support opaque types in method selection
See my notes in https://hackmd.io/4ILASx3mQ3u_gW9r1JyqCw.
This PR builds on https://github.com/rust-lang/rust/pull/145993 and allows not-yet defined opaque types as self types in the `method_autoderef_chain`.
E.g. for `Box<impl Deref<impl Foo>>` this results in the autoderef chain `Box<impl Deref> -> ?deref_hidden_ty -> ?foo_hidden_ty`. Method selection stays ambiguous if the final autoderef step is still an infer var unless that var is an opaque.
TODO: treating opaques as rigid jank.
r? `@BoxyUwU`
We'll still error due to the `opt_bad_ty` of `method_autoderef_steps`.
This slightly worsens the span of `infer_var.method()` which is now the
same as for `Box::new(infer_var).method()`.
Unlike `structurally_resolve_type`, `probe_op` does not check whether
the infcx is already tainted, so this results in 2 previously not emitted
errors.
Make cargo test work for bootstrap self test
This PR enables the bootstrap self-test to run via cargo test. I have removed the detect_src_and_out test for now, but it will be reintroduced in a follow-up PR where all bootstrap tests will be migrated to use testCtx.
r? `@Kobzol`
try-job: aarch64-apple
Rollup of 7 pull requests
Successful merges:
- rust-lang/rust#146283 (Resolve: (Ref)Cell wrappers to deny mutation during spec resolution.)
- rust-lang/rust#146453 (Add general arm-linux.md platform doc.)
- rust-lang/rust#146991 (const_caller_location to use real Span instead of `DUMMY_SP`)
- rust-lang/rust#146994 (Add `clippy::unconditional_recursion` to `./x clippy ci`)
- rust-lang/rust#147038 (Rename verbosity functions in bootstrap)
- rust-lang/rust#147047 (rustdoc: put the toolbar on the all item index)
- rust-lang/rust#147049 (std: fix warning in VEXos stdio module)
r? `@ghost`
`@rustbot` modify labels: rollup
std: fix warning in VEXos stdio module
Fixes building `std` on the `armv7a-vex-v5` target due to an unnecessarily mutable argument in `Stdin`.
This was a stupid oversight on my part towards the end of rust-lang/rust#145973's review process. Missed a warning and had a bad bootstrap config that didn't tell me about it when testing changes.
Rename verbosity functions in bootstrap
Just a small cleanup, these function names have been bothering me for a while. I realized that we can delete some of them outright, rather than just renaming them.
r? ``@jieyouxu``
Add `clippy::unconditional_recursion` to `./x clippy ci`
The clippy lint catches some things that rustc's equivalent builtin lint
does not, for example rust-lang/rust#146940:
error: function cannot return without recursing
--> library/std/src/path.rs:3428:5
|
3428 | / fn eq(&self, other: &String) -> bool {
3429 | | self == &*other
3430 | | }
| |_____^
|
note: recursive call site
--> library/std/src/path.rs:3429:9
|
3429 | self == &*other
| ^^^^^^^^^^^^^^^
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#unconditional_recursion
= note: requested on the command line with `-D clippy::unconditional-recursion`
Add general arm-linux.md platform doc.
Adds a new page that covers all 32-bit Arm Linux systems. This means that we can reduce the amount of information required in the target specific pages to just the Tier level, the maintainer, and any specific details for that target.
I have no changed those pages yet, though. Let's start with this.
Resolve: (Ref)Cell wrappers to deny mutation during spec resolution.
Introduces wrappers around `Cell` and `RefCell` that only allow mutation when we are not in speculative resolution. This is preparatory work for rust-lang/rust#145108.
It would allow us to make `ImportData` and `ModuleData` sync and send safe.
r? ``@petrochenkov``
remove incorrect fast path
Using `tcx.is_copy_modulo_regions` drops information from the current `typing_env`. Writing a regression test for this is really hard. We need to prove `Copy` of something that doesn't directly reference a coroutine or an opaque, but does so indirectly.
cc rust-lang/rust#146813.
The clippy lint catches some things that rustc's equivalent builtin lint
does not, for example rust-lang/rust#146940:
error: function cannot return without recursing
--> library/std/src/path.rs:3428:5
|
3428 | / fn eq(&self, other: &String) -> bool {
3429 | | self == &*other
3430 | | }
| |_____^
|
note: recursive call site
--> library/std/src/path.rs:3429:9
|
3429 | self == &*other
| ^^^^^^^^^^^^^^^
= help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#unconditional_recursion
= note: requested on the command line with `-D clippy::unconditional-recursion`
Make `render_example_with_highlighting` return an `impl fmt::Display`
Removes some more usages of `write_str`. Shouldn't affect runtime, but makes the codebase a bit more consistent.
Each commit can be reviewed separately
Add doc for `NonZero*` const creation
I ran into trouble using `NonZero*` values because I didn’t see any clear way to create them at compile time. At first I ended up using `NonZero*::new_unchecked` a lot, until I realized that `Option::unwrap` and `Option::expect` are `const` and can be used in a `const` context. With that, you can create non-zero values at compile time safely, without touching `unsafe`. This wasn’t obvious to me and my peers who’ve been using Rust for a while, so I thought adding a note to the docs would make it easier for others to discover.
If this should be worded differently or placed in another location, we can do that. I just want to make this more obvious.
BTreeMap: Don't leak allocators when initializing nodes
Memory was allocated via `Box::leak` and thence intended to be tracked and deallocated manually, but the allocator was also leaked, not tracked, and never dropped. Now it is dropped immediately.
According to my reading of the `Allocator` trait, if a copy of the allocator remains live, then its allocations must remain live. Since the B-tree has a copy of the allocator that will only be dropped after the nodes, it's safe to not store the allocator in each node (which would be a much more intrusive change).
Fixes: rust-lang/rust#106203
avoid violating `slice::from_raw_parts` safety contract in `Vec::extract_if`
The implementation of the `Vec::extract_if` iterator violates the safety contract adverized by `slice::from_raw_parts` by always constructing a mutable slice for the entire length of the vector even though that span of memory can contain holes from items already drained. The safety contract of `slice::from_raw_parts` requires that all elements must be properly
initialized.
As an example we can look at the following code:
```rust
let mut v = vec![Box::new(0u64), Box::new(1u64)];
for item in v.extract_if(.., |x| **x == 0) {
drop(item);
}
```
In the second iteration a `&mut [Box<u64>]` slice of length 2 will be constructed. The first slot of the slice contains the bitpattern of an already deallocated box, which is invalid.
This fixes the issue by only creating references to valid items and using pointer manipulation for the rest. I have also taken the liberty to remove the big `unsafe` blocks in place of targetted ones with a SAFETY comment. The approach closely mirrors the implementation of `Vec::retain_mut`.
**Note to reviewers:** The diff is easier to follow with whitespace hidden.
rustdoc: hide `#[repr]` if it isn't part of the public ABI
> [!IMPORTANT]
> Temporarily stacked on top of PR https://github.com/rust-lang/rust/pull/146527; only the last commit is relevant!
Follow-up to rust-lang/rust#115439.
Unblocks rust-lang/rust#116743, CC ``@dtolnay.``
Fixesrust-lang/rust#66401.
Fixesrust-lang/rust#128364.
Fixesrust-lang/rust#137440.
Only display the representation `#[repr(REPR)]` (where `REPR` is not `Rust` or `transparent`) of a given type if none of its variants (incl. the synthetic variants of structs) are `#[doc(hidden)]` and all of its fields are public and not `#[doc(hidden)]` since it's likely not meant to be considered part of the public ABI otherwise.
`--document-{private,hidden}-items` works as expected in this context, too.
Moreover, we now also factor in the presence of `#[doc(hidden)]` when checking whether to show `repr(transparent)` or not.
Use `LLVMDisposeTargetMachine`
After bumping the minimum LLVM version to 20 (rust-lang/rust#145071), we no longer need to run any custom code when disposing of a TargetMachine, so we can just use the upstream LLVM-C function.
bootstrap.py: Respect build.jobs while building bootstrap tool
On resource-constrained systems, it is vital to respect the value of build.jobs, in order to avoid overwhelming the available memory.
Explicitly note `&[SocketAddr]` impl of `ToSocketAddrs`
Although the examples below this list do imply that there's an impl of `ToSocketAddrs` for `&[SocketAddr]`, it's not actually noted in the list of default implementations.
Small string formatting cleanup
This PR is mostly useless. I was going through this file, saw that and corrected it. That's pretty much it. Feel free to close it if it's a bother.
llvm: update remarks support on LLVM 22
LLVM change dfbd76bda01e removed separate remark support entirely, but
it turns out we can just drop the parameter and everything appears to
work fine.
Fixesrust-lang/rust#146912 as far as I can tell (the test passes.)
