The optimization only makes sense for non-drop elements anyway.
Use the default implementation for items that are Drop instead.
It also simplifies the implementation.
core: VaArgSafe is an unsafe trait
`T: VaArgSafe` is relied on for soundness. Safe impls promise nothing. Therefore this must be an unsafe trait. Slightly pedantic, as only core can impl this, but we *could* choose to unseal the trait. That would allow soundly (but unsafely) implementing this for e.g. a `#[repr(C)] struct` that should be passable by varargs.
Relates to https://github.com/rust-lang/rust/issues/44930
Detect unused structs which derived Default
<!--
If this PR is related to an unstable feature or an otherwise tracked effort,
please link to the relevant tracking issue here. If you don't know of a related
tracking issue or there are none, feel free to ignore this.
This PR will get automatically assigned to a reviewer. In case you would like
a specific user to review your work, you can assign it to them by using
r? <reviewer name>
-->
Fixes#98871
Currently, `std` performs an atomic load to get the OS key on every access to `StaticKey` even when the key is already known. This PR thus replaces `StaticKey` with the platform-specific `get` and `set` function and a new `LazyKey` type that acts as a `LazyLock<Key>`, allowing the reuse of the retreived key for multiple accesses.
This suite tests all library functions that are now available for the
types. Tests are only run on certain platforms where `f16` and `f128`
are known to work (have symbols available and don't crash LLVM).
This adds everything that was directly or transitively blocked on const
arithmetic for these types, which was recently merged.
Since const arithmetic is recent, most of these need to be gated by
`bootstrap`.
Anything that relies on intrinsics that are still missing is excluded.
The symbols that these tests rely on are not available on all platforms
and some ABIs are buggy, tests that rely on external functions are
configured to only run on x86 (`f128`) or aarch64 (`f16`).
There are some complexities about what platforms we can test f16 and
f128 on. Put this in build.rs so we have an easy way to configure tests
with a single attribute, and keep it up to date.
This was added (with a different name) to improve an error message. It
is no longer needed -- removing it changes the error message, but overall
I think the new message is no worse:
- the mention of `#` in the first line is a little worse,
- but the extra context makes it very clear what the problem is, perhaps
even clearer than the old message,
- and the removal of the note about the `expr` fragment (an internal
detail of `__rust_force_expr`) is an improvement.
Overall I think the error is quite clear and still far better than the
old message that prompted #61933, which didn't even mention patterns.
The motivation for this is #124141, which will cause pasted
metavariables to be tokenized and reparsed instead of the AST node being
cached. This change in behaviour occasionally has a non-zero perf cost,
and `__rust_force_expr` causes the tokenize/reparse step to occur twice.
Removing `__rust_force_expr` greatly reduces the extra overhead for the
`deep-vector` benchmark.
`T: VaArgSafe` is relied on for soundness. Safe impls promise nothing.
Therefore this must be an unsafe trait. Slightly pedantic, as
only core can impl this, but we could choose to unseal the trait.
That would allow soundly (but unsafely) implementing this for e.g.
a `#[repr(C)] struct` that should be passable by varargs.
SmartPointer derive-macro
<!--
If this PR is related to an unstable feature or an otherwise tracked effort,
please link to the relevant tracking issue here. If you don't know of a related
tracking issue or there are none, feel free to ignore this.
This PR will get automatically assigned to a reviewer. In case you would like
a specific user to review your work, you can assign it to them by using
r? <reviewer name>
-->
Possibly replacing #123472 for continued upkeep of the proposal rust-lang/rfcs#3621 and implementation of the tracking issue #123430.
cc `@Darksonn` `@wedsonaf`
Remove `MaybeUninit::uninit_array()` and replace it with inline const blocks.
\[This PR originally contained the changes in #125995 too. See edit history for the original PR description.]
The documentation of `MaybeUninit::uninit_array()` says:
> Note: in a future Rust version this method may become unnecessary when Rust allows [inline const expressions](https://github.com/rust-lang/rust/issues/76001). The example below could then use `let mut buf = [const { MaybeUninit::<u8>::uninit() }; 32];`.
The PR adding it also said: <https://github.com/rust-lang/rust/pull/65580#issuecomment-544200681>
> if it’s stabilized soon enough maybe it’s not worth having a standard library method that will be replaceable with `let buffer = [MaybeUninit::<T>::uninit(); $N];`
That time has come to pass — inline const expressions are stable — so `MaybeUninit::uninit_array()` is now unnecessary. The only remaining question is whether it is an important enough *convenience* to keep it around.
I believe it is net good to remove this function, on the principle that it is better to compose two orthogonal features (`MaybeUninit` and array construction) than to have a specific function for the specific combination, now that that is possible.
This is possible now that inline const blocks are stable; the idea was
even mentioned as an alternative when `uninit_array()` was added:
<https://github.com/rust-lang/rust/pull/65580#issuecomment-544200681>
> if it’s stabilized soon enough maybe it’s not worth having a
> standard library method that will be replaceable with
> `let buffer = [MaybeUninit::<T>::uninit(); $N];`
Const array repetition and inline const blocks are now stable (in the
next release), so that circumstance has come to pass, and we no longer
have reason to want `uninit_array()` other than convenience. Therefore,
let’s evaluate the inconvenience by not using `uninit_array()` in
the standard library, before potentially deleting it entirely.
std: refactor the TLS implementation
As discovered by Mara in #110897, our TLS implementation is a total mess. In the past months, I have simplified the actual macros and their expansions, but the majority of the complexity comes from the platform-specific support code needed to create keys and register destructors. In keeping with #117276, I have therefore moved all of the `thread_local_key`/`thread_local_dtor` modules to the `thread_local` module in `sys` and merged them into a new structure, so that future porters of `std` can simply mix-and-match the existing code instead of having to copy the same (bad) implementation everywhere. The new structure should become obvious when looking at `sys/thread_local/mod.rs`.
Unfortunately, the documentation changes associated with the refactoring have made this PR rather large. That said, this contains no functional changes except for two small ones:
* the key-based destructor fallback now, by virtue of sharing the implementation used by macOS and others, stores its list in a `#[thread_local]` static instead of in the key, eliminating one indirection layer and drastically simplifying its code.
* I've switched over ZKVM (tier 3) to use the same implementation as WebAssembly, as the implementation was just a way worse version of that
Please let me know if I can make this easier to review! I know these large PRs aren't optimal, but I couldn't think of any good intermediate steps.
`@rustbot` label +A-thread-locals
Update docs for AtomicBool/U8/I8 with regard to alignment
Fixes#126084.
Since `AtomicBool`/`AtomicU8`/`AtomicI8` are guaranteed to have size == 1, and Rust guarantees that `size % align == 0`, they also must have alignment equal to 1, so some current docs are contradictory/confusing when describing their alignment requirements.
Specifically:
* Fix `AtomicBool::from_ptr` claiming that `align_of::<AtomicBool>() > align_of::<bool>()` on some platforms. (same for `AtomicU8::from_ptr`/`AtomicI8::from_ptr`)
* Explicitly state that `AtomicU8`/`AtomicI8` have the same alignment as `u8`/`i8` (in addition to size and bit validity)
* (internal) Change the `if_not_8_bit` macro to be `if_8_bit` and to allow an "if-else"-like structure, instead of just "if"-like.
---
I opted to leave the "`ptr` must be aligned" wording in `from_ptr`'s docs and just clarify that it is always satsified, instead of just removing the wording entirely. If that is instead preferred I can do that.
std::unix::fs: copy simplification for apple.
since we do support from macOs Sierra, we avoid the little runtime overhead with the fclonefileat symbol check.
Rollup of 3 pull requests
Successful merges:
- #126140 (Rename `std::fs::try_exists` to `std::fs::exists` and stabilize fs_try_exists)
- #126318 (Add a `x perf` command for integrating bootstrap with `rustc-perf`)
- #126552 (Remove use of const traits (and `feature(effects)`) from stdlib)
r? `@ghost`
`@rustbot` modify labels: rollup
Remove use of const traits (and `feature(effects)`) from stdlib
The current uses are already unsound because they are using non-const impls in const contexts. We can reintroduce them by reverting the commit in this PR, after #120639 lands.
Also, make `effects` an incomplete feature.
cc `@rust-lang/project-const-traits`
r? `@compiler-errors`
Generalize `{Rc,Arc}::make_mut()` to unsized types.
* `{Rc,Arc}::make_mut()` now accept any type implementing the new unstable trait `core::clone::CloneToUninit`.
* `CloneToUninit` is implemented for `T: Clone` and for `[T] where T: Clone`.
* `CloneToUninit` is a generalization of the existing internal trait `alloc::alloc::WriteCloneIntoRaw`.
* New feature gate: `clone_to_uninit`
This allows performing `make_mut()` on `Rc<[T]>` and `Arc<[T]>`, which was not previously possible.
---
Previous PR description, now obsolete:
> Add `{Rc, Arc}::make_mut_slice()`
>
> These functions behave identically to `make_mut()`, but operate on `Arc<[T]>` instead of `Arc<T>`.
>
> This allows performing the operation on slices, which was not previously possible because `make_mut()` requires `T: Clone` (and slices, being `!Sized`, do not and currently cannot implement `Clone`).
>
> Feature gate: `make_mut_slice`
try-job: test-various
This requires introducing a new internal type `RcUninit` (and
`ArcUninit`), which can own an `RcBox<T>` without requiring it to be
initialized, sized, or a slice. This is similar to `UniqueRc`, but
`UniqueRc` doesn't support the allocator parameter, and there is no
`UniqueArc`.
This trait allows cloning DSTs, but is unsafe to implement and use
because it writes to possibly-uninitialized memory which must be of the
correct size, and must initialize that memory.
It is only implemented for `T: Clone` and `[T] where T: Clone`, but
additional implementations could be provided for specific `dyn Trait`
or custom-DST types.
Stop using `unlikely` in `strict_*` methods
The `strict_*` methods don't need (un)likely, because the `overflow_panic` calls are all `#[cold]`, [meaning](https://llvm.org/docs/LangRef.html#function-attributes) that LLVM knows any branch to them is unlikely without us needing to say so.
r? libs
Add PidFd::{kill, wait, try_wait}
#117957 changed `Child` kill/wait/try_wait to use its pidfd instead of the pid, when one is available.
This PR extracts those implementations and makes them available on `PidFd` directly.
The `PidFd` implementations differ significantly from the corresponding `Child` methods:
* the methods can be called after the child has been reaped, which will result in an error but will be safe. This state is not observable in `Child` unless something stole the zombie child
* the `ExitStatus` is not kept, meaning that only the first time a wait succeeds it will be returned
* `wait` does not close stdin
* `wait` only requires `&self` instead of `&mut self` since there is no state to maintain and subsequent calls are safe
Tracking issue: #82971
As long as a pidfd is on a child it can be safely reaped. Taking it
would mean the child would now have to be awaited through its pid, but could also
be awaited through the pidfd. This could then suffer from a recycling race.
Rollup of 8 pull requests
Successful merges:
- #126125 (Improve conflict marker recovery)
- #126481 (Add `powerpc-unknown-openbsd` maintenance status)
- #126613 (Print the tested value in int_log tests)
- #126617 (Expand `avx512_target_feature` to include VEX variants)
- #126700 (Make edition dependent `:expr` macro fragment act like the edition-dependent `:pat` fragment does)
- #126707 (Pass target to inaccessible-temp-dir rmake test)
- #126767 (`StaticForeignItem` and `StaticItem` are the same)
- #126774 (Fix another assertion failure for some Expect diagnostics.)
r? `@ghost`
`@rustbot` modify labels: rollup
Print the tested value in int_log tests
Tiny change - from the failures in https://github.com/rust-lang/rust/pull/125016, it would have been nice to see what the tested values were. Update the assertion messages.
It's unnecessary when that arm leads to a `#[cold]` panic anyway, since controlling branch likihood is what `#[cold]` is all about.
(And, well, it's unclear whether `unlikely!` even works these days anyway.)
Account for things that optimize out in inlining costs
This updates the MIR inlining `CostChecker` to have both bonuses and penalties, rather than just penalties.
That lets us add bonuses for some things where we want to encourage inlining without risking wrapping into a gigantic cost. For example, `switchInt(const …)` we give an inlining bonus because codegen will actually eliminate the branch (and associated dead blocks) once it's monomorphized, so measuring both sides of the branch gives an unrealistically-high cost to it. Similarly, an `unreachable` terminator gets a small bonus, because whatever branch leads there doesn't actually exist post-codegen.
Replace sort implementations
This PR replaces the sort implementations with tailor-made ones that strike a balance of run-time, compile-time and binary-size, yielding run-time and compile-time improvements. Regressing binary-size for `slice::sort` while improving it for `slice::sort_unstable`. All while upholding the existing soft and hard safety guarantees, and even extending the soft guarantees, detecting strict weak ordering violations with a high chance and reporting it to users via a panic.
* `slice::sort` -> driftsort [design document](https://github.com/Voultapher/sort-research-rs/blob/main/writeup/driftsort_introduction/text.md), includes detailed benchmarks and analysis.
* `slice::sort_unstable` -> ipnsort [design document](https://github.com/Voultapher/sort-research-rs/blob/main/writeup/ipnsort_introduction/text.md), includes detailed benchmarks and analysis.
#### Why should we change the sort implementations?
In the [2023 Rust survey](https://blog.rust-lang.org/2024/02/19/2023-Rust-Annual-Survey-2023-results.html#challenges), one of the questions was: "In your opinion, how should work on the following aspects of Rust be prioritized?". The second place was "Runtime performance" and the third one "Compile Times". This PR aims to improve both.
#### Why is this one big PR and not multiple?
* The current documentation gives performance recommendations for `slice::sort` and `slice::sort_unstable`. If for example only one of them were to be changed, this advice would be misleading for some Rust versions. By replacing them atomically, the advice remains largely unchanged, and users don't have to change their code.
* driftsort and ipnsort share a substantial part of their implementations.
* The implementation of `select_nth_unstable` uses internals of `slice::sort_unstable`, which makes it impractical to split changes.
---
This PR is a collaboration with `@orlp.`
Remove `feature(const_closures)` from libcore
This is an incomplete feature and apparently it has no uses in `core`. Incomplete features should generally not be used in our standard library.
Clean up some comments near `use` declarations
#125443 will reformat all `use` declarations in the repository. There are a few edge cases involving comments on `use` declarations that require care. This PR cleans up some clumsy comment cases, taking us a step closer to #125443 being able to merge.
r? ``@lqd``
Make Option::as_[mut_]slice const
These two functions can both be made `const`. I have added them to the `const_option_ext` feature, #91930. I don't believe there is anything blocking stabilization of `as_slice`, but `as_mut_slice` contains mutable references so depends on `const_mut_refs`.
Stabilise `c_unwind`
Fix#74990Fix#115285 (that's also where FCP is happening)
Marking as draft PR for now due to `compiler_builtins` issues
r? `@Amanieu`
reword the hint::blackbox non-guarantees
People were tripped up by the "precludes", interpreting it that this function must not ever be used in cryptographic contexts rather than the std lib merely making zero promises about it being fit-for-purpose.
What remains unchanged is that if someone does try to use it *despite the warnings* then it is on them to pin their compiler versions and verify the assembly of every single binary build they do.
Most modules have such a blank line, but some don't. Inserting the blank
line makes it clearer that the `//!` comments are describing the entire
module, rather than the `use` declaration(s) that immediately follows.
This makes their intent and expected location clearer. We see some
examples where these comments were not clearly separate from `use`
declarations, which made it hard to understand what the comment is
describing.
People were tripped up by the "precludes", interpreting it that this function
must not ever be used in cryptographic contexts rather than the std lib merely
making zero promises about it being fit-for-purpose.
What remains unchanged is that if someone does try to use it *despite the warnings*
then it is on them to pin their compiler versions and verify the assembly of every
single binary build they do.
Return opaque type from PanicInfo::message()
This changes the return type of the (unstable) PanicInfo::message() method to an opaque type (that implements Display). This allows for a bit more flexibility in the future.
See https://github.com/rust-lang/rust/issues/66745
Due to refactoring the const_trait usage, the CopyMarker impl was
accidentally deleted, which had the consequence that the Copy
specialization for the small-sort was never picked.
The changes made only a limited improvement for the current small
miri coverage and in general test coverage of the sort implementations.
But they exploded test times from ~13s to ~240s, which is not deemed
worth it.
Freeze + Copy types should be allowed to choose between all three
small-sort variants. With the recent changes to small-sort selection,
a regression was added that only let such types choose between network
and fallback. It can now also choose general where appropriate.
Rollup of 8 pull requests
Successful merges:
- #125258 (Resolve elided lifetimes in assoc const to static if no other lifetimes are in scope)
- #126250 (docs(change): Don't mention a Cargo 2024 edition change for 1.79)
- #126288 (doc: Added commas where needed)
- #126346 (export std::os::fd module on HermitOS)
- #126468 (div_euclid, rem_euclid: clarify/extend documentation)
- #126531 (Add codegen test for `Request::provide_*`)
- #126535 (coverage: Arrange span extraction/refinement as a series of passes)
- #126538 (coverage: Several small improvements to graph code)
r? `@ghost`
`@rustbot` modify labels: rollup
Add codegen test for `Request::provide_*`
Codegen before & after https://github.com/rust-lang/rust/pull/126242: https://gist.github.com/slanterns/3789ee36f59ed834e1a6bd4677b68ed4.
Also adjust an outdated comment since `tag_id` is no longer attached to `TaggedOption` via `Erased`, but stored next to it in `Tagged` under the new implementation.
My first time writing FileCheck xD. Correct me if there is anything that should be amended.
r? libs
export std::os::fd module on HermitOS
The HermitOS' IO interface is similiar to Unix. Consequently, this PR synchronize the FD implementation between both.
closes#126198
doc: Added commas where needed
<!--
If this PR is related to an unstable feature or an otherwise tracked effort,
please link to the relevant tracking issue here. If you don't know of a related
tracking issue or there are none, feel free to ignore this.
This PR will get automatically assigned to a reviewer. In case you would like
a specific user to review your work, you can assign it to them by using
r? <reviewer name>
-->
