, T> Bar<'a> for W<'b, T> {}
impl<'a, 'b, T> Foo<'a, T> for W<'b, T> {
fn method(&self) {}
}
```
Specifically, I don't believe this is really going to be encountered in practice. For this to fail, there must be a where clause in the *trait method* that would shadow an impl or built-in (non-`Sized`) candidate in the trait, and this shadowing would need to be encountered when solving a nested WF goal from the impl self type.
See #108544 for the original regression. Crater run is clean!
r? lcnr
Including:
- Infer `label {}` and `const` operands.
- Correctly handle unsafe check inside `label {}`.
- Fix an embarrassing parser typo that cause labels to never be part of the AST
Use the new solver in the `impossible_predicates`
The old solver is unsound for many reasons. One of which was weaponized by `@lcnr` in #140212, where the old solver was incompletely considering a dyn vtable method to be impossible and replacing its vtable entry with a null value. This null function could be called post-mono.
The new solver is expected to be less incomplete due to its correct handling of higher-ranked aliases in relate. This PR switches the `impossible_predicates` query to use the new solver, which patches this UB.
r? lcnr
For some reason we had them in some projects, I'm not sure why. But this caused cache priming to appear stuck - because it uses a set of crate IDs for the actual work, but for the number of crates to index it just uses `db.all_crates().len()`.
