Commit Graph

212496 Commits

Author SHA1 Message Date
Matthias Krüger 15b9e20640 Rollup merge of #105500 - oli-obk:unhide_unknown_spans, r=estebank
Make some diagnostics not depend on the source of what they reference being available

r? `@estebank`

follow up to https://github.com/rust-lang/rust/pull/104449
2022-12-13 19:57:11 +01:00
Matthias Krüger dcdbbd0471 Rollup merge of #105476 - estebank:moves-n-borrows, r=compiler-errors
Change pattern borrowing suggestions to be verbose and remove invalid suggestion

Synthesize a more accurate span and use verbose suggestion output to
make the message clearer.

Do not suggest borrowing binding in pattern in let else. Fix #104838.
2022-12-13 19:57:11 +01:00
Matthias Krüger 5e38e702aa Rollup merge of #105464 - nbdd0121:hir, r=compiler-errors
Support #[track_caller] on async closures

Follow up on #105180

r? ```@compiler-errors```
cc ```@cjgillot```
2022-12-13 19:57:10 +01:00
Matthias Krüger 51375dd7a8 Rollup merge of #105438 - compiler-errors:move-methods, r=estebank
Move some codegen-y methods from `rustc_hir_analysis::collect` -> `rustc_codegen_ssa`

Unclear if they should live here, but they seem codegen-y enough, and `rustc_hir_analysis::collect` is extremely long, so it should probably lose some methods.
2022-12-13 19:57:10 +01:00
Matthias Krüger 7357cfbf3d Rollup merge of #105147 - nbdd0121:inline_const_unsafe, r=oli-obk
Allow unsafe through inline const

Handle similar to closures.

Address https://github.com/rust-lang/rust/pull/104087#issuecomment-1324173328

Note that this PR does not fix the issue for `unsafe { [0; function_requiring_unsafe()] }`. This is fundamentally unfixable for MIR unsafeck IMO.

This PR also does not fix unsafety checking for inline const in pattern position. It actually breaks it, allowing unsafe functions to be used in inline const in pattern position without unsafe blocks. Inline const in pattern position is not visible in MIR so ignored by MIR unsafety checking (currently it is also not checked by borrow checker, which is the reason why it's considered an incomplete feature).

`@rustbot` label: +T-lang +F-inline_const
2022-12-13 19:57:09 +01:00
Esteban Küber cf0b6b9337 Account for dereference expressions 2022-12-13 10:06:15 -08:00
Esteban Küber b8bd1d0826 Fix span for &mut removal suggestion 2022-12-13 10:06:14 -08:00
Esteban Küber dd72b1a0df Suggest ref for some patterns as a fallback 2022-12-13 10:06:14 -08:00
Esteban Küber 4f7c257fd8 Do not suggest borrowing binding in pattern in let else
Fix #104838.
2022-12-13 10:06:13 -08:00
Esteban Küber e46416eed6 Change pattern borrowing suggestions to be verbose
Synthesize a more accurate span and use verbose suggestion output to
make the message clearer.
2022-12-13 10:06:13 -08:00
bors aa5b179599 Auto merge of #105646 - lqd:revert-thinlto, r=Mark-Simulacrum
Revert "enable ThinLTO for rustc on x86_64-apple-darwin dist builds"

Apparently ThinLTO on x64 mac can regress some of the ICEs' output. This reverts #103647 to allow for investigation, and helps with #105637 in the meantime.
2022-12-13 15:50:16 +00:00
bors ed620cf969 Auto merge of #105612 - oli-obk:bind_rustdoc, r=GuillaumeGomez
use ty::Binder in rustdoc instead of `skip_binder`

r? `@GuillaumeGomez`

this is a preliminary cleanup required to be able to normalize correctly/conveniently in rustdoc
2022-12-13 13:09:40 +00:00
Oli Scherer 82ce70af62 bless fulldeps tests 2022-12-13 10:11:19 +00:00
Oli Scherer 1f5cb9e50e Use a label instead of a note for the drop site to create denser diagnostics 2022-12-13 10:08:46 +00:00
Oli Scherer 6984085088 Stop pointing to operators if their libcore method source is not available 2022-12-13 10:06:11 +00:00
Oli Scherer 59554a2b54 Avoid rendering empty annotations 2022-12-13 10:06:08 +00:00
Oli Scherer 2e2a4797a2 Don't emit empty notes 2022-12-13 09:53:36 +00:00
Oli Scherer 063b1675b2 Clarify what "this" means 2022-12-13 09:51:03 +00:00
Oli Scherer 717294fa04 Inform the user which trait is meant in the diagnostic itself instead of relying on the span making it obvious 2022-12-13 09:48:26 +00:00
Oli Scherer cb26b35b12 Make some diagnostics not depend on the source of what they reference being available 2022-12-13 09:48:22 +00:00
bors 109cccbe4f Auto merge of #105350 - compiler-errors:faster-binder-relate, r=oli-obk
Fast-path some binder relations

A simpler approach than #104598

Fixes #104583

r? types
2022-12-13 07:10:53 +00:00
Michael Goulet a8a45100a0 Move some codegen-y methods from rustc_hir_analysis::collect -> rustc_codegen_ssa 2022-12-13 05:01:36 +00:00
bors 71ec1457ee Auto merge of #105436 - nnethercote:inline-place_contents_drop_state_cannot_differ, r=spastorino
Inline and remove `place_contents_drop_state_cannot_differ`.

It has a single call site and is hot enough to be worth inlining. And make sure `is_terminal_path` is inlined, too.

r? `@ghost`
2022-12-13 03:28:57 +00:00
Michael Goulet 2025a96ee1 Fast path some binder relations 2022-12-13 03:17:14 +00:00
Gary Guo d6dc9124b7 Extract shared logic into a new function 2022-12-13 02:34:43 +00:00
Gary Guo d2830cb204 Add tests (currently broken) for unsafe + inline const pat 2022-12-13 01:38:38 +00:00
Gary Guo 5c58a1b003 Remove unnecessary recursive call to parent unsafeck
All bodies are unsafe checked anyway. Current MIR unsafeck also just
returns for closures.
2022-12-13 01:38:38 +00:00
Gary Guo adf1717219 Ensure valid local_data is set for custom mir building
MIR unsafety checking requires this to be valid
2022-12-13 01:38:38 +00:00
Gary Guo f03c2aef47 Add tests 2022-12-13 01:38:38 +00:00
Gary Guo aa5af2a003 Allow unsafe through inline const for THIR unsafety checker
The closure handling code is changed slightly to avoid allocation
when THIR building failed.
2022-12-13 01:38:38 +00:00
Gary Guo 9342d1e73e Allow unsafe through inline const
This is handled similar to closures
2022-12-13 01:38:38 +00:00
Rémy Rakic ddb6fe2e1d Revert "enable ThinLTO for rustc on x86_64-apple-darwin dist builds"
This reverts commit 3a085f7695.
2022-12-13 01:11:54 +00:00
bors b96d9e0e20 Auto merge of #105644 - matthiaskrgr:rollup-qc6hlzq, r=matthiaskrgr
Rollup of 8 pull requests

Successful merges:

 - #104864 (Account for item-local in inner scope for E0425)
 - #105332 (Point out the type of associated types in every method call of iterator chains)
 - #105620 (Remove unnecessary uses of `clone`)
 - #105625 (minor code cleanups)
 - #105629 (rustdoc: stop treating everything in a trait item as a method)
 - #105636 (Add check for local-storage value when changing "display line numbers" settings)
 - #105639 (rustdoc: remove `type="text/css" from stylesheet links)
 - #105640 (Adjust miri to still be optional)

Failed merges:

r? `@ghost`
`@rustbot` modify labels: rollup
2022-12-13 00:29:28 +00:00
Matthias Krüger 5af04471aa Rollup merge of #105640 - lukas-code:miri-beta, r=Mark-Simulacrum
Adjust miri to still be optional

r? `@pietroalbini`
2022-12-13 01:17:12 +01:00
Matthias Krüger 84a725e4e3 Rollup merge of #105639 - notriddle:notriddle/text-css, r=GuillaumeGomez
rustdoc: remove `type="text/css" from stylesheet links

MDN directly recommends this in <https://developer.mozilla.org/en-US/docs/Web/HTML/Element/link>, since "CSS is the only stylesheet language used on the web."

Like 07a243b2a4, but a few places that were missed the first time.
2022-12-13 01:17:12 +01:00
Matthias Krüger 8ba9c211bb Rollup merge of #105636 - GuillaumeGomez:extend-gui-test, r=notriddle
Add check for local-storage value when changing "display line numbers" settings

r? `@notriddle`
2022-12-13 01:17:11 +01:00
Matthias Krüger 105e398349 Rollup merge of #105629 - notriddle:notriddle/method-toggle-trait, r=GuillaumeGomez
rustdoc: stop treating everything in a trait item as a method

This was added in 0b9b4b7068 to fix the spacing on trait pages, but stopped being needed because 791f04e5a4 stopped styling method-toggle. By only putting the method-toggle class on actual methods, the JS setting does the right thing.
2022-12-13 01:17:11 +01:00
Matthias Krüger 2707801858 Rollup merge of #105625 - matthiaskrgr:clippy_dec12, r=compiler-errors
minor code cleanups

r? `@compiler-errors`
2022-12-13 01:17:10 +01:00
Matthias Krüger 4069792d73 Rollup merge of #105620 - TaKO8Ki:remove-unnecessary-uses-of-clone, r=compiler-errors
Remove unnecessary uses of `clone`
2022-12-13 01:17:10 +01:00
Matthias Krüger 8917cc0f23 Rollup merge of #105332 - estebank:iterator-chains, r=oli-obk
Point out the type of associated types in every method call of iterator chains

Partially address #105184 by pointing out the type of associated types in every method call of iterator chains:

```
note: the expression is of type `Map<std::slice::Iter<'_, {integer}>, [closure@src/test/ui/iterators/invalid-iterator-chain.rs:12:18: 12:21]>`
    --> src/test/ui/iterators/invalid-iterator-chain.rs:12:14
     |
10   |         vec![0, 1]
     |         ---------- this expression has type `Vec<{integer}>`
11   |             .iter()
     |              ------ associated type `std::iter::Iterator::Item` is `&{integer}` here
12   |             .map(|x| { x; })
     |              ^^^^^^^^^^^^^^^ associated type `std::iter::Iterator::Item` is `()` here
```

We also reduce the number of impls we mention when any of the candidates is an "exact match". This benefits the output of cases with numerics greatly.

Outstanding work would be to provide a structured suggestion for appropriate changes, like in this case detecting the spurious `;` in the closure.
2022-12-13 01:17:09 +01:00
Matthias Krüger 6dbaf86672 Rollup merge of #104864 - chenyukang:yukang/fix-104700-binding, r=estebank
Account for item-local in inner scope for E0425

Fixes #104700
2022-12-13 01:17:08 +01:00
Mark Rousskov 4ac8190287 Adjust miri to still be optional
We don't distribute a miri build for beta/stable so it needs to be kept
optional. In the future it likely makes sense to switch the miri
*artifacts* to always be built, but the rustup component to not be
included -- this will avoid some of this pain.
2022-12-13 00:01:20 +01:00
Michael Howell 46b4a3b831 rustdoc: remove `type="text/css" from stylesheet links
MDN directly recommends this in <https://developer.mozilla.org/en-US/docs/Web/HTML/Element/link>,
since "CSS is the only stylesheet language used on the web."

Like 07a243b2a4, but a few places that were
missed the first time.
2022-12-12 15:17:49 -07:00
Guillaume Gomez 21a2b64209 Add check for local-storage value when changing "display line numbers" settings 2022-12-12 22:49:19 +01:00
bors ed9749324a Auto merge of #105622 - matthiaskrgr:rollup-a6i81q7, r=matthiaskrgr
Rollup of 8 pull requests

Successful merges:

 - #104405 (1.66.0 release notes)
 - #105561 (Normalize receiver substs and erase the regions)
 - #105593 (Fix typo in comment: length_limit)
 - #105597 (Correct typos in `core::sync::Exclusive::get_{pin_mut, mut}`)
 - #105614 (delete mentions of type ascription from lint descriptions)
 - #105615 (Fixup method doc that mentions removed param)
 - #105616 (Add a "the" to proc_macro documentation)
 - #105619 (rustdoc: remove no-op CSS `.source pre { overflow: auto }`)

Failed merges:

r? `@ghost`
`@rustbot` modify labels: rollup
2022-12-12 21:34:28 +00:00
Esteban Küber f4ed2d1cca Do not skip_binders 2022-12-12 11:53:39 -08:00
Michael Howell 7e64cebf97 rustdoc: stop treating everything in a trait item as a method
This was added in 0b9b4b7068 to fix the
spacing on trait pages, but stopped being needed because
791f04e5a4 stopped styling method-toggle.
By only putting the method-toggle class on actual methods, the JS setting
does the right thing.
2022-12-12 12:49:29 -07:00
Matthias Krüger 2ea368e53c minor code cleanups 2022-12-12 19:49:53 +01:00
Matthias Krüger ab29c0459c Rollup merge of #105619 - notriddle:notriddle/source-pre-overflow, r=GuillaumeGomez
rustdoc: remove no-op CSS `.source pre { overflow: auto }`

Since source pages use the `example-wrap` wrapper, this rule became redundant because there is already an `overflow-x: auto` rule.
2022-12-12 19:20:38 +01:00
Matthias Krüger f5852c41a0 Rollup merge of #105616 - est31:add_the, r=Dylan-DPC
Add a "the" to proc_macro documentation
2022-12-12 19:20:37 +01:00