This is a newer version of the same action. None of the uses here were particularly special (no complex features of v3 were used) so this is a straightforward as-is upgrade.
Add diagnostic item for `std::iter::Enumerate`
This adds a diagnostic item for `std::iter::Enumerate`. The change will be used by the clippy `unused_enumerate_index` lint to move away from type paths to using diagnostic items.
see: https://github.com/rust-lang/rust-clippy/issues/5393
windows fill_utf16_buf: explain the expected return value
The comment just says "return what the syscall returns", but that doesn't work for all syscalls as the Windows API is not consistent in how buffer size is negotiated. For instance, GetUserProfileDirectoryW works a bit differently, and so home_dir_crt has to translate this to the usual protocol itself. So it's worth describing that protocol.
r? ``@ChrisDenton``
Bump Fuchsia versions
This updates the Fuchsia commit used in `auto - x86_64-gnu-integration` CI bot to use the Rust commit 703dc9ce64. This should help improve the coverage of this builder.
It also updates the SDK version to F20.20240412.3.1, and the Fuchsia Clang version to c777c011a709dffd4fa5e79cad7947b7c3405d02.
r? ``@tmandry``
More DefineOpaqueTypes::Yes
This accepts more code on stable. It is now possible to have match arms return a function item `foo::<ConcreteType>` and a function item `foo::<OpaqueTypeInDefiningScope>` in another, and that will constrain `OpaqueTypeInDefiningScope` to have the hidden type `ConcreteType`. So the following function will now compile, but on master it errors with a type mismatch on the second match arm
```rust
// The function item whose generic params we want to merge.
fn foo<T>(t: T) -> T { t }
// Helper ensuring we can constrain `T` on `F` without explicitly specifying it
fn bind<T, F: FnOnce(T) -> T>(_: T, f: F) -> F { f }
fn k() -> impl Sized {
let x = match true {
true => {
// `f` is `FnDef(foo, [infer_var])`
let f = foo;
// Get a value of an opaque type on stable
let t = k();
// this returns `FnDef(foo, [k::return])`
bind(t, f)
}
false => foo::<()>,
};
todo!()
}
```
r? ``@compiler-errors``
cc https://github.com/rust-lang/rust/issues/116652
Test `#[unix_sigpipe = "inherit"]` with both `SIG_DFL` and `SIG_IGN`
Extend our `#[unix_sigpipe = "inherit"]` test so that it detects if `SIGPIPE` wrongly ends up being `SIG_DFL` when the parent has `SIG_IGN`. We have no current test for this particular case.
Tracking issue: https://github.com/rust-lang/rust/issues/97889
delegation: Support renaming, and async, const, extern "ABI" and C-variadic functions
Also allow delegating to functions with opaque types (`impl Trait`).
The delegation item will refer to the original opaque type from the callee, fresh opaque type won't be created, which seems like a reasonable behavior.
(Such delegation items will cause query cycles when used in trait impls, but it can be fixed later.)
Part of https://github.com/rust-lang/rust/issues/118212.
Require explicitly marking closures as coroutines
instead of relying on patching up the closure to be a coroutine if it happens to contain a `yield` expression.
I only do this in the 2024 edition, as the `gen` keyword is only available there.
CI: dynamic jobs
This PR modifies our CI workflows to be dynamic. This means that when a GitHub event is generated, we will run a Python script (`calculate-job-matrix.py`), which decides which CI jobs should be generated. These jobs are defined in `src/ci/github-actions/jobs.yml`).
This should provide a few benefits:
- Once the migration to dynamic jobs is complete, we shouldn't need `expand-yaml-anchors` anymore.
- The job table on PRs (and also the left job column on auto/try builds) should be much cleaner and contain only the jobs that are actually relevant/executed.
- It should be much easier to support dynamic try builds, i.e. to run an arbitrary CI job on a try build.
See [this Zulip discussion](https://rust-lang.zulipchat.com/#narrow/stream/242791-t-infra/topic/job.20matrix.20re-ordered.20PR.20list) for more context.
r? `@ghost`
It's a highly misleading name, because it's completely different to
`MetaItem::name_value_literal`. Specifically, it doesn't match
`MetaItemKind::NameValue` (e.g. `#[foo = 3]`), it matches
`MetaItemKind::List` (e.g. `#[foo(3)]`).
Stop using LLVM struct types for alloca
The alloca type has no semantic meaning, only the size (and alignment, but we specify it explicitly) matter. Using `[N x i8]` is a more direct way to specify that we want `N` bytes, and avoids relying on LLVM's struct layout. It is likely that a future LLVM version will change to an untyped alloca representation.
Split out from #121577.
r? `@ghost`
