This commit provides a safe, but unstable interface for the `try` functionality
of running a closure and determining whether it panicked or not.
There are two primary reasons that this function was previously marked `unsafe`:
1. A vanilla version of this function exposes the problem of exception safety by
allowing a bare try/catch in the language. It is not clear whether this
concern should be directly tied to `unsafe` in Rust at the API level. At this
time, however, the bounds on `ffi::try` require the closure to be both
`'static` and `Send` (mirroring those of `thread::spawn`). It may be possible
to relax the bounds in the future, but for now it's the level of safety that
we're willing to commit to.
2. Panicking while panicking will leak resources by not running destructors.
Because panicking is still controlled by the standard library, safeguards
remain in place to prevent this from happening.
The new API is now called `catch_panic` and is marked as `#[unstable]` for now.
closes#23620
This PR patches the issue mentioned in #23620, but there is also an ICE for invalid escape sequences in byte literals. This is due to the fact that the `scan_byte` function returns ` token::intern("??") ` for invalid bytes, resulting in an ICE later on. Is there a reason for this behavior? Shouldn't `scan_byte` fail when it encounters an invalid byte?
And I noticed a small inconsistency in the documentation. According to the formal byte literal definition in http://doc.rust-lang.org/reference.html#byte-and-byte-string-literals , a byte string literal contains `string_body *`, but according to the text (and the behavior of the lexer) it should not accept unicode escape sequences. Hence it should be replaced by `byte_body *`. If this is valid, I can add this fix to this PR.
Replace zeroing-on-drop with filling-on-drop.
This is meant to set the stage for removing *all* zeroing and filling (on drop) in the future.
Note that the code is meant to be entirely abstract with respect to the particular values used for the drop flags: the final commit demonstrates how to go from zeroing-on-drop to filling-on-drop by changing the value of three constants (in two files).
See further discussion on the internals thread:
http://internals.rust-lang.org/t/attention-hackers-filling-drop/1715/11
[breaking-change] especially for structs / enums using `#[unsafe_no_drop_flag]`.
This PR introduces a `Reflect` marker trait which is a supertrait of `Any`. The idea is that `Reflect` is defined for all concrete types, but is not defined for type parameters unless there is a `T:Reflect` bound. This is intended to preserve the parametricity property. This allows the `Any` interface to be stabilized without committing us to unbounded reflection that is not easily detectable by the caller.
The implementation of `Reflect` relies on an experimental variant of OIBIT. This variant behaves differently for objects, since it requires that all types exposed as part of the object's *interface* are `Reflect`, but isn't concerned about other types that may be closed over. In other words, you don't have to write `Foo+Reflect` in order for `Foo: Reflect` to hold (where `Foo` is a trait).
Given that `Any` is slated to stabilization and hence that we are committed to some form of reflection, the goal of this PR is to leave our options open with respect to parametricity. I see the options for full stabilization as follows (I think an RFC would be an appropriate way to confirm whichever of these three routes we take):
1. We make `Reflect` a lang-item.
2. We stabilize some version of the OIBIT variation I implemented as a general mechanism that may be appropriate for other use cases.
3. We give up on preserving parametricity here and just have `impl<T> Reflect for T` instead. In that case, `Reflect` is a harmless but not especially useful trait going forward.
cc @aturon
cc @alexcrichton
cc @glaebhoerl (this is more-or-less your proposal, as I understood it)
cc @reem (this is more-or-less what we discussed on IRC at some point)
cc @FlaPer87 (vaguely pertains to OIBIT)
closes#23620
This PR patches the issue mentioned in #23620, but there is also an ICE for invalid escape sequences in byte literals. This is due to the fact that the `scan_byte` function returns ` token::intern(\"??\") ` for invalid bytes, resulting in an ICE later on. Is there a reason for this behavior? Shouldn't `scan_byte` fail when it encounters an invalid byte?
And I noticed a small inconsistency in the documentation. According to the formal byte literal definition in http://doc.rust-lang.org/reference.html#byte-and-byte-string-literals , a byte string literal contains `string_body *`, but according to the text (and the behavior of the lexer) it should not accept unicode escape sequences. Hence it should be replaced by `byte_body *`. If this is valid, I can add this fix to this PR.
Windows gets quite unhappy when a thread fails while the main thread is exiting,
frequently leading to process deadlock. This has been causing quite a few
deadlocks on the windows bots recently. The child threads are presumably failing
because the `println!` is failing due to the main thread being shut down.
RFC pending, but this is the patch that does it.
Totally untested. Likely needs some removed imports. std::collections docs should also be updated to provide better examples.
Closes#23508
Marks as `#[stable}`:
* `ok_or`
* `ok_or_else`
* `iter_mut`
* `cloned`
Similarly to `IteratorExt::cloned`, the `cloned` method is pared down to
work only on `Option<&T>`. Thus, this is a:
[breaking-change]
Now that `<[_]>::split` is an inherent method, it will trump `BufRead::split`
when `BufRead` is in scope, so there is no longer a conflict. As a result,
calling `slice.split()` will probably always give you precisely what you want!
Found a few 404s that seemed like simple fixes:
The Result docs use old_io Writer as an example. Fix the link to old_io Writer. There's probably an effort to update the example away from a deprecated api but this was a simple fix.
rustc/plugin was pointing at the old guide and it was a broken link anyways (plugin vs plugins). Point at the book instead.
The main page of the API docs referenced c_{str,vec}. Looks like these were deleted in 25d5a3a194. Point at ffi docs instead.
This commit revises `path` and `os_str` to use blanket impls for `From`
on reference types. This both cuts down on the number of required impls,
and means that you can pass through e.g. `T: AsRef<OsStr>` to
`PathBuf::from` without an intermediate call to `as_ref`.
It also makes a FIXME note for later generalizing the blanket impls for
`AsRef` and `AsMut` to use `Deref`/`DerefMut`, once it is possible to do
so.
This attribute has been deprecated in favor of #[should_panic]. This also
updates rustdoc to no longer accept the `should_fail` directive and instead
renames it to `should_panic`.
This commit removes compiler support for the `old_impl_check` attribute which
should in theory be entirely removed now. The last remaining use of it in the
standard library has been updated by moving the type parameter on the
`old_io::Acceptor` trait into an associated type. As a result, this is a
breaking change for all current users of the deprecated `old_io::Acceptor`
trait. Code can be migrated by using the `Connection` associated type instead.
[breaking-change]
This is technically a breaking change as it deprecates and unstables
some previously stable apis that were missed in the last round of
deprecations.
[breaking change]
This PR adds support for associated types to the `#[derive(...)]` syntax extension. In order to do this, it switches over to using where predicates to apply the type constraints. So now this:
```rust
type Trait {
type Type;
}
#[derive(Clone)]
struct Foo<A> where A: Trait {
a: A,
b: <A as Trait>::Type,
}
```
Gets expended into this impl:
```rust
impl<A: Clone> Clone for Foo<A> where
A: Trait,
<A as Trait>::Type: Clone,
{
fn clone(&self) -> Foo<T> {
Foo {
a: self.a.clone(),
b: self.b.clone(),
}
}
}
```
