Switch `x86_64-msvc-{1,2}` back to Windows Server 2025 images
New Windows Server 2025 images have been released (**20250527.1.0**). New images appear to not exhibit the lack-of-disk-space problem as tracked by rust-lang/rust#141022, and the new runner image's storage capacity appears to be configured correctly.
Windows Server 2025 image version **20250527.1.0** release notes: <https://github.com/actions/runner-images/releases/tag/win25%2F20250527.1>.
Resolvesrust-lang/rust#141022.
tools-aux ci runner: also cross-test doctests in Miri
Miri now supports running doctests across different targets. Let's use that to run the std doctests on aarch64-apple-darwin, i686-pc-windows-msvc.
try-job: x86_64-gnu-aux
`UsePath` contains a `SmallVec<[Res; 3]>`. This holds up to three `Res`
results, one per namespace (type, value, or macro). `lower_import_res`
takes a `PerNS<Option<Res<NodeId>>>` result and lowers it into the
`SmallVec`. This is pretty weird. The input `PerNS` makes it clear which
`Res` belongs to which namespace, but the `SmallVec` throws that
information away.
And code that operates on the `SmallVec` tends to use iteration (or even
just grabbing the first entry!) without knowing which namespace the
`Res` belongs to. Even weirder! Also, `SmallVec` is an overly flexible
type to use here, because it can contain any number of elements (even
though it's optimized for 3 in this case).
This commit changes `UsePath` so it also contains a
`PerNS<Option<Res<HirId>>>`. This type preserves more information and is
more self-documenting. The commit also changes a lot of the use sites to
access the result for a particular namespace. E.g. if you're looking up
a trait, it will be in the `Res` for the type namespace if it's present;
it's silly to look in the `Res` for the value namespace or macro
namespace. Overall I find the new code much easier to understand.
However, some use sites still iterate. These now use `present_items`
because that filters out the `None` results.
Also, `redundant_pub_crate.rs` gets a bigger change. A
`UseKind:ListStem` item gets no `Res` results, which means the old `all`
call in `is_not_macro_export` would succeed (because `all` succeeds on
an empty iterator) and the `ListStem` would be ignored. This is what we
want, but was more by luck than design. The new code detects `ListStem`
explicitly. The commit generalizes the name of that function
accordingly.
Finally, the commit also removes the `use_path` arena, because
`PerNS<Option<Res>>` impls `Copy` (unlike `SmallVec`) and it can be
allocated in the arena shared by all `Copy` types.
test-float-parse: apply `cfg(not(bootstrap))`
Prior to stage 0 redesign, `test-float-parse` ran against in-tree std but now it runs against beta std. `f16::FromStr` is only present in in-tree std and not yet beta std, so apply `cfg(not(bootstrap))` gating to unbrick `./x check --stage=0`.
Fixesrust-lang/rust#141900.
`./x check --stage=0` in CI is intended for follow-up.
r? `@Mark-Simulacrum` (or bootstrap/libs)
cc `@tgross35`
Optionally don't steal the THIR
The THIR being stolen is a recurrent pain for authors of rustc drivers. This makes it optional, so that the `thir_body` query can still be used after analysis of the crate has completed.
Fix citool tests when executed locally
They couldn't be executed locally before due to some additional environment reads.
I also investigated the annoying rebuilds that we see on CI all the time, and they are caused by `ring`'s build script. It should be fixed in the next ring release (https://github.com/briansmith/ring/issues/2525), so we can just wait for that and then update `ring`.
r? `@marcoieni`
Prior to stage 0 redesign, `test-float-parse` ran against in-tree std
but now it runs against beta std. `f16::FromStr` were only present in
in-tree std and not yet beta std, so apply `cfg(not(bootstrap))` gating
to unbrick `./x check --stage=0`.
Fix TLS model on bootstrap for cygwin
There aren't other targets that both use emutls and enable `has_thread_local`, so cygwin triggers this bug first.
r? mati865
See: https://github.com/rust-lang/rust/pull/141719#issuecomment-2925445263
``@jeremyd2019`` Could you check if this PR fixes the issue? I just found my pre-built stage-0 rustc was too old to build the current rustc :(
Exclude `CARGO_HOME` from `generate-copyright` in-tree determination
On Ferrocene, we noticed that in our releases the out-of-tree notices were not being included. When `x.py run generate-copyright` was ran on local development machines, it worked fine.
After some investigations ``@tshepang`` and I determined that the problem was that the cargo registry (located in `CARGO_HOME`) started with the source directory on CI jobs, and was being excluded by this line:
15825b7161/src/tools/generate-copyright/src/cargo_metadata.rs (L85-L88)
In Ferrocene's `run.sh` we set `CARGO_HOME` to be `build/cargo-home`: 96a45dd9a1/ferrocene/ci/run.sh (L34-L46) which caused this issue.
This PR passes the `CARGO_HOME` variable to the `generate-copyright` tool and expands the consideration of in-tree-ness to be aware of `CARGO_HOME`. It is an upstreaming of https://github.com/ferrocene/ferrocene/pull/1491.
## Testing
Run `CARGO_HOME=build/cargo-home ./x.py run generate-copyright` on `master`, then check `build/host/doc/COPYRIGHT` and look for out of tree dependencies (at the bottom).
Then, try running the same command in this branch.
Tokio `AsyncWriteExt::write` doesn't actually ensure that the contents
have written, it just *starts* the write operation. To ensure that the
file has actually been written, we need to `sync_all` first.
The test did `write` and `read` and hoped that it would read/write
everything, which doesn't always happen and caused CI failures.
Switch to `write_all` and `read_to_end` to make it more reliable.
Add `const` support for float rounding methods
# Add `const` support for float rounding methods
This PR makes the following float rounding methods `const`:
- `f64::{floor, ceil, trunc, round, round_ties_even}`
- and the corresponding methods for `f16`, `f32` and `f128`
Tracking issue: https://github.com/rust-lang/rust/issues/141555
## Procedure
I followed c09ed3e767 as closely as I could in making float methods `const`, and also received great guidance from https://internals.rust-lang.org/t/const-rounding-methods-in-float-types/22957/3?u=ruancomelli.
## Note
This is my first code contribution to the Rust project, so please let me know if I missed anything - I'd be more than happy to revise and learn more. Thank you for taking the time to review it!
Miri CI: test aarch64-apple-darwin in PRs instead of the x86_64 target
The aarch64 target is more important, and also this ensures we cover all main architectures (x86_64, i686, aarch64) in PR CI.
