rustdoc: fix & clean up handling of cross-crate higher-ranked parameters
Preparatory work for the refactoring planned in #113015 (for correctness & maintainability).
---
1. Render the higher-ranked parameters of cross-crate function pointer types **(*)**.
2. Replace occurrences of `collect_referenced_late_bound_regions()` (CRLBR) with `bound_vars()`.
The former is quite problematic and the use of the latter allows us to yank a lot of hacky code **(†)**
as you can tell from the diff! :)
3. Add support for cross-crate higher-ranked types (`#![feature(non_lifetime_binders)]`).
We were previously ICE'ing on them (see `inline_cross/non_lifetime_binders.rs`).
---
**(*)**: Extracted from test `inline_cross/fn-type.rs`:
```diff
- fn(_: &'z fn(_: &'b str), _: &'a ()) -> &'a ()
+ for<'z, 'a, '_unused> fn(_: &'z for<'b> fn(_: &'b str), _: &'a ()) -> &'a ()
```
**(†)**: It returns an `FxHashSet` which isn't *predictable* or *stable* wrt. source code (`.rmeta`) changes. To elaborate, the ordering of late-bound regions doesn't necessarily reflect the ordering found in the source code. It does seem to be stable across compilations but modifying the source code of the to-be-documented crates (like adding or renaming items) may result in a different order:
<details><summary>Example</summary>
Let's assume that we're documenting the cross-crate re-export of `produce` from the code below. On `master`, rustdoc would render the list of binders as `for<'x, 'y, 'z>`. However, once you add back the functions `a`–`l`, it would be rendered as `for<'z, 'y, 'x>` (reverse order)! Results may vary. `bound_vars()` fixes this as it returns them in source order.
```rs
// pub fn a() {}
// pub fn b() {}
// pub fn c() {}
// pub fn d() {}
// pub fn e() {}
// pub fn f() {}
// pub fn g() {}
// pub fn h() {}
// pub fn i() {}
// pub fn j() {}
// pub fn k() {}
// pub fn l() {}
pub fn produce() -> impl for<'x, 'y, 'z> Trait<'z, 'y, 'x> {}
pub trait Trait<'a, 'b, 'c> {}
impl Trait<'_, '_, '_> for () {}
```
</details>
Further, as the name suggests, CRLBR only collects *referenced* regions and thus we drop unused binders. `bound_vars()` contains unused binders on the other hand. Let's stay closer to the source where possible and keep unused binders.
Lastly, using `bound_vars()` allows us to get rid of
* the deduplication and alphabetical sorting hack in `simplify.rs`
* the weird field `bound_params` on `EqPredicate`
both of which were introduced by me in #102707 back when I didn't know better.
To illustrate, let's look at the cross-crate bound `T: for<'a, 'b> Trait<A<'a> = (), B<'b> = ()>`.
* With CRLBR + `EqPredicate.bound_params`, *before* bounds simplification we would have the bounds `T: Trait`, `for<'a> <T as Trait>::A<'a> == ()` and `for<'b> <T as Trait>::B<'b> == ()` which required us to merge `for<>`, `for<'a>` and `for<'b>` into `for<'a, 'b>` in a deterministic manner and without introducing duplicate binders.
* With `bound_vars()`, we now have the bounds `for<'a, b> T: Trait`, `<T as Trait>::A<'a> == ()` and `<T as Trait>::B<'b> == ()` before bound simplification similar to rustc itself. This obviously no longer requires any funny merging of `for<>`s. On top of that `for<'a, 'b>` is guaranteed to be in source order.
coverage: Allow each coverage statement to have multiple code regions
The original implementation of coverage instrumentation was built around the assumption that a coverage counter/expression would be associated with *up to one* code region. When it was discovered that *multiple* regions would sometimes need to share a counter, a workaround was found: for the remaining regions, the instrumentor would create a fresh expression that adds zero to the existing counter/expression.
That got the job done, but resulted in some awkward code, and produces unnecessarily complicated coverage maps in the final binary.
---
This PR removes that tension by changing `StatementKind::Coverage`'s code region field from `Option<CodeRegion>` to `Vec<CodeRegion>`.
The changes on the codegen side are fairly straightforward. As long as each `CoverageKind::Counter` only injects one `llvm.instrprof.increment`, the rest of coverage codegen is happy to handle multiple regions mapped to the same counter/expression, with only minor option-to-vec adjustments.
On the instrumentor/mir-transform side, we can get rid of the code that creates extra (x + 0) expressions. Instead we gather all of the code regions associated with a single BCB, and inject them all into one coverage statement.
---
There are several patches here but they can be divided in to three phases:
- Preparatory work
- Actually switching over to multiple regions per coverage statement
- Cleaning up
So viewing the patches individually may be easier.
non_lifetime_binders: fix ICE in lint opaque-hidden-inferred-bound
Opaque types like `impl for<T> Trait<T>` would previously lead to an ICE.
r? `@compiler-errors`
Suggest `pin!()` instead of `Pin::new()` when appropriate
When encountering a type that needs to be pinned but that is `!Unpin`, suggest using the `pin!()` macro.
Fix#57994.
Rollup of 5 pull requests
Successful merges:
- #115863 (Add check_unused_messages in tidy)
- #116210 (Ensure that `~const` trait bounds on associated functions are in const traits or impls)
- #116358 (Rename both of the `Match` relations)
- #116371 (Remove unused features from `rustc_llvm`.)
- #116374 (Print normalized ty)
r? `@ghost`
`@rustbot` modify labels: rollup
Don't suggest nonsense suggestions for unconstrained type vars in `note_source_of_type_mismatch_constraint`
The way we do type inference for suggestions in `note_source_of_type_mismatch_constraint` is a bit strange. We compute the "ideal" method signature, which takes the receiver that we *want* and uses it to compute the types of the arguments that would have given us that receiver via type inference, and use *that* to suggest how to change an argument to make sure our receiver type is inferred correctly.
The problem is that sometimes we have totally unconstrained arguments (well, they're constrained by things outside of the type checker per se, like associated types), and therefore type suggestions are happy to coerce anything to that unconstrained argument. This leads to bogus suggestions, like #116155. This is partly due to above, and partly due to the fact that `emit_type_mismatch_suggestions` doesn't double check that its suggestions are actually compatible with the program other than trying to satisfy the type mismatch.
This adds a hack to make sure that at least the types are fully constrained, but I guess I could also rip out this logic altogether. There would be some sad diagnostics regressions though, such as `tests/ui/type/type-check/point-at-inference-4.rs`.
Fixes#116155
For a single impl candidate, try to unify it with error trait ref
This allows us to point out an exact type mismatch when there's only one applicable impl.
cc `@asquared31415`
r? `@estebank`
If a BCB has more than one code region, those extra regions can now all be
stored in the same coverage statement, instead of being stored in additional
statements.
Cleanup number handling in match exhaustiveness
Doing a little bit of cleanup; handling number constants was somewhat messy. In particular, this:
- evals float consts once instead of repetitively
- reduces `Constructor` from 88 bytes to 56 (`mir::Const` is big!)
The `fast_try_eval_bits` function was mostly constructed from inlining existing code but I don't fully understand it; I don't follow how consts work and are evaluated very well.
coverage: Regression test for functions with unreachable bodies
This is a regression test for the coverage issue that was addressed temporarily by #116166, and is tracked by #116171.
---
If we instrument a function for coverage, but all of its counter-increment statements are removed by MIR optimizations, LLVM will think it isn't instrumented and it will disappear from coverage maps and coverage reports.
Most MIR opts won't cause this because they tend not to remove statements from bb0, but `UnreachablePropagation` can do so if it sees that bb0 ends with `TerminatorKind::Unreachable`.
Currently we have worked around this by turning off `UnreachablePropagation` when coverage instrumentation is enabled, which is why this test is able to pass.
---
`@rustbot` label +A-code-coverage
resolve: skip underscore character during candidate lookup
Fixes#116164
In use statement, an underscore is merely a placeholder symbol and does not bind to any name. Therefore, it can be safely ignored.
Previously, any associated function could have `~const` trait bounds on
generic parameters, which could lead to ICEs when these bounds were used
on associated functions of non-`#[const_trait] trait` or
non-`impl const` blocks.
Includes changes as per @fee1-dead's comments in #116210.
Prototype using const generic for simd_shuffle IDX array
cc https://github.com/rust-lang/rust/issues/85229
r? `@workingjubilee` on the design
TLDR: there is now a `fn simd_shuffle_generic<T, U, const IDX: &'static [u32]>(x: T, y: T) -> U;` intrinsic that allows replacing
```rust
simd_shuffle(a, b, const { stuff })
```
with
```rust
simd_shuffle_generic::<_, _, {&stuff}>(a, b)
```
which makes the compiler implementations much simpler, if we manage to at some point eliminate `simd_shuffle`.
There are some issues with this today though (can't do math without bubbling it up in the generic arguments). With this change, we can start porting the simple cases and get better data on the others.
More fixes for running the test suite on a bare metal target
This PR adds more fixes needed to run the test suite on bare metal targets (in this case, without unwinding and with static relocations). There is no CI job exercising tests without unwinds, but I can confirm this worked in Ferrocene's CI.
fix(suggestion): insert projection to associated types
Fixes#98562
This PR has fixed some help suggestions for unsupported syntax, such as `fn f<T>(_:T) where T: IntoIterator, std::iter::IntoIterator::Item = () {}` to `fn f<T: IntoIterator<Item = ()>>(_T) {}`.
stabilize combining +bundle and +whole-archive link modifiers
Per discussion on https://github.com/rust-lang/rust/issues/108081 combining +bundle and +whole-archive already works and can be stabilized independently of other aspects of the packed_bundled_libs feature. There is no risk of regression because this was not previously allowed.
r? `@petrochenkov`
Reveal opaque types before drop elaboration
fixes https://github.com/rust-lang/rust/issues/113594
r? `@cjgillot`
cc `@JakobDegen`
This pass was introduced in https://github.com/rust-lang/rust/pull/110714
I moved it before drop elaboration (which only cares about the hidden types of things, not the opaque TAIT or RPIT type) and set it to run unconditionally (instead of depending on the optimization level and whether the inliner is active)
Make `adt_const_params` feature suggestion consistent with other features and improve when it is emitted
Makes the suggestion to add `adt_const_params` formatted like every other feature gate (notably this makes it such that the playground recognizes it). Additionally improves the situations in which that help is emitted so that it's only emitted when the type would be valid or the type *could* be valid (using a slightly incorrect heuristic that favors suggesting the feature over not) instead of, for example, implying that adding the feature would allow the use of `String`.
Also adds the "the only supported types are integers, `bool` and `char`" note to the errors on fn and raw pointers.
r? `@compiler-errors`
Fix `noop_method_call` detection
This needs to be merged before #116198 can compile. The error occurs before the compiler is built so this needs to be a separate PR.
new solver: remove provisional cache
The provisional cache is a performance optimization if there are large, interleaving cycles. Such cycles generally do not exist. It is incredibly complex and unsound in all trait solvers which have one: the old solver, chalk, and the new solver ([link](https://github.com/rust-lang/rust/blob/master/tests/ui/traits/new-solver/cycles/inductive-not-on-stack.rs)).
Given the assumption that it is not perf-critical and also incredibly complex, remove it from the new solver, only checking whether a goal is on the stack. While writing this, I uncovered two additional soundness bugs, see the inline comments for them.
r? `@compiler-errors`
