Lint on invalid values passed to x.py --warnings
This also introduces support for `--warnings allow` and fixes --warnings
being overridden by the configuration file, config.toml.
Fixes#62402
r? @RalfJung
remove Scalar::is_null_ptr
Comparing pointers should be done more carefully than that. With https://github.com/rust-lang/miri/pull/825, Miri does not need it any more and it is otherwise unused.
rustc_target: avoid negative register counts in the SysV x86_64 ABI.
Because `needed_{int,sse}` and `{int,sse}_regs` were only used with integer literals, they were inferred to `i32` and `{int,sse}_regs` could therefore be negative.
There was a check which prevented that, but *only* for aggregate arguments, not scalars.
Fixes#62350.
r? @nagisa or @rkruppe
Remove `compile-pass` from compiletest
This is a part of #62277.
Removes `compile-pass` from compiletest (and modify some tests' annotations).
r? @Centril
Create async version of the dynamic-drop test
Some of the tests in dynamic-drop have been cut:
* The tests that are just simpler versions of other tests - these tests are already fairly slow due to all of the unwinding and async functions have more control flow paths than normal functions.
* The union test - it's for an unstable feature that has an RFC to remove it.
* The generator test - there aren't async generators yet.
* The tests that show values being leaked - these can be added once the issue is fixed.
r? @Centril
cc #62121 @cramertj
The (almost) culmination of HirIdification
It's finally over.
This PR removes old `FIXME`s and renames some functions so that the `HirId` variant has the shorter name.
All that remains (and rightfully so) is stuff in `resolve`, `save_analysis` and (as far as I can tell) in a few places where we can't replace `NodeId` with `HirId`.
Add MemoryExtra in InterpretCx constructor params
This is to avoid modifying `MemoryExtra` inside `InterpretCx` after initialization. Related miri PR: https://github.com/rust-lang/miri/pull/792
r? @RalfJung
Implement another internal lints
cc #49509
This adds ~~two~~ one internal lint~~s~~:
1. LINT_PASS_IMPL_WITHOUT_MACRO: Make sure, that the `{declare,impl}_lint_pass` macro is used to implement lint passes. cc #59669
2. ~~USAGE_OF_TYCTXT_AND_SPAN_ARGS: item 2 on the list in #49509~~
~~With 2. I wasn't sure, if this lint should be applied everywhere. That means a careful review of 0955835 would be great. Also 73fb9b4 allows this lint on some functions. Should I also apply this lint there?~~
TODO (not directly relevant for review):
- [ ] https://github.com/rust-lang/rust/pull/59316#discussion_r280186517 (not sure yet, if this works or how to query for `rustc_private`, since it's not in [`Features`](https://doc.rust-lang.org/nightly/nightly-rustc/syntax/feature_gate/struct.Features.html) 🤔 cc @eddyb)
- [x] https://github.com/rust-lang/rust/pull/61735#discussion_r292389870
- [x] Check explicitly for the `{declare,impl}_lint_pass!` macros
r? @oli-obk
Rollup of 10 pull requests
Successful merges:
- #62123 ( Remove needless lifetimes (std))
- #62150 (Implement mem::{zeroed,uninitialized} in terms of MaybeUninit.)
- #62169 (Derive which queries to save using the proc macro)
- #62238 (Fix code block information icon position)
- #62292 (Move `async || ...` closures into `#![feature(async_closure)]`)
- #62323 (Clarify unaligned fields in ptr::{read,write}_unaligned)
- #62324 (Reduce reliance on `await!(...)` macro)
- #62371 (Add tracking issue for Box::into_pin)
- #62383 (Improve error span for async type inference error)
- #62388 (Break out of the correct number of scopes in loops)
Failed merges:
r? @ghost
Break out of the correct number of scopes in loops
We were incorrectly breaking out of one too many drop scopes when
generating MIR for loops and breakable blocks, resulting in use after
free and associated borrow checker warnings.
This wasn't noticed because the scope that we're breaking out of twice
is only used for temporaries that are created for adjustments applied to
the loop. Since loops generally propagate coercions to the `break`
expressions, the only case we see this is when the type of the loop is a
smart pointer to a trait object.
Closes#62312
Improve error span for async type inference error
Fixes#62382
Previously, we would point at the spawn of the 'await' expression,
instead of the actual expression with an unknown type.
Before this change `SliceConcatExt` was an unstable extension trait
with stable methods. It was in the libstd prelude, so that its methods
could be used on the stable channel.
This replaces it with inherent methods,
which can be used without any addition to the prelude.
Since the methods are stable and very generic
(with for example a return type that depends on the types of parameters),
an helper trait is still needed.
But now that trait does not need to be in scope for the methods to be used.
Removing this depedency on the libstd prelude allows the methods to be used
in `#![no_std]` crate that use liballoc, which does not have its own
implicitly-imported prelude.
Remove io::Result from syntax::print
Since we're now writing directly to the vector, there's no need to
thread results through the whole printing infrastructure
This is useful for getting the total compilation time at the end.
To do this, the patch changes `print_time_passes_entry` to not increment
the depth, which means that `print_time_passes_entry_internal` is no
longer needed.
Update the `rust-installer` submodule
This pulls in a commit which uses parallel xz encoding which should
hopefully help shave some time off the dist builders which spend an
inordinate amount of time compressing this data.
We were incorrectly breaking out of one too many drop scopes when
generating MIR for loops and breakable blocks, resulting in use after
free and associated borrow checker warnings.
This wasn't noticed because the scope that we're breaking out of twice
is only used for temporaries that are created for adjustments applied to
the loop. Since loops generally propagate coercions to the `break`
expressions, the only case we see this is when the type of the loop is a
smart pointer to a trait object.
This pulls in a commit which uses parallel xz encoding which should
hopefully help shave some time off the dist builders which spend an
inordinate amount of time compressing this data.
