Tweak relations to no longer rely on `TypeTrace`
Remove `At::trace`, and inline all of the `Trace::equate`,etc methods into `At`.
The only nontrivial change is that we use `AliasTerm` to relate two unevaluated consts in the old-solver impl of `ConstEquate`, since `AliasTerm` does implement `ToTrace` and will relate the args structurally (shallowly).
r? lcnr
drop region constraints for ambiguous goals
See the comment in `compute_external_query_constraints`. While the underlying issue is preexisting, this fixes a bug introduced by #125343.
It slightly weakens the leak chec, even if we didn't have any test which was affected. I want to write such a test before merging this PR.
r? `@compiler-errors`
Uplift `EarlyBinder` into `rustc_type_ir`
We also need to give `EarlyBinder` a `'tcx` param, so that we can carry the `Interner` in the `EarlyBinder` too. This is necessary because otherwise we have an unconstrained `I: Interner` parameter in many of the `EarlyBinder`'s inherent impls.
I also generally think that this is desirable to have, in case we later want to track some state in the `EarlyBinder`.
r? lcnr
[perf] Delay the construction of early lint diag structs
Attacks some of the perf regressions from https://github.com/rust-lang/rust/pull/124417#issuecomment-2123700666.
See individual commits for details. The first three commits are not strictly necessary.
However, the 2nd one (06bc4fc671, *Remove `LintDiagnostic::msg`*) makes the main change way nicer to implement.
It's also pretty sweet on its own if I may say so myself.
Remove `DefId` from `EarlyParamRegion`
Currently we represent usages of `Region` parameters via the `ReEarlyParam` or `ReLateParam` variants. The `ReEarlyParam` is effectively equivalent to `TyKind::Param` and `ConstKind::Param` (i.e. it stores a `Symbol` and a `u32` index) however it also stores a `DefId` for the definition of the lifetime parameter.
This was used in roughly two places:
- Borrowck diagnostics instead of threading the appropriate `body_id` down to relevant locations. Interestingly there were already some places that had to pass down a `DefId` manually.
- Some opaque type checking logic was using the `DefId` field to track captured lifetimes
I've split this PR up into a commit for generate rote changes to diagnostics code to pass around a `DefId` manually everywhere, and another commit for the opaque type related changes which likely require more careful review as they might change the semantics of lints/errors.
Instead of manually passing the `DefId` around everywhere I previously tried to bundle it in with `TypeErrCtxt` but ran into issues with some call sites of `infcx.err_ctxt` being unable to provide a `DefId`, particularly places involved with trait solving and normalization. It might be worth investigating adding some new wrapper type to pass this around everywhere but I think this might be acceptable for now.
This pr also has the effect of reducing the size of `EarlyParamRegion` from 16 bytes -> 8 bytes. I wouldn't expect this to have any direct performance improvement however, other variants of `RegionKind` over `8` bytes are all because they contain a `BoundRegionKind` which is, as far as I know, mostly there for diagnostics. If we're ever able to remove this it would shrink the `RegionKind` type from `24` bytes to `12` (and with clever bit packing we might be able to get it to `8` bytes). I am curious what the performance impact would be of removing interning of `Region`'s if we ever manage to shrink `RegionKind` that much.
Sidenote: by removing the `DefId` the `Debug` output for `Region` has gotten significantly nicer. As an example see this opaque type debug print before vs after this PR:
`Opaque(DefId(0:13 ~ impl_trait_captures[aeb9]::foo::{opaque#0}), [DefId(0:9 ~ impl_trait_captures[aeb9]::foo::'a)_'a/#0, T, DefId(0:9 ~ impl_trait_captures[aeb9]::foo::'a)_'a/#0])`
`Opaque(DefId(0:13 ~ impl_trait_captures[aeb9]::foo::{opaque#0}), ['a/#0, T, 'a/#0])`
r? `@compiler-errors` (I would like someone who understands the opaque type setup to atleast review the type system commit, but the rest is likely reviewable by anyone)
remove proof tree formatting, make em shallow
Debugging via tracing `RUSTC_LOG=rustc_trait_selection::solve=debug` is now imo slightly more readable then the actual proof tree formatter. Removing everything that's not needed for the `analyse` visitor allows us to remove a bunch of code.
I personally believe that we should continue to use tracing over proof trees for debugging:
- it eagerly prints, allowing us to debug ICEs
- the proof tree builder ends up going out of sync with the actual runtime behavior, which is confusing
- using shallow proof trees is a lot more performant as we frequently do not recurse into all nested goals when using an analyse visitor
- this allows us to clean up the implementation and remove some code
r? ```@compiler-errors```
We already handle this case this way on the coherence side, and it matches the new solver's behaviour. While there is some breakage around type-alias-impl-trait (see new "type annotations needed" in tests/ui/type-alias-impl-trait/issue-84660-unsoundness.rs), no stable code breaks, and no new stable code is accepted.
* instead simply set the primary message inside the lint decorator functions
* it used to be this way before [#]101986 which introduced `msg` to prevent
good path delayed bugs (which no longer exist) from firing under certain
circumstances when lints were suppressed / silenced
* this is no longer necessary for various reasons I presume
* it shaves off complexity and makes further changes easier to implement
An async closure may implement `FnMut`/`Fn` if it has no self-borrows
There's no reason that async closures may not implement `FnMut` or `Fn` if they don't actually borrow anything with the closure's env lifetime. Specifically, #123660 made it so that we don't always need to borrow captures from the closure's env.
See the doc comment on `should_reborrow_from_env_of_parent_coroutine_closure`:
c00957a3e2/compiler/rustc_hir_typeck/src/upvar.rs (L1777-L1823)
If there are no such borrows, then we are free to implement `FnMut` and `Fn` as permitted by our closure's inferred `ClosureKind`.
As far as I can tell, this change makes `async || {}` work in precisely the set of places they used to work before #120361.
Fixes#125247.
r? oli-obk
Move `#[do_not_recommend]` to the `#[diagnostic]` namespace
This commit moves the `#[do_not_recommend]` attribute to the `#[diagnostic]` namespace. It still requires
`#![feature(do_not_recommend)]` to work.
r? `@compiler-errors`
Uplift `RegionVid`, `TermKind` to `rustc_type_ir`, and `EagerResolver` to `rustc_next_trait_solver`
- Uplift `RegionVid`. This was complicated due to the fact that we implement `polonius_engine::Atom` for `RegionVid` -- but I just separated that into `PoloniusRegionVid`, and added `From`/`Into` impls so it can be defined in `rustc_borrowck` separately. Coherence 😵
- Change `InferCtxtLike` to expose `opportunistically_resolve_{ty,ct,lt,int,float}_var` so that we can uplift `EagerResolver` for use in the canonicalization methods.
- Uplift `TermKind` much like `GenericArgKind`
All of this is miscellaneous dependencies for making more `EvalCtxt` methods generic.
This change tweaks the error message generation to actually use the
`#[do_not_recommend]` attribute if present by just skipping the marked
trait impl in favour of the parent impl. It also adds a compile test for
this behaviour. Without this change the test would output the following
error:
```
error[E0277]: the trait bound `&str: Expression` is not satisfied
--> /home/weiznich/Documents/rust/rust/tests/ui/diagnostic_namespace/do_not_recommend.rs:53:15
|
LL | SelectInt.check("bar");
| ^^^^^ the trait `Expression` is not implemented for `&str`, which is required by `&str: AsExpression<Integer>`
|
= help: the following other types implement trait `Expression`:
Bound<T>
SelectInt
note: required for `&str` to implement `AsExpression<Integer>`
--> /home/weiznich/Documents/rust/rust/tests/ui/diagnostic_namespace/do_not_recommend.rs:26:13
|
LL | impl<T, ST> AsExpression<ST> for T
| ^^^^^^^^^^^^^^^^ ^
LL | where
LL | T: Expression<SqlType = ST>,
| ------------------------ unsatisfied trait bound introduced here
```
Note how that mentions `&str: Expression` before and now mentions `&str:
AsExpression<Integer>` instead which is much more helpful for users.
Open points for further changes before stabilization:
* We likely want to move the attribute to the `#[diagnostic]` namespace
to relax the guarantees given?
* How does it interact with the new trait solver?
Uplift more query stuff
- Uplift various query input/response internals
- Uplift the `ProofTree` structures and make the `ProofTreeBuilder` stuff (mostly) generic over `Interner`
- Stop using `TyCtxt::def_kind` in favor of `AliasTerm::kind`
r? lcnr
Only make GAT ambiguous in `match_projection_projections` considering shallow resolvability
In #123537, I tweaked the hack from #93892 to use `resolve_vars_if_possible` instead of `shallow_resolve`. This considers more inference guidance ambiguous. This resulted in crater regressions in #125196.
I've effectively reverted the change to the old behavior. That being said, I don't *like* this behavior, but I'd rather keep it for now since #123537 was not meant to make any behavioral changes. See the attached example.
This also affects the new solver, for the record, which doesn't have any rules about not guiding inference from param-env candidates which may constrain GAT args as a side-effect.
r? `@lcnr` or `@jackh726`
Rename Unsafe to Safety
Alternative to #124455, which is to just have one Safety enum to use everywhere, this opens the posibility of adding `ast::Safety::Safe` that's useful for unsafe extern blocks.
This leaves us today with:
```rust
enum ast::Safety {
Unsafe(Span),
Default,
// Safe (going to be added for unsafe extern blocks)
}
enum hir::Safety {
Unsafe,
Safe,
}
```
We would convert from `ast::Safety::Default` into the right Safety level according the context.
Split out `ty::AliasTerm` from `ty::AliasTy`
Splitting out `AliasTerm` (for use in project and normalizes goals) and `AliasTy` (for use in `ty::Alias`)
r? lcnr
solve: all "non-structural" logging to trace
This enables us to start with `RUSTC_LOG=rustc_trait_selection::solve=debug` to figure out *where* something went wrong, to then separately use `trace` to get to the details.
r? ``@compiler-errors``
Eliminate some `FIXME(lcnr)` comments
In some cases this involved changing code. In some cases the comment was able to removed or replaced.
r? ``@lcnr``
`InferCtxt::next_{ty,const}_var*` all take an origin, but the
`param_def_id` is almost always `None`. This commit changes them to just
take a `Span` and build the origin within the method, and adds new
methods for the rare cases where `param_def_id` might not be `None`.
This avoids a lot of tedious origin building.
Specifically:
- next_ty_var{,_id_in_universe,_in_universe}: now take `Span` instead of
`TypeVariableOrigin`
- next_ty_var_with_origin: added
- next_const_var{,_in_universe}: takes Span instead of ConstVariableOrigin
- next_const_var_with_origin: added
- next_region_var, next_region_var_in_universe: these are unchanged,
still take RegionVariableOrigin
The API inconsistency (ty/const vs region) seems worth it for the
large conciseness improvements.
Don't ICE when we cannot eval a const to a valtree in the new solver
Use `const_eval_resolve` instead of `try_const_eval_resolve` because naming aside, the former doesn't ICE when a value can't be evaluated to a valtree.
r? lcnr
borrowck: prepopulate opaque storage more eagerly
otherwise we ICE due to ambiguity when normalizing while computing implied bounds.
r? ``@compiler-errors``
Record impl args in the proof tree in new solver
Rather than rematching them during select.
Also use `ImplSource::Param` instead of `ImplSource::Builtin` for alias-bound candidates, so we don't ICE in `Instance::resolve`.
r? lcnr
Don't consider candidates with no failing where clauses when refining obligation causes in new solver
Improves error messages when we have param-env candidates that don't deeply unify (i.e. after alias-bounds).
r? lcnr
Prefer lower vtable candidates in select in new solver
Also, adjust the select visitor to only winnow when the *parent* goal is `Certainty::Yes`. This means that we won't winnow in cases when we have any ambiguous inference guidance from two candidates.
r? lcnr
Implement `do_not_recommend` in the new solver
Put the test into `diagnostic_namespace` test folder even though it's not in the diagnostic namespace, because it should be soon.
r? lcnr
cc `@weiznich`
Use a proof tree visitor to refine the `Obligation` for error reporting in new solver
With the magic of `ProofTreeVisitor`, we can close the gap that we have on `ObligationCause`s being not as descriptive in the new trait solver.
r? lcnr
Needs some work and obviously documentation.
Use `tcx.types.unit` instead of `Ty::new_unit(tcx)`
I don't think there is any need for the function, given that we can just access the `.types`, similarly to all other primitives?
Rewrite select (in the new solver) to use a `ProofTreeVisitor`
We can use a proof tree visitor rather than collecting and recomputing all the nested goals ourselves.
Based on #124415
Cleanup: Replace item names referencing GitHub issues or error codes with something more meaningful
**lcnr** in https://github.com/rust-lang/rust/pull/117164#pullrequestreview-1969935387:
> […] while I know that there's precendent to name things `Issue69420`, I really dislike this as it requires looking up the issue to figure out the purpose of such a variant. Actually referring to the underlying issue, e.g. `AliasMayNormToUncovered` or whatever and then linking to the issue in a doc comment feels a lot more desirable to me. We should ideally rename all the functions and enums which currently use issue numbers.
I've grepped through `compiler/` like crazy and think that I've found all instances of this pattern.
However, I haven't renamed `compute_2229_migrations_*`. Should I?
The first commit introduces an abhorrent and super long name for an item because naming is hard but also scary looking / unwelcoming names are good for things related to temporary-ish backcompat hacks. I'll let you discover it by yourself.
Contains a bit of drive-by cleanup and a diag migration bc that was the simplest option.
r? lcnr or compiler
Lazily normalize inside trait ref during orphan check & consider ty params in rigid alias types to be uncovered
Fixes#99554, fixesrust-lang/types-team#104.
Fixes#114061.
Supersedes #100555.
Tracking issue for the future compatibility lint: #124559.
r? lcnr
Remove many `#[macro_use] extern crate foo` items
This requires the addition of more `use` items, which often make the code more verbose. But they also make the code easier to read, because `#[macro_use]` obscures where macros are defined.
r? `@fee1-dead`
`obligations_for_self_ty`: use `ProofTreeVisitor` for nested goals
As always, dealing with proof trees continues to be a hacked together mess. After this PR and #124380 the only remaining blocker for core is https://github.com/rust-lang/trait-system-refactor-initiative/issues/90. There is also a `ProofTreeVisitor` issue causing an ICE when compiling `alloc` which I will handle in a separate PR. This issue likely affects coherence diagnostics more generally.
The core idea is to extend the proof tree visitor to support visiting nested candidates without using a `probe`. We then simply recurse into nested candidates if they are the only potentially applicable candidate for a given goal and check whether the self type matches the expected one.
For that to work, we need to improve `CanonicalState` to also handle unconstrained inference variables created inside of the trait solver. This is done by extending the `var_values` of `CanoncalState` with each fresh inference variables. Furthermore, we also store the state of all inference variables at the end of each probe. When recursing into `InspectCandidates` we then unify the values of all these states.
r? `@compiler-errors`
Remove special-casing for `SimplifiedType` for next solver
It's unnecessary due to the way that we fully normalize the self type before assembly begins.
r? lcnr
uses a `ProofTreeVisitor` to look into nested
goals when looking at the pending obligations
during hir typeck. Used by closure signature
inference, coercion, and for async functions.
```
error[E0507]: cannot move out of `bar`, a captured variable in an `FnMut` closure
--> $DIR/borrowck-move-by-capture.rs:9:29
|
LL | let bar: Box<_> = Box::new(3);
| --- captured outer variable
LL | let _g = to_fn_mut(|| {
| -- captured by this `FnMut` closure
LL | let _h = to_fn_once(move || -> isize { *bar });
| ^^^^^^^^^^^^^^^^ ----
| | |
| | variable moved due to use in closure
| | move occurs because `bar` has type `Box<isize>`, which does not implement the `Copy` trait
| `bar` is moved here
|
help: clone the value before moving it into the closure
|
LL ~ let value = bar.clone();
LL ~ let _h = to_fn_once(move || -> isize { value });
|
```
Start pointing to where bindings were declared when they are captured in closures:
```
error[E0597]: `x` does not live long enough
--> $DIR/suggest-return-closure.rs:23:9
|
LL | let x = String::new();
| - binding `x` declared here
...
LL | |c| {
| --- value captured here
LL | x.push(c);
| ^ borrowed value does not live long enough
...
LL | }
| -- borrow later used here
| |
| `x` dropped here while still borrowed
```
Suggest cloning in more cases involving closures:
```
error[E0507]: cannot move out of `foo` in pattern guard
--> $DIR/issue-27282-move-ref-mut-into-guard.rs:11:19
|
LL | if { (|| { let mut bar = foo; bar.take() })(); false } => {},
| ^^ --- move occurs because `foo` has type `&mut Option<&i32>`, which does not implement the `Copy` trait
| |
| `foo` is moved here
|
= note: variables bound in patterns cannot be moved from until after the end of the pattern guard
help: consider cloning the value if the performance cost is acceptable
|
LL | if { (|| { let mut bar = foo.clone(); bar.take() })(); false } => {},
| ++++++++
```
Use `DefiningOpaqueTypes::Yes` in rustdoc, where the `InferCtxt` is guaranteed to have no opaque types it can define
r? `@lcnr`
I manually checked there it's always `tcx.infer_ctxt().build()`
Add simple async drop glue generation
This is a prototype of the async drop glue generation for some simple types. Async drop glue is intended to behave very similar to the regular drop glue except for being asynchronous. Currently it does not execute synchronous drops but only calls user implementations of `AsyncDrop::async_drop` associative function and awaits the returned future. It is not complete as it only recurses into arrays, slices, tuples, and structs and does not have same sensible restrictions as the old `Drop` trait implementation like having the same bounds as the type definition, while code assumes their existence (requires a future work).
This current design uses a workaround as it does not create any custom async destructor state machine types for ADTs, but instead uses types defined in the std library called future combinators (deferred_async_drop, chain, ready_unit).
Also I recommend reading my [explainer](https://zetanumbers.github.io/book/async-drop-design.html).
This is a part of the [MCP: Low level components for async drop](https://github.com/rust-lang/compiler-team/issues/727) work.
Feature completeness:
- [x] `AsyncDrop` trait
- [ ] `async_drop_in_place_raw`/async drop glue generation support for
- [x] Trivially destructible types (integers, bools, floats, string slices, pointers, references, etc.)
- [x] Arrays and slices (array pointer is unsized into slice pointer)
- [x] ADTs (enums, structs, unions)
- [x] tuple-like types (tuples, closures)
- [ ] Dynamic types (`dyn Trait`, see explainer's [proposed design](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#async-drop-glue-for-dyn-trait))
- [ ] coroutines (https://github.com/rust-lang/rust/pull/123948)
- [x] Async drop glue includes sync drop glue code
- [x] Cleanup branch generation for `async_drop_in_place_raw`
- [ ] Union rejects non-trivially async destructible fields
- [ ] `AsyncDrop` implementation requires same bounds as type definition
- [ ] Skip trivially destructible fields (optimization)
- [ ] New [`TyKind::AdtAsyncDestructor`](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#adt-async-destructor-types) and get rid of combinators
- [ ] [Synchronously undroppable types](https://github.com/zetanumbers/posts/blob/main/async-drop-design.md#exclusively-async-drop)
- [ ] Automatic async drop at the end of the scope in async context
Rollup of 7 pull requests
Successful merges:
- #123673 (Don't ICE for kind mismatches during error rendering)
- #123675 (Taint const qualifs if a static is referenced that didn't pass wfcheck)
- #123975 (Port the 2 `rust-lld` run-make tests to `rmake`)
- #124000 (Use `/* value */` as a placeholder)
- #124013 (Box::into_raw: make Miri understand that this is a box-to-raw cast)
- #124027 (Prefer identity equality over equating types during coercion.)
- #124036 (Remove `default_hidden_visibility: false` from wasm targets)
r? `@ghost`
`@rustbot` modify labels: rollup
Use `/* value */` as a placeholder
The expression `value` isn't a valid suggestion; let's use `/* value */` as a placeholder (which is also invalid) since it more clearly signals to the user that they need to fill it in with something meaningful. This parallels the suggestions we have in a couple other places, like arguments.
We could also print the type name instead of `/* value */`, especially if it's suggestable, but I don't care strongly about that.
Trait predicates for types which have errors may still
evaluate to OK leading to downstream ICEs. Now we return
a selection error for such types in candidate assembly and
thereby prevent such issues
Remove `TypeVariableOriginKind` and `ConstVariableOriginKind`
It's annoying to have to import `TypeVariableOriginKind` just to fill it with `MiscVariable` for almost every use. Every other usage other than `TypeParameterDefinition` wasn't even used -- I can see how it may have been useful once for debugging, but I do quite a lot of typeck debugging and I've never really needed it.
So let's just remove it, and keep around the only useful thing which is the `DefId` of the param for `var_for_def`.
This is based on #123006, which removed the special use of `TypeVariableOriginKind::OpaqueInference`, which I'm pretty sure I was the one that added.
r? lcnr or re-roll to types
Fix various bugs in `ty_kind_suggestion`
Consolidates two implementations of `ty_kind_suggestion`
Fixes some misuse of the empty param-env
Fixes a problem where we suggested `(42)` instead of `(42,)` for tuple suggestions
Suggest a value when `return;`, making it consistent with `break;`
Fixes#123906
Discard overflow obligations in `impl_may_apply`
Hacky fix for #123493. Throws away obligations that are overflowing in `impl_may_apply` when we recompute if an impl applies, since those will lead to fatal overflow if processed during fulfillment.
Something about #114811 (I think it's the predicate reordering) caused us to evaluate predicates differently in error reporting leading to fatal overflow, though I believe the underlying overflow is possible to hit since this code was rewritten to use fulfillment.
Fixes#123493
Don't rely on upvars being assigned just because coroutine-closure kind is assigned
Previously, code relied on the implicit assumption that if a coroutine-closure's kind variable was constrained, then its upvars were also constrained. This is because we assign all of them at once at the end up upvar analysis.
However, there's another way that a coroutine-closure's kind can be constrained: from a signature hint in closure signature deduction. After #123350, we use these hints, which means the implicit assumption above no longer holds.
This PR adds the necessary checks so that we don't ICE.
r? oli-obk
Pass list of defineable opaque types into canonical queries
This eliminates `DefiningAnchor::Bubble` for good and brings the old solver closer to the new one wrt cycles and nested obligations. At that point the difference between `DefiningAnchor::Bind([])` and `DefiningAnchor::Error` was academic. We only used the difference for some sanity checks, which actually had to be worked around in places, so I just removed `DefiningAnchor` entirely and just stored the list of opaques that may be defined.
fixes#108498
fixes https://github.com/rust-lang/rust/issues/116877
* [x] run crater
- https://github.com/rust-lang/rust/pull/122077#issuecomment-2013293931
Safe Transmute: Compute transmutability from `rustc_target::abi::Layout`
In its first step of computing transmutability, `rustc_transmutability` constructs a byte-level representation of type layout (`Tree`). Previously, this representation was computed for ADTs by inspecting the ADT definition and performing our own layout computations. This process was error-prone, verbose, and limited our ability to analyze many types (particularly default-repr types).
In this PR, we instead construct `Tree`s from `rustc_target::abi::Layout`s. This helps ensure that layout optimizations are reflected our analyses, and increases the kinds of types we can now analyze, including:
- default repr ADTs
- transparent unions
- `UnsafeCell`-containing types
Overall, this PR expands the expressvity of `rustc_transmutability` to be much closer to the transmutability analysis performed by miri. Future PRs will work to close the remaining gaps (e.g., support for `Box`, raw pointers, `NonZero*`, coroutines, etc.).
r? `@compiler-errors`
In its first step of computing transmutability, `rustc_transmutability`
constructs a byte-level representation of type layout (`Tree`). Previously, this
representation was computed for ADTs by inspecting the ADT definition and
performing our own layout computations. This process was error-prone, verbose,
and limited our ability to analyze many types (particularly default-repr types).
In this PR, we instead construct `Tree`s from `rustc_target::abi::Layout`s. This
helps ensure that layout optimizations are reflected our analyses, and increases
the kinds of types we can now analyze, including:
- default repr ADTs
- transparent unions
- `UnsafeCell`-containing types
Overall, this PR expands the expressvity of `rustc_transmutability` to be much
closer to the transmutability analysis performed by miri. Future PRs will work
to close the remaining gaps (e.g., support for `Box`, raw pointers, `NonZero*`,
coroutines, etc.).
Check def id before calling `match_projection_projections`
When I "inlined" `assemble_candidates_from_predicates` into `for_each_item_bound` in #120584, I forgot to copy over the check that actually made sure the def id of the candidate was equal to the def id of the obligation. This means that we normalize goal a bit too often even if it's not productive to do so.
This PR adds that def id check back.
Fixes#123448
do not ICE in `fn forced_ambiguity` if we get an error
see the comment. currently causing an ICE in typenum which we've been unable to minimize.
r? `@compiler-errors`
Cleanup: Rename `HAS_PROJECTIONS` to `HAS_ALIASES` etc.
The name of the bitflag `HAS_PROJECTIONS` and of its corresponding method `has_projections` is quite historical dating back to a time when projections were the only kind of alias type.
I think it's time to update it to clear up any potential confusion for newcomers and to reduce unnecessary friction during contributor onboarding.
r? types
change `NormalizesTo` to fully structurally normalize
notes in https://hackmd.io/wZ016dE4QKGIhrOnHLlThQ
need to also update the dev-guide once this PR lands. in short, the setup is now as follows:
`normalizes-to` internally implements one step normalization, applying that normalization to the `goal.predicate.term` causes the projected term to get recursively normalized. With this `normalizes-to` normalizes until the projected term is rigid, meaning that we normalize as many steps necessary, but at least 1.
To handle rigid aliases, we add another candidate only if the 1 to inf step normalization failed. With this `normalizes-to` is now full structural normalization. We can now change `AliasRelate` to simply emit `normalizes-to` goals for the rhs and lhs.
This avoids the concerns from https://github.com/rust-lang/trait-system-refactor-initiative/issues/103 and generally feels cleaner
some smaller DefiningOpaqueTypes::No -> Yes switches
r? `@compiler-errors`
These are some easy cases, so let's get them out of the way first.
I added tests exercising the specialization code paths that I believe weren't tested so far.
follow-up to https://github.com/rust-lang/rust/pull/117348
The reason is that in specialization graph computation we use `DefiningAnchor::Error`, so there's no difference anyway. And in the other use cases, we
* already errored in the specialization_graph computation, or
* already errored in coherence, or
* are comparing opaque types with inference variables already, or
* there are no opaque types involved
check `FnDef` return type for WF
better version of #106807, fixes#84533 (mostly). It's not perfect given that we still ignore WF requirements involving bound regions but I wasn't able to quickly write an example, so even if theoretically exploitable, it should be far harder to trigger.
This is strictly more restrictive than checking the return type for WF as part of the builtin `FnDef: FnOnce` impl (#106807) and moving to this approach in the future will not break any code.
~~It also agrees with my theoretical view of how this should behave~~
r? types
instantiate higher ranked goals outside of candidate selection
This PR modifies `evaluate` to more eagerly instantiate higher-ranked goals, preventing the `leak_check` during candidate selection from detecting placeholder errors involving that binder.
For a general background regarding higher-ranked region solving and the leak check, see https://hackmd.io/qd9Wp03cQVy06yOLnro2Kg.
> The first is something called the **leak check**. You can think of it as a "quick and dirty" approximation for the region check, which will come later. The leak check detects some kinds of errors early, essentially deciding between "this set of outlives constraints are guaranteed to result in an error eventually" or "this set of outlives constraints may be solvable".
## The ideal future
We would like to end up with the following idealized design to handle universal binders:
```rust
fn enter_forall<'tcx, T, R>(
forall: Binder<'tcx, T>,
f: impl FnOnce(T) -> R,
) -> R {
let new_universe = infcx.increment_universe_index();
let value = instantiate_binder_with_placeholders_in(new_universe, forall);
let result = f(value);
eagerly_handle_higher_ranked_region_constraints_in(new_universe);
infcx.decrement_universe_index();
assert!(!result.has_placeholders_in_or_above(new_universe));
result
}
```
That is, when universally instantiating a binder, anything using the placeholders has to happen inside of a limited scope (the closure `f`). After this closure has completed, all constraints involving placeholders are known.
We then handle any *external constraints* which name these placeholders. We destructure `TypeOutlives` constraints involving placeholders and eagerly handle any region constraints involving these placeholders. We do not return anything mentioning the placeholders created inside of this function to the caller.
Being able to eagerly handle *all* region constraints involving placeholders will be difficult due to complex `TypeOutlives` constraints, involving inference variables or alias types, and higher ranked implied bounds. The exact issues and possible solutions are out of scope of this FCP.
#### How does the leak check fit into this
The `leak_check` is an underapproximation of `eagerly_handle_higher_ranked_region_constraints_in`. It detects some kinds of errors involving placeholders from `new_universe`, but not all of them.
It only looks at region outlives constraints, ignoring `TypeOutlives`, and checks whether one of the following two conditions are met for **placeholders in or above `new_universe`**, in which case it results in an error:
- `'!p1: '!p2` a placeholder `'!p2` outlives a different placeholder `'!p1`
- `'!p1: '?2` an inference variable `'?2` outlives a placeholder `'!p1` *which it cannot name*
It does not handle all higher ranked region constraints, so we still return constraints involving placeholders from `new_universe` which are then (re)checked by `lexical_region_resolve` or MIR borrowck.
As we check higher ranked constraints in the full regionck anyways, the `leak_check` is not soundness critical. It's current only purpose is to move some higher ranked region errors earlier, enabling it to guide type inference and trait solving. Adding additional uses of the `leak_check` in the future would only strengthen inference and is therefore not breaking.
## Where do we use currently use the leak check
The `leak_check` is currently used in two places:
Coherence does not use a proper regionck, only relying on the `leak_check` called [at the end of the implicit negative overlap check](8b94152af6/compiler/rustc_trait_selection/src/traits/coherence.rs (L235-L238)). During coherence all parameters are instantiated with inference variables, so the only possible region errors are higher-ranked. We currently also sometimes make guesses when destructuring `TypeOutlives` constraints which can theoretically result in incorrect errors. This could result in overlapping impls.
We also use the `leak_check` [at the end of `fn evaluation_probe`](8b94152af6/compiler/rustc_trait_selection/src/traits/select/mod.rs (L607-L610)). This function is used during candidate assembly for `Trait` goals. Most notably we use [inside of `evaluate_candidate` during winnowing](0e4243538b/compiler/rustc_trait_selection/src/traits/select/mod.rs (L491-L502)). Conceptionally, it is as if we compute each candidate in a separate `enter_forall`.
## The current use in `fn evaluation_probe` is undesirable
Because we only instantiate a higher-ranked goal once inside of `fn evaluation_probe`, errors involving placeholders from that binder can impact selection. This results in inconsistent behavior ([playground](
*[playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=dac60ebdd517201788899ffa77364831)*)):
```rust
trait Leak<'a> {}
impl Leak<'_> for Box<u32> {}
impl Leak<'static> for Box<u16> {}
fn impls_leak<T: for<'a> Leak<'a>>() {}
trait IndirectLeak<'a> {}
impl<'a, T: Leak<'a>> IndirectLeak<'a> for T {}
fn impls_indirect_leak<T: for<'a> IndirectLeak<'a>>() {}
fn main() {
// ok
//
// The `Box<u16>` impls fails the leak check,
// meaning that we apply the `Box<u32>` impl.
impls_leak::<Box<_>>();
// error: type annotations needed
//
// While the `Box<u16>` impl would fail the leak check
// we have already instantiated the binder while applying
// the generic `IndirectLeak` impl, so during candidate
// selection of `Leak` we do not detect the placeholder error.
// Evaluation of `Box<_>: Leak<'!a>` is therefore ambiguous,
// resulting in `for<'a> Box<_>: Leak<'a>` also being ambiguous.
impls_indirect_leak::<Box<_>>();
}
```
We generally prefer `where`-bounds over implementations during candidate selection, both for [trait goals](11f32b73e0/compiler/rustc_trait_selection/src/traits/select/mod.rs (L1863-L1887)) and during [normalization](11f32b73e0/compiler/rustc_trait_selection/src/traits/project.rs (L184-L198)). However, we currently **do not** use the `leak_check` during candidate assembly in normalizing. This can result in inconsistent behavior:
```rust
trait Trait<'a> {
type Assoc;
}
impl<'a, T> Trait<'a> for T {
type Assoc = usize;
}
fn trait_bound<T: for<'a> Trait<'a>>() {}
fn projection_bound<T: for<'a> Trait<'a, Assoc = usize>>() {}
// A function with a trivial where-bound which is more
// restrictive than the impl.
fn function<T: Trait<'static, Assoc = usize>>() {
// ok
//
// Proving `for<'a> T: Trait<'a>` using the where-bound results
// in a leak check failure, so we use the more general impl,
// causing this to succeed.
trait_bound::<T>();
// error
//
// Proving the `Projection` goal `for<'a> T: Trait<'a, Assoc = usize>`
// does not use the leak check when trying the where-bound, causing us
// to prefer it over the impl, resulting in a placeholder error.
projection_bound::<T>();
// error
//
// Trying to normalize the type `for<'a> fn(<T as Trait<'a>>::Assoc)`
// only gets to `<T as Trait<'a>>::Assoc` once `'a` has been already
// instantiated, causing us to prefer the where-bound over the impl
// resulting in a placeholder error. Even if were were to also use the
// leak check during candidate selection for normalization, this
// case would still not compile.
let _higher_ranked_norm: for<'a> fn(<T as Trait<'a>>::Assoc) = |_| ();
}
```
This is also likely to be more performant. It enables more caching in the new trait solver by simply [recursively calling the canonical query][new solver] after instantiating the higher-ranked goal.
It is also unclear how to add the leak check to normalization in the new solver. To handle https://github.com/rust-lang/trait-system-refactor-initiative/issues/1 `Projection` goals are implemented via `AliasRelate`. This again means that we instantiate the binder before ever normalizing any alias. Even if we were to avoid this, we lose the ability to [cache normalization by itself, ignoring the expected `term`](5bd5d214ef/compiler/rustc_trait_selection/src/solve/normalizes_to/mod.rs (L34-L49)). We cannot replace the `term` with an inference variable before instantiating the binder, as otherwise `for<'a> T: Trait<Assoc<'a> = &'a ()>` breaks. If we only replace the term after instantiating the binder, we cannot easily evaluate the goal in a separate context, as [we'd then lose the information necessary for the leak check](11f32b73e0/compiler/rustc_next_trait_solver/src/canonicalizer.rs (L230-L232)). Adding this information to the canonical input also seems non-trivial.
## Proposed solution
I propose to instantiate the binder outside of candidate assembly, causing placeholders from higher-ranked goals to get ignored while selecting their candidate. This mostly[^1] matches the [current behavior of the new solver][new solver]. The impact of this change is therefore as follows:
```rust
trait Leak<'a> {}
impl Leak<'_> for Box<u32> {}
impl Leak<'static> for Box<u16> {}
fn impls_leak<T: for<'a> Leak<'a>>() {}
trait IndirectLeak<'a> {}
impl<'a, T: Leak<'a>> IndirectLeak<'a> for T {}
fn impls_indirect_leak<T: for<'a> IndirectLeak<'a>>() {}
fn guide_selection() {
// ok -> ambiguous
impls_leak::<Box<_>>();
// ambiguous
impls_indirect_leak::<Box<_>>();
}
trait Trait<'a> {
type Assoc;
}
impl<'a, T> Trait<'a> for T {
type Assoc = usize;
}
fn trait_bound<T: for<'a> Trait<'a>>() {}
fn projection_bound<T: for<'a> Trait<'a, Assoc = usize>>() {}
// A function which a trivial where-bound which is more
// restrictive than the impl.
fn function<T: Trait<'static, Assoc = usize>>() {
// ok -> error
trait_bound::<T>();
// error
projection_bound::<T>();
// error
let _higher_ranked_norm: for<'a> fn(<T as Trait<'a>>::Assoc) = |_| ();
}
```
This does not change the behavior if candidates have higher ranked nested goals, as in this case the `leak_check` causes the nested goal to result in an error ([playground](https://play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=a74c25300b23db9022226de99d8a2fa6)):
```rust
trait LeakCheckFailure<'a> {}
impl LeakCheckFailure<'static> for () {}
trait Trait<T> {}
impl Trait<u32> for () where for<'a> (): LeakCheckFailure<'a> {}
impl Trait<u16> for () {}
fn impls_trait<T: Trait<U>, U>() {}
fn main() {
// ok
//
// It does not matter whether candidate assembly
// considers the placeholders from higher-ranked goal.
//
// Either `for<'a> (): LeakCheckFailure<'a>` has no
// applicable candidate or it has a single applicable candidate
// when then later results in an error. This allows us to
// infer `U` to `u16`.
impls_trait::<(), _>()
}
```
## Impact on existing crates
This is a **breaking change**. [A crater run](https://github.com/rust-lang/rust/pull/119820#issuecomment-1926862174) found 17 regressed crates with 7 root causes.
For a full analysis of all affected crates, see https://gist.github.com/lcnr/7c1c652f30567048ea240554a36ed95c.
---
I believe this breakage to be acceptable and would merge this change. I am confident that the new position of the leak check matches our idealized future and cannot envision any other consistent alternative. Where possible, I intend to open PRs fixing/avoiding the regressions before landing this PR.
I originally intended to remove the `coherence_leak_check` lint in the same PR. However, while I am confident in the *position* of the leak check, deciding on its exact behavior is left as future work, cc #112999. This PR therefore only moves the leak check while keeping the lint when relying on it in coherence.
[new solver]: https://github.com/rust-lang/rust/blob/master/compiler/rustc_trait_selection/src/solve/eval_ctxt/mod.rs#L479-L484
[^1]: the new solver has a separate cause of inconsistent behavior rn https://github.com/rust-lang/trait-system-refactor-initiative/issues/53#issuecomment-1914310171
r? `@nikomatsakis`
Check `x86_64` size assertions on `aarch64`, too
(Context: https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/Checking.20size.20assertions.20on.20aarch64.3F)
Currently the compiler has around 30 sets of `static_assert_size!` for various size-critical data structures (e.g. various IR nodes), guarded by `#[cfg(all(target_arch = "x86_64", target_pointer_width = "64"))]`.
(Presumably this cfg avoids having to maintain separate size values for 32-bit targets and unusual 64-bit targets. Apparently it may have been necessary before the i128/u128 alignment changes, too.)
This is slightly incovenient for people on aarch64 workstations (e.g. Macs), because the assertions normally aren't checked until we push to a PR. So this PR adds `aarch64` to the `#[cfg(..)]` guarding all of those assertions in the compiler.
---
Implemented with a simple find/replace. Verified by manually inspecting each `static_assert_size!` in `compiler/`, and checking that either the replacement succeeded, or adding aarch64 wouldn't have been appropriate.
Assert that args are actually compatible with their generics, rather than just their count
Right now we just check that the number of args is right, rather than actually checking the kinds. Uplift a helper fn that I wrote from trait selection to do just that. Found a couple bugs along the way.
r? `@lcnr` or `@fmease` (or anyone really lol)
Make inductive cycles always ambiguous
This makes inductive cycles always result in ambiguity rather than be treated like a stack-dependent error.
This has some interactions with specialization, and so breaks a few UI tests that I don't agree should've ever worked in the first place, and also breaks a handful of crates in a way that I don't believe is a problem.
On the bright side, it puts us in a better spot when it comes to eventually enabling coinduction everywhere.
## Results
This was cratered in https://github.com/rust-lang/rust/pull/116494#issuecomment-2008657494, which boils down to two regressions:
* `lu_packets` - This code should have never compiled in the first place. More below.
* **ALL** other regressions are due to `commit_verify@0.11.0-beta.1` (edit: and `commit_verify@0.10.x`) - This actually seems to be fixed in version `0.11.0-beta.5`, which is the *most* up to date version, but it's still prerelease on crates.io so I don't think cargo ends up picking `beta.5` when building dependent crates.
### `lu_packets`
Firstly, this crate uses specialization, so I think it's automatically worth breaking. However, I've minimized [the regression](https://crater-reports.s3.amazonaws.com/pr-116494-3/try%23d614ed876e31a5f3ad1d0fbf848fcdab3a29d1d8/gh/lcdr.lu_packets/log.txt) to:
```rust
// Upstream crate
pub trait Serialize {}
impl Serialize for &() {}
impl<S> Serialize for &[S] where for<'a> &'a S: Serialize {}
// ----------------------------------------------------------------------- //
// Downstream crate
#![feature(specialization)]
#![allow(incomplete_features, unused)]
use upstream::Serialize;
trait Replica {
fn serialize();
}
impl<T> Replica for T {
default fn serialize() {}
}
impl<T> Replica for Option<T>
where
for<'a> &'a T: Serialize,
{
fn serialize() {}
}
```
Specifically this fails when computing the specialization graph for the `downstream` crate.
The code ends up cycling on `&[?0]: Serialize` when we equate `&?0 = &[?1]` during impl matching, which ends up needing to prove `&[?1]: Serialize`, which since cycles are treated like ambiguity, ends up in a **fatal overflow**. For some reason this requires two crates, squashing them into one crate doesn't work.
Side-note: This code is subtly order dependent. When minimizing, I ended up having the code start failing on `nightly` very easily after removing and reordering impls. This seems to me all the more reason to remove this behavior altogether.
## Side-note: Item Bounds (edit: this was fixed independently in #121123)
Due to the changes in #120584 where we now consider an alias's item bounds *and* all the item bounds of the alias's nested self type aliases, I've had to add e6b64c6194 which is a hack to make sure we're not eagerly normalizing bounds that have nothing to do with the predicate we're trying to solve, and which result in.
This is fixed in a more principled way in #121123.
---
r? lcnr for an initial review
rustdoc: heavily simplify the synthesis of auto trait impls
`gd --numstat HEAD~2 HEAD src/librustdoc/clean/auto_trait.rs`
**+315 -705** 🟩🟥🟥🟥⬛
---
As outlined in issue #113015, there are currently 3[^1] large separate routines that “clean” `rustc_middle::ty` data types related to generics & predicates to rustdoc data types. Every single one has their own kinds of bugs. While I've patched a lot of bugs in each of the routines in the past, it's about time to unify them. This PR is only the first in a series. It completely **yanks** the custom “bounds cleaning” of mod `auto_trait` and reuses the routines found in mod `simplify`. As alluded to, `simplify` is also flawed but it's still more complete than `auto_trait`'s routines. [See also my review comment over at `tests/rustdoc/synthetic_auto/bounds.rs`](https://github.com/rust-lang/rust/pull/123340#discussion_r1546900539).
This is preparatory work for rewriting “bounds cleaning” from scratch in follow-up PRs in order to finally [fix] #113015.
Apart from that, I've eliminated all potential sources of *instability* in the rendered output.
See also #119597. I'm pretty sure this fixes#119597.
This PR does not attempt to fix [any other issues related to synthetic auto trait impls](https://github.com/rust-lang/rust/issues?q=is%3Aissue+is%3Aopen+label%3AA-synthetic-impls%20label%3AA-auto-traits).
However, it's definitely meant to be a *stepping stone* by making `auto_trait` more contributor-friendly.
---
* Replace `FxHash{Map,Set}` with `FxIndex{Map,Set}` to guarantee a stable iteration order
* Or as a perf opt, `UnordSet` (a thin wrapper around `FxHashSet`) in cases where we never iterate over the set.
* Yes, we do make use of `swap_remove` but that shouldn't matter since all the callers are deterministic. It does make the output less “predictable” but it's still better than before. Ofc, I rely on `rustc_infer` being deterministic. I hope that holds.
* Utilizing `clean::simplify` over the custom “bounds cleaning” routines wipes out the last reference to `collect_referenced_late_bound_regions` in rustdoc (`simplify` uses `bound_vars`) which was a source of instability / unpredictability (cc #116388)
* Remove the types `RegionTarget` and `RegionDeps` from `librustdoc`. They were duplicates of the identical types found in `rustc`. Just import them from `rustc`. For some reason, they were duplicated when splitting `auto_trait` in two in #49711.
* Get rid of the useless “type namespace” `AutoTraitFinder` in `librustdoc`
* The struct only held a `DocContext`, it was over-engineered
* Turn the associated functions into free ones
* Eliminates rightward drift; increases legibility
* `rustc` also contains a useless `AutoTraitFinder` struct but I plan on removing that in a follow-up PR
* Rename a bunch of methods to be way more descriptive
* Eliminate `use super::*;`
* Lead to `clean/mod.rs` accumulating a lot of unnecessary imports
* Made `auto_traits` less modular
* Eliminate a custom `TypeFolder`: We can just use the rustc helper `fold_regions` which does that for us
I plan on adding extensive documentation to `librustdoc`'s `auto_trait` in follow-up PRs.
I don't want to do that in this PR because further refactoring & bug fix PRs may alter the overall structure of `librustdoc`'s & `rustc`'s `auto_trait` modules to a great degree. I'm slowly digging into the dark details of `rustc`'s `auto_trait` module again and once I have the full picture I will be able to provide proper docs.
---
While this PR does indeed touch `rustc`'s `auto_trait` — mostly tiny refactorings — I argue this PR doesn't need any compiler reviewers next to rustdoc ones since that module falls under the purview of rustdoc — it used to be part of `librustdoc` after all (#49711).
Sorry for not having split this into more commits. If you'd like me to I can try to split it into more atomic commits retroactively. However, I don't know if that would actually make reviewing easier. I think the best way to review this might just be to place the master version of `auto_trait` on the left of your screen and the patched one on the right, not joking.
r? `@GuillaumeGomez`
[^1]: Or even 4 depending on the way you're counting.
