summary refs log tree commit diff
path: root/library/std/src/sys/unix
AgeCommit message (Collapse)AuthorLines
2022-05-11to_timespec could be unused by some targetsSébastien Marie-0/+1
2022-05-11avoid using both Some() and ? on linux/android/freebsd codeSébastien Marie-1/+1
2022-05-11openbsd: convert futex timeout managment to Timespec usageSébastien Marie-9/+7
2022-05-09Use Rust 2021 prelude in std itself.Mara Bos-13/+0
2022-05-06Share more unix SystemTime codeJosh Stone-101/+57
2022-05-06Use statx's 64-bit times on 32-bit linux-gnuJosh Stone-38/+79
2022-05-06Use __clock_gettime64 on 32-bit linux-gnuJosh Stone-0/+27
2022-05-06unix: always use 64-bit TimespecJosh Stone-67/+65
2022-05-03Round timeouts up to infinite in futex_wait on DragonFlyBSD.Mara Bos-9/+8
2022-05-03Add #[cfg] in cfg_if for linux in unix/futex.Mara Bos-3/+5
2022-05-03Don't use futexes on netbsd.Mara Bos-43/+5
The latest NetBSD release doesn't include the futex syscall yet.
2022-04-29Disable pthread thread parker on futex platforms.Mara Bos-1/+5
2022-04-29Always return false in futex_wake on {Free,DragonFly}BSD.Mara Bos-16/+13
2022-04-29Use futex-based locks and thread parker on FreeBSD.Mara Bos-4/+59
2022-04-29Use futex-based locks and thread parker on DragonFlyBSD.Mara Bos-1/+43
2022-04-29Use futex-based locks and thread parker on NetBSD.Mara Bos-24/+52
2022-04-29Use futex-based locks and thread parker on OpenBSD.Mara Bos-1/+52
2022-04-28Auto merge of #96393 - joboet:pthread_parker, r=thomccbors-2/+268
std: directly use pthread in UNIX parker implementation `Mutex` and `Condvar` are being replaced by more efficient implementations, which need thread parking themselves (see #93740). Therefore we should use the `pthread` synchronization primitives directly. Also, we can avoid allocating the mutex and condition variable because the `Parker` struct is being placed in an `Arc` anyways. This basically is just a copy of the current `Mutex` and `Condvar` code, which will however be removed (again, see #93740). An alternative implementation could be to use dedicated private `OsMutex` and `OsCondvar` types, but all the other platforms supported by std actually have their own thread parking primitives. I used `Pin` to guarantee a stable address for the `Parker` struct, while the current implementation does not, rather using extra unsafe declaration. Since the thread struct is shared anyways, I assumed this would not add too much clutter while being clearer.
2022-04-28std: simplify UNIX parker timeoutsjoboet-59/+21
2022-04-25Remove references to git.ioEric Huss-1/+2
2022-04-25std: directly use pthread in UNIX parker implementationjoboet-0/+304
Mutex and Condvar are being replaced by more efficient implementations, which need thread parking themselves (see #93740). Therefore use the pthread synchronization primitives directly. Also, avoid allocating because the Parker struct is being placed in an Arc anyways.
2022-04-24Auto merge of #94609 - esp-rs:esp-idf-stat-type-fixes, r=Mark-Simulacrumbors-3/+18
espidf: fix stat Marking as draft as currently dependant on [a libc fix](https://github.com/rust-lang/libc/pull/2708) and release.
2022-04-21Rollup merge of #96193 - djkoloski:fuchsia_current_exe, r=tmandryDylan DPC-1/+21
[fuchsia] Add implementation for `current_exe` This implementation returns a best attempt at the current exe path. On fuchsia, fdio will always use `argv[0]` as the process name and if it is not set then an error will be returned. Because this is not guaranteed to be the case, this implementation returns an error if `argv` does not contain any elements.
2022-04-20Rollup merge of #96234 - goffrie:eloop, r=thomccDylan DPC-1/+2
remove_dir_all_recursive: treat ELOOP the same as ENOTDIR On older Linux kernels (I tested on 4.4, corresponding to Ubuntu 16.04), opening a symlink using `O_DIRECTORY | O_NOFOLLOW` returns `ELOOP` instead of `ENOTDIR`. We should handle it the same, since a symlink is still not a directory and needs to be `unlink`ed.
2022-04-20Rollup merge of #96167 - CAD97:weak-dlsym-less-ptr-crime, r=thomccDylan DPC-17/+15
Replace sys/unix/weak AtomicUsize with AtomicPtr Should fix #96163. Can't easily test on Windows though...
2022-04-20remove_dir_all_recursive: treat ELOOP the same as ENOTDIRGeoffry Song-1/+2
2022-04-19[fuchsia] Add implementation for `current_exe`David Koloski-1/+21
This implementation returns a best attempt at the current exe path. On fuchsia, fdio will always use `argv[0]` as the process name and if it is not set then an error will be returned. Because this is not guaranteed to be the case, this implementation returns an error if `argv` does not contain any elements.
2022-04-19espidf: fix statScott Mabin-3/+18
* corect type usage with new type definitions in libc
2022-04-19Use futex locks on emscripten.Mara Bos-0/+1
2022-04-19Make std::sys::unix::futex consistent on emscripten.Mara Bos-22/+25
2022-04-18Auto merge of #96042 - m-ou-se:one-reentrant-mutex, r=Amanieubors-146/+2
Use a single ReentrantMutex implementation on all platforms. This replaces all platform specific ReentrantMutex implementations by the one I added in #95727 for Linux, since that one does not depend on any platform specific details. r? `@Amanieu`
2022-04-17Replace sys/unix/weak AtomicUsize with AtomicPtrCAD97-17/+15
2022-04-17Auto merge of #93530 - anonion0:pthread_sigmask_fix, r=JohnTitorbors-4/+5
fix error handling for pthread_sigmask(3) Errors from `pthread_sigmask(3)` were handled using `cvt()`, which expects a return value of `-1` on error and uses `errno`. However, `pthread_sigmask(3)` returns `0` on success and an error number otherwise. Fix it by replacing `cvt()` with `cvt_nz()`.
2022-04-17move import to fix warning with emscripten targetRalf Sager-1/+2
2022-04-16Use a single ReentrantMutex implementation on all platforms.Mara Bos-146/+2
2022-04-15Rollup merge of #96040 - m-ou-se:futex-u32, r=AmanieuDylan DPC-42/+40
Use u32 instead of i32 for futexes. This changes futexes from i32 to u32. The [Linux man page](https://man7.org/linux/man-pages/man2/futex.2.html) uses `uint32_t` for them, so I'm not sure why I used i32 for them. Maybe because I first used them for thread parkers, where I used -1, 0, and 1 as the states. (Wasm's `memory.atomic.wait32` does use `i32`, because wasm doesn't support `u32`.) It doesn't matter much, but using the unsigned type probably results in fewer surprises when shifting bits around or using comparison operators. r? ```@Amanieu```
2022-04-14library: Remove definitions and reexports of `strlen` from libstdVadim Petrochenkov-1/+0
2022-04-14Use u32 instead of i32 for futexes.Mara Bos-42/+40
2022-04-12Allow cvt_nz to be unused on some platforms.Mara Bos-0/+1
2022-04-12Add debug asserts to futex ReentrantMutex impl.Mara Bos-0/+2
2022-04-12Initialize thread local with const{}.Mara Bos-1/+1
2022-04-12Move current_thread_unique_ptr to the only module that uses it.Mara Bos-1/+9
2022-04-12Add futex-based ReentrantMutex on Linux.Mara Bos-6/+88
2022-04-11Use is_ or has_ prefix for pure `-> bool` functions.Mara Bos-23/+25
2022-04-11Use compare_exchange_weak in futex rwlock implementation.Mara Bos-4/+11
2022-04-11Add comments to futex rwlock implementation.Mara Bos-1/+12
2022-04-11Add doc comments to futex operations.Mara Bos-0/+10
2022-04-08Fix typo in futex rwlock.Mara Bos-1/+1
Co-authored-by: Amanieu d'Antras <amanieu@gmail.com>
2022-04-08Add futex-based RwLock on Linux.Mara Bos-2/+295
2022-04-07Return status from futex_wake().Mara Bos-6/+4