This adapts llvm-wrapper to use the new alternative where available, following https://rust-lang.zulipchat.com/#narrow/stream/187780-t-compiler.2Fwg-llvm/topic/LLVMConstExtractValue.20removal.
Matthias Krüger [Thu, 30 Jun 2022 17:55:51 +0000 (19:55 +0200)]
Rollup merge of #98503 - RalfJung:scope-race, r=m-ou-se
fix data race in thread::scope
Puts the `ScopeData` into an `Arc` so it sticks around as long as we need it.
This means one extra `Arc::clone` per spawned scoped thread, which I hope is fine.
`Exclusive` is a wrapper that exclusively allows mutable access to the inner value if you have exclusive access to the wrapper. It acts like a compile time mutex, and hold an unconditional `Sync` implementation.
## Justification for inclusion into std
- This wrapper unblocks actual problems:
- The example that I hit was a vector of `futures::future::BoxFuture`'s causing a central struct in a script to be non-`Sync`. To work around it, you either write really difficult code, or wrap the futures in a needless mutex.
- Easy to maintain: this struct is as simple as a wrapper can get, and its `Sync` implementation has very clear reasoning
- Fills a gap: `&/&mut` are to `RwLock` as `Exclusive` is to `Mutex`
## Public Api
```rust
// core::sync
#[derive(Default)]
struct Exclusive<T: ?Sized> { ... }
impl<T> From<T> for Exclusive<T> { ... }
impl<T: ?Sized> Debug for Exclusive { ... }
```
## Naming
This is a big bikeshed, but I felt that `Exclusive` captured its general purpose quite well.
## Stability and location
As this is so simple, it can be in `core`. I feel that it can be stabilized quite soon after it is merged, if the libs teams feels its reasonable to add. Also, I don't really know how unstable feature work in std/core's codebases, so I might need help fixing them
## Tips for review
The docs probably are the thing that needs to be reviewed! I tried my best, but I'm sure people have more experience than me writing docs for `Core`
### Implementation:
The API is mostly pulled from https://docs.rs/sync_wrapper/latest/sync_wrapper/struct.SyncWrapper.html (which is apache 2.0 licenesed), and the implementation is trivial:
- its an unsafe justification for pinning
- its an unsafe justification for the `Sync` impl (mostly reasoned about by ````@danielhenrymantilla```` here: https://github.com/Actyx/sync_wrapper/pull/2)
- and forwarding impls, starting with derivable ones and `Future`
Yiming Lei [Wed, 29 Jun 2022 16:09:34 +0000 (09:09 -0700)]
For diagnostic information of Boolean, remind it as use the type: 'bool'
It helps programmers coming from other languages
modified: compiler/rustc_resolve/src/late/diagnostics.rs
bors [Thu, 30 Jun 2022 09:20:52 +0000 (09:20 +0000)]
Auto merge of #98377 - davidv1992:add-lifetimes-to-argument-temporaries, r=oli-obk
Added llvm lifetime annotations to function call argument temporaries.
The goal of this change is to ensure that llvm will do stack slot
optimization on these temporaries. This ensures that in code like:
```rust
const A: [u8; 1024] = [0; 1024];
fn copy_const() {
f(A);
f(A);
}
```
we only use 1024 bytes of stack space, instead of 2048 bytes.
I am new to developing for the rust compiler, and as such not entirely sure, but I believe this should be sufficient to close #98156.
Also, this does not contain a test case to ensure this keeps working, primarily because I am not sure how to go about testing this. I would love some suggestions as to how that could be approached.
bors [Thu, 30 Jun 2022 03:50:35 +0000 (03:50 +0000)]
Auto merge of #98649 - RalfJung:guardians-of-mir, r=oli-obk
move MIR syntax into a dedicated file and ping some people whenever it changes
Adding or changing MIR operations/statements/whatever should be under significant scrutiny wrt their wider impact, specified semantics, and so on. So let's start by putting all that into a dedicated file and pinging some people whenever that file changes.
This PR only moves definitions around, and then fiddles with imports until it all works again.
bors [Thu, 30 Jun 2022 01:02:24 +0000 (01:02 +0000)]
Auto merge of #98691 - matthiaskrgr:rollup-ymsa64p, r=matthiaskrgr
Rollup of 6 pull requests
Successful merges:
- #96727 (Make TAIT behave exactly like RPIT)
- #98681 (rustdoc-json: Make default value of blanket impl assoc types work)
- #98682 (add tests for ICE 94432)
- #98683 (add test for ice 68875)
- #98685 (Replace `sort_modules_alphabetically` boolean with enum)
- #98687 (add test for 47814)
Matthias Krüger [Wed, 29 Jun 2022 22:23:54 +0000 (00:23 +0200)]
Rollup merge of #98685 - camelid:sorting-flag, r=GuillaumeGomez
Replace `sort_modules_alphabetically` boolean with enum
This fixes the long-standing FIXME there and makes the code easier to
understand. The reference to modules in both the old and new names seems
potentially wrong since I believe it applies to all items.
This makes type-alias-impl-trait behave like return-position-impl-trait. Unfortunately it also causes some cases to stop compiling due to "needing type annotations" and makes panicking cause fallback for the hidden type to `()`.
All of these are addressable, but we should probably address them for RPIT and TAIT together
Noah Lev [Wed, 29 Jun 2022 20:04:43 +0000 (13:04 -0700)]
Replace `sort_modules_alphabetically` boolean with enum
This fixes the long-standing FIXME there and makes the code easier to
understand. The reference to modules in both the old and new names seems
potentially wrong since I believe it applies to all items.
bors [Wed, 29 Jun 2022 18:42:19 +0000 (18:42 +0000)]
Auto merge of #98680 - matthiaskrgr:rollup-1bkrrn9, r=matthiaskrgr
Rollup of 10 pull requests
Successful merges:
- #98434 (Ensure that `static_crt` is set in the bootstrapper whenever using `cc-rs` to get a compiler command line.)
- #98636 (Triagebot: Fix mentions word wrapping.)
- #98642 (Fix #98260)
- #98643 (Improve pretty printing of valtrees for references)
- #98646 (rustdoc: fix bugs in main.js popover help and settings)
- #98647 (Update cargo)
- #98652 (`alloc`: clean and ensure `no_global_oom_handling` builds are warning-free)
- #98660 (Unbreak stage1 tests via ignore-stage1 in `proc-macro/invalid-punct-ident-1.rs`.)
- #98665 (Use verbose help for deprecation suggestion)
- #98668 (Avoid some `&str` to `String` conversions with `MultiSpan::push_span_label`)
Matthias Krüger [Wed, 29 Jun 2022 18:35:05 +0000 (20:35 +0200)]
Rollup merge of #98660 - eddyb:invalid-punct-stage1, r=lqd
Unbreak stage1 tests via ignore-stage1 in `proc-macro/invalid-punct-ident-1.rs`.
#98188 broke `./x.py test --stage 1` (which I thought we ran in PR CI, cc `@rust-lang/infra)` i.e. the default `./x.py test` in dev checkouts, as the panic in `src/test/ui/proc-macro/invalid-punct-ident-1.rs` moved from the server (`rustc`) to the client (proc macro), and that means it's now affected by #59998.
I made the test look like `src/test/ui-fulldeps/issue-76270-panic-in-libproc-macro.rs` tho I'm a bit confused why that one is in `src/test/ui-fulldeps`, it should still work in `src/test/ui`, no? (cc `@Aaron1011)`
warning: associated function `shrink` is never used
--> library/alloc/src/raw_vec.rs:424:8
|
424 | fn shrink(&mut self, cap: usize) -> Result<(), TryReserveError> {
| ^^^^^^
|
= note: `#[warn(dead_code)]` on by default
warning: associated function `forget_remaining_elements` is never used
--> library/alloc/src/vec/into_iter.rs:126:19
|
126 | pub(crate) fn forget_remaining_elements(&mut self) {
| ^^^^^^^^^^^^^^^^^^^^^^^^^
```
</details>
This PR cleans them and ensures no new ones are introduced
so that projects compiling `alloc` without infallible allocations
do not see them (and may want to enable `-Dwarnings`).
The couple `dead_code` ones may be reverted when some fallible
allocation support starts using them.
Matthias Krüger [Wed, 29 Jun 2022 18:34:59 +0000 (20:34 +0200)]
Rollup merge of #98636 - ehuss:mentions-wrapping, r=Mark-Simulacrum
Triagebot: Fix mentions word wrapping.
I forgot that GitHub's markdown treats newlines as hard breaks. This was causing some ugly-looking word wrapping in the triagebot mention messages. This fixes it so that the lines are not hard-wrapped.
Dylan DPC [Wed, 29 Jun 2022 12:29:37 +0000 (17:59 +0530)]
Rollup merge of #98625 - RalfJung:retag, r=oli-obk
emit Retag for compound types with reference fields
I want to add an option to Miri to do retagging inside reference fields. But that means we first have to even emit `Retag` for types that *contain* references (rather than being of reference types). :)
Stacked Borrows originally did that, but we stopped doing it when hitting bunch of issues in the standard library. However I have since realized that we actually do emit `noalias` for newtypes references, which means for soundness we should recurse into fields. Also it'd probably be bad news if newtypes lose out on optimizations (and they don't, for anything else). I want to add an option for that to Miri so that we can start experimenting with those semantics.
Dylan DPC [Wed, 29 Jun 2022 12:29:36 +0000 (17:59 +0530)]
Rollup merge of #98607 - compiler-errors:tuple-wrap-suggestion, r=oli-obk
Clean up arg mismatch diagnostic, generalize tuple wrap suggestion
This is based on top of #97542, so just look at the last commit which contains the relevant changes.
1. Remove `final_arg_types` which was one of the last places we were using raw (`usize`) indices instead of typed indices in the arg mismatch suggestion code.
2. Improve the tuple wrap suggestion, now we suggest things like `call(a, b, c, d)` -> `call(a, (b, c), d)` :smiley_cat:
3. Folded in fix #98645
Dylan DPC [Wed, 29 Jun 2022 12:29:33 +0000 (17:59 +0530)]
Rollup merge of #98499 - JulianKnodt:erase_lifetime, r=lcnr
Erase regions in New Abstract Consts
When an abstract const is constructed, we previously included lifetimes in the set of substitutes, so it was not able to unify two abstract consts if their lifetimes did not match but the values did, despite the values not depending on the lifetimes. This caused code that should have compiled to not compile.
Dylan DPC [Wed, 29 Jun 2022 12:29:32 +0000 (17:59 +0530)]
Rollup merge of #98479 - leocth:atomic-bool-fetch-not, r=joshtriplett
Add `fetch_not` method on `AtomicBool`
This PR adds a `fetch_not` method on `AtomicBool` performs the NOT operation on the inner value.
Internally, this just calls the `fetch_xor` method with the value `true`.
[See this IRLO discussion](https://internals.rust-lang.org/t/could-we-have-fetch-not-for-atomicbool-s/16881)
Dylan DPC [Wed, 29 Jun 2022 04:58:24 +0000 (10:28 +0530)]
Rollup merge of #98603 - compiler-errors:minor-borrowck-diagnostic-fixes, r=davidtwco
Some borrowck diagnostic fixes
1. Remove some redundant `.as_ref` suggestion logic from borrowck, this has the consequence of also not suggesting `.as_ref` after `Option` methods, but (correctly) before.
2. Fix a bug where we were replacing a binding's name with a type. Instead, make it a note.
This is somewhat incomplete. See `src/test/ui/borrowck/suggest-as-ref-on-mut-closure.rs` for more improvements.
Dylan DPC [Wed, 29 Jun 2022 04:58:23 +0000 (10:28 +0530)]
Rollup merge of #98549 - RalfJung:interpret-stacktraces, r=oli-obk
interpret: do not prune requires_caller_location stack frames quite so early
https://github.com/rust-lang/rust/pull/87000 made the interpreter skip `caller_location` frames for its stacktraces and `cur_span`. However, those functions are used for much more than just panic reporting, and e.g. when Miri reports UB somewhere, it probably wants to point inside `caller_location` frames. (And if it did not, it would want to have its own logic to decide that, not be forced into it by the core interpreter engine.) This fixes some rare ICEs in Miri that say "we should never pop more than one frame at once".
So let's remove all `caller_location` logic from the core interpreter, and instead move it to CTFE error reporting. This does not change user-visible behavior. That's the first commit.
We might additionally want to change CTFE error reporting to treat panics differently from other errors: only prune `caller_location` frames for panics. The second commit does that. But honestly I am not sure if this is an improvement.
Dylan DPC [Wed, 29 Jun 2022 04:58:21 +0000 (10:28 +0530)]
Rollup merge of #98277 - compiler-errors:issue-93596, r=estebank
Fix trait object reborrow suggestion
Fixes #93596
Slightly generalizes the logic we use to suggest fix first implemented in #95609, specifically when we have a `Sized` obligation that comes from a struct's unsized tail.
Dylan DPC [Wed, 29 Jun 2022 04:58:20 +0000 (10:28 +0530)]
Rollup merge of #97786 - ferrocene:pa-fix-simulate-remap-prefix, r=Mark-Simulacrum
Account for `-Z simulate-remapped-rust-src-base` when resolving remapped paths
Discovered in #97682, `-Z simulate-remapped-rust-src-base` only partially simulated the behavior of `remap-debuginfo = true`. While the flag successfully simulates the remapping when stdlib's `rmeta` file is loaded, the simulated prefix was not accounted for when the remapped path's local path was being discovered. This caused the flag to not fully simulate the behavior of `remap-debuginfo = true`, leading to inconsistent behaviors.
This PR fixes https://github.com/rust-lang/rust/issues/97682 by also accounting for the simulated path.
Dylan DPC [Wed, 29 Jun 2022 04:58:19 +0000 (10:28 +0530)]
Rollup merge of #97542 - compiler-errors:arg-mismatch, r=jackh726
Use typed indices in argument mismatch algorithm
I kinda went overboard with the renames, but in general, "arg" is renamed to "expected", and "input" is renamed to "provided", and we use new typed indices to make sure we're indexing into the right sized array.
Other drive-by changes:
1. Factor this logic into a new function, so we don't need to `break 'label` to escape it.
1. Factored out dependence on `final_arg_types`, which is never populated for arguments greater than the number of expected args. Instead, we just grab the final coerced expression type from `in_progress_typeck_results`.
1. Adjust the criteria we use to print (provided) type names, before we didn't suggest anything that had infer vars, but now we suggest thing that have infer vars but aren't `_`.
~Also, sorry in advance, I kinda want to backport this but I know I have folded in a lot of unnecessary drive-by changes that might discourage that. I would be open to brainstorming how to get some of these changes on beta at least.~ edit: Minimized the ICE-fixing changes to #97557
cc `@jackh726` as author of #92364, and `@estebank` as reviewer of the PR.
fixes #97484
- `seqcst` is no longer implied
- The failure ordering on chxchg is no longer implied in some cases, but now always explicitly part of the name.
- `release` is no longer shortened to just `rel`. That was especially confusing, since `relaxed` also starts with `rel`.
- `acquire` is no longer shortened to just `acq`, such that the names now all match the `std::sync::atomic::Ordering` variants exactly.
- This now allows for more combinations on the compare exchange operations, such as `atomic_cxchg_acquire_release`, which is necessary for #68464.
- This PR only exposes the new possibilities through unstable intrinsics, but not yet through the stable API. That's for [a separate PR](https://github.com/rust-lang/rust/pull/98383) that requires an FCP.
Suffixes for operations with a single memory order:
Miguel Ojeda [Wed, 29 Jun 2022 02:02:21 +0000 (04:02 +0200)]
alloc: ensure `no_global_oom_handling` builds are warning-free
Rust 1.62.0 introduced a couple new `unused_imports` warnings
in `no_global_oom_handling` builds, making a total of 5 warnings.
To avoid accumulating more over time, let's keep the builds
warning-free. This ensures projects compiling `alloc` without
infallible allocations do not see the warnings in the future
and that they can keep enabling `-Dwarnings`.
warning: associated function `shrink` is never used
--> library/alloc/src/raw_vec.rs:424:8
|
424 | fn shrink(&mut self, cap: usize) -> Result<(), TryReserveError> {
| ^^^^^^
|
= note: `#[warn(dead_code)]` on by default
warning: associated function `forget_remaining_elements` is never used
--> library/alloc/src/vec/into_iter.rs:126:19
|
126 | pub(crate) fn forget_remaining_elements(&mut self) {
| ^^^^^^^^^^^^^^^^^^^^^^^^^
```
This patch cleans them so that projects compiling `alloc` without
infallible allocations do not see the warnings. It also enables
the use of `-Dwarnings`.
The couple `dead_code` ones may be reverted when some fallible
allocation support starts using them.
Michael Howell [Tue, 28 Jun 2022 21:56:24 +0000 (14:56 -0700)]
rustdoc: fix keyboard shortcuts bug in settings menu
This commit fixes the keyboard shorts code to call localStorage every time a
key is pressed. This matters because you're supposed to be able to change a
setting and have it immediately take effect.
projects / rust.git / log