| Age | Commit message (Collapse) | Author | Lines |
|---|
|
|
|
Make sure the implementation of TcpStream::as_raw_fd is fully inlined
Currently the following function:
```rust
use std::os::fd::{AsRawFd, RawFd};
use std::net::TcpStream;
pub fn as_raw_fd(socket: &TcpStream) -> RawFd {
socket.as_raw_fd()
}
```
Is optimized to the following:
```asm
example::as_raw_fd:
push rax
call qword ptr [rip + <std::net::tcp::TcpStream as std::sys_common::AsInner<std::sys_common::net::TcpStream>>::as_inner@GOTPCREL]
mov rdi, rax
call qword ptr [rip + std::sys_common::net::TcpStream::socket@GOTPCREL]
mov rdi, rax
pop rax
jmp qword ptr [rip + _ZN73_$LT$std..sys..unix..net..Socket$u20$as$u20$std..os..fd..raw..AsRawFd$GT$9as_raw_fd17h633bcf7e481df8bbE@GOTPCREL]
```
I think it would make more sense to inline trivial functions used within `TcpStream::AsRawFd`.
|
|
update wasi_clock_time_api ref.
Closes #110809
>Preview0 corresponded to the import module name wasi_unstable. It was also called snapshot_0 in some places. It was short-lived, and the changes to preview1 were minor, so the focus here is on preview1.
we use the `preview1` doc according to the above quote form [WASI legacy Readme](https://github.com/WebAssembly/WASI/blob/main/legacy/README.md) .
|
|
Add 64-bit `time_t` support on 32-bit glibc Linux to `set_times`
Add support to `set_times` for 64-bit `time_t` on 32-bit glibc Linux platforms which have a 32-bit `time_t`. Split from #109773.
Tracking issue: #98245
|
|
std docs: edit `PathBuf::set_file_name` example
To make explicit that `set_file_name` might replace or remove the
extension, not just the file stem.
Also edit docs for `Path::with_file_name`, which calls `set_file_name`.
|
|
|
|
|
|
|
|
Signed-off-by: clundro <859287553@qq.com>
|
|
|
|
|
|
|
|
Rollup of 6 pull requests
Successful merges:
- #110766 (More core::fmt::rt cleanup.)
- #110873 (Migrate trivially translatable `rustc_parse` diagnostics)
- #110904 (rustdoc: rebind bound vars to type-outlives predicates)
- #110913 (Add some missing built-in lints)
- #110918 (`remove_dir_all`: try deleting the directory even if `FILE_LIST_DIRECTORY` access is denied)
- #110920 (Fix unavailable url)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
Remove unused std::sys_common::thread_local_key::Key
Part of https://github.com/rust-lang/rust/issues/110897
This `Key` type seems unused. Let's remove it and see if anything explodes. :)
|
|
Document `const {}` syntax for `std::thread_local`.
It exists and is pretty cool. More people should use it.
It was added in #83416 and stabilized in #91355 with the tracking issue #84223.
|
|
If opening a directory with `FILE_LIST_DIRECTORY` access fails then we should try opening without requesting that access. We may still be able to delete it if it's empty or a link.
|
|
Change memory ordering in System wrapper example
Currently, the `SeqCst` ordering is used, which seems unnecessary:
+ Even `Relaxed` ordering guarantees that all updates are atomic and are executed in total order
+ User code only reads atomic for monitoring purposes, no "happens-before" relationships with actual allocations and deallocations are needed for this
If argumentation above is correct, I propose changing ordering to `Relaxed` to clarify that no synchronization is required here, and improve performance (if somebody copy-pastes this example into their code).
|
|
Correct `std::prelude` comment
(Read the changed file first for context.)
First, `alloc` has no prelude.
Second, the docs for `v1` don't matter since the [prelude module] already has all the doc links. The `rust_2021` module for instance also doesnt have a convenient doc page. However as I understand glob imports still cant be used because the items dont have the same stabilisation versions.
[prelude module]: https://doc.rust-lang.org/std/prelude/index.html
|
|
docs(std): clarify remove_dir_all errors
When using `remove_dir_all`, I assumed that the function was idempotent and that I could always call it to remove a directory if it existed. That's not the case and it bit me in production, so I figured I'd submit this to clarify the docs.
|
|
It exists and is pretty cool. More people should use it.
|
|
|
|
|
|
|
|
I think it was left there by mistake after previous refactoring.
Signed-off-by: Ayush Singh <ayushsingh1325@gmail.com>
|
|
|
|
Add definitions for riscv64gc-unknown-fuchsia
To compile, also requires a libc update with https://github.com/rust-lang/libc/pull/3204
|
|
|
|
Restructure and rename std thread_local internals to make it less of a maze
Every time I try to work on std's thread local internals, it feels like I'm trying to navigate a confusing maze made of macros, deeply nested modules, and types with multiple names/aliases. Time to clean it up a bit.
This PR:
- Exports `Key` with its own name (`Key`), instead of `__LocalKeyInner`
- Uses `pub macro` to put `__thread_local_inner` into a (unstable, hidden) module, removing `#[macro_export]`, removing it from the crate root.
- Removes the `__` from `__thread_local_inner`.
- Removes a few unnecessary `allow_internal_unstable` features from the macros
- Removes the `libstd_thread_internals` feature. (Merged with `thread_local_internals`.)
- And removes it from the unstable book
- Gets rid of the deeply nested modules for the `Key` definitions (`mod fast` / `mod os` / `mod statik`).
- Turns a `#[cfg]` mess into a single `cfg_if`, now that there's no `#[macro_export]` anymore that breaks with `cfg_if`.
- Simplifies the `cfg_if` conditions to not repeat the conditions.
- Removes useless `normalize-stderr-test`, which were left over from when the `Key` types had different names on different platforms.
- Removes a seemingly unnecessary `realstd` re-export on `cfg(test)`.
This PR changes nothing about the thread local implementation. That's for a later PR. (Which should hopefully be easier once all this stuff is a bit cleaned up.)
|
|
|
|
Spelling library
Split per https://github.com/rust-lang/rust/pull/110392
I can squash once people are happy w/ the changes. It's really uncommon for large sets of changes to be perfectly acceptable w/o at least some changes.
I probably won't have time to respond until tomorrow or the next day
|
|
Fix `std` compilation error for wasi+atomics
Fix https://github.com/rust-lang/rust/issues/109727
It seems that the `unsupported/once.rs` module isn't meant to exist at the same time as the `futex` module, as they have conflicting definitions.
I've solved this by defining the `once` module only if `not(target_feature = "atomics")`.
The `wasm32-unknown-unknown` target [similarly only defines the `once` module if `not(target_feature = "atomics")`](https://github.com/tomaka/rust/blob/01c4f319276da912dd2be768ae0ce9857ad6bb63/library/std/src/sys/wasm/mod.rs#L69-L70).
As show in [this block of code](https://github.com/tomaka/rust/blob/01c4f319276da912dd2be768ae0ce9857ad6bb63/library/std/src/sys_common/once/mod.rs#L10-L34), the `sys::once` module doesn't need to exist if `all(target_arch = "wasm32", target_feature = "atomics")`.
|
|
Update documentation wording on path 'try_exists' functions
Just eliminate the quadruple negation in `doesn't silently ignore errors unrelated to ... not existing.`
|
|
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
|
|
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
|
|
* advance
* aligned
* borrowed
* calculate
* debugable
* debuggable
* declarations
* desugaring
* documentation
* enclave
* ignorable
* initialized
* iterator
* kaboom
* monomorphization
* nonexistent
* optimizer
* panicking
* process
* reentrant
* rustonomicon
* the
* uninitialized
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
|
|
|
|
Use libc::cpuset_getaffinity to determine the CPUs available to the current process.
The existing sysconf and sysctl paths are left as fallback.
|
|
Revert panic oom
This temporarily reverts https://github.com/rust-lang/rust/pull/109507 until https://github.com/rust-lang/rust/issues/110771 is addressed
r? `@Amanieu`
|
|
Add LazyCell::into_inner
This enables uses cases that need to extract the evaluated value and do something owned with it.
|
|
This reverts commit c9a6e41026d7aa27d897fb83e995447719753076.
|
|
This reverts commit abc0660118cc95f47445fd33502a11dd448f5968.
|
|
Limit read size in `File::read_to_end` loop
Fixes #110650.
Windows file reads have perf overhead that's proportional to the buffer size. When we have a reasonable expectation that we know the file size, we can set a reasonable upper bound for the size of the buffer in one read call.
|
|
Report allocation errors as panics
OOM is now reported as a panic but with a custom payload type (`AllocErrorPanicPayload`) which holds the layout that was passed to `handle_alloc_error`.
This should be review one commit at a time:
- The first commit adds `AllocErrorPanicPayload` and changes allocation errors to always be reported as panics.
- The second commit removes `#[alloc_error_handler]` and the `alloc_error_hook` API.
ACP: https://github.com/rust-lang/libs-team/issues/192
Closes #51540
Closes #51245
|
|
This works around performance issues on Windows by limiting reads the size of reads when the expected size is known.
|
|
More `mem::take` in `library`
A bunch of places were using `replace(…, &mut [])`, but that can just be `take`.
|
|
A bunch of places were using `replace(…, &mut [])`, but that can just be `take`.
|
|
`&[u8]`
|
|
|
|
|
|
Support AIX-style archive type
Reading facility of AIX big archive has been supported by `object` since 0.30.0.
Writing facility of AIX big archive has already been supported by `ar_archive_writer`, but we need to bump the version to support the new archive type enum.
|
