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
Migrate to modern datetime API
# PR Summary
This small PR resolves the `datetime` library warnings:
```python
DeprecationWarning: datetime.datetime.utcnow() is deprecated and scheduled for removal in a future version. Use timezone-aware objects to represent datetimes in UTC: datetime.datetime.now(datetime.UTC). or datetime.datetime.utcnow()
```
Note that `.replace(tzinfo=None)` allows to keep the original behavior where the time appears as a naive UTC timestamp (i.e., without any timezone offset). Comparision:
```python
# With .utcnow() or .now(datetime.timezone.utc).replace(tzinfo=None)
Time,Idle
2025-05-14T15:40:25.013414,98.73417721518987
# With .now(datetime.timezone.utc)
Time,Idle
2025-05-14T15:40:25.013414+00:00,98.73417721518987
```
Add negative test coverage for `-Clink-self-contained` and `-Zlinker-features`
Noticed while reviewing stabilization #140525 that we don't have any negative test coverage for these flags. Feel free to cherry-pick these tests into the stabilization PR, or we can land these before separately.
r? `@lqd`
std: explain prefer `TryInto` over `TryFrom` when specifying traits bounds on generic function
Fixes#140761
This PR keeps the explanations of `Into` and `From` consistent and adds explanations for `TryInto` and `TryFrom`.
r? libs
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
remove `RustfmtState` to reduce `initial_rustfmt` complexity
The current use of `RustfmtState` doesn't serve its main purpose as it never does the lazy evaulation since `Build::build` forces it to be ready on the early stage. If we want rustfmt to be ready on the early stage, we don't need to have `RustfmtState` complexity at all.
Improve `dangerous_implicit_aurorefs` diagnostic output
This PR *greatly* improves the `dangerous_implicit_aurorefs` lint diagnostic output.
Kind of related to #140721.
r? ```@jieyouxu``` (maybe)
docs(library/core/src/pin): fix typo "necessarily" -> "necessary"
Fix a typo in [`library/core/src/pin.rs`](14662fabeb/library/core/src/pin.rs), from
> As we'll see later, this is **necessarily** from the time the value is first pinned until the end of its lifespan.
to
> As we'll see later, this is **necessary** from the time the value is first pinned until the end of its lifespan.
(my emphasis).
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
Rollup of 7 pull requests
Successful merges:
- #140827 (Do not ICE when reassigning in GatherLocalsVisitor on the bad path)
- #140904 (Add an issue template for future-incompatible lints)
- #140953 (Fix a compiletest blessing message)
- #140973 (Update rustix to 1.0.7 for bootstrap)
- #140976 (Add `Ipv4Addr` and `Ipv6Addr` diagnostic items)
- #140988 (MaybeUninit::write: fix doc)
- #140989 (Suggest replace f with f: Box<f> when expr field is short hand)
r? `@ghost`
`@rustbot` modify labels: rollup
The current use of `RustfmtState` doesn't serve its main purpose as it
never does the lazy evaulation since `Build::build` forces it to be ready
on the early stage. If we want rustfmt to be ready on the early stage, we
don't need to have `RustfmtState` complexity at all.
Signed-off-by: onur-ozkan <work@onurozkan.dev>
Temporarily use Windows Server 2022 instead of Windows Server 2025 images
Tracking issue: https://github.com/rust-lang/rust/issues/141022
At the moment, it seems like Windows Server 2025 20250504.1.0 is misconfigured, causing insufficient disk space failures. Temporarily go back to Windows Server 2022 in the hope that those are not also misconfigured to try to unblock the queue.
r? `@marcoieni` (or infra-ci)
At the moment, it seems like Windows Server 2025 20250504.1.0 is
misconfigured causing insufficient disk space failures. Temporarily go
back to Windows Server 2022 in the hope that those are not also
misconfigured.
