Commit graph

6702 commits

Author SHA1 Message Date
Michael Goulet
803772e81d Enable new capture rules by default on edition 2024 2023-12-05 19:53:59 +00:00
Michael Goulet
acba7efe1b Add test for implicitly capturing late-bound var with new capture rules 2023-12-05 19:53:59 +00:00
Michael Goulet
0ad160a585 Add lifetime_capture_rules_2024 2023-12-05 19:53:59 +00:00
Michael Goulet
dbcde57171
Rollup merge of #118585 - sjwang05:issue-118564, r=compiler-errors
Fix parser ICE when recovering `dyn`/`impl` after `for<...>`

Fixes #118564
2023-12-05 14:52:43 -05:00
Michael Goulet
598ca0ea3f
Rollup merge of #118346 - compiler-errors:deeply-normalize-for-diagnostic, r=lcnr
Add `deeply_normalize_for_diagnostics`, use it in coherence

r? lcnr

Normalize trait refs used for coherence error reporting with `-Ztrait-solver=next-coherence`.

Two things:
1. I said before that we can't add this to `TyErrCtxt` because we compute `OverlapResult`s even if there are no diagnostics being emitted, e.g. for a reservation impl.
2. I didn't want to add this to an `InferCtxtExt` trait because I felt it was unnecessary. I don't particularly care about the API though.
2023-12-05 14:52:42 -05:00
Michael Goulet
ad23f30b1d
Rollup merge of #118268 - compiler-errors:pretty-print, r=estebank
Pretty print `Fn<(..., ...)>` trait refs with parentheses (almost) always

It's almost always better, at least in diagnostics, to print `Fn(i32, u32)` instead of `Fn<(i32, u32)>`.

Related to but doesn't fix #118225. That needs a separate fix.
2023-12-05 14:52:41 -05:00
Michael Goulet
19bf749560
Rollup merge of #118123 - RalfJung:internal-lib-features, r=compiler-errors
Add support for making lib features internal

We have the notion of an "internal" lang feature: a feature that is never intended to be stabilized, and using which can cause ICEs and other issues without that being considered a bug.

This extends that idea to lib features as well. It is an alternative to https://github.com/rust-lang/rust/pull/115623: instead of using an attribute to declare lib features internal, we simply do this based on the name. Everything ending in `_internals` or `_internal` is considered internal.

Then we rename `core_intrinsics` to `core_intrinsics_internal`, which fixes https://github.com/rust-lang/rust/issues/115597.
2023-12-05 14:52:41 -05:00
Trevor Gross
b225aab5ef Stabilize type_name_of_val
Make the following API stable:

    // in core::any
    pub fn type_name_of_val<T: ?Sized>(_val: &T) -> &'static str

Const stability is not added because this relies on `type_name` which is also
not const. That has a blocking issue.

Fixes #66359
2023-12-05 14:49:09 -05:00
bors
56278a6e28 Auto merge of #118457 - eholk:genfn, r=compiler-errors
Add support for `gen fn`

This builds on #116447 to add support for `gen fn` functions. For the most part we follow the same approach as desugaring `async fn`, but replacing `Future` with `Iterator` and `async {}` with `gen {}` for the body.

The version implemented here uses the return type of a `gen fn` as the yield type. For example:

```rust
gen fn count_to_three() -> i32 {
    yield 1;
    yield 2;
    yield 3;
}
```

In the future, I think we should experiment with a syntax like `gen fn count_to_three() yield i32 { ... }`, but that can go in another PR.

cc `@oli-obk` `@compiler-errors`
2023-12-05 18:37:15 +00:00
Michael Goulet
b97ff8eb16 Add print_trait_sugared 2023-12-05 17:15:46 +00:00
Michael Goulet
3448284f8d Continue folding if deep normalizer fails 2023-12-05 16:55:10 +00:00
Michael Goulet
334577f091 Add deeply_normalize_for_diagnostics, use it in coherence 2023-12-05 16:33:37 +00:00
Matthias Krüger
c6c0c658df
Rollup merge of #118608 - fee1-dead-contrib:backdoor-in-askconv, r=compiler-errors
Use default params until effects in desugaring

See the comment and [this section](https://fee1-dead.github.io/devlog001/#effects-desugaring-with-default-type-params) from my blog post

r? ``@compiler-errors``
2023-12-05 16:08:37 +01:00
Matthias Krüger
81b6263dd0
Rollup merge of #118598 - Nadrieril:remove_precise_pointer_size_matching, r=davidtwco
Remove the `precise_pointer_size_matching` feature gate

`usize` and `isize` are special for pattern matching because their range might depend on the platform. To make code portable across platforms, the following is never considered exhaustive:
```rust
let x: usize = ...;
match x {
    0..=18446744073709551615 => {}
}
```
Because of how rust handles constants, this also unfortunately counts `0..=usize::MAX` as non-exhaustive. The [`precise_pointer_size_matching`](https://github.com/rust-lang/rust/issues/56354) feature gate was introduced both for this convenience and for the possibility that the lang team could decide to allow the above.

Since then, [half-open range patterns](https://github.com/rust-lang/rust/issues/67264) have been implemented, and since #116692 they correctly support `usize`/`isize`:
```rust
match 0usize { // exhaustive!
    0..5 => {}
    5.. => {}
}
```
I believe this subsumes all the use cases of the feature gate. Moreover no attempt has been made to stabilize it in the 5 years of its existence. I therefore propose we retire this feature gate.

Closes https://github.com/rust-lang/rust/issues/56354
2023-12-05 16:08:35 +01:00
Matthias Krüger
fddda14ac0
Rollup merge of #118594 - hdost:patch-1, r=fmease
Remove mention of rust to make the error message generic.

The deprecation notice is used when in crates as well. This applies to versions Rust or Crates.

Relates #118148
2023-12-05 16:08:35 +01:00
Matthias Krüger
2d01eeeeac
Rollup merge of #117922 - estebank:unclosed-generics, r=b-naber
Tweak unclosed generics errors

Remove unnecessary span label for parse errors that already have a suggestion.

Provide structured suggestion to close generics in more cases.
2023-12-05 16:08:34 +01:00
Urgau
3f0369e0f2 Remove deprecated --check-cfg names() and values() syntax 2023-12-05 13:25:11 +01:00
Harold Dost
1b503042b8 Remove mention of rust to make the error message generic.
The deprecation notice is used when in crates as well. This applies to versions Rust or Crates.

Fixes #118148

Signed-off-by: Harold Dost <h.dost@criteo.com>
2023-12-05 09:18:41 +01:00
bors
f67523d0b3 Auto merge of #118076 - estebank:issue-109429, r=davidtwco
Tweak `.clone()` suggestion to work in more cases

When going through auto-deref, the `<T as Clone>` impl sometimes needs to be specified for rustc to actually clone the value and not the reference.

```
error[E0507]: cannot move out of dereference of `S`
  --> $DIR/needs-clone-through-deref.rs:15:18
   |
LL |         for _ in self.clone().into_iter() {}
   |                  ^^^^^^^^^^^^ ----------- value moved due to this method call
   |                  |
   |                  move occurs because value has type `Vec<usize>`, which does not implement the `Copy` trait
   |
note: `into_iter` takes ownership of the receiver `self`, which moves value
  --> $SRC_DIR/core/src/iter/traits/collect.rs:LL:COL
help: you can `clone` the value and consume it, but this might not be your desired behavior
   |
LL |         for _ in <Vec<usize> as Clone>::clone(&self.clone()).into_iter() {}
   |                  ++++++++++++++++++++++++++++++            +
```

When encountering a move error, look for implementations of `Clone` for the moved type. If there is one, check if all its obligations are met. If they are, we suggest cloning without caveats. If they aren't, we suggest cloning while mentioning the unmet obligations, potentially suggesting `#[derive(Clone)]` when appropriate.

```
error[E0507]: cannot move out of a shared reference
  --> $DIR/suggest-clone-when-some-obligation-is-unmet.rs:20:28
   |
LL |     let mut copy: Vec<U> = map.clone().into_values().collect();
   |                            ^^^^^^^^^^^ ------------- value moved due to this method call
   |                            |
   |                            move occurs because value has type `HashMap<T, U, Hash128_1>`, which does not implement the `Copy` trait
   |
note: `HashMap::<K, V, S>::into_values` takes ownership of the receiver `self`, which moves value
  --> $SRC_DIR/std/src/collections/hash/map.rs:LL:COL
help: you could `clone` the value and consume it, if the `Hash128_1: Clone` trait bound could be satisfied
   |
LL |     let mut copy: Vec<U> = <HashMap<T, U, Hash128_1> as Clone>::clone(&map.clone()).into_values().collect();
   |                            ++++++++++++++++++++++++++++++++++++++++++++           +
help: consider annotating `Hash128_1` with `#[derive(Clone)]`
   |
LL + #[derive(Clone)]
LL | pub struct Hash128_1;
   |
```

Fix #109429.

When encountering multiple mutable borrows, suggest cloning and adding
derive annotations as needed.

```
error[E0596]: cannot borrow `sm.x` as mutable, as it is behind a `&` reference
  --> $DIR/accidentally-cloning-ref-borrow-error.rs:32:9
   |
LL |     foo(&mut sm.x);
   |         ^^^^^^^^^ `sm` is a `&` reference, so the data it refers to cannot be borrowed as mutable
   |
help: `Str` doesn't implement `Clone`, so this call clones the reference `&Str`
  --> $DIR/accidentally-cloning-ref-borrow-error.rs:31:21
   |
LL |     let mut sm = sr.clone();
   |                     ^^^^^^^
help: consider annotating `Str` with `#[derive(Clone)]`
   |
LL + #[derive(Clone)]
LL | struct Str {
   |
help: consider specifying this binding's type
   |
LL |     let mut sm: &mut Str = sr.clone();
   |               ++++++++++
```

Fix #34629. Fix #76643. Fix #91532.
2023-12-05 06:34:44 +00:00
bors
9358642e3b Auto merge of #118066 - estebank:structured-use-suggestion, r=b-naber
Structured `use` suggestion on privacy error

When encoutering a privacy error on an item through a re-export that is accessible in an alternative path, provide a structured suggestion with that path.

```
error[E0603]: module import `mem` is private
  --> $DIR/private-std-reexport-suggest-public.rs:4:14
   |
LL |     use foo::mem;
   |              ^^^ private module import
   |
note: the module import `mem` is defined here...
  --> $DIR/private-std-reexport-suggest-public.rs:8:9
   |
LL |     use std::mem;
   |         ^^^^^^^^
note: ...and refers to the module `mem` which is defined here
  --> $SRC_DIR/std/src/lib.rs:LL:COL
   |
   = note: you could import this
help: import `mem` through the re-export
   |
LL |     use std::mem;
   |         ~~~~~~~~
```

Fix #42909.
2023-12-05 04:34:57 +00:00
Deadbeef
65212a07e7 Remove #[rustc_host], use internal desugaring 2023-12-05 01:15:21 +00:00
bors
25dca40751 Auto merge of #117088 - lcnr:generalize-alias, r=compiler-errors
generalize: handle occurs check failure in aliases

mostly fixes #105787, except for the `for<'a> fn(<<?x as OtherTrait>::Assoc as Trait<'a>>::Assoc) eq ?x` case in https://github.com/rust-lang/trait-system-refactor-initiative/issues/8.

r? `@compiler-errors`
2023-12-05 00:37:58 +00:00
Eric Holk
50ef8006eb
Address code review feedback 2023-12-04 14:33:46 -08:00
Esteban Küber
beaf31581a Structured use suggestion on privacy error
When encoutering a privacy error on an item through a re-export that is
accessible in an alternative path, provide a structured suggestion with
that path.

```
error[E0603]: module import `mem` is private
  --> $DIR/private-std-reexport-suggest-public.rs:4:14
   |
LL |     use foo::mem;
   |              ^^^ private module import
   |
note: the module import `mem` is defined here...
  --> $DIR/private-std-reexport-suggest-public.rs:8:9
   |
LL |     use std::mem;
   |         ^^^^^^^^
note: ...and refers to the module `mem` which is defined here
  --> $SRC_DIR/std/src/lib.rs:LL:COL
   |
   = note: you could import this
help: import `mem` through the re-export
   |
LL |     use std::mem;
   |         ~~~~~~~~
```

Fix #42909.
2023-12-04 22:26:08 +00:00
Esteban Küber
a8faa8288c Fix rebase 2023-12-04 22:00:34 +00:00
Esteban Küber
cc80106cb5 Provide more suggestions for cloning immutable bindings
When encountering multiple mutable borrows, suggest cloning and adding
derive annotations as needed.

```
error[E0596]: cannot borrow `sm.x` as mutable, as it is behind a `&` reference
  --> $DIR/accidentally-cloning-ref-borrow-error.rs:32:9
   |
LL |     foo(&mut sm.x);
   |         ^^^^^^^^^ `sm` is a `&` reference, so the data it refers to cannot be borrowed as mutable
   |
help: `Str` doesn't implement `Clone`, so this call clones the reference `&Str`
  --> $DIR/accidentally-cloning-ref-borrow-error.rs:31:21
   |
LL |     let mut sm = sr.clone();
   |                     ^^^^^^^
help: consider annotating `Str` with `#[derive(Clone)]`
   |
LL + #[derive(Clone)]
LL | struct Str {
   |
help: consider specifying this binding's type
   |
LL |     let mut sm: &mut Str = sr.clone();
   |               ++++++++++
```

```
error[E0596]: cannot borrow `*inner` as mutable, as it is behind a `&` reference
  --> $DIR/issue-91206.rs:14:5
   |
LL |     inner.clear();
   |     ^^^^^ `inner` is a `&` reference, so the data it refers to cannot be borrowed as mutable
   |
help: you can `clone` the `Vec<usize>` value and consume it, but this might not be your desired behavior
  --> $DIR/issue-91206.rs:11:17
   |
LL |     let inner = client.get_inner_ref();
   |                 ^^^^^^^^^^^^^^^^^^^^^^
help: consider specifying this binding's type
   |
LL |     let inner: &mut Vec<usize> = client.get_inner_ref();
   |              +++++++++++++++++
```
2023-12-04 21:54:34 +00:00
Esteban Küber
a754fcaa43 On "this .clone() is on the reference", provide more info
When encountering a case where `let x: T = (val: &T).clone();` and
`T: !Clone`, already mention that the reference is being cloned. We now
also suggest `#[derive(Clone)]` not only on `T` but also on type
parameters to satisfy blanket implementations.

```
error[E0308]: mismatched types
  --> $DIR/assignment-of-clone-call-on-ref-due-to-missing-bound.rs:17:39
   |
LL |             let mut x: HashSet<Day> = v.clone();
   |                        ------------   ^^^^^^^^^ expected `HashSet<Day>`, found `&HashSet<Day>`
   |                        |
   |                        expected due to this
   |
   = note: expected struct `HashSet<Day>`
           found reference `&HashSet<Day>`
note: `HashSet<Day>` does not implement `Clone`, so `&HashSet<Day>` was cloned instead
  --> $DIR/assignment-of-clone-call-on-ref-due-to-missing-bound.rs:17:39
   |
LL |             let mut x: HashSet<Day> = v.clone();
   |                                       ^
   = help: `Clone` is not implemented because the trait bound `Day: Clone` is not satisfied
help: consider annotating `Day` with `#[derive(Clone)]`
   |
LL + #[derive(Clone)]
LL | enum Day {
   |
```

Case taken from # #41825.
2023-12-04 21:54:33 +00:00
Esteban Küber
1c69c6ab08 Mark more tests as run-rustfix 2023-12-04 21:54:32 +00:00
Esteban Küber
90db536741 Tweak output on specific case 2023-12-04 21:54:32 +00:00
Esteban Küber
98cfed7b97 Suggest cloning and point out obligation errors on move error
When encountering a move error, look for implementations of `Clone` for
the moved type. If there is one, check if all its obligations are met.
If they are, we suggest cloning without caveats. If they aren't, we
suggest cloning while mentioning the unmet obligations, potentially
suggesting `#[derive(Clone)]` when appropriate.

```
error[E0507]: cannot move out of a shared reference
  --> $DIR/suggest-clone-when-some-obligation-is-unmet.rs:20:28
   |
LL |     let mut copy: Vec<U> = map.clone().into_values().collect();
   |                            ^^^^^^^^^^^ ------------- value moved due to this method call
   |                            |
   |                            move occurs because value has type `HashMap<T, U, Hash128_1>`, which does not implement the `Copy` trait
   |
note: `HashMap::<K, V, S>::into_values` takes ownership of the receiver `self`, which moves value
  --> $SRC_DIR/std/src/collections/hash/map.rs:LL:COL
help: you could `clone` the value and consume it, if the `Hash128_1: Clone` trait bound could be satisfied
   |
LL |     let mut copy: Vec<U> = <HashMap<T, U, Hash128_1> as Clone>::clone(&map.clone()).into_values().collect();
   |                            ++++++++++++++++++++++++++++++++++++++++++++           +
help: consider annotating `Hash128_1` with `#[derive(Clone)]`
   |
LL + #[derive(Clone)]
LL | pub struct Hash128_1;
   |
```

Fix #109429.
2023-12-04 21:54:32 +00:00
Esteban Küber
03c88aaf21 Tweak .clone() suggestion to work in more cases
When going through auto-deref, the `<T as Clone>` impl sometimes needs
to be specified for rustc to actually clone the value and not the
reference.

```
error[E0507]: cannot move out of dereference of `S`
  --> $DIR/needs-clone-through-deref.rs:15:18
   |
LL |         for _ in self.clone().into_iter() {}
   |                  ^^^^^^^^^^^^ ----------- value moved due to this method call
   |                  |
   |                  move occurs because value has type `Vec<usize>`, which does not implement the `Copy` trait
   |
note: `into_iter` takes ownership of the receiver `self`, which moves value
  --> $SRC_DIR/core/src/iter/traits/collect.rs:LL:COL
help: you can `clone` the value and consume it, but this might not be your desired behavior
   |
LL |         for _ in <Vec<usize> as Clone>::clone(&self.clone()).into_iter() {}
   |                  ++++++++++++++++++++++++++++++            +
```

CC #109429.
2023-12-04 21:54:32 +00:00
Eric Holk
26f9954971
Fix some broken tests 2023-12-04 13:03:41 -08:00
Eric Holk
48d5f1f0f2
Merge Async and Gen into CoroutineKind 2023-12-04 12:48:01 -08:00
Guillaume Gomez
baa3f96b42
Rollup merge of #118600 - GuillaumeGomez:fields-heading, r=notriddle
[rustdoc] Don't generate the "Fields" heading if there is no field displayed

Fixes https://github.com/rust-lang/rust/issues/118195.

If no field is displayed, we should not generate the `Fields` heading in enum struct variants.

r? ``@notriddle``
2023-12-04 20:46:10 +01:00
Eric Holk
3887b1645a
Add a basic test for gen fn 2023-12-04 11:23:08 -08:00
Eric Holk
f29b36d03e
Make async gen fn an error 2023-12-04 11:23:07 -08:00
Eric Holk
7c43784cb0
gate gen fn behind gen_blocks 2023-12-04 11:23:06 -08:00
sjwang05
d627e2a4e8
Fix parser ICE when recovering dyn/impl after for<...> 2023-12-04 10:40:09 -08:00
Matthew Jasper
43adf41ca6 Don't include destruction scopes in THIR
They are not used by anyone, and add memory/performance overhead.
2023-12-04 16:31:13 +00:00
Deadbeef
e6a14c0336 Use default params until effects in desugaring 2023-12-04 15:08:14 +00:00
DianQK
9ed0d11efb
Avoid adding compiler-used functions to symbols.o 2023-12-04 22:28:00 +08:00
Takayuki Maeda
87625dbf2b
Rollup merge of #118540 - RalfJung:unsized-packed-offset, r=TaKO8Ki
codegen, miri: fix computing the offset of an unsized field in a packed struct

`#[repr(packed)]`  strikes again.

Fixes https://github.com/rust-lang/rust/issues/118537
Fixes https://github.com/rust-lang/miri/issues/3200

`@bjorn3` I assume cranelift needs the same fix.
2023-12-04 21:19:44 +09:00
Takayuki Maeda
da30882eb3
Rollup merge of #118495 - weiznich:more_tests_for_on_unimplemented, r=compiler-errors
Restrict what symbols can be used in `#[diagnostic::on_unimplemented]` format strings

This commit restricts what symbols can be used in a format string for any option of the `diagnostic::on_unimplemented` attribute. We previously allowed all the ad-hoc options supported by the internal `#[rustc_on_unimplemented]` attribute. For the stable attribute we only want to support generic parameter names and `{Self}` as parameters.  For any other parameter an warning is emitted and the parameter is replaced by the literal parameter string, so for example `{integer}` turns into `{integer}`. This follows the general design of attributes in the `#[diagnostic]` attribute namespace, that any syntax "error" is treated as warning and subsequently ignored.

r? `@compiler-errors`
2023-12-04 21:19:43 +09:00
Guillaume Gomez
8e53edb2ec Add regression test for #118195 2023-12-04 12:13:24 +01:00
Nadrieril
5e470db05c Remove the precise_pointer_size_matching feature gate 2023-12-04 11:56:21 +01:00
lcnr
cf8a2bdd81 rebase 2023-12-04 10:48:00 +01:00
lcnr
407c117e88 cleanup and comments 2023-12-04 10:40:36 +01:00
lcnr
f69d67221e generalize: handle occurs check failure in aliases 2023-12-04 10:39:00 +01:00
Georg Semmler
1a1cd6e6db
Restrict what symbols can be used in #[diagnostic::on_unimplemented] format strings
This commit restricts what symbols can be used in a format string for
any option of the `diagnostic::on_unimplemented` attribute. We
previously allowed all the ad-hoc options supported by the internal
`#[rustc_on_unimplemented]` attribute. For the stable attribute we only
want to support generic parameter names and `{Self}` as parameters.  For
any other parameter an warning is emitted and the parameter is replaced
by the literal parameter string, so for example `{integer}` turns into
`{integer}`. This follows the general design of attributes in the
`#[diagnostic]` attribute namespace, that any syntax "error" is treated
as warning and subsequently ignored.
2023-12-04 10:00:33 +01:00
bohan
199098b71b dedup for duplicate suggestions 2023-12-04 10:43:10 +08:00