This is where our Windows API bindings previously (and incorrectly) used `*mut` instead of `*const` pointers. Now that the bindings have been corrected, the mutable references (which auto-convert to `*mut`) are unnecessary and we can use shared references.
feat: specialize `SpecFromElem` for `()`
# Description
This PR adds a specialization `SpecFromElem for ()` which allows to significantly reduce `vec![(), N]` time in debug builds (specifically, tests) turning it from observable $O(n)$ to $O(1)$.
# Observing the change
The problem this PR aims to fix explicitly is slow `vec![(), N]` on big `N`s which may appear in tests (see [Background section](#Background) for more details).
The minimal example to see the problem:
```rust
#![feature(test)]
extern crate test;
#[cfg(test)]
mod tests {
const HUGE_SIZE: usize = i32::MAX as usize + 1;
#[bench]
fn bench_vec_literal(b: &mut test::Bencher) {
b.iter(|| vec![(); HUGE_SIZE]);
}
#[bench]
fn bench_vec_repeat(b: &mut test::Bencher) {
b.iter(|| [(); 1].repeat(HUGE_SIZE));
}
}
```
<details><summary>Output</summary>
<p>
```bash
cargo +nightly test -- --report-time -Zunstable-options
Compiling huge-zst-vec-literal-bench v0.1.0 (/home/progrm_jarvis/RustroverProjects/huge-zst-vec-literal-bench)
Finished test [unoptimized + debuginfo] target(s) in 0.31s
Running unittests src/lib.rs (target/debug/deps/huge_zst_vec_literal_bench-e43b1ef287ba8b36)
running 2 tests
test tests::bench_vec_repeat ... ok <0.000s>
test tests::bench_vec_literal ... ok <14.382s>
test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 14.38s
Doc-tests huge-zst-vec-literal-bench
running 0 tests
test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s
```
</p>
</details>
> [!IMPORTANT]
> This problem is only observable in Debug (unoptimized) builds, while Release (optimized) ones do not observe this problem. It still is worth fixing it, IMO, since the original scenario observes the problem in tests for which optimizations are disabled by default and it seems unreasonable to override this for the whole project while the problem is very local.
# Background
While working on a crate for a custom data format which has an `i32::MAX` limit on its list's sizes, I wrote the following test to ensure that this invariant is upheld:
```rust
#[test]
fn lists_should_have_i32_size() {
assert!(
RawNbtList::try_from(vec![(); i32::MAX as usize]).is_ok(),
"lists should permit the size up to {}",
i32::MAX
);
assert!(
RawNbtList::try_from(vec![(); i32::MAX as usize + 1]).is_err(),
"lists should have the size of at most {}",
i32::MAX
);
}
```
Soon I discovered that this takes $\approx 3--4s$ per assertion on my machine, almost all of which is spent on `vec![..]`.
While this would be logical for a non-ZST vector (which would require actual $O(n)$ allocation), here `()` was used intentionally considering that for ZSTs size-changing operations should anyway be $O(1)$ (at least from allocator perspective). Yet, this "overhead" is logical if we consider that in general case `clone()` (which is used by `Vec` literal) may have a non-trivial implementation and thus each element has to actually be visited (even if they occupy no space).
In my specific case, the following (contextual) equivalent solved the issue:
```rust
#[test]
fn lists_should_have_i32_size() {
assert!(
RawNbtList::try_from([(); 1].repeat(i32::MAX as usize)).is_ok(),
"lists should permit the size up to {}",
i32::MAX
);
assert!(
RawNbtList::try_from([(); 1].repeat(i32::MAX as usize + 1)).is_err(),
"lists should have the size of at most {}",
i32::MAX
);
}
```
which works since `repeat` explicitly uses `T: Copy` and so does not have to think about non-trivial `Clone`.
But it still may be counter-intuitive for users to observe such long time on the "canonical" vec literal thus the PR.
# Generic solution
This change is explicitly non-generic. Initially I considered it possible to implement in generically, but this would require the specialization to have the following type requirements:
- ✅ the type must be a ZST: easily done via
```rust
if core::mem::size_of::<T>() == 0 {
todo!("specialization")
}
```
or
```rust
use core::mem::SizedTypeProperties;
if T::IS_ZST {
todo!("specialization")
}
```
- :white_check_mark`: the type must implement `Copy`: implementable non-conflictable via a separate specialization:
```rust
trait IsCopyZst: Sized {
fn is_copy_zst() -> bool;
}
impl<T> IsCopyZst for T {
fn is_copy_zst() -> bool {
false
}
}
impl<T: Copy> IsCopyZst for T {
fn is_copy_zst() -> bool {
Self::IS_ZST
}
}
```
- ❌ the type should have a trivial `Clone` implementation: since `vec![t; n]` is specified to use `clone()`, we can only use this "performance optimization" when we are guaranteed that `clone()` does nothing except for copying.
The latter is the real blocker for a generic fix since I am unaware of any way to get this information in a compiler-guaranteed way.
While there may be a fix for this (my friend `@CertainLach` has suggested a potential solution by an perma-unstable fn in `Clone` like `is_trivially_cloneable()` defaulting to `false` and only overridable by `rustc` on derive), this is surely out of this PRs scope.
While a better approach would be to implement it for all ZSTs
which are `Copy` and have trivial `Clone`,
the last property cannot be detected for now.
Signed-off-by: Petr Portnov <me@progrm-jarvis.ru>
While a bare "NUL" *should* be redirected to the NUL device, especially in this simple case, let's be explicit that we aren't opening a file called "NUL" and instead open it directly.
This will also set a good example for people copying std code.
Adjust frame IP in backtraces relative to image base for SGX target
This is followup to https://github.com/rust-lang/backtrace-rs/pull/566.
The backtraces printed by `panic!` or generated by `std::backtrace::Backtrace` in SGX target are not usable. The frame addresses need to be relative to image base address so they can be used for symbol resolution. Here's an example panic backtrace generated before this change:
```
$ cargo r --target x86_64-fortanix-unknown-sgx
...
stack backtrace:
0: 0x7f8fe401d3a5 - <unknown>
1: 0x7f8fe4034780 - <unknown>
2: 0x7f8fe401c5a3 - <unknown>
3: 0x7f8fe401d1f5 - <unknown>
4: 0x7f8fe401e6f6 - <unknown>
```
Here's the same panic after this change:
```
$ cargo +stage1 r --target x86_64-fortanix-unknown-sgx
stack backtrace:
0: 0x198bf - <unknown>
1: 0x3d181 - <unknown>
2: 0x26164 - <unknown>
3: 0x19705 - <unknown>
4: 0x1ef36 - <unknown>
```
cc `@jethrogb` and `@workingjubilee`
Add Seek::seek_relative
The `BufReader` struct has a `seek_relative` method because its `Seek::seek` implementation involved dumping the internal buffer (https://github.com/rust-lang/rust/issues/31100).
Unfortunately, there isn't really a good way to take advantage of that method in generic code. This PR adds the same method to the main `Seek` trait with the straightforward default method, and an override for `BufReader` that calls its implementation.
_Also discussed in [this](https://internals.rust-lang.org/t/add-seek-seek-relative/19546) internals.rust-lang.org thread._
Remove option_payload_ptr; redundant to offset_of
The `option_payload_ptr` intrinsic is no longer required as `offset_of` supports traversing enums (#114208). This PR removes it in order to dogfood offset_of (as suggested at https://github.com/rust-lang/rust/issues/106655#issuecomment-1790907626). However, it will not build until those changes reach beta (which I think is within the next 8 days?) so I've opened it as a draft.
Expose tests for {f32,f64}.total_cmp in docs
Expose tests for {f32,f64}.total_cmp in docs
Uncomment the helpful `assert_eq!` line, which is stripped out completely in docs, and leaves the reader to mentally play through the algorithm, or go to the playground and add a println!, to see what the result will be.
(If these tests are known to fail on some platforms, is there some mechanism to conditionalize this or escape the test so the `assert_eq!` source will be visible on the web? I am a newbie, which is why I was reading docs ;)
impl more traits for ptr::Alignment, add mask method
Changes:
* Adds `rustc_const_unstable` attributes where missing
* Makes `log2` method const
* Adds `mask` method
* Implements `Default`, which is equivalent to `Alignment::MIN`
No longer included in PR:
* Removes indirection of `AlignmentEnum` type alias (this was intentional)
* Implements `Display`, `Binary`, `Octal`, `LowerHex`, and `UpperHex` (should go through libs-api instead)
* Controversially implements `LowerExp` and `UpperExp` using `p` instead of `e` to indicate a power of 2 (also should go through libs-api)
Tracking issue for `ptr::Alignment`: #102070
Reenable effects in libcore
With #116670, #117531, and #117171, I think we would be comfortable with re-enabling the effects feature for more testing in libcore.
r? `@oli-obk`
cc `@fmease`
cc #110395
Add T: ?Sized to `RwLockReadGuard` and `RwLockWriteGuard`'s Debug impls.
For context, `MutexGuard` has `+ ?Sized` on its `Debug` impl, and all three have `+ ?Sized` on their `Display` impls.
It looks like the `?Sized` was just missed when the impls were added (the impl for `MutexGuard` was added in the same PR (https://github.com/rust-lang/rust/pull/38006) with support for `T: Debug + ?Sized`, and `RwLock*Guard`s did allow `T: ?Sized` types already); the `Display` impls were added later (https://github.com/rust-lang/rust/pull/42822) with support for `T: Debug + ?Sized` types.
I think this needs a T-libs-api FCP? I'm not sure if this also needs an ACP. If so I can make one.
These are changes to (stable) trait impls on stable types so will be insta-stable.
`@rustbot` label +T-libs-api
Remove asmjs
Fulfills [MCP 668](https://github.com/rust-lang/compiler-team/issues/668).
`asmjs-unknown-emscripten` does not work as-specified, and lacks essential upstream support for generating asm.js, so it should not exist at all.
feat: implement `DoubleEndedSearcher` for `CharArray[Ref]Searcher`
This PR implements `DoubleEndedSearcher` for both `CharArraySearcher` and `CharArrayRefSearcher`. I'm not sure whether this was just overlooked or if there is a reason for it, but since it behaves exactly like `CharSliceSearcher`, I think the implementations should be appropriate.
document ABI compatibility
I don't think we have any central place where we document our ABI compatibility rules, so let's create one. The `fn()` pointer type seems like a good place since ABI questions can only become relevant when invoking a function through a function pointer.
This will likely need T-lang FCP.
avoid exhaustive i16 test in Miri
https://github.com/rust-lang/rust/pull/116301 added a test that is way too slow to be running in Miri. So let's only test a few hopefully representative cases.
Xous as an operating system is compiled with gcc-type personalities when
it comes to unwinding. This enables unwinding inside panics on Xous,
which enables Rust tests.
Signed-off-by: Sean Cross <sean@xobs.io>
The `unwinding` crate supports processing unwinding data, and is written
entirely in Rust. This allows it to be ported to new platforms more
easily than using the llvm-based `libunwind`.
While `libunwind` is very well supported on major targets, it is
difficult to use on other targets. SGX is an example of this where Rust
carries custom patches in order to enable backtrace support.
This adds an alternative for supported architectures. Rather than
providing a custom target, `unwinding` allows for a solution that is
completely written in Rust.
This adds `xous` as the first consumer, and forthcoming patches will
modify libstd to take advantage of this.
Signed-off-by: Sean Cross <sean@xobs.io>
Custom MIR: Support cleanup blocks
Cleanup blocks are declared with `bb (cleanup) = { ... }`.
`Call` and `Drop` terminators take an additional argument describing the unwind action, which is one of the following:
* `UnwindContinue()`
* `UnwindUnreachable()`
* `UnwindTerminate(reason)`, where reason is `ReasonAbi` or `ReasonInCleanup`
* `UnwindCleanup(block)`
Also support unwind resume and unwind terminate terminators:
* `UnwindResume()`
* `UnwindTerminate(reason)`
Cleanup blocks are declared with `bb (cleanup) = { ... }`.
`Call` and `Drop` terminators take an additional argument describing the
unwind action, which is one of the following:
* `UnwindContinue()`
* `UnwindUnreachable()`
* `UnwindTerminate(reason)`, where reason is `ReasonAbi` or `ReasonInCleanup`
* `UnwindCleanup(block)`
Also support unwind resume and unwind terminate terminators:
* `UnwindResume()`
* `UnwindTerminate(reason)`
Add `std:#️⃣:{DefaultHasher, RandomState}` exports (needs FCP)
This implements rust-lang/libs-team#267 to move the libstd hasher types to `std::hash` where they belong, instead of `std::collections::hash_map`.
<details><summary>The below no longer applies, but is kept for clarity.</summary>
This is a small refactor for #27242, which moves the definitions of `RandomState` and `DefaultHasher` into `std::hash`, but in a way that won't be noticed in the public API.
I've opened rust-lang/libs-team#267 as a formal ACP to move these directly into the root of `std::hash`, but for now, they're at least separated out from the collections code in a way that will make moving that around easier.
I decided to simply copy the rustdoc for `std::hash` from `core::hash` since I think it would be ideal for the two to diverge longer-term, especially if the ACP is accepted. However, I would be willing to factor them out into a common markdown document if that's preferred.
</details>
Clarify UB in `get_unchecked(_mut)`
Inspired by #116915, it was unclear to me what exactly "out-of-bounds index" means in `get_unchecked`.
One could [potentially](https://rust.godbolt.org/z/hxM764orW) interpret it that `get_unchecked` is just another way to write `offset`, but I think `get_unchecked(len)` is supposed to be UB even though `.offet(len)` is well-defined (as is `.get_unchecked(..len)`), so write that more directly in the docs.
**libs-api folks**: Can you confirm whether this is what you expect this to mean? And is the situation any different for `<*const [T]>::get_unchecked`?
patterns: reject raw pointers that are not just integers
Matching against `0 as *const i32` is fine, matching against `&42 as *const i32` is not.
This extends the existing check against function pointers and wide pointers: we now uniformly reject all these pointer types during valtree construction, and then later lint because of that. See [here](https://github.com/rust-lang/rust/pull/116930#issuecomment-1784654073) for some more explanation and context.
Also fixes https://github.com/rust-lang/rust/issues/116929.
Cc `@oli-obk` `@lcnr`
Refactor the if/else checking on cmp::Ordering variants to a
"branchless" reassignment of left and right. This change results
in fewer branches and instructions.
Removes the private type `std::sys::solid::net::FileDesc`, replacing its
only usage in `std::sys::solid::net::Socket` with `std::os::solid::io::
OwnedFd`.
Documentation cleanup for core::error::Request.
This part of the documentation currently render like this:
The new version renders like this:
Fixes:
* Add missing closing back tick.
* Remove spurious double back ticks.
* Add missing newline to render bullet point correctly.
* Fix grammar "there are methods calledrequest_ref and request_value are available" -> "there are methods calledrequest_ref and request_value".
* Change "methods" to "functions", which seems more appropriate for free functions.
document that the null pointer has the 0 address
Fixes https://github.com/rust-lang/rust/issues/116895
Will need t-lang FCP, but I think this is fairly uncontroversial -- there's probably already tons of code out there that relies on this.
Override `Waker::clone_from` to avoid cloning `Waker`s unnecessarily
This would be very useful for futures — I think it’s pretty much always what they want to do instead of `*waker = cx.waker().clone()`.
Tracking issue: https://github.com/rust-lang/rust/issues/98287
r? rust-lang/libs-api `@rustbot` label +T-libs-api -T-libs
Avoid unnecessary comparison in partition_equal
The branchy Hoare partition `partition_equal` as part of `slice::sort_unstable` has a bug that makes it perform a comparison of the last element twice.
Measuring inputs with a Zipfian distribution with characterizing exponent s == 1.0, yields a ~0.05% reduction in the total number of comparisons performed.
Rollup of 4 pull requests
Successful merges:
- #116017 (Don't pass `-stdlib=libc++` when building C files on macOS)
- #117524 (bootstrap/setup: create hooks directory if non-existing)
- #117588 (Remove unused LoadResult::DecodeIncrCache variant)
- #117596 (Add diagnostic items for a few of core's builtin macros)
r? `@ghost`
`@rustbot` modify labels: rollup
Add diagnostic items for a few of core's builtin macros
Specifically, `env`, `option_env`, and `include`. There are a number of reasons why people might want to look at these in lints (For example, to ensure that things behave consistently, detect things that might make builds less reproducible, etc).
Concretely, in PL/Rust (well, `plrustc`) we have lints that forbid these (which I'd like to [add to clippy as restriction lints](https://rust-lang.zulipchat.com/#narrow/stream/257328-clippy/topic/Landing.20a.20flotilla.20of.20lints.3F) eventually), and `dylint` also has [lints that look for `env!`/`option_env!`](109a07e9f2/examples/general/env_cargo_path/src/lib.rs) (although perhaps not `include`), which would benefit from this.
My experience is that it's pretty annoying to (robustly) check uses of builtin macros without these IME, although that's perhaps just my own fault (e.g. I could be doing it wrong).
At `@Nilstrieb's` suggestion, I've added a comment that explains why these are here, even though they are not used in the compiler. This is mostly to discourage removal, although it's not a big deal if it happens (I'm certainly not suggesting the presence of these be in any way stable).
---
In theory this is a library PR (in that it's in library/core), but I'm going to roll compiler because the existence of this or not is much more likely something they care about rather than libs. Hopefully nobody objects to this.
r? compiler
Remove obsolete support for linking unwinder on Android
Linking libgcc is no longer supported (see #103673), so remove the related link attributes and the check in unwind's build.rs. The check was the last remaining significant piece of logic in build.rs, so remove build.rs as well.
Stabilize `const_maybe_uninit_zeroed` and `const_mem_zeroed`
Make `MaybeUninit::zeroed` and `mem::zeroed` const stable. Newly stable API:
```rust
// core::mem
pub const unsafe fn zeroed<T>() ->;
impl<T> MaybeUninit<T> {
pub const fn zeroed() -> MaybeUninit<T>;
}
```
This relies on features based around `const_mut_refs`. Per `@RalfJung,` this should be OK since we do not leak any `&mut` to the user.
For this to be possible, intrinsics `assert_zero_valid` and `assert_mem_uninitialized_valid` were made const stable.
Tracking issue: #91850
Zulip discussion: https://rust-lang.zulipchat.com/#narrow/stream/146212-t-compiler.2Fconst-eval/topic/.60const_mut_refs.60.20dependents
r? libs-api
`@rustbot` label -T-libs +T-libs-api +A-const-eval
cc `@RalfJung` `@oli-obk` `@rust-lang/wg-const-eval`
Hint optimizer about try-reserved capacity
This is #116568, but limited only to the less-common `try_reserve` functions to reduce bloat in debug binaries from debug info, while still addressing the main use-case #116570
Deref docs: expand and remove "smart pointer" qualifier
**Ready for review**
~~This is an unpolished draft to be sanity-checked~~
Fixes#91004
~~Comments on substance and content of this are welcome. This is deliberately unpolished until ready to review so please try to stay focused on the big-picture.~~
~~Once this has been sanity checked, I will similarly update `DerefMut` and polish for review.~~
Make `core::mem::zeroed` const stable. Newly stable API:
// core::mem
pub const unsafe fn zeroed<T>() -> T;
This is stabilized with `const_maybe_uninit_zeroed` since it is a simple
wrapper.
In order to make this possible, intrinsics `assert_zero_valid` was made
const stable under `const_assert_type2`.
`assert_mem_uninitialized_valid` was also made const stable since it is
under the same gate.
Make `MaybeUninit::zeroed` const stable. Newly stable API:
// core::mem
impl<T> MaybeUninit<T> {
pub const fn zeroed() -> MaybeUninit<T>;
}
Use of `const_mut_refs` should be acceptable since we do not leak the
mutability.
Tracking issue: #91850
Re-draft Deref docs
Make general advice more explicit and note the difference between
generic and specific implementations.
Re-draft DerefMut docs in-line with Deref
Fix Deref docs typos
Fix broken links
Clarify advice for specific-over-generic impls
Add comment addressing Issue #73682
x fmt
Copy faillibility warning to DerefMut
The initial probe-for-empty-source by stack_buffer_copy only detected EOF
if the source was empty but not when it was merely small which lead to
additional calls to read() after Ok(0) had already been returned
in the stack copy routine
It now keeps track of initialized bytes to avoid reinitialization.
It also keeps track of read sizes to avoid initializing more bytes
than the reader needs. This is important when passing a huge vector to a
Read that only has a few bytes to offer and doesn't implement read_buf().
Remove support for compiler plugins.
They've been deprecated for four years.
This commit includes the following changes.
- It eliminates the `rustc_plugin_impl` crate.
- It changes the language used for lints in `compiler/rustc_driver_impl/src/lib.rs` and `compiler/rustc_lint/src/context.rs`. External lints are now called "loaded" lints, rather than "plugins" to avoid confusion with the old plugins. This only has a tiny effect on the output of `-W help`.
- E0457 and E0498 are no longer used.
- E0463 is narrowed, now only relating to unfound crates, not plugins.
- The `plugin` feature was moved from "active" to "removed".
- It removes the entire plugins chapter from the unstable book.
- It removes quite a few tests, mostly all of those in `tests/ui-fulldeps/plugin/`.
Closes#29597.
r? `@ghost`
They've been deprecated for four years.
This commit includes the following changes.
- It eliminates the `rustc_plugin_impl` crate.
- It changes the language used for lints in
`compiler/rustc_driver_impl/src/lib.rs` and
`compiler/rustc_lint/src/context.rs`. External lints are now called
"loaded" lints, rather than "plugins" to avoid confusion with the old
plugins. This only has a tiny effect on the output of `-W help`.
- E0457 and E0498 are no longer used.
- E0463 is narrowed, now only relating to unfound crates, not plugins.
- The `plugin` feature was moved from "active" to "removed".
- It removes the entire plugins chapter from the unstable book.
- It removes quite a few tests, mostly all of those in
`tests/ui-fulldeps/plugin/`.
Closes#29597.
Guarantee representation of None in NPO
This allows users to soundly transmute zeroes into `Option` types subject to the null pointer optimization (NPO). It unblocks https://github.com/google/zerocopy/issues/293.
Add track_caller to transmute_copy
Currently if `size_of::<Src>() < size_of::<Dst>()` you will see the following error:
```rust
thread 'test' panicked at /rustc/cc66ad468955717ab92600c770da8c1601a4ff33/library/core/src/mem/mod.rs:1056:5:
cannot transmute_copy if Dst is larger than Src
```
This fixes it so it will show the invocation location
Linking libgcc is no longer supported (see #103673), so remove the
related link attributes and the check in unwind's build.rs. The check
was the last remaining significant piece of logic in build.rs, so
remove build.rs as well.
Clarify `Unsize` documentation
The documentation erroneously says that:
```rust
/// - Types implementing a trait `Trait` also implement `Unsize<dyn Trait>`.
/// - Structs `Foo<..., T, ...>` implement `Unsize<Foo<..., U, ...>>` if all of these conditions
/// are met:
/// - `T: Unsize<U>`.
/// - Only the last field of `Foo` has a type involving `T`.
/// - `Bar<T>: Unsize<Bar<U>>`, where `Bar<T>` stands for the actual type of that last field.
```
Specifically, `T: Unsize<U>` is not required to hold -- only the final field must implement `FinalField<T>: Unsize<FinalField<U>>`. This can be demonstrated by the test I added.
---
Second commit fleshes out the documentation a lot more.
Fix switch_stdout_to on Windows7
The `switch_stdout_to` test was broken on Windows7, as deleting the temporary test folder would fail since the `switch-stdout-output` file we redirected the stdout to is never closed, and it's impossible on Win7 to delete an opened file.
To fix this issue, we make `switch_stdout_to` return the previous handle. Using this, we add a new `switch_stdout_to` call at the end of the test to return the stdio handles to their original state, and recover the handle to the file we opened. This handle is automatically closed at the end of the function, which should allow the temporary test folder to be deleted properly.
Support enum variants in offset_of!
This MR implements support for navigating through enum variants in `offset_of!`, placing the enum variant name in the second argument to `offset_of!`. The RFC placed it in the first argument, but I think it interacts better with nested field access in the second, as you can then write things like
```rust
offset_of!(Type, field.Variant.field)
```
Alternatively, a syntactic distinction could be made between variants and fields (e.g. `field::Variant.field`) but I'm not convinced this would be helpful.
[RFC 3308 # Enum Support](https://rust-lang.github.io/rfcs/3308-offset_of.html#enum-support-offset_ofsomeenumstructvariant-field_on_variant)
Tracking Issue #106655.
Clean up unchecked_math, separate out unchecked_shifts
Tracking issue: #85122
Changes:
1. Remove `const_inherent_unchecked_arith` flag and make const-stability flags the same as the method feature flags. Given the number of other unsafe const fns already stabilised, it makes sense to just stabilise these in const context when they're stabilised.
2. Move `unchecked_shl` and `unchecked_shr` into a separate `unchecked_shifts` flag, since the semantics for them are unclear and they'll likely be stabilised separately as a result.
3. Add an `unchecked_neg` method exclusively to signed integers, under the `unchecked_neg` flag. This is because it's a new API and probably needs some time to marinate before it's stabilised, and while it *would* make sense to have a similar version for unsigned integers since `checked_neg` also exists for those there is absolutely no case where that would be a good idea, IMQHO.
The longer-term goal here is to prepare the `unchecked_math` methods for an FCP and stabilisation since they've existed for a while, their semantics are clear, and people seem in favour of stabilising them.
Rollup of 5 pull requests
Successful merges:
- #113241 (rustdoc: Document lack of object safety on affected traits)
- #117388 (Turn const_caller_location from a query to a hook)
- #117417 (Add a stable MIR visitor)
- #117439 (prepopulate opaque ty storage before using it)
- #117451 (Add support for pre-unix-epoch file dates on Apple platforms (#108277))
r? `@ghost`
`@rustbot` modify labels: rollup
Time in UNIX system calls counts from the epoch, 1970-01-01. The timespec
struct used in various system calls represents this as a number of seconds and
a number of nanoseconds. Nanoseconds are required to be between 0 and
999_999_999, because the portion outside that range should be represented in
the seconds field; if nanoseconds were larger than 999_999_999, the seconds
field should go up instead.
Suppose you ask for the time 1969-12-31, what time is that? On UNIX systems
that support times before the epoch, that's seconds=-86400, one day before the
epoch. But now, suppose you ask for the time 1969-12-31 23:59:00.1. In other
words, a tenth of a second after one minute before the epoch. On most UNIX
systems, that's represented as seconds=-60, nanoseconds=100_000_000. The macOS
bug is that it returns seconds=-59, nanoseconds=-900_000_000.
While that's in some sense an accurate description of the time (59.9 seconds
before the epoch), that violates the invariant of the timespec data structure:
nanoseconds must be between 0 and 999999999. This causes this assertion in the
Rust standard library.
So, on macOS, if we get a Timespec value with seconds less than or equal to
zero, and nanoseconds between -999_999_999 and -1 (inclusive), we can add
1_000_000_000 to the nanoseconds and subtract 1 from the seconds, and then
convert. The resulting timespec value is still accepted by macOS, and when fed
back into the OS, produces the same results. (If you set a file's mtime with
that timestamp, then read it back, you get back the one with negative
nanoseconds again.)
Co-authored-by: Josh Triplett <josh@joshtriplett.org>
The switch_stdout_to test was broken on Windows7, as the test
infrastructure would refuse to delete the temporary test folder because
the switch-stdout-output file we redirected the stdout to was still
opened.
To fix this issue, we make switch_stdout_to return the previous handle,
and add a new switch_stdout_to call at the end of the test to return the
stdio handles to their original state. The handle the second
switch_stdout_to returns will be automatically closed, which should
allow the temporary test folder to be deleted properly.
Allows `#[diagnostic::on_unimplemented]` attributes to have multiple
notes
This commit extends the `#[diagnostic::on_unimplemented]` (and `#[rustc_on_unimplemented]`) attributes to allow multiple `note` options. This enables emitting multiple notes for custom error messages. For now I've opted to not change any of the existing usages of `#[rustc_on_unimplemented]` and just updated the relevant compile tests.
r? `@compiler-errors`
I'm happy to adjust any of the existing changed location to emit the old error message if that's desired.
Increase the reach of panic_immediate_abort
I wanted to use/abuse this recently as part of another project, and I was surprised how many panic-related things were left in my binaries if I built a large crate with the feature enabled along with LTO. These changes get all the panic-related symbols that I could find out of my set of locally installed Rust utilities.
memcpy assumptions: link to source showing that GCC makes the same assumption
I finally stumbled upon a source showing that GCC also generates overlapping `memcpy`. So if we're linking major C compilers making such assumptions here, let's have both clang and GCC.
Don't use LFS64 symbols on musl
Supersedes #106246
~~Note to packagers: If your distro's musl package has already been updated, then you won't be able to build a newer version of rust until a new rust release is made with these changes merged (which can be used to bootstrap). I'm using a super hacky method to bypass this by creating a stub library with LFS64 symbols and building a patched rust, so the symbols satisfy the build requirements while the final compiler build has no references to LFS64 symbols, example: https://codeberg.org/kiss-community/repo/pulls/160/files~~ Doesn't seem to be necessary with new rustup nightly builds, likely due to updates to vendored crates
cc ```@alyssais```
Bump stdarch submodule and remove special handling for LLVM intrinsics that are no longer needed
Bumps stdarch to pull https://github.com/rust-lang/stdarch/pull/1477, which reimplemented some functions with portable SIMD intrinsics instead of arch specific LLVM intrinsics.
Handling of those LLVM intrinsics is removed from cranelift codegen and miri.
cc `@RalfJung` `@bjorn3`
Create `windows/api.rs` for safer FFI
FFI is inherently unsafe. For memory safety we need to assert that some contract is being upheld on both sides of the FFI, though of course we can only ever check our side. In Rust, `unsafe` blocks are used to assert safety and `// SAFETY` comments describing why it is safe. Currently in sys/windows we have a lot of this unsafety spread all over the place, with variations on the same unsafe patterns repeated. And because of the repitition and frequency, we're a bit lax with the safety comments.
This PR aims to fix this and to make FFI safety more auditable by creating an `api` module with the goal of centralising and consolidating this unsafety. It contains thin wrappers around the Windows API that make most functions safe to call or, if that's not possible, then at least safer. Note that its goal is *only* to address safety. It does not stray far from the Windows API and intentionally does not attempt to make higher lever wrappers around, for example, file handles. This is better left to the existing modules. The windows/api.rs file has a top level comment to help future contributors understand the intent of the module and the design decisions made.
I chose two functions as a first tentative step towards the above goal:
- [`GetLastError`](https://learn.microsoft.com/en-us/windows/win32/api/errhandlingapi/nf-errhandlingapi-getlasterror) is trivially safe. There's no reason to wrap it in an `unsafe` block every time. So I simply created a safe `get_last_error` wrapper.
- [`SetFileInformationByHandle`](https://learn.microsoft.com/en-us/windows/win32/api/fileapi/nf-fileapi-setfileinformationbyhandle) is more complex. It essentially takes a generic type but over a C API which necessitates some amount of ceremony. Rather than implementing similar unsafe patterns in multiple places, I provide a safe `set_file_information_by_handle` that takes a Rusty generic type and handles converting that to the form required by the C FFI.
r? libs
Hide internal methods from documentation
The two methods here are perma-unstable and only made public for technical reasons. There is no reason to show them in documentation.
`@rustbot` label +A-docs
notes
This commit extends the `#[diagnostic::on_unimplemented]` (and
`#[rustc_on_unimplemented]`) attributes to allow multiple `note`
options. This enables emitting multiple notes for custom error messages.
For now I've opted to not change any of the existing usages of
`#[rustc_on_unimplemented]` and just updated the relevant compile tests.
Rollup of 6 pull requests
Successful merges:
- #114998 (feat(docs): add cargo-pgo to PGO documentation 📝)
- #116868 (Tweak suggestion span for outer attr and point at item following invalid inner attr)
- #117240 (Fix documentation typo in std::iter::Iterator::collect_into)
- #117241 (Stash and cancel cycle errors for auto trait leakage in opaques)
- #117262 (Create a new ConstantKind variant (ZeroSized) for StableMIR)
- #117266 (replace transmute by raw pointer cast)
r? `@ghost`
`@rustbot` modify labels: rollup
Refactor some `char`, `u8` ASCII functions to be branchless
Extract conditions in singular `matches!` with or-patterns to individual `matches!` statements which enables branchless code output. The following functions were changed:
- `is_ascii_alphanumeric`
- `is_ascii_hexdigit`
- `is_ascii_punctuation`
Added codegen tests
---
Continued from https://github.com/rust-lang/rust/pull/103024.
Based on the comment from `@scottmcm` https://github.com/rust-lang/rust/pull/103024#pullrequestreview-1248697206.
The unmodified `is_ascii_*` functions didn't seem to benefit from extracting the conditions.
I've never written a codegen test before, but I tried to check that no branches were emitted.
Decompose singular `matches!` with or-patterns to individual `matches!`
statements to enable branchless code output. The following functions
were changed:
- `is_ascii_alphanumeric`
- `is_ascii_hexdigit`
- `is_ascii_punctuation`
Add codegen tests
Co-authored-by: George Bateman <george.bateman16@gmail.com>
Co-authored-by: scottmcm <scottmcm@users.noreply.github.com>
Chris Denton
coekjan