Dylan DPC [Tue, 30 Aug 2022 05:56:48 +0000 (11:26 +0530)]
Rollup merge of #100092 - compiler-errors:issue-100075, r=oli-obk
Fall back when relating two opaques by substs in MIR typeck
This is certainly _one_ way to fix #100075. Not really confident it's the _best_ way to do it, though.
The root cause of this issue is that during MIR type-check, we end up trying to equate an opaque against the same opaque def-id but with different substs. Because of the way that we replace RPITs during (HIR) typeck with an inference variable, we don't end up emitting a type-checking error, so the delayed MIR bug causes an ICE.
See the `src/test/ui/impl-trait/issue-100075-2.rs` test below to make that clear -- in that example, we try to equate `{impl Sized} substs=[T]` and `{impl Sized} substs=[Option<T>]`, which causes an ICE. This new logic will instead cause us to infer `{impl Sized} substs=[Option<T>]` as the hidden type for `{impl Sized} substs=[T]`, which causes a proper error to be emitted later on when we check that an opaque isn't recursive.
I'm open to closing this in favor of something else. Ideally we'd fix this in typeck, but the thing we do to ensure backwards compatibility with weird RPIT cases makes that difficult. Also open to discussing this further.
impl<T, F, A: Allocator> DrainFilter<'_, T, F, A>
where
F: FnMut(&mut T) -> bool,
{
pub fn keep_rest(self);
}
```
Both these methods cancel draining of elements that were not yet yielded from the iterators. While this needs more testing & documentation, I want at least start the discussion. This may be a potential way out of the "should `DrainFilter` exhaust itself on drop?" argument.
Jason Newcomb [Tue, 30 Aug 2022 04:33:56 +0000 (00:33 -0400)]
Fix the order of `trait_duplication_in_bounds`
* Emit the lint in source order
* Make suggestions with multiple traits be in source order rather than alphabetical
Michael Howell [Tue, 30 Aug 2022 00:25:43 +0000 (17:25 -0700)]
rustdoc: remove unused CSS selectors for `.table-display`
This class was added to support the function signature [src] lockup. That
lockup was changed in 34bd2b845b3acd84c5a9bddae3ff8081c19ec5e9 to use flexbox
instead, leaving these selectors unused.
Tyler Mandry [Mon, 29 Aug 2022 23:43:47 +0000 (16:43 -0700)]
Don't --bless in pre-push hook
Running with --bless causes the push to succeed if there are fixable
formatting changes, but the changes don't make it into the push.
We should have the user rerun with --bless (or x.py fmt) and commit the
changes themselves (they might want to amend a particular commit, for
instance).
bors [Mon, 29 Aug 2022 22:49:04 +0000 (22:49 +0000)]
Auto merge of #101167 - matthiaskrgr:rollup-yt3jdmp, r=matthiaskrgr
Rollup of 7 pull requests
Successful merges:
- #100898 (Do not report too many expr field candidates)
- #101056 (Add the syntax of references to their documentation summary.)
- #101106 (Rustdoc-Json: Retain Stripped Modules when they are imported, not when they have items)
- #101131 (CTFE: exposing pointers and calling extern fn is just impossible)
- #101141 (Simplify `get_trait_ref` fn used for `virtual_function_elimination`)
- #101146 (Various changes to logging of borrowck-related code)
- #101156 (Remove `Sync` requirement from lint pass objects)
Matthias Krüger [Mon, 29 Aug 2022 19:13:00 +0000 (21:13 +0200)]
Rollup merge of #101156 - Jarcho:remove_sync_lint_pass, r=compiler-errors
Remove `Sync` requirement from lint pass objects
This is blocking the clippy sync (#101140). One of the lint passes contains a `Cell` in order to make lifetimes work. It could be worked around, but this is the easier change to make if there are no objections.
Rational for removing the requirement
* All lint pass methods take `&mut self` arguments.
* Many passes depend on running is visitor order.
* Lint passes are created on demand so they're only ever stored in a local.
* `Send` is enough to lint different passes in parallel.
`LintStore` remains `Sync` with this. The constructor functions it contains still maintain their `Sync` requirement.
Matthias Krüger [Mon, 29 Aug 2022 19:12:58 +0000 (21:12 +0200)]
Rollup merge of #101141 - compiler-errors:get-trait-ref-is-a-misleading-name, r=oli-obk
Simplify `get_trait_ref` fn used for `virtual_function_elimination`
1. The name `get_trait_ref` is misleading, so I renamed it to something more like `expect_...` because it ICEs if used incorrectly.
2. No need to manually go through the existential trait refs, we already have `.principal()` for that.
Matthias Krüger [Mon, 29 Aug 2022 19:12:57 +0000 (21:12 +0200)]
Rollup merge of #101131 - RalfJung:ctfe-no-needs-rfc, r=oli-obk
CTFE: exposing pointers and calling extern fn is just impossible
The remaining "needs RFC" errors are just needlessly confusing, I think -- time to get rid of that error variant. They are anyway only reachable with miri-unleashed (if at all).
Matthias Krüger [Mon, 29 Aug 2022 19:12:55 +0000 (21:12 +0200)]
Rollup merge of #101056 - kpreid:prim-doc, r=JohnTitor
Add the syntax of references to their documentation summary.
Without this change, in <https://doc.rust-lang.org/1.63.0/std/#primitives>, `reference` is the only entry in that list which does not contain the syntax by which the type is named in source code. With this change, it contains them, in roughly the same way as the `pointer` entry does.
Matthias Krüger [Mon, 29 Aug 2022 19:12:54 +0000 (21:12 +0200)]
Rollup merge of #100898 - compiler-errors:too-many-expr-fields, r=spastorino
Do not report too many expr field candidates
When considering "this expressions' field has a {field/method}" suggestions:
1. Don't report methods that are out of scope
2. Use `span_suggestions` instead of reporting each field candidate, which caps the number of suggestions to 4
4. Blacklist some common traits like `Clone` and `Deref`
bors [Mon, 29 Aug 2022 18:56:59 +0000 (18:56 +0000)]
Auto merge of #101152 - Dylan-DPC:rollup-v4iw8ux, r=Dylan-DPC
Rollup of 8 pull requests
Successful merges:
- #98304 (Add MaybeUninit memset test)
- #98801 (Add a `File::create_new` constructor)
- #99821 (Remove separate indexing of early-bound regions)
- #100239 (remove an ineffective check in const_prop)
- #100337 (Stabilize `std::io::read_to_string`)
- #100819 (Make use of `[wrapping_]byte_{add,sub}`)
- #100934 (Remove a panicking branch from `fmt::builders::PadAdapter`)
- #101000 (Separate CountIsStar from CountIsParam in rustc_parse_format.)
bors [Mon, 29 Aug 2022 18:07:50 +0000 (18:07 +0000)]
Auto merge of #9247 - clubby789:raw_slice_pointer_cast, r=Alexendoo
New lint: Raw slice pointer cast
Adds a lint to check for a raw slice being created and cast back to a pointer, suggesting `ptr::slice_from_raw_parts`, to identify UB such as https://github.com/SimonSapin/rust-typed-arena/pull/54.
```
changelog: [`cast_slice_from_raw_parts`]: Add lint to check for `slice::from_raw_parts(.., ..) as *const _`
```
bors [Mon, 29 Aug 2022 16:04:11 +0000 (16:04 +0000)]
Auto merge of #101147 - weihanglo:update-cargo, r=Mark-Simulacrum
Update cargo
5 commits in 6da726708a4406f31f996d813790818dce837161..4ed54cecce3ce9ab6ff058781f4c8a500ee6b8b5
2022-08-23 21:39:56 +0000 to 2022-08-27 18:41:39 +0000
- doc: pause, for readability (rust-lang/cargo#11027)
- Bump git2 to 0.15 and libgit2-sys to 0.14 (rust-lang/cargo#11004)
- Fix typo (rust-lang/cargo#11025)
- Update cargo-toml-vs-cargo-lock.md (rust-lang/cargo#11021)
- Apply GitHub fast path even for partial hashes (rust-lang/cargo#10807)
Dylan DPC [Mon, 29 Aug 2022 11:19:45 +0000 (16:49 +0530)]
Rollup merge of #101000 - m-ou-se:count-is-star, r=nagisa
Separate CountIsStar from CountIsParam in rustc_parse_format.
`rustc_parse_format`'s parser would result in the exact same output for `{:.*}` and `{:.0$}`, making it hard for diagnostics to handle these cases properly.
This splits those cases by adding a new `CountIsStar` enum variant.
This fixes #100995
Prerequisite for https://github.com/rust-lang/rust/pull/100996
Dylan DPC [Mon, 29 Aug 2022 11:19:43 +0000 (16:49 +0530)]
Rollup merge of #100819 - WaffleLapkin:use_ptr_byte_methods, r=scottmcm
Make use of `[wrapping_]byte_{add,sub}`
These new methods trivially replace old `.cast().wrapping_offset().cast()` & similar code.
Note that [`arith_offset`](https://doc.rust-lang.org/std/intrinsics/fn.arith_offset.html) and `wrapping_offset` are the same thing.
Dylan DPC [Mon, 29 Aug 2022 11:19:40 +0000 (16:49 +0530)]
Rollup merge of #100239 - RalfJung:const-prop-uninit, r=oli-obk
remove an ineffective check in const_prop
Based on https://github.com/rust-lang/rust/pull/100043, only the last two commits are new.
ConstProp has a special check when reading from a local that prevents reading uninit locals. However, if that local flows into `force_allocation`, then no check fires and evaluation proceeds. So this check is not really effective at preventing accesses to uninit locals.
With https://github.com/rust-lang/rust/pull/100043, `read_immediate` and friends always fail when reading uninit locals, so I don't see why ConstProp would need a separate check. Thus I propose we remove it. This is needed to be able to do https://github.com/rust-lang/rust/pull/100085.
Dylan DPC [Mon, 29 Aug 2022 11:19:39 +0000 (16:49 +0530)]
Rollup merge of #99821 - cjgillot:ast-lifetimes-2, r=compiler-errors
Remove separate indexing of early-bound regions
~Based on https://github.com/rust-lang/rust/pull/99728.~
This PR copies some modifications from https://github.com/rust-lang/rust/pull/97839 around object lifetime defaults.
These modifications allow to stop counting generic parameters during lifetime resolution, and rely on the indexing given by `rustc_typeck::collect`.
Dylan DPC [Mon, 29 Aug 2022 11:19:38 +0000 (16:49 +0530)]
Rollup merge of #98801 - joshtriplett:file-create-new, r=thomcc
Add a `File::create_new` constructor
We have `File::create` for creating a file or opening an existing file,
but the secure way to guarantee creating a new file requires a longhand
invocation via `OpenOptions`.
Add `File::create_new` to handle this case, to make it easier for people
to do secure file creation.
bors [Mon, 29 Aug 2022 07:54:06 +0000 (07:54 +0000)]
Auto merge of #100786 - sunshowers:macos-posix-chdir, r=sunshowers
Use posix_spawn for absolute paths on macOS
Currently, on macOS, Rust never uses the fast posix_spawn path if a
directory change is requested, due to a bug in Apple's libc. However, the
bug is only triggered if the program is a relative path.
This PR makes it so that the fast path continues to work if the program
is an absolute path or a lone filename.
This was an alternative proposed in https://github.com/rust-lang/rust/pull/80537#issue-776674009, and it makes a measurable performance difference in some of my code that spawns thousands of processes.
Weihang Lo [Mon, 29 Aug 2022 06:15:36 +0000 (07:15 +0100)]
Update cargo
5 commits in 6da726708a4406f31f996d813790818dce837161..4ed54cecce3ce9ab6ff058781f4c8a500ee6b8b5
2022-08-23 21:39:56 +0000 to 2022-08-27 18:41:39 +0000
- doc: pause, for readability (rust-lang/cargo#11027)
- Bump git2 to 0.15 and libgit2-sys to 0.14 (rust-lang/cargo#11004)
- Fix typo (rust-lang/cargo#11025)
- Update cargo-toml-vs-cargo-lock.md (rust-lang/cargo#11021)
- Apply GitHub fast path even for partial hashes (rust-lang/cargo#10807)
Replace `rustc_data_structures::thin_vec::ThinVec` with `thin_vec::ThinVec`.
`rustc_data_structures::thin_vec::ThinVec` looks like this:
```
pub struct ThinVec<T>(Option<Box<Vec<T>>>);
```
It's just a zero word if the vector is empty, but requires two
allocations if it is non-empty. So it's only usable in cases where the
vector is empty most of the time.
This commit removes it in favour of `thin_vec::ThinVec`, which is also
word-sized, but stores the length and capacity in the same allocation as
the elements. It's good in a wider variety of situation, e.g. in enum
variants where the vector is usually/always non-empty.
The commit also:
- Sorts some `Cargo.toml` dependency lists, to make additions easier.
- Sorts some `use` item lists, to make additions easier.
- Changes `clean_trait_ref_with_bindings` to take a
`ThinVec<TypeBinding>` rather than a `&[TypeBinding]`, because this
avoid some unnecessary allocations.
bors [Mon, 29 Aug 2022 05:12:53 +0000 (05:12 +0000)]
Auto merge of #101143 - matthiaskrgr:rollup-g8y5k0g, r=matthiaskrgr
Rollup of 9 pull requests
Successful merges:
- #94890 (Support parsing IP addresses from a byte string)
- #96334 (socket `set_mark` addition.)
- #99027 (Replace `Body::basic_blocks()` with field access)
- #100437 (Improve const mismatch `FulfillmentError`)
- #100843 (Migrate part of rustc_infer to session diagnostic)
- #100897 (extra sanity check against consts pointing to mutable memory)
- #100959 (translations: rename warn_ to warning)
- #101111 (Use the declaration's SourceInfo for FnEntry retags, not the outermost)
- #101116 ([rustdoc] Remove Attrs type alias)
Matthias Krüger [Mon, 29 Aug 2022 04:34:49 +0000 (06:34 +0200)]
Rollup merge of #101116 - GuillaumeGomez:rm-attrs-ty-alias, r=notriddle
[rustdoc] Remove Attrs type alias
When working on https://github.com/rust-lang/rust/pull/101006, I was quite confused because of this type alias as I'm used to having rustdoc types into `clean/types.rs`. Anyway, considering how few uses of it we have, I simply removed it.
Matthias Krüger [Mon, 29 Aug 2022 04:34:48 +0000 (06:34 +0200)]
Rollup merge of #101111 - saethlin:better-fnentry-spans, r=RalfJung
Use the declaration's SourceInfo for FnEntry retags, not the outermost
This addresses a long-standing `// FIXME` in the pass that adds retags.
The changes to Miri's UI tests will look like this:
```
--> $DIR/aliasing_mut1.rs:LL:CC
|
LL | pub fn safe(_x: &mut i32, _y: &mut i32) {}
< | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ not granting access to tag <TAG> because incompatible item [Unique for <TAG>] is protected by call ID
> | ^^ not granting access to tag <TAG> because incompatible item [Unique for <TAG>] is protected by call ID
|
```
Matthias Krüger [Mon, 29 Aug 2022 04:34:47 +0000 (06:34 +0200)]
Rollup merge of #100959 - LuisCardosoOliveira:translation-rename-attr-warning, r=davidtwco
translations: rename warn_ to warning
## Description
This MR renames the the macro `warn_` to `warning`.
To give a little bit of context, as [explained](https://rust-lang.zulipchat.com/#narrow/stream/336883-i18n/topic/.23100717.20diag.20translation/near/295074146) by ```````@davidtwco``````` in the Zulip channel, `warn_` was named like that because the keyword `warn` is a built-in attribute and at the time this macro was created the word `warning` was also
taken.
However, it is no longer the case and we can rename `warn_` to `warning`.
Matthias Krüger [Mon, 29 Aug 2022 04:34:46 +0000 (06:34 +0200)]
Rollup merge of #100897 - RalfJung:const-not-to-mutable, r=lcnr
extra sanity check against consts pointing to mutable memory
This should be both unreachable and redundant (since we already ensure that validation only reads from read-only memory, when validating consts), but I feel like we cannot be paranoid enough here, and also if this ever fails it'll be a nicer error than the "cannot read from mutable memory" error.
Matthias Krüger [Mon, 29 Aug 2022 04:34:43 +0000 (06:34 +0200)]
Rollup merge of #99027 - tmiasko:basic-blocks, r=oli-obk
Replace `Body::basic_blocks()` with field access
Since the refactoring in #98930, it is possible to borrow the basic blocks
independently from other parts of MIR by accessing the `basic_blocks` field
directly.
Replace unnecessary `Body::basic_blocks()` method with a direct field access,
which has an additional benefit of borrowing the basic blocks only.
Matthias Krüger [Mon, 29 Aug 2022 04:34:41 +0000 (06:34 +0200)]
Rollup merge of #94890 - marmeladema:ip-addr-try-from-bytes, r=joshtriplett
Support parsing IP addresses from a byte string
Fixes #94821
The goal is to be able to parse addresses from a byte string without requiring to do any utf8 validation. Since internally the parser already works on byte strings, this should be possible and I personally already needed this in the past too.
~~I used the proposed approach from the issue by implementing `TryFrom<&'a [u8]>` for all 6 address types (3 ip address types and 3 socket address types). I believe implementing stable traits for stable types is insta-stable so this will probably need an FCP?~~
Switched to an unstable inherent method approach called `parse_ascii` as requested.
Rain [Sat, 20 Aug 2022 01:24:21 +0000 (18:24 -0700)]
Use posix_spawn for absolute paths on macOS
Currently, on macOS, Rust never uses the fast posix_spawn path if a
directory change is requested due to a bug in Apple's libc. However, the
bug is only triggered if the program is a relative path.
This PR makes it so that the fast path continues to work if the program
is an absolute path or a lone filename.
This was an alternative proposed in
https://github.com/rust-lang/rust/pull/80537#issue-776674009, and it
makes a measurable performance difference in some of my code that spawns
thousands of processes.
bors [Sun, 28 Aug 2022 22:31:19 +0000 (22:31 +0000)]
Auto merge of #100578 - Urgau:float-next-up-down, r=scottmcm
Add next_up and next_down for f32/f64 - take 2
This is a revival of https://github.com/rust-lang/rust/pull/88728 which staled due to inactivity of the original author. I've address the last review comment.
---
This is a pull request implementing the features described at https://github.com/rust-lang/rfcs/pull/3173.
`@rustbot` label +T-libs-api -T-libs
r? `@scottmcm`
cc `@orlp`
bors [Sun, 28 Aug 2022 15:12:31 +0000 (15:12 +0000)]
Auto merge of #100201 - RalfJung:thread-local-key, r=thomcc
std: use realstd fast key when building tests
Under `cfg(test)`, the `std` crate is not the actual standard library, just any old crate we are testing. It imports the real standard library as `realstd`, and then does some careful `cfg` magic so that the crate built for testing uses the `realstd` global state rather than having its own copy of that.
However, this was not done for all global state hidden in std: the 'fast' version of thread-local keys, at least on some platforms, also involves some global state. Specifically its macOS version has this [`static REGISTERED`](https://github.com/rust-lang/rust/blob/bc63d5a26a65752fb105957d3235cc9c8cb0767f/library/std/src/sys/unix/thread_local_dtor.rs#L62) that would get duplicated. So this PR imports the 'fast' key type from `realstd` rather than using the local copy, to ensure its internal state (and that of the functions it calls) does not get duplicated.
I also noticed that the `__OsLocalKeyInner` is unused under `cfg(target_thread_local)`, so I removed it for that configuration. There was a comment saying macOS picks between `__OsLocalKeyInner` and `__FastLocalKeyInner` at runtime, but I think that comment is outdated -- I found no trace of such a runtime switching mechanism, and the library still check-builds on apple targets with this PR. (I don't have a Mac so I cannot actually run it.)
bors [Sun, 28 Aug 2022 10:59:16 +0000 (10:59 +0000)]
Auto merge of #9389 - lukaslueg:penmacro, r=llogiq
Don't lint literal `None` from expansion
This addresses https://github.com/rust-lang/rust-clippy/pull/9288#issuecomment-1229398524: If the literal `None` is from expansion, we never lint. This is correct because e.g. replacing the call to `option_env!` with whatever that macro expanded to at the time of linting is certainly wrong.
changelog: Don't lint [`partialeq_to_none`] for macro-expansions
bors [Sun, 28 Aug 2022 09:56:06 +0000 (09:56 +0000)]
Auto merge of #13133 - Veykril:diag-hack, r=Veykril
Move empty diagnostics workaround back into the server
This only touches on the diagnostics in one place instead of multiple as was previously done, since all published diagnostics will go through this code path anyways.