Dylan DPC [Tue, 23 Aug 2022 15:10:04 +0000 (20:40 +0530)]
Rollup merge of #100768 - Facel3ss1:plugin-impl-translation, r=davidtwco
Migrate `rustc_plugin_impl` to `SessionDiagnostic`
Migration of the `rustc_plugin_impl` crate.
~Draft PR because it is blocked on #100694 for `#[fatal(...)]` support~ (this has been merged, and I've changed over to `#[diag(...)]` now too), but I would also like to know if what I did with `LoadPluginError` is okay, because all it does is display the error message from `libloading` ([See conversation on zulip](https://rust-lang.zulipchat.com/#narrow/stream/147480-t-compiler.2Fwg-diagnostics/topic/.23100717.20diagnostic.20translation/near/294327843)). This crate is apparently for a deprecated feature which is used by servo, so I don't know how much this matters anyway.
Dylan DPC [Tue, 23 Aug 2022 15:10:01 +0000 (20:40 +0530)]
Rollup merge of #99249 - cjgillot:no-reparse-fn, r=fee1-dead
Do not re-parse function signatures to suggest generics
This PR uses the existing resolution rib infrastructure to channel the correct span information to suggest generic parameters. This allows to avoid re-parsing a function's source code.
Drive-by cleanup: this removes useless `FnItemRibKind` from late resolution ribs. All the use cases are already covered by `ItemRibKind` and `AssocItemRibKind` which have more precise semantics.
Afonso Bordado [Tue, 23 Aug 2022 10:42:35 +0000 (11:42 +0100)]
Use native scalar `fma` instruction (#1267)
Cranelift 0.87 now supports lowering `fma` as a libcall on x86 [0].
With 0.88 enabling the native x86 instruction under the `has_fma` flag.
aarch64 and s390x already support this as a native instruction, so it's
nice that we emit it for those.
We can't lower the SIMD version using the `fma` instruction since the
lowering can fail if the x86 `has_fma` flag is not enabled. Cranelift
doesn't yet know how to fallback for these cases
bors [Tue, 23 Aug 2022 08:09:19 +0000 (08:09 +0000)]
Auto merge of #100782 - thomcc:fix-android-sigaddset, r=Mark-Simulacrum
Align android `sigaddset` impl with the reference impl from Bionic
In https://github.com/rust-lang/rust/pull/100737 I noticed we were treating the sigset_t as an array of bytes, while referencing code from android (https://github.com/aosp-mirror/platform_bionic/blob/ad8dcd6023294b646e5a8288c0ed431b0845da49/libc/include/android/legacy_signal_inlines.h) which treats it as an array of unsigned long.
That said, the behavior difference is so subtle here that it's not hard to see why nobody noticed. This fixes the implementation to be equivalent to the one in bionic.
bors [Tue, 23 Aug 2022 05:22:05 +0000 (05:22 +0000)]
Auto merge of #100904 - matthiaskrgr:rollup-z3yzivj, r=matthiaskrgr
Rollup of 9 pull requests
Successful merges:
- #100382 (Make the GATS self outlives error take into GATs in the inputs)
- #100565 (Suggest adding a missing semicolon before an item)
- #100641 (Add the armv4t-none-eabi target to the supported_targets)
- #100789 (Use separate infcx to solve obligations during negative coherence)
- #100832 (Some small bootstrap cleanup)
- #100861 (fix ICE with extra-const-ub-checks)
- #100862 (tidy: remove crossbeam-utils)
- #100887 (Refactor part of codegen_call_terminator)
- #100893 (Remove out-of-context comment in `mem::MaybeUninit` documentation)
Matthias Krüger [Tue, 23 Aug 2022 04:55:30 +0000 (06:55 +0200)]
Rollup merge of #100887 - eholk:codegen_call_terminator-cleanup, r=fee1-dead
Refactor part of codegen_call_terminator
I was reading through this code and found the chain of `if let` and a nested match on the same value that was matched in the `if let` to be kind of hard to follow. This PR cleans it up by flattening the `if let` chain and nested match into a single `match` expression.
Matthias Krüger [Tue, 23 Aug 2022 04:55:27 +0000 (06:55 +0200)]
Rollup merge of #100832 - ehuss:bootstrap-cleanup, r=jyn514
Some small bootstrap cleanup
This is a collection of a few small cleanups. See commits for more details.
* Remove some unused fields from the tool_extended macro.
* Remove rustfmt from publish_toolstate.
* Remove Steve from toolstate failure notices.
* Don't allow rustfmt to fail on dist.
Matthias Krüger [Tue, 23 Aug 2022 04:55:23 +0000 (06:55 +0200)]
Rollup merge of #100382 - jackh726:gat-self-outlives-input, r=compiler-errors
Make the GATS self outlives error take into GATs in the inputs
Before, the reasoning was that outlives should factor in to the outlives error, because that value is produced and inputs aren't. However, this is potentially confusing, and we can just require this for now and relax it later if we need. GATs in where clauses still don't count for the self outlives error, and I've added a test for that.
This now errors:
```rust
trait Input {
type Item<'a>;
//~^ missing required
fn takes_item<'a>(&'a self, item: Self::Item<'a>);
}
```
I've also added a test that this does not:
```rust
trait WhereClause {
type Item<'a>;
fn takes_item<'a>(&'a self) where Self::Item<'a>: ;
}
```
bors [Mon, 22 Aug 2022 20:28:16 +0000 (20:28 +0000)]
Auto merge of #99762 - Nilstrieb:unreachable-prop, r=oli-obk
UnreachableProp: Preserve unreachable branches for multiple targets
Before, UnreachablePropagation removed all unreachable branches. This was a pessimization, as it removed information about reachability that was used later in the optimization pipeline.
For example, this code
```rust
pub enum Two { A, B }
pub fn identity(x: Two) -> Two {
match x {
Two::A => Two::A,
Two::B => Two::B,
}
}
```
basically has `switchInt() -> [0: 0, 1: 1, otherwise: unreachable]` for the match. This allows it to be transformed into a simple `x`. If we remove the unreachable branch, the transformation becomes illegal.
This was the problem keeping `UnreachablePropagation` from being enabled, so we can enable it now.
Something similar already happened in #77800, but it did not show a perf improvement there. Let's try it again anyways!
Fixes #68105, although that issue has been fixed for a long time (see #77680).
bors [Mon, 22 Aug 2022 17:46:57 +0000 (17:46 +0000)]
Auto merge of #100881 - Dylan-DPC:rollup-q9rr658, r=Dylan-DPC
Rollup of 8 pull requests
Successful merges:
- #98200 (Expand potential inner `Or` pattern for THIR)
- #99770 (Make some const prop mir-opt tests `unit-test`s)
- #99957 (Rework Ipv6Addr::is_global to check for global reachability rather than global scope - rebase)
- #100331 (Guarantee `try_reserve` preserves the contents on error)
- #100336 (Fix two const_trait_impl issues)
- #100713 (Convert diagnostics in parser/expr to SessionDiagnostic)
- #100820 (Use pointer `is_aligned*` methods)
- #100872 (Add guarantee that Vec::default() does not alloc)
Dylan DPC [Mon, 22 Aug 2022 15:04:16 +0000 (20:34 +0530)]
Rollup merge of #100872 - JanBeh:PR_vec_default_alloc_doc, r=fee1-dead
Add guarantee that Vec::default() does not alloc
Currently `Vec::new()` is guaranteed to not allocate until elements are pushed onto the `Vec`, but such a guarantee is missing for `Vec`'s implementation of `Default::default`.
This adds such a guarantee for `Vec::default()` to the API reference.
See also [this discussion on URLO](https://users.rust-lang.org/t/guarantee-that-vec-default-does-not-allocate/79903).
Dylan DPC [Mon, 22 Aug 2022 15:04:14 +0000 (20:34 +0530)]
Rollup merge of #100713 - Xiretza:parser-expr-session-diagnostics, r=estebank
Convert diagnostics in parser/expr to SessionDiagnostic
This migrates all the easy cases in `rustc_parse::parser::expr` to `SessionDiagnostic`s, I've left things such as `multipart_suggestion`s out for now in the hopes of a derive API being developed soon.
Dylan DPC [Mon, 22 Aug 2022 15:04:09 +0000 (20:34 +0530)]
Rollup merge of #99770 - Nilstrieb:mir-pass-unit-test, r=oli-obk
Make some const prop mir-opt tests `unit-test`s
Most of these have no or only tiny diffs beyond line numbers being changed (would it make sense to not have line numbers in mir-opt tests?). Some things changed a bit, but I think it should all be fine, not sure though.
Dylan DPC [Mon, 22 Aug 2022 15:04:08 +0000 (20:34 +0530)]
Rollup merge of #98200 - ouz-a:issue-98177, r=oli-obk
Expand potential inner `Or` pattern for THIR
Code assumed there wouldn't be a deeper `Or` pattern inside expanded `PatStack` this fixes it by looking for the `Or` pattern inside expanded `PatStack`.
A more ideal solution would be recursively doing this but I haven't found a good way to do that.
_fixes #97898_
bors [Mon, 22 Aug 2022 14:59:21 +0000 (14:59 +0000)]
Auto merge of #99908 - Nilstrieb:mir-opt-span, r=oli-obk
Show absolute line numbers if span is outside relative span
In the MIR pretty printing, it can sometimes happen that the span of the statement is outside the span of the body (for example through inlining). In this case, don't display a relative span but an absolute span. This will make the mir-opt-tests a little more prone to diffs again, but the impact should be small.
Show absolute line numbers if span is outside relative span
In the MIR pretty printing, it can sometimes happen that the span of the
statement is outside the span of the body (for example through
inlining). In this case, don't display a relative span but an absolute
span. This will make the mir-opt-tests a little more prone to diffs
again, but the impact should be small.
bors [Mon, 22 Aug 2022 11:39:53 +0000 (11:39 +0000)]
Auto merge of #99702 - SparrowLii:transtive_relation, r=oli-obk
get rid of `RefCell` in `TransitiveRelation`
This is one of the jobs in `Pending refactorings` in #48685. The parallel-compiler's work has been suspended for quite some time, but I think I can pick it up gradually. I think this PR should be a start.
Regarding the refactoring of `TransitiveRelation`, `@nikomatsakis` has proposed [two(three?) schemes](https://github.com/rust-lang/rust/pull/48587#issuecomment-369336651). In order to satisfy both compilation efficiency and robustness, I think adding the `freeze` method may be the best solution, although it requires relatively more code changes.
Jan Behrens [Mon, 22 Aug 2022 10:36:44 +0000 (12:36 +0200)]
Add guarantee that Vec::default() does not alloc
Currently `Vec::new()` is guaranteed to not allocate until elements are
pushed onto the `Vec`, but such a guarantee is missing for `Vec`'s
implementation of `Default::default`. This adds such a guarantee for
`Vec::default()` to the API reference.