Rollup of 3 pull requests
Successful merges:
- #124003 (Dellvmize some intrinsics (use `u32` instead of `Self` in some integer intrinsics))
- #124169 (Don't fatal when calling `expect_one_of` when recovering arg in `parse_seq`)
- #124286 (Subtree sync for rustc_codegen_cranelift)
r? `@ghost`
`@rustbot` modify labels: rollup
Subtree sync for rustc_codegen_cranelift
This fixes a crash when compiling the standard library. In addition the Cranelift update fixes all the 128bit int abi incompatibility between cg_clif and cg_llvm.
r? ``@ghost``
``@rustbot`` label +A-codegen +A-cranelift +T-compiler
Don't fatal when calling `expect_one_of` when recovering arg in `parse_seq`
In `parse_seq`, when parsing a sequence of token-separated items, if we don't see a separator, we try to parse another item eagerly in order to give a good diagnostic and recover from a missing separator:
d1a0fa5ed3/compiler/rustc_parse/src/parser/mod.rs (L900-L901)
If parsing the item itself calls `expect_one_of`, then we will fatal because of #58903:
d1a0fa5ed3/compiler/rustc_parse/src/parser/mod.rs (L513-L516)
For `precise_capturing` feature I implemented, we do end up calling `expected_one_of`:
d1a0fa5ed3/compiler/rustc_parse/src/parser/ty.rs (L712-L714)
This leads the compiler to fatal *before* having emitted the first error, leading to absolutely no useful information for the user about what happened in the parser.
This PR makes it so that we stop doing that.
Fixes#124195
Dellvmize some intrinsics (use `u32` instead of `Self` in some integer intrinsics)
This implements https://github.com/rust-lang/compiler-team/issues/693 minus what was implemented in #123226.
Note: I decided to _not_ change `shl`/... builder methods, as it just doesn't seem worth it.
r? ``@scottmcm``
Rollup of 7 pull requests
Successful merges:
- #120929 (Wrap dyn type with parentheses in suggestion)
- #122591 (Suggest using type args directly instead of equality constraint)
- #122598 (deref patterns: lower deref patterns to MIR)
- #123048 (alloc::Layout: explicitly document size invariant on the type level)
- #123993 (Do `check_coroutine_obligations` once per typeck root)
- #124218 (Allow nesting subdiagnostics in #[derive(Subdiagnostic)])
- #124285 (Mark ``@RUSTC_BUILTIN`` search path usage as unstable)
r? `@ghost`
`@rustbot` modify labels: rollup
Do `check_coroutine_obligations` once per typeck root
We only need to do `check_coroutine_obligations` once per typeck root, especially since the new solver can't really (easily) associate which obligations correspond to which coroutines.
This requires us to move the checks for sized coroutine fields into `mir_coroutine_witnesses`, but that's fine imo.
r? lcnr
alloc::Layout: explicitly document size invariant on the type level
https://github.com/rust-lang/rust/pull/95295 added this to the constructor, but it seems worth documenting the type invariant at the type level.
deref patterns: lower deref patterns to MIR
This lowers deref patterns to MIR. This is a bit tricky because this is the first kind of pattern that requires storing a value in a temporary. Thanks to https://github.com/rust-lang/rust/pull/123324 false edges are no longer a problem.
The thing I'm not confident about is the handling of fake borrows. This PR ignores any fake borrows inside a deref pattern. We are guaranteed to at least fake borrow the place of the first pointer value, which could be enough, but I'm not certain.
Suggest using type args directly instead of equality constraint
When type arguments are written erroneously using an equality constraint we suggest specifying them directly without the equality constraint.
Fixes#122162
Changes the diagnostic in the issue from:
```rust
error[E0229]: associated type bindings are not allowed here
9 | impl std::cmp::PartialEq<Rhs = T> for S {
| ^^^^^^^ associated type not allowed here
|
```
to
```rust
error[E0229]: associated type bindings are not allowed here
9 | impl std::cmp::PartialEq<Rhs = T> for S {
| ^^^^^^^ associated type not allowed here
|
help: to use `T` as a generic argument specify it directly
|
| impl std::cmp::PartialEq<T> for S {
| ~
```
Use fulfillment in method probe, not evaluation
This PR reworks method probing to use fulfillment instead of a `for`-loop of `evaluate_predicate` calls, and moves normalization from method candidate assembly into the `consider_probe`, where it's applied to *all* candidates. This last part coincidentally fixes https://github.com/rust-lang/rust/issues/121643#issuecomment-1975371248.
Regarding *why* this large rewrite is done: In general, it's an anti-pattern to do `for o in obligations { evaluate(o); }` because it's not compatible with the way that the new solver emits alias-relate obligations which constrain variables that may show up in other predicates.
r? lcnr
Rollup of 5 pull requests
Successful merges:
- #123050 (panic_str only exists for the migration to 2021 panic macros)
- #124067 (weak lang items are not allowed to be #[track_caller])
- #124099 (Disallow ambiguous attributes on expressions)
- #124284 (parser: remove unused(no reads) max_angle_bracket_count field)
- #124288 (remove `push_trait_bound_inner`)
r? `@ghost`
`@rustbot` modify labels: rollup
Disallow ambiguous attributes on expressions
This implements the suggestion in [#15701](https://github.com/rust-lang/rust/issues/15701#issuecomment-2033124217) to disallow ambiguous outer attributes on expressions. This should resolve one of the concerns blocking the stabilization of `stmt_expr_attributes`.
weak lang items are not allowed to be #[track_caller]
For instance the panic handler will be called via this import
```rust
extern "Rust" {
#[lang = "panic_impl"]
fn panic_impl(pi: &PanicInfo<'_>) -> !;
}
```
A `#[track_caller]` would add an extra argument and thus make this the wrong signature.
The 2nd commit is a consistency rename; based on the docs [here](https://doc.rust-lang.org/unstable-book/language-features/lang-items.html) and [here](https://rustc-dev-guide.rust-lang.org/lang-items.html) I figured "lang item" is more widely used. (In the compiler output, "lang item" and "language item" seem to be pretty even.)
panic_str only exists for the migration to 2021 panic macros
The only caller is `expect_failed`, which is already a cold inline(never) function, so inlining into that function should be fine. (And indeed `panic_str` was `#[inline]` anyway.)
The existence of panic_str risks someone calling it when they should call `panic` instead, and I can't see a reason why this footgun should exist.
I also extended the comment in `panic` to explain why it needs a `'static` string -- I know I've wondered about this in the past and it took me quite a while to understand.
Rollup of 7 pull requests
Successful merges:
- #123680 (Deny gen keyword in `edition_2024_compat` lints)
- #124057 (Fix ICE when ADT tail has type error)
- #124168 (Use `DefiningOpaqueTypes::Yes` in rustdoc, where the `InferCtxt` is guaranteed to have no opaque types it can define)
- #124197 (Move duplicated code in functions in `tests/rustdoc-gui/notable-trait.goml`)
- #124200 (Improve handling of expr->field errors)
- #124220 (Miri: detect wrong vtables in wide pointers)
- #124266 (remove an unused type from the reentrant lock tests)
r? `@ghost`
`@rustbot` modify labels: rollup
remove an unused type from the reentrant lock tests
At least it seems unused. This was added back in 45aa6c8d1b together with a test related to poisoning; when the test got removed, it seems like it was forgotten to also remove this type.
Improve handling of expr->field errors
The current message for "`->` used for field access" is the following:
```rust
error: expected one of `!`, `.`, `::`, `;`, `?`, `{`, `}`, or an operator, found `->`
--> src/main.rs:2:6
|
2 | a->b;
| ^^ expected one of 8 possible tokens
```
([playground link](https://play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=7f8b6f4433aa7866124123575456f54e))
This PR tries to address this by adding a dedicated error message and recovery. The proposed error message is:
```
error: `->` used for field access or method call
--> ./tiny_test.rs:2:6
|
2 | a->b;
| ^^ help: try using `.` instead
|
= help: the `.` operator will dereference the value if needed
```
(feel free to bikeshed it as much as necessary)
Move duplicated code in functions in `tests/rustdoc-gui/notable-trait.goml`
It also allowed me to add some new common test like:
```
assert-position: (
"//*[`@class='tooltip` popover']",
{"x": |popover_x|}
)
```
r? `@notriddle`
Use `DefiningOpaqueTypes::Yes` in rustdoc, where the `InferCtxt` is guaranteed to have no opaque types it can define
r? `@lcnr`
I manually checked there it's always `tcx.infer_ctxt().build()`
Deny gen keyword in `edition_2024_compat` lints
Splits the `keyword_idents` lint into two -- `keyword_idents_2018` and `keyword_idents_2024` -- since each corresponds to a future-compat warning in a different edition. Group these together into a new `keyword_idents` lint group, and add the latter to the `rust_2024_compatibility` so that `gen` is ready for the 2024 edition.
cc `@traviscross` `@ehuss`
Add simple async drop glue generation
This is a prototype of the async drop glue generation for some simple types. Async drop glue is intended to behave very similar to the regular drop glue except for being asynchronous. Currently it does not execute synchronous drops but only calls user implementations of `AsyncDrop::async_drop` associative function and awaits the returned future. It is not complete as it only recurses into arrays, slices, tuples, and structs and does not have same sensible restrictions as the old `Drop` trait implementation like having the same bounds as the type definition, while code assumes their existence (requires a future work).
This current design uses a workaround as it does not create any custom async destructor state machine types for ADTs, but instead uses types defined in the std library called future combinators (deferred_async_drop, chain, ready_unit).
Also I recommend reading my [explainer](https://zetanumbers.github.io/book/async-drop-design.html).
This is a part of the [MCP: Low level components for async drop](https://github.com/rust-lang/compiler-team/issues/727) work.
Feature completeness:
- [x] `AsyncDrop` trait
- [ ] `async_drop_in_place_raw`/async drop glue generation support for
- [x] Trivially destructible types (integers, bools, floats, string slices, pointers, references, etc.)
- [x] Arrays and slices (array pointer is unsized into slice pointer)
- [x] ADTs (enums, structs, unions)
- [x] tuple-like types (tuples, closures)
- [ ] Dynamic types (`dyn Trait`, see explainer's [proposed design](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#async-drop-glue-for-dyn-trait))
- [ ] coroutines (https://github.com/rust-lang/rust/pull/123948)
- [x] Async drop glue includes sync drop glue code
- [x] Cleanup branch generation for `async_drop_in_place_raw`
- [ ] Union rejects non-trivially async destructible fields
- [ ] `AsyncDrop` implementation requires same bounds as type definition
- [ ] Skip trivially destructible fields (optimization)
- [ ] New [`TyKind::AdtAsyncDestructor`](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#adt-async-destructor-types) and get rid of combinators
- [ ] [Synchronously undroppable types](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#exclusively-async-drop)
- [ ] Automatic async drop at the end of the scope in async context
miri libstd tests: test windows-msvc instead of windows-gnu
This avoids https://github.com/rust-lang/rust/issues/123583. MSVC is the more widely used target anyway AFAIK, so probably makes more sense to cover that.
Update books
## rust-lang/book
41 commits in 3131aa4642c627a24f523c82566b94a7d920f68c..d207d894cc5e1d496ab99beeacd1a420e5d4d238
2024-04-19 19:48:19 UTC to 2024-04-08 19:34:07 UTC
- Ch. 2: intentionally use `{}` at first (rust-lang/book#3898)
- Inline fmt lints for `redirects/` (rust-lang/book#3897)
- ch12-03: `expect` is not used multitude times by this chapter (rust-lang/book#3287)
- Inline format args to `tools/src` (rust-lang/book#3896)
- Improve awkward phrasing around the kinds of closures (rust-lang/book#3290)
- Ch. 14: clarify workspace dependency reuse (rust-lang/book#3890)
- Edit on line no. 153. Replaced fn add_to_waitlist with fn eat_at_restaurant. (rust-lang/book#3281)
- Ch. 2: less ambiguity about numbers (rust-lang/book#3895)
- Clarified first slices paragraph (rust-lang/book#3363)
- Inline all format arguments (rust-lang/book#3385)
- Use v1.58 captured ident formatting in examples (rust-lang/book#3048)
- Revert "ch18-03: Guarded match arm exhaustivness clarification" (rust-lang/book#3894)
- Ch. 18: further clarify about irrefutable patterns (rust-lang/book#3893)
- ch18-02: Minor corrections (rust-lang/book#3381)
- Make style of listings 9-7 and 9-8 consistent with 9-6 (rust-lang/book#3217)
- ch18-03: Guarded match arm exhaustivness clarification (rust-lang/book#3397)
- ch11-03: Unnecessary import removal (rust-lang/book#3284)
- Update ch20-01-single-threaded.md (rust-lang/book#3394)
- In Appendix D, adds hyperlink to Appendix E (rust-lang/book#3418)
- Chapter 13-03 fixed Iterator 'type' to 'trait' (rust-lang/book#3433)
- Fix the expected substring in ch11-01 (rust-lang/book#3534)
- Ch. 17: clarify 'in order to compile' (rust-lang/book#3892)
- Revert "Chapter 15-05: Fix incorrect interpretation of compiler error" (rust-lang/book#3891)
- Chapter 15-05: Fix incorrect interpretation of compiler error (rust-lang/book#3501)
- Remove direction to authenticate with `cargo login {API token}` (rust-lang/book#3506)
- Update syn dependencies in Chapter 19 to 2.0 (rust-lang/book#3888)
- Chapter 3: Clarify `return` vs. `break`. (rust-lang/book#3889)
- Fix "`Display` type" to "`Display` trait" in ch19-03 (rust-lang/book#3835)
- Update listings in ch 11-01 to reflect current output from cargo new (rust-lang/book#3533)
- Fix "`test` module" to "`tests` module" in ch11-03 (rust-lang/book#3579)
- Ch. 2: fix wording about `cargo update` (rust-lang/book#3882)
- Remove hebrew niqqud in 8-14 String::from example (rust-lang/book#3648)
- Ch. 16: fix a grammar issue (rust-lang/book#3881)
- Update MSVC instructions (rust-lang/book#3616)
- Replace RLS with rust-analyzer in ch00 (rust-lang/book#3546)
- Fix em dashes (rust-lang/book#3570)
- Revert #3711 (rust-lang/book#3880)
- Procedure Macro: Update dependencies (rust-lang/book#3631)
- Parity with rustup's help (rust-lang/book#3642)
- Fix confusing sentence (rust-lang/book#3877)
- Ch. 7: add a note about the need for `pub` in one more place (rust-lang/book#3879)
## rust-lang/edition-guide
2 commits in eb3eb80e106d03250c1fb7c5666b1c8c59672862..0c68e90acaae5a611f8f5098a3c2980de9845ab2
2024-04-16 21:23:55 UTC to 2024-04-15 16:05:32 UTC
- Document RustcEncodable/RustcDecodable (rust-lang/edition-guide#298)
- Flesh out detail behind disallowing refs to static mut (rust-lang/edition-guide#299)
## rust-lang/reference
5 commits in 55694913b1301cc809f9bf4a1ad1b3d6920efbd9..5854fcc286557ad3ab34d325073d11d8118096b6
2024-04-21 13:47:07 UTC to 2024-04-15 16:12:40 UTC
- Stabilize Wasm target features that are in phase 4 and 5 (rust-lang/reference#1420)
- Expand and clarify primitive alignment (rust-lang/reference#1493)
- Update clone reference to include closures (rust-lang/reference#1492)
- Document how `non_exhaustive` interacts with tuple and unit-like structs. (rust-lang/reference#1491)
- Fix link to RISC-V Zkt spec; it was pointing to Zkr (rust-lang/reference#1490)
## rust-lang/rustc-dev-guide
4 commits in b77a34bd46399687b4ce6a17198e9f316c988794..07425fed36b00e60341c5e29e28d37d40cbd4451
2024-04-13 15:16:44 UTC to 2024-04-08 19:00:54 UTC
- Add a note about checks with multiple revisions (rust-lang/rustc-dev-guide#1963)
- compiletest: clarify that headers must be one header per line (rust-lang/rustc-dev-guide#1962)
- chore: fix some typos in conments (rust-lang/rustc-dev-guide#1961)
- compiletest: specify which special env var and which particular CI job (rust-lang/rustc-dev-guide#1960)
