| Age | Commit message (Collapse) | Author | Lines |
|---|
|
|
|
|
|
see https://godbolt.org/z/Wx1KWezbe to see that these in fact generate equivalent assembly
|
|
|
|
|
|
rename sub_ptr to offset_from_unsigned
i also made `byte_sub_ptr` `byte_offset_from_unsigned`
fixes #137121
tracking issue #95892
|
|
Update `compiler-builtins` to 0.1.147
Removes an ABI hack that used `<2 x i64>` to return `i128` in `xmm0` on Windows [1].
[1]: https://github.com/rust-lang/compiler-builtins/pull/759
Link: https://github.com/rust-lang/rust/issues/116558
Link: https://github.com/rust-lang/compiler-builtins/issues/758
try-job: x86_64-mingw-1
try-job: x86_64-mingw-2
|
|
intrinsics: unify rint, roundeven, nearbyint in a single round_ties_even intrinsic
LLVM has three intrinsics here that all do the same thing (when used in the default FP environment). There's no reason Rust needs to copy that historically-grown mess -- let's just have one intrinsic and leave it up to the LLVM backend to decide how to lower that.
Suggested by `@hanna-kruppe` in https://github.com/rust-lang/rust/issues/136459; Cc `@tgross35`
try-job: test-various
|
|
This will be faster, and also it deduplicates the code so win/win
The dup() is actually infallible here. But whatever.
Before:
poll([{fd=0, events=0}, {fd=1, events=0}, {fd=2, events=0}], 3, 0) = 1 ([{fd=2, revents=POLLNVAL}])
openat(AT_FDCWD, "/dev/null", O_RDWR) = 2
rt_sigaction(SIGPIPE, {sa_handler=SIG_IGN, sa_mask=[PIPE], sa_flags=SA_RESTORER|SA_RESTART, sa_restorer=0x7f5749313050}, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
poll([{fd=0, events=0}, {fd=1, events=0}, {fd=2, events=0}], 3, 0) = 2 ([{fd=0, revents=POLLNVAL}, {fd=2, revents=POLLNVAL}])
openat(AT_FDCWD, "/dev/null", O_RDWR) = 0
openat(AT_FDCWD, "/dev/null", O_RDWR) = 2
rt_sigaction(SIGPIPE, {sa_handler=SIG_IGN, sa_mask=[PIPE], sa_flags=SA_RESTORER|SA_RESTART, sa_restorer=0x7efe12006050}, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
poll([{fd=0, events=0}, {fd=1, events=0}, {fd=2, events=0}], 3, 0) = 3 ([{fd=0, revents=POLLNVAL}, {fd=1, revents=POLLNVAL}, {fd=2, revents=POLLNVAL}])
openat(AT_FDCWD, "/dev/null", O_RDWR) = 0
openat(AT_FDCWD, "/dev/null", O_RDWR) = 1
openat(AT_FDCWD, "/dev/null", O_RDWR) = 2
rt_sigaction(SIGPIPE, {sa_handler=SIG_IGN, sa_mask=[PIPE], sa_flags=SA_RESTORER|SA_RESTART, sa_restorer=0x7fc2dc7ca050}, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
After:
poll([{fd=0, events=0}, {fd=1, events=0}, {fd=2, events=0}], 3, 0) = 1 ([{fd=1, revents=POLLNVAL}])
openat(AT_FDCWD, "/dev/null", O_RDWR) = 1
rt_sigaction(SIGPIPE, {sa_handler=SIG_IGN, sa_mask=[PIPE], sa_flags=SA_RESTORER|SA_RESTART, sa_restorer=0x7f488a3fb050}, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
poll([{fd=0, events=0}, {fd=1, events=0}, {fd=2, events=0}], 3, 0) = 2 ([{fd=1, revents=POLLNVAL}, {fd=2, revents=POLLNVAL}])
openat(AT_FDCWD, "/dev/null", O_RDWR) = 1
dup(1) = 2
rt_sigaction(SIGPIPE, {sa_handler=SIG_IGN, sa_mask=[PIPE], sa_flags=SA_RESTORER|SA_RESTART, sa_restorer=0x7f1a8943c050}, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
poll([{fd=0, events=0}, {fd=1, events=0}, {fd=2, events=0}], 3, 0) = 3 ([{fd=0, revents=POLLNVAL}, {fd=1, revents=POLLNVAL}, {fd=2, revents=POLLNVAL}])
openat(AT_FDCWD, "/dev/null", O_RDWR) = 0
dup(0) = 1
dup(0) = 2
rt_sigaction(SIGPIPE, {sa_handler=SIG_IGN, sa_mask=[PIPE], sa_flags=SA_RESTORER|SA_RESTART, sa_restorer=0x7f4e3a4c7050}, {sa_handler=SIG_DFL, sa_mask=[], sa_flags=0}, 8) = 0
|
|
|
|
|
|
|
|
|
|
`File::open_native`
|
|
There is a time window during which the OS can terminate a thread before stdlib
can retreive its `Packet`. Currently the `Thread::join` panics with no message
in such an event, which makes debugging difficult; fixes #124466.
|
|
Master bootstrap update
https://forge.rust-lang.org/release/process.html#master-bootstrap-update-tuesday
r? `@Mark-Simulacrum`
|
|
|
|
Since [1], the issue should be resolved so the workaround can be
dropped.
This reverts commit 88e83b96ad09f3cf9e2d1b4543a7d43f9c5a77c0.
[1]: https://github.com/rust-lang/compiler-builtins/pull/759
|
|
More const {} init in thread_local
`const {}` in `thread_local!` gets an optimization just based on the syntax, rather than the expression being const-compatible. This is easy to miss, so I've added more examples to the docs.
I've also added `const {}` in a couple of places in std where this optimization has been missed.
|
|
Replace mem::zeroed with mem::MaybeUninit::uninit for large struct in Unix
As discussion in #136737.
- Replace `mem::zeroed()` with `MaybeUninit::uninit()` for `sockaddr_storage` in `accept()` and `recvfrom()` since these functions fill in the address structure
- Replace `mem::zeroed()` with `MaybeUninit::uninit()` for `pthread_attr_t` in thread-related functions since `pthread_attr_init()` initializes the structure
- Add references to man pages to document this behavior
|
|
illumos has `posix_spawn`, and the very newest versions also have `_addchdir`,
so use that.
This is a nice ~4x performance improvement for process creation. My go-to as
usual is nextest against the clap repo, which acts as a stress test for process
creation -- with [this commit]:
```console
$ cargo nextest run -E 'not test(ui_tests) and not test(example_tests)'
before: Summary [ 1.747s] 879 tests run: 879 passed, 2 skipped
after: Summary [ 0.445s] 879 tests run: 879 passed, 2 skipped
```
[this commit]: https://github.com/clap-rs/clap/commit/fde45f9aea766fb8de46e3d46e6575f393c3b6b9
|
|
Removes an ABI hack that used `<2 x i64>` to return `i128` in `xmm0` on
Windows [1].
[1]: https://github.com/rust-lang/compiler-builtins/pull/759
Link: https://github.com/rust-lang/rust/issues/116558
Link: https://github.com/rust-lang/compiler-builtins/issues/758
|
|
r=Noratrieb
stabilize `unsigned_is_multiple_of`
tracking issue: https://github.com/rust-lang/rust/issues/128101
fcp completed in: https://github.com/rust-lang/rust/issues/128101#issuecomment-2674880635
### Public API
A version of this for all the unsigned types
```rust
fn is_multiple_of(lhs: u64, rhs: u64) -> bool {
match rhs {
// prevent division by zero
0 => lhs == 0,
_ => lhs % rhs == 0,
}
}
```
|
|
stabilize `(const_)ptr_sub_ptr`
Tracking issue: #95892
Closes #95892
FCP Completed: https://github.com/rust-lang/rust/issues/95892#issuecomment-2561139730
r? ````@Noratrieb````
|
|
Inject `compiler_builtins` during postprocessing and ensure it is made private
Follow up of https://github.com/rust-lang/rust/pull/135278
Do the following:
* Inject `compiler_builtins` during postprocessing, rather than injecting `extern crate compiler_builtins as _` into the AST
* Do not make dependencies of `std` private by default (this was added in #135278)
* Make sure sysroot crates correctly mark their dependencies private/public
* Ensure that marking a dependency private makes its dependents private by default as well, unless otherwise specified
* Do the `compiler_builtins` update that has been blocked on this
There is more detail in the commit messages. This includes the changes I was working on in https://github.com/rust-lang/rust/pull/136226.
try-job: test-various
try-job: x86_64-msvc-1
try-job: x86_64-msvc-2
try-job: i686-mingw-1
try-job: i686-mingw-2
|
|
|
|
|
|
|
|
PaulDance:disable-rename-posix-semantics-tests-under-win7, r=ChrisDenton
Fix(lib/fs/tests): Disable rename POSIX semantics FS tests under Windows 7
Would otherwise fail there. The Windows7-specific parts were left pretty much untouched by the changes introduced by
51df98ddb094b39b2e17d24f887cd66c52560ef6, so it is expected that these tests fail under Windows 7 as they were probably written to run under Windows 10+ only.
|
|
Implement feature `isolate_most_least_significant_one` for integer types
Accepted ACP - https://github.com/rust-lang/libs-team/issues/467
Tracking issue - #136909
Implement ACP for functions that isolate the most significant set bit and least significant set bit on unsigned, signed, and `NonZero` integers.
Add function `isolate_most_significant_one`
Add function `isolate_least_significant_one`
---
This PR adds the following impls
```rust
impl {u8, u16, u32, u64, u128, usize} {
const fn isolate_most_significant_one(self) -> Self;
const fn isolate_least_significant_one(self) -> Self;
}
impl {i8, i16, i32, i64, i128, isize} {
const fn isolate_most_significant_one(self) -> Self;
const fn isolate_least_significant_one(self) -> Self;
}
impl NonZeroT {
const fn isolate_most_significant_one(self) -> Self;
const fn isolate_least_significant_one(self) -> Self;
}
```
Example behavior
```rust
assert_eq!(u8::isolate_most_significant_one(0b01100100), 0b01000000);
assert_eq!(u8::isolate_least_significant_one(0b01100100), 0b00000100);
```
|
|
r=jhpratt
Add #[track_caller] to Duration Div impl
Previously the location of the divide-by-zero error condition would be attributed to the code in the rust standard library, eg:
thread 'main' panicked at /home/user/.rustup/toolchains/nightly-x86_64-unknown-linux-gnu/lib/rustlib/src/rust/library/core/src/time.rs:1172:31:
divide by zero error when dividing duration by scalar
With #[track_caller] the error is correctly attributed to the callee.
|
|
Rollup of 10 pull requests
Successful merges:
- #132876 (rustdoc book: acknowledge --document-hidden-items)
- #136148 (Optionally add type names to `TypeId`s.)
- #136609 (libcore/net: `IpAddr::as_octets()`)
- #137336 (Stabilise `os_str_display`)
- #137350 (Move methods from Map to TyCtxt, part 3.)
- #137353 (Implement `read_buf` for WASI stdin)
- #137361 (Refactor `OperandRef::extract_field` to prep for MCP838)
- #137367 (Do not exempt nonexistent platforms from platform policy)
- #137374 (Stacker now handles miri using a noop impl itself)
- #137392 (remove few unused fields)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
Would otherwise fail there. The Windows7-specific parts were left pretty
much untouched by the changes introduced by
51df98ddb094b39b2e17d24f887cd66c52560ef6, so it is expected that these
tests fail under Windows 7 as they were probably written to run under
Windows 10+ only.
Signed-off-by: Paul Mabileau <paul.mabileau@harfanglab.fr>
|
|
|
|
r=alexcrichton
Implement `read_buf` for WASI stdin
`WasiFd::read_buf` already exists. Simply use it in `Stdin`.
cc `@alexcrichton`
Tracked in https://github.com/rust-lang/rust/issues/136756
|
|
Stabilise `os_str_display`
Closes #120048.
|
|
libcore/net: `IpAddr::as_octets()`
[ACP](https://github.com/rust-lang/libs-team/issues/535)
[Tracking issue](https://github.com/rust-lang/rust/issues/137259)
Adds `const` `core::net::IpAddr{,v4,v6}::as_octets()` methods to provide reference access to IP address contents.
The concrete usecase for me is allowing the `IpAddr` to provide an extended lifetime in contexts that want a `&[u8]`:
```rust
trait AddrSlice {
fn addr_slice(&self) -> &[u8];
}
impl AddrSlice for IpAddrV4 {
fn addr_slice(&self) -> &[u8] {
// self.octets() doesn't help us here, because we can't return a reference to the owned array.
// Instead we want the IpAddrV4 to continue owning the memory:
self.as_octets()
}
}
```
(Notably, in this case we can't parameterize `AddrSlice` by a `const N: usize` (such that `fn addr_slice(&self) -> [u8; N]`) and maintain object-safety.)
|
|
Optionally add type names to `TypeId`s.
This feature is intended to provide expensive but thorough help for developers who have an unexpected `TypeId` value and need to determine what type it actually is. It causes `impl Debug for TypeId` to print the type name in addition to the opaque ID hash, and in order to do so, adds a name field to `TypeId`. The cost of this is the increased size of `TypeId` and the need to store type names in the binary; therefore, it is an optional feature. It does not expose any new public API, only change the `Debug` implementation.
It may be enabled via `cargo -Zbuild-std -Zbuild-std-features=debug_typeid`. (Note that `-Zbuild-std-features` disables default features which you may wish to reenable in addition; see
<https://doc.rust-lang.org/cargo/reference/unstable.html#build-std-features>.)
Example usage and output:
```
fn main() {
use std::any::{Any, TypeId};
dbg!(TypeId::of::<usize>(), drop::<usize>.type_id());
}
```
```
TypeId::of::<usize>() = TypeId(0x763d199bccd319899208909ed1a860c6 = usize)
drop::<usize>.type_id() = TypeId(0xe6a34bd13f8c92dd47806da07b8cca9a = core::mem::drop<usize>)
```
Also added feature declarations for the existing `debug_refcell` feature so it is usable from the `rust.std-features` option of `config.toml`.
Related issues:
* #68379
* #61533
|
|
The recent fixes to private dependencies exposed some cases in the UEFI
module where private dependencies are exposed in a public interface.
These do not need to be crate-public, so change them to `pub(crate)`.
|
|
In [1], most dependencies of `std` and other sysroot crates were marked
private, but this did not happen for `alloc` and `test`. Update these
here, marking public standard library crates as the only non-private
dependencies.
[1]: https://github.com/rust-lang/rust/pull/111076
|
|
|
|
|
|
Rollup of 8 pull requests
Successful merges:
- #128080 (Specify scope in `out_of_scope_macro_calls` lint)
- #135630 (add more `s390x` target features)
- #136089 (Reduce `Box::default` stack copies in debug mode)
- #137204 (Clarify MIR dialects and phases)
- #137299 (Simplify `Postorder` customization.)
- #137302 (Use a probe to avoid registering stray region obligations when re-checking drops in MIR typeck)
- #137305 (Tweaks in and around `rustc_middle`)
- #137313 (Some codegen_llvm cleanups)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
Fixes #83999.
|
|
|
|
|
|
|
|
Reduce `Box::default` stack copies in debug mode
The `Box::new(T::default())` implementation of `Box::default` only
had two stack copies in debug mode, compared to the current version,
which has four. By avoiding creating any `MaybeUninit<T>`'s and just writing
`T` directly to the `Box` pointer, the stack usage in debug mode remains
the same as the old version.
Another option would be to mark `Box::write` as `#[inline(always)]`,
and change it's implementation to to avoid calling `MaybeUninit::write`
(which creates a `MaybeUninit<T>` on the stack) and to use `ptr::write` instead.
Fixes: #136043
|
|
Co-authored-by: Jubilee <workingjubilee@gmail.com> and jmaargh
|
|
Remove obsolete Windows ThinLTO+TLS workaround
The bug #109797 has been fixed by #129079, so this workaround is no longer needed.
|
