Commit Graph

200329 Commits

Author SHA1 Message Date
Oli Scherer
90ec6f847f Show errors instead of hiding them due to an earlier error 2022-09-23 07:12:59 +00:00
Oli Scherer
5ace12c409 Showcase a broken diagnostic 2022-09-23 07:10:11 +00:00
Oli Scherer
86f1ca812b Improve diagnostic for adding more bounds to opaque types 2022-09-21 13:15:24 +00:00
Oli Scherer
4b323e62ba Reproduce sad diagnostic 2022-09-21 13:15:24 +00:00
Oli Scherer
36cb01deb3 Try out a perf optimization 2022-09-21 13:15:24 +00:00
Oli Scherer
20d962cfa5 Simplify insert_outlives_predicate opaque type logic 2022-09-21 13:15:24 +00:00
Oli Scherer
615c9e8647 Inline substitution logic into declared_region_bounds 2022-09-21 13:11:54 +00:00
Oli Scherer
a8c9784336 Add regression test 2022-09-21 13:11:54 +00:00
Oli Scherer
90b6744af7 Also collect bounds from the ParamEnv for opaque types 2022-09-21 13:11:54 +00:00
Oli Scherer
37928f5986 Neither require nor imply lifetime bounds on opaque type for well formedness 2022-09-21 13:11:54 +00:00
Oli Scherer
c8ecf09a25 Generalize projection_must_outlive 2022-09-21 12:51:33 +00:00
Oli Scherer
4d4cc4fe53 Generalize a helper to be useful for types other than projections 2022-09-21 12:51:33 +00:00
Oli Scherer
96b819a456 Inline a trivial function 2022-09-21 12:51:33 +00:00
Oli Scherer
759c04a00b Some tracing instrumentation cleanups 2022-09-21 11:31:55 +00:00
Oli Scherer
3d69d23a25 Split match out into a helper function 2022-09-21 11:31:55 +00:00
bors
1de00d1ac5 Auto merge of #100096 - compiler-errors:fn-return-must-be-sized, r=jackh726
a fn pointer doesn't implement `Fn`/`FnMut`/`FnOnce` if its return type isn't sized

I stumbled upon #83915 which hasn't received much attention recently, and I wanted to revive it since this is one existing soundness hole that seems pretty easy to fix.

I'm not actually sure that the [alternative approach described here](https://github.com/rust-lang/rust/pull/83915#issuecomment-823643322) is sufficient, given the `src/test/ui/function-pointer/unsized-ret.rs` example I provided below. Rebasing the branch mentioned in that comment and testing that UI test, it seems that we actually end up only observing that `str: !Sized` during monomorphization, whereupon we ICE. Even if we were to fix that ICE, ideally we'd be raising an error that a fn pointer is being used badly during _typecheck_ instead of monomorphization, hence adapting the original approach in #83915.

I am happy to close this if people would prefer we rebase the original PR and land that -- I am partly opening to be annoying and get people thinking about this unsoundness again ❤️ 😸

cc: `@estebank` and `@nikomatsakis`
r? types

Here's a link to the thread: https://rust-lang.zulipchat.com/#narrow/stream/144729-t-types/topic/PR.20.2383915/near/235421351 for more context.
2022-09-21 04:35:20 +00:00
bors
4ecfdfac51 Auto merge of #100214 - scottmcm:strict-range, r=thomcc
Optimize `array::IntoIter`

`.into_iter()` on arrays was slower than it needed to be (especially compared to slice iterator) since it uses `Range<usize>`, which needs to handle degenerate ranges like `10..4`.

This PR adds an internal `IndexRange` type that's like `Range<usize>` but with a safety invariant that means it doesn't need to worry about those cases -- it only handles `start <= end` -- and thus can give LLVM more information to optimize better.

I added one simple demonstration of the improvement as a codegen test.

(`vec::IntoIter` uses pointers instead of indexes, so doesn't have this problem, but that only works because its elements are boxed.  `array::IntoIter` can't use pointers because that would keep it from being movable.)
2022-09-21 00:41:33 +00:00
bors
7743aa836e Auto merge of #100581 - joboet:sync_rwlock_everywhere, r=thomcc
std: use `sync::RwLock` for internal statics

Since `sync::RwLock` is now `const`-constructible, it can be used for internal statics, removing the need for `sys_common::StaticRwLock`. This adds some extra allocations on platforms which need to box their locks (currently SGX and some UNIX), but these will become unnecessary with the lock improvements tracked in #93740.
2022-09-20 22:00:08 +00:00
bors
432abd86f2 Auto merge of #102061 - notriddle:rollup-kwu9vp8, r=notriddle
Rollup of 12 pull requests

Successful merges:

 - #100250 (Manually cleanup token stream when macro expansion aborts.)
 - #101014 (Fix -Zmeta-stats ICE by giving `FileEncoder` file read permissions)
 - #101958 (Improve error for when query is unsupported by crate)
 - #101976 (MirPhase: clarify that linting is not a semantic change)
 - #102001 (Use LLVM C-API to build atomic cmpxchg and fence)
 - #102008 (Add GUI test for notable traits element position)
 - #102013 (Simplify rpitit handling on lower_fn_decl)
 - #102021 (some post-valtree cleanup)
 - #102027 (rustdoc: remove `docblock` class from `item-decl`)
 - #102034 (rustdoc: remove no-op CSS `h1-6 { border-bottom-color }`)
 - #102038 (Make the `normalize-overflow` rustdoc test actually do something)
 - #102053 (⬆️ rust-analyzer)

Failed merges:

r? `@ghost`
`@rustbot` modify labels: rollup
2022-09-20 19:10:11 +00:00
Michael Howell
25f5483f58
Rollup merge of #102053 - lnicola:rust-analyzer-2022-09-20, r=lnicola
⬆️ rust-analyzer

r? ``@ghost``
2022-09-20 10:13:03 -07:00
Michael Howell
3fac4bbdda
Rollup merge of #102038 - jyn514:rustdoc-normalize-test, r=compiler-errors
Make the `normalize-overflow` rustdoc test actually do something

Since https://github.com/rust-lang/rust/pull/88679, rustdoc doesn't load crates eagerly. Add an explicit `extern crate` item to make sure the crate is loaded and the bug reproduces.
You can verify this fix by adding `// compile-flags: -Znormalize-docs` and running the test to make sure it gives an error.
2022-09-20 10:13:02 -07:00
Michael Howell
128733d742
Rollup merge of #102034 - notriddle:notriddle/border-bottom, r=GuillaumeGomez
rustdoc: remove no-op CSS `h1-6 { border-bottom-color }`

For this rule to have an actual effect, the border-bottom width needs specified, elsewhere, without also specifying the color. This doesn't happen. Ever since 88b137d5fe, every spot where headers get a border assigned to them also assigns the color.

Preview: https://notriddle.com/notriddle-rustdoc-test/border-bottom/rustc_monomorphize/collector/index.html
2022-09-20 10:13:02 -07:00
Michael Howell
1b54ee38a3
Rollup merge of #102027 - notriddle:notriddle/docblock-item-decl, r=GuillaumeGomez
rustdoc: remove `docblock` class from `item-decl`

This class was originally added in 73b97c7e7c to support hiding and showing the item, because `main.js` went through all `docblock` elements in the DOM and added toggles to them.

73b97c7e7c/src/librustdoc/html/static/main.js (L1856-L1867)

The `item-decl` is no longer auto-hidden since c96f86de30 removed it.

`item-decl` used to be called `type-decl`: that name was changed in 8b7a2dd462.

The `docblock` class is no longer used for implementing toggles, since rustdoc switched to using `<details>` elements.

Preview: https://notriddle.com/notriddle-rustdoc-test/docblock-item-decl/rustdoc/clean/types/enum.Type.html
2022-09-20 10:13:01 -07:00
Michael Howell
b149c48186
Rollup merge of #102021 - lcnr:tyConst-fun, r=b-naber,BoxyUwU
some post-valtree cleanup

r? project-const-generics cc ```@b-naber```
2022-09-20 10:13:01 -07:00
Michael Howell
08b369a501
Rollup merge of #102013 - spastorino:rpitit-lower-fn-decl, r=compiler-errors
Simplify rpitit handling on lower_fn_decl

r? `````@compiler-errors`````
2022-09-20 10:13:00 -07:00
Michael Howell
09221722af
Rollup merge of #102008 - GuillaumeGomez:notable-trait-gui-test, r=notriddle
Add GUI test for notable traits element position

Fixes #101891.

r? ````@notriddle````
2022-09-20 10:12:59 -07:00
Michael Howell
2c57a5b166
Rollup merge of #102001 - cuviper:build-atomic-capi, r=nikic
Use LLVM C-API to build atomic cmpxchg and fence

We don't need C++ wrappers because the LLVM C API can build these.
2022-09-20 10:12:59 -07:00
Michael Howell
0d2e57cb04
Rollup merge of #101976 - RalfJung:mir-semantic-phases, r=oli-obk
MirPhase: clarify that linting is not a semantic change

r? ```@JakobDegen```
2022-09-20 10:12:58 -07:00
Michael Howell
2c893737a6
Rollup merge of #101958 - hanar3:101666/enhance-error-message, r=oli-obk
Improve error for when query is unsupported by crate

This is an improvement to the error message mentioned on #101666.  It seems like a good idea to also add [this link to the rustc-dev-guide](https://rustc-dev-guide.rust-lang.org/query.html), if explaining the query system in detail is a concern here, but I'm unsure if there is any restrictions on adding links to error messages.
2022-09-20 10:12:58 -07:00
Michael Howell
3f377d3f96
Rollup merge of #101014 - isikkema:fix-zmeta-stats-file-encoder-no-read-perms, r=isikkema
Fix -Zmeta-stats ICE by giving `FileEncoder` file read permissions

Fixes #101001

As far as I can tell, #101001 is caused because the file is being created with write-only permissions here: https://github.com/rust-lang/rust/blob/master/compiler/rustc_serialize/src/opaque.rs#L196

but it is trying to be read here: https://github.com/rust-lang/rust/blob/master/compiler/rustc_metadata/src/rmeta/encoder.rs#L780

This PR attempts to fix this by creating/opening the file with the same permissions as `File::create()` with the addition of read.
2022-09-20 10:12:57 -07:00
Michael Howell
14b27cfd11
Rollup merge of #100250 - cjgillot:recover-token-stream, r=Aaron1011
Manually cleanup token stream when macro expansion aborts.

In case of syntax error in macro expansion, the expansion code can decide to stop processing anything. In that case, the token stream is malformed. This makes downstream users, like derive macros, ICE.

In this case, this PR manually cleans up the token stream by closing all currently open delimiters.

Fixes https://github.com/rust-lang/rust/issues/96818.
Fixes https://github.com/rust-lang/rust/issues/80447.
Fixes https://github.com/rust-lang/rust/issues/81920.
Fixes https://github.com/rust-lang/rust/issues/91023.
2022-09-20 10:12:56 -07:00
bors
cd8cc91045 Auto merge of #101989 - fee1-dead-contrib:const_trait_impl-assoc-caller-bounds, r=oli-obk
make projection bounds with const bounds satisfy const

Fixes #101982.
2022-09-20 16:24:36 +00:00
Laurențiu Nicola
9dcd19bd2e ⬆️ rust-analyzer 2022-09-20 17:39:17 +03:00
Sikkema, Isaac
a2cb8a4949 add comment explaining read permissions 2022-09-20 10:06:50 -04:00
bors
4136b59b7d Auto merge of #99806 - oli-obk:unconstrained_opaque_type, r=estebank
Allow patterns to constrain the hidden type of opaque types

fixes #96572

reverts a revert as original PR was a perf regression that was fixed by reverting it: https://github.com/rust-lang/rust/pull/99368#issuecomment-1186587864)

TODO:

* check if https://github.com/rust-lang/rust/issues/99685 is avoided
2022-09-20 12:09:52 +00:00
bors
8fd6d03e22 Auto merge of #101806 - BelovDV:issue-fix-fn-find_library, r=petrochenkov
fix verbatim with upstream dependencies

https://github.com/rust-lang/rust/issues/99425#issuecomment-1207224161

r? `@petrochenkov`
2022-09-20 07:10:18 +00:00
bors
a38a082afb fix verbatim with upstream dependencies
https://github.com/rust-lang/rust/issues/99425#issuecomment-1207224161

r? `@petrochenkov`
2022-09-20 07:10:17 +00:00
Scott McMurray
6dbd9a29c2 Optimize array::IntoIter
`.into_iter()` on arrays was slower than it needed to be (especially compared to slice iterator) since it uses `Range<usize>`, which needs to handle degenerate ranges like `10..4`.

This PR adds an internal `IndexRange` type that's like `Range<usize>` but with a safety invariant that means it doesn't need to worry about those cases -- it only handles `start <= end` -- and thus can give LLVM more information to optimize better.

I added one simple demonstration of the improvement as a codegen test.
2022-09-19 23:24:34 -07:00
Joshua Nelson
690c467799 Make the normalize-overflow rustdoc test actually do something
Since https://github.com/rust-lang/rust/pull/88679, rustdoc doesn't load crates eagerly.
Add an explicit `extern crate` item to make sure the crate is loaded and the bug reproduces.
You can verify this fix by adding `// compile-flags: -Znormalizing-docs` and running the test.
2022-09-19 22:49:37 -05:00
Michael Howell
17259c3995 Remove useless color assertions on 0px borders 2022-09-19 20:14:09 -07:00
Michael Howell
78d8ce3faa rustdoc: remove no-op CSS h1-6 { border-bottom-color }
For this rule to have an actual effect, the border-bottom width needs
specified, elsewhere, without also specifying the color. This doesn't
happen. Ever since 88b137d5fe, every spot
where headers get a border assigned to them also assigns the color.
2022-09-19 20:14:09 -07:00
bors
acb8934fd5 Auto merge of #102022 - matthiaskrgr:rollup-emwfjd1, r=matthiaskrgr
Rollup of 11 pull requests

Successful merges:

 - #101389 (Tone down explanation on RefCell::get_mut)
 - #101798 (Make `from_waker`, `waker` and `from_raw` unstably `const`)
 - #101881 (Remove an unused struct field `late_bound`)
 - #101904 (Add help for invalid inline argument)
 - #101966 (Add unit test for identifier Unicode emoji diagnostics)
 - #101979 (Update release notes for 1.64)
 - #101985 (interpret: expose generate_stacktrace without full InterpCx)
 - #102004 (Try to clarify what's new in 1.64.0 ffi types)
 - #102005 (rustdoc: remove unused CSS `td.summary-column`)
 - #102017 (Add all submodules to the list of directories tidy skips)
 - #102019 (Remove backed off PRs from relnotes)

Failed merges:

r? `@ghost`
`@rustbot` modify labels: rollup
2022-09-20 02:20:15 +00:00
bors
a57a6ac2eb Auto merge of #101909 - weihanglo:update-cargo, r=ehuss
Update cargo (CVE fixes included)

8 commits in 082503982ea0fb7a8fd72210427d43a2e2128a63..73ba3f35e0205844418260722c11602113179c4a
2022-09-13 17:49:38 +0000 to 2022-09-18 06:38:16 +0000

- Revert "Clarify when cargo detects changes" (https://github.com/rust-lang/cargo/pull/11107)
- Fix links to workspace inheritance headings in workspace docs (https://github.com/rust-lang/cargo/pull/11103)
- docs(ref): Clarify workspace settings (https://github.com/rust-lang/cargo/pull/11082)
- Update comment about ResolveVersion default version (https://github.com/rust-lang/cargo/pull/11095)
- [master] Run `reach_max_unpack_size` test only on debug build (https://github.com/rust-lang/cargo/pull/11091)
- Clarify when cargo detects changes (https://github.com/rust-lang/cargo/pull/11092)
- [master] Fix for https://github.com/advisories/GHSA-rfj2-q3h3-hm5j and https://github.com/advisories/GHSA-2hvr-h6gw-qrxp (https://github.com/rust-lang/cargo/pull/11089)
- Expose cargo add internals as edit API (https://github.com/rust-lang/cargo/pull/11059)
2022-09-19 23:18:12 +00:00
joboet
be09a4a8b2
std: use sync::RwLock for internal statics 2022-09-19 23:27:26 +02:00
Michael Howell
8e6cf7d840 rustdoc: update test cases for changed item-decl HTML 2022-09-19 13:39:31 -07:00
Michael Howell
9457380ac9 rustdoc: remove docblock class from item-decl
This class was originally added in 73b97c7e7c
to support hiding and showing the item, because `main.js` went through all
`docblock` elements in the DOM and added toggles to them.

73b97c7e7c/src/librustdoc/html/static/main.js (L1856-L1867)

The `item-decl` is no longer auto-hidden since
c96f86de30 removed it.

`item-decl` used to be called `type-decl`: that name was changed in
8b7a2dd462.

The `docblock` class is no longer used for implementing toggles, since
rustdoc switched to using `<details>` elements.
2022-09-19 10:52:51 -07:00
bors
2019147c56 Auto merge of #101894 - dingxiangfei2009:let-else-avoid-duplicate-storage-live, r=oli-obk
Avoid duplicating StorageLive in let-else

cc `@est31`

Fix #101867
Fix #101932

#101410 introduced directives to activate storages of bindings in let-else earlier. However, since it is using the machinery of `match` and friends for pattern matching and binding, those storages are activated for the second time. This PR adjusts this behavior and avoid the duplicated activation for let-else statements.
2022-09-19 17:27:30 +00:00
Matthias Krüger
683cf99f1d
Rollup merge of #102019 - pietroalbini:pa-relnotes-revert, r=pietroalbini
Remove backed off PRs from relnotes

Backport of 7358a9b03e to master.
2022-09-19 17:55:24 +02:00
Matthias Krüger
ade6505976
Rollup merge of #102017 - est31:rustc_dev_guide_tidy_skip, r=Mark-Simulacrum
Add all submodules to the list of directories tidy skips

Tidy contains a blacklist of directories that it is not visiting. This list is also used by the `replace-version-placeholder` tool added by #100591 , to determine the directories to do its replacement from. Generally, tidy does not check submodules, but this is not done consistently for all submodules. This PR adds the submodules that were previously missing, so that the `replace-version-placeholder` tool does not attempt to change content of the books. This was needed because `rustc-dev-guide` contains the placeholder, leading to #102014.

Fixes #102014
2022-09-19 17:55:24 +02:00
Matthias Krüger
548c3eaa24
Rollup merge of #102005 - notriddle:notriddle/td-summary-column, r=Dylan-DPC
rustdoc: remove unused CSS `td.summary-column`

It was added in 2a1bad70dd to go with this module summary function:

2a1bad70dd/src/librustdoc/html/format.rs (L767-L780)

The corresponding HTML was removed in 0a46933c4d.
2022-09-19 17:55:23 +02:00