8526: fix: Do not show flyimports in trait or impl declarations r=SomeoneToIgnore a=SomeoneToIgnore
Part of https://github.com/rust-analyzer/rust-analyzer/issues/8518
Removes autoimport suggestions for the case:
> inside trait definitions / impls (trait Trait {$0} / impl Foo {$0}), nothing except the fn, type and const keywords (and the full item completions for trait impls) should appear (currently many types and autoimport suggestions)
Co-authored-by: Kirill Bulatov <mail4score@gmail.com>
8514: Add more profiling spans into type inference r=matklad a=SomeoneToIgnore
Fixes https://github.com/rust-analyzer/rust-analyzer/issues/8503
I've added a minimal set of spans to remove `???` and showcase the underlying issue:
<img width="1552" alt="image" src="https://user-images.githubusercontent.com/2690773/114722983-f2181900-9d42-11eb-821d-9e93ded5f81f.png">
`db.trait_solve` reliably produces the same timings for the same input in the same place for me, despite supposedly being cached by Salsa.
In my perception, this is a bit odd, so I've decided to stop at this point and discuss it with people with better knowledge on the topic.
Co-authored-by: Kirill Bulatov <mail4score@gmail.com>
8432: decl_check: consider outer scopes' allows r=jonas-schievink a=lf-
Fix#8417. Also makes it less noisy about no_mangle annotated stuff the
user can do nothing about.
Note: this still is broken with bitfield! macros. A repro in an ignore
test is included here. I believe this bug is elsewhere, and I don't
think I can work around it here.
I would like help filing the remaining bug, as it does actually affect
users, but I don't know how to describe the behaviour (or even if it
is unintended).
Co-authored-by: Jade <software@lfcode.ca>
8354: Distinguishing between different operators in semantic highlighting r=matklad a=chetankhilosiya
Co-authored-by: Chetan Khilosiya <chetan.khilosiya@gmail.com>
8415: Fix faulty assertion when extracting function with macro call r=matklad a=brandondong
**Reproduction:**
```rust
fn main() {
let n = 1;
let k = n * n;
dbg!(n);
}
```
1. Select the second and third lines of the main function. Use the "Extract into function" code assist.
2. Panic occurs in debug, error is logged in release: "[ERROR ide_assists::handlers::extract_function] assertion failed: matches!(path, ast :: Expr :: PathExpr(_))".
3. Function generates successfully on release where the panic was bypassed.
```rust
fn fun_name(n: i32) {
let k = n * n;
dbg!(n);
}
```
**Cause:**
- The generated function will take `n` as a parameter. The extraction logic needs to search the usages of `n` to determine whether it is used mutably or not. The helper `path_element_of_reference` is called for each usage but the second usage is a macro call and fails the `Expr::PathExpr(_)` match assertion.
- The caller of `path_element_of_reference` does implicitly assume it to be a `Expr::PathExpr(_)` in how it looks at its parent node for determining whether it is used mutably. This logic will not work for macros.
- I'm not sure if there are any other cases besides macros where it could be something other than a `Expr::PathExpr(_)`. I tried various examples and could not find any.
**Fix:**
- Update assertion to include the macro case.
- Add a FIXME to properly handle checking if a macro usage requires mutable access. For now, return false instead of running the existing logic that is tailored for `Expr::PathExpr(_)`'s.
Co-authored-by: Brandon <brandondong604@hotmail.com>
8496: Exclude nightly tag from git describe to fix version string r=lnicola a=lnicola
Otherwise if we run `git describe` on the release day we pick up the `nightly` tag from the previous release.
changelog fix
bors r+
Co-authored-by: Laurențiu Nicola <lnicola@dend.ro>
Conceptually, using a *message* here is wrong, because this is a
"status", rather than "point in time" thing. But statuses are an LSP
extension, while messages are stable. As a compromise, send message only
for more critical `metadata` failures, and only once per state change.
