ASCII methods on OsStr
Would close#69566
I don't know enough about encodings to know if this is a valid change, however the comment on the issue suggests it could be.
This does two things:
1. Makes ASCII methods available on OsStr
2. Makes it possible to obtain a `&mut OsStr`. This is necessary to actually use `OsStr::make_ascii_*case` methods since they modify the underlying value. As far as I can tell, the only way to modify a `&mut OsStr` is via the methods I just added.
My original hope was to have these methods on `OsStrExt` for Windows, since the standard library already assumes `make_ascii_uppercase` is valid in Windows (see the change I made to windows/process.rs). If it is found these are not valid changes on non-Windows platforms, I can move the methods to the ext trait instead.
Mozilla's IRC service was shut down in March 2020. The official
instant messaging variant has been Discord for a while, and most of
the links were already replaced by #61524.
This was the last line that came up with `irc.mozilla.org` or any
combination of "irc.*#[a-z]+" in a `git grep`:
git grep -i -E "irc.*#[a-z]+"
As there is only one other link directly to Rust's discord, I used the
same Markdown link `[rust-discord]` as in `bootstrap/README.md` to
stay consistent. This might come in handy if the chat platform changes
at a later point again.
As an aside: for those interested in the use of IRC, Mozilla's [wiki]
still offers a lot of in-depth knowledge.
[wiki]: https://wiki.mozilla.org/IRC
Rename asm! to llvm_asm!
As per https://github.com/rust-lang/rfcs/pull/2843, this PR renames `asm!` to `llvm_asm!`. It also renames the compiler's internal `InlineAsm` data structures to `LlvmInlineAsm` in preparation for the new `asm!` functionality specified in https://github.com/rust-lang/rfcs/pull/2850.
This PR doesn't actually deprecate `asm!` yet, it just makes it redirect to `llvm_asm!`. This is necessary because we first need to update the submodules (in particular stdarch) to use `llvm_asm!`.
Allow obtaining &mut OsStr
```rust
impl DerefMut for OsString {...} // type Target = OsStr
impl IndexMut<RangeFull> for OsString {...} // type Output = OsStr
```
---
This change is pulled out of #69937 per @dtolnay
This implements `DerefMut for OsString` to allow obtaining a `&mut OsStr`. This also implements `IndexMut for OsString`, which is used by `DerefMut`. This pattern is the same as is used by `Deref`.
This is necessary to for methods like `make_ascii_lowercase` which need to mutate the underlying value.
They used to be covered by `optin_builtin_traits` but negative impls
are now applicable to all traits, not just auto traits.
This also adds docs in the unstable book for the current state of auto traits.
Update backtrace crate to 0.3.46
* Support line-tables-only when using libbacktrace
* Update libbacktrace to latest master
* Define HAVE_KERN_PROC on FreeBSD to fixrust-lang/rust#54434
Currently, constructing a waker requires calling the unsafe
`Waker::from_raw` API. This API requires the user to manually construct
a vtable for the waker themself - which is both cumbersome and very
error prone. This API would provide an ergonomic, straightforward and
guaranteed memory-safe way of constructing a waker.
It has been our longstanding intention that the `Waker` type essentially
function as an `Arc<dyn Wake>`, with a `Wake` trait as defined here. Two
considerations prevented the original API from being shipped as simply
an `Arc<dyn Wake>`:
- We want to support futures on embedded systems, which may not have an
allocator, and in optimized executors for which this API may not be
best-suited. Therefore, we have always explicitly supported the
maximally-flexible (but also memory-unsafe) `RawWaker` API, and
`Waker` has always lived in libcore.
- Because `Waker` lives in libcore and `Arc` lives in liballoc, it has
not been feasible to provide a constructor for `Waker` from `Arc<dyn
Wake>`.
Therefore, the Wake trait was left out of the initial version of the
task waker API.
However, as Rust 1.41, it is possible under the more flexible orphan
rules to implement `From<Arc<W>> for Waker where W: Wake` in liballoc.
Therefore, we can now define this constructor even though `Waker` lives
in libcore.
This PR adds these APIs:
- A `Wake` trait, which contains two methods
- A required method `wake`, which is called by `Waker::wake`
- A provided method `wake_by_ref`, which is called by
`Waker::wake_by_ref` and which implementors can override if they
can optimize this use case.
- An implementation of `From<Arc<W>> for Waker where W: Wake + Send +
Sync + 'static`
- A similar implementation of `From<Arc<W>> for RawWaker`.
Return NonZeroU64 from ThreadId::as_u64.
As discussed in #67939, this allows turning Option<ThreadId> into Option<NonZeroU64> which
can then be stored inside an AtomicU64.
For issue 53957: revise unit test to focus on underlying bug of 23076.
Fix#53957 by revising unit test to focus on underlying bug of #23076.
Namely, this version focuses on the end-to-end behavior that the attempt to create the UDP binding will fail, regardless of the semantics of how particular DNS servers handle junk inputs.
(I spent some time trying to create a second more-focused test that would sidestep the DNS resolution, but this is not possible without more invasive changes to the internal infrastructure of `ToSocketAddrs` and what not. It is not worth it.)
Fix abort-on-eprintln during process shutdown
This commit fixes an issue where if `eprintln!` is used in a TLS
destructor it can accidentally cause the process to abort. TLS
destructors are executed after `main` returns on the main thread, and at
this point we've also deinitialized global `Lazy` values like those
which store the `Stderr` and `Stdout` internals. This means that despite
handling TLS not being accessible in `eprintln!`, we will fail due to
not being able to call `stderr()`. This means that we'll double-panic
quickly because panicking also attempt to write to stderr.
The fix here is to reimplement the global stderr handle to avoid the
need for destruction. This avoids the need for `Lazy` as well as the
hidden panic inside of the `stderr` function.
Overall this should improve the robustness of printing errors and/or
panics in weird situations, since the `stderr` accessor should be
infallible in more situations.
Use generator resume arguments in the async/await lowering
This removes the TLS requirement from async/await and enables it in `#![no_std]` crates.
Closes https://github.com/rust-lang/rust/issues/56974
I'm not confident the HIR lowering is completely correct, there seem to be quite a few undocumented invariants in there. The `async-std` and tokio test suites are passing with these changes though.
Namely, this version focuses on the end-to-end behavior that the attempt to
create the UDP binding will fail, regardless of the semantics of how particular
DNS servers handle junk inputs.
(I spent some time trying to create a second more-focused test that would
sidestep the DNS resolution, but this is not possible without more invasive
changes to the internal infrastructure of `ToSocketAddrs` and what not. It is
not worth it.)
This commit fixes an issue where if `eprintln!` is used in a TLS
destructor it can accidentally cause the process to abort. TLS
destructors are executed after `main` returns on the main thread, and at
this point we've also deinitialized global `Lazy` values like those
which store the `Stderr` and `Stdout` internals. This means that despite
handling TLS not being accessible in `eprintln!`, we will fail due to
not being able to call `stderr()`. This means that we'll double-panic
quickly because panicking also attempt to write to stderr.
The fix here is to reimplement the global stderr handle to avoid the
need for destruction. This avoids the need for `Lazy` as well as the
hidden panic inside of the `stderr` function.
Overall this should improve the robustness of printing errors and/or
panics in weird situations, since the `stderr` accessor should be
infallible in more situations.
unix: Set a guard page at the end of signal stacks
This mitigates possible issues when signal stacks overflow, which could
manifest as segfaults or in unlucky circumstances possible clobbering of
other memory values as stack overflows tend to enable.
I went ahead and made a PR for this because it's a pretty small change, though if I should open an issue/RFC for this and discuss there first I'll happily do so. I've also added some example programs that demonstrate the uncomfortably clobber-happy behavior we currently have, and the segfaults that could/should result instead, [here](https://github.com/iximeow/jubilant-train).
std: Don't abort process when printing panics in tests
This commit fixes an issue when using `set_print` and friends, notably
used by libtest, to avoid aborting the process if printing panics. This
previously panicked due to borrowing a mutable `RefCell` twice, and this
is worked around by borrowing these cells for less time, instead
taking out and removing contents temporarily.
Closes#69558
This commit fixes an issue when using `set_print` and friends, notably
used by libtest, to avoid aborting the process if printing panics. This
previously panicked due to borrowing a mutable `RefCell` twice, and this
is worked around by borrowing these cells for less time, instead
taking out and removing contents temporarily.
Closes#69558
expand: Implement something similar to `#[cfg(accessible(path))]`
cc https://github.com/rust-lang/rust/issues/64797
The feature is implemented as a `#[cfg_accessible(path)]` attribute macro rather than as `#[cfg(accessible(path))]` because it needs to wait until `path` becomes resolvable, and `cfg` cannot wait, but macros can wait.
Later we can think about desugaring or not desugaring `#[cfg(accessible(path))]` into `#[cfg_accessible(path)]`.
This implementation is also incomplete in the sense that it never returns "false" from `cfg_accessible(path)`, it requires some tweaks to resolve, which is not quite ready to answer queries like this during early resolution.
However, the most important part of this PR is not `cfg_accessible` itself, but expansion infrastructure for retrying expansions.
Before this PR we could say "we cannot resolve this macro path, let's try it later", with this PR we can say "we cannot expand this macro, let's try it later" as well.
This is a pre-requisite for
- turning `#[derive(...)]` into a regular attribute macro,
- properly supporting eager expansion for macros that cannot yet be resolved like
```
fn main() {
println!(not_available_yet!());
}
macro_rules! make_available {
() => { #[macro_export] macro_rules! not_available_yet { () => { "Hello world!" } }}
}
make_available!();
```
Implement a feature for a sound specialization subset
This implements a new feature (`min_specialization`) that restricts specialization to a subset that is reasonable for the standard library to use.
The plan is to then:
* Update `libcore` and `liballoc` to compile with `min_specialization`.
* Add a lint to forbid use of `feature(specialization)` (and other unsound, type system extending features) in the standard library.
* Fix the soundness issues around `specialization`.
* Remove `min_specialization`
The rest of this is an overview from a comment in this PR
## Basic approach
To enforce this requirement on specializations we take the following approach:
1. Match up the substs for `impl2` so that the implemented trait and self-type match those for `impl1`.
2. Check for any direct use of `'static` in the substs of `impl2`.
3. Check that all of the generic parameters of `impl1` occur at most once in the *unconstrained* substs for `impl2`. A parameter is constrained if its value is completely determined by an associated type projection predicate.
4. Check that all predicates on `impl1` also exist on `impl2` (after matching substs).
## Example
Suppose we have the following always applicable impl:
```rust
impl<T> SpecExtend<T> for std::vec::IntoIter<T> { /* specialized impl */ }
impl<T, I: Iterator<Item=T>> SpecExtend<T> for I { /* default impl */ }
```
We get that the subst for `impl2` are `[T, std::vec::IntoIter<T>]`. `T` is constrained to be `<I as Iterator>::Item`, so we check only `std::vec::IntoIter<T>` for repeated parameters, which it doesn't have. The predicates of `impl1` are only `T: Sized`, which is also a predicate of impl2`. So this specialization is sound.
## Extensions
Unfortunately not all specializations in the standard library are allowed by this. So there are two extensions to these rules that allow specializing on some traits.
### rustc_specialization_trait
If a trait is always applicable, then it's sound to specialize on it. We check trait is always applicable in the same way as impls, except that step 4 is now "all predicates on `impl1` are always applicable". We require that `specialization` or `min_specialization` is enabled to implement these traits.
### rustc_specialization_marker
There are also some specialization on traits with no methods, including the `FusedIterator` trait which is advertised as allowing optimizations. We allow marking marker traits with an unstable attribute that means we ignore them in point 3 of the checks above. This is unsound but we allow it in the short term because it can't cause use after frees with purely safe code in the same way as specializing on traits methods can.
r? @nikomatsakis
cc #31844#67194
Dylan DPC