Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[perf] Try revert of #102472 #109780

Closed
wants to merge 1 commit into from
Closed

[perf] Try revert of #102472 #109780

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 14 additions & 3 deletions compiler/rustc_infer/src/infer/freshen.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,16 +43,18 @@ pub struct TypeFreshener<'a, 'tcx> {
const_freshen_count: u32,
ty_freshen_map: FxHashMap<ty::InferTy, Ty<'tcx>>,
const_freshen_map: FxHashMap<ty::InferConst<'tcx>, ty::Const<'tcx>>,
keep_static: bool,
}

impl<'a, 'tcx> TypeFreshener<'a, 'tcx> {
pub fn new(infcx: &'a InferCtxt<'tcx>) -> TypeFreshener<'a, 'tcx> {
pub fn new(infcx: &'a InferCtxt<'tcx>, keep_static: bool) -> TypeFreshener<'a, 'tcx> {
TypeFreshener {
infcx,
ty_freshen_count: 0,
const_freshen_count: 0,
ty_freshen_map: Default::default(),
const_freshen_map: Default::default(),
keep_static,
}
}

Expand Down Expand Up @@ -119,9 +121,18 @@ impl<'a, 'tcx> TypeFolder<TyCtxt<'tcx>> for TypeFreshener<'a, 'tcx> {
| ty::ReFree(_)
| ty::ReVar(_)
| ty::RePlaceholder(..)
| ty::ReStatic
| ty::ReError(_)
| ty::ReErased => self.interner().lifetimes.re_erased,
| ty::ReErased => {
// replace all free regions with 'erased
self.interner().lifetimes.re_erased
}
ty::ReStatic => {
if self.keep_static {
r
} else {
self.interner().lifetimes.re_erased
}
}
}
}

Expand Down
7 changes: 6 additions & 1 deletion compiler/rustc_infer/src/infer/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -713,7 +713,12 @@ impl<'tcx> InferCtxt<'tcx> {
}

pub fn freshener<'b>(&'b self) -> TypeFreshener<'b, 'tcx> {
freshen::TypeFreshener::new(self)
freshen::TypeFreshener::new(self, false)
}

/// Like `freshener`, but does not replace `'static` regions.
pub fn freshener_keep_static<'b>(&'b self) -> TypeFreshener<'b, 'tcx> {
freshen::TypeFreshener::new(self, true)
}

pub fn unsolved_variables(&self) -> Vec<Ty<'tcx>> {
Expand Down
164 changes: 78 additions & 86 deletions compiler/rustc_trait_selection/src/traits/select/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -210,7 +210,7 @@ impl<'cx, 'tcx> SelectionContext<'cx, 'tcx> {
pub fn new(infcx: &'cx InferCtxt<'tcx>) -> SelectionContext<'cx, 'tcx> {
SelectionContext {
infcx,
freshener: infcx.freshener(),
freshener: infcx.freshener_keep_static(),
intercrate_ambiguity_causes: None,
query_mode: TraitQueryMode::Standard,
}
Expand Down Expand Up @@ -769,16 +769,14 @@ impl<'cx, 'tcx> SelectionContext<'cx, 'tcx> {
}

ty::PredicateKind::Clause(ty::Clause::TypeOutlives(pred)) => {
// A global type with no free lifetimes or generic parameters
// outlives anything.
if pred.0.has_free_regions()
|| pred.0.has_late_bound_regions()
|| pred.0.has_non_region_infer()
|| pred.0.has_non_region_infer()
{
Ok(EvaluatedToOkModuloRegions)
} else {
// A global type with no late-bound regions can only
// contain the "'static" lifetime (any other lifetime
// would either be late-bound or local), so it is guaranteed
// to outlive any other lifetime
if pred.0.is_global() && !pred.0.has_late_bound_vars() {
Ok(EvaluatedToOk)
} else {
Ok(EvaluatedToOkModuloRegions)
}
}

Expand Down Expand Up @@ -1826,12 +1824,6 @@ enum DropVictim {
No,
}

impl DropVictim {
fn drop_if(should_drop: bool) -> DropVictim {
if should_drop { DropVictim::Yes } else { DropVictim::No }
}
}

/// ## Winnowing
///
/// Winnowing is the process of attempting to resolve ambiguity by
Expand Down Expand Up @@ -1897,7 +1889,11 @@ impl<'tcx> SelectionContext<'_, 'tcx> {
// or the current one if tied (they should both evaluate to the same answer). This is
// probably best characterized as a "hack", since we might prefer to just do our
// best to *not* create essentially duplicate candidates in the first place.
DropVictim::drop_if(other.bound_vars().len() <= victim.bound_vars().len())
if other.bound_vars().len() <= victim.bound_vars().len() {
DropVictim::Yes
} else {
DropVictim::No
}
} else if other.skip_binder().trait_ref == victim.skip_binder().trait_ref
&& victim.skip_binder().constness == ty::BoundConstness::NotConst
&& other.skip_binder().polarity == victim.skip_binder().polarity
Expand Down Expand Up @@ -1927,13 +1923,17 @@ impl<'tcx> SelectionContext<'_, 'tcx> {
| ObjectCandidate(_)
| ProjectionCandidate(..),
) => {
// We have a where clause so don't go around looking
// for impls. Arbitrarily give param candidates priority
// over projection and object candidates.
//
// Global bounds from the where clause should be ignored
// here (see issue #50825).
DropVictim::drop_if(!is_global(other_cand))
if is_global(other_cand) {
DropVictim::No
} else {
// We have a where clause so don't go around looking
// for impls. Arbitrarily give param candidates priority
// over projection and object candidates.
//
// Global bounds from the where clause should be ignored
// here (see issue #50825).
DropVictim::Yes
}
}
(ObjectCandidate(_) | ProjectionCandidate(..), ParamCandidate(ref victim_cand)) => {
// Prefer these to a global where-clause bound
Expand All @@ -1955,16 +1955,18 @@ impl<'tcx> SelectionContext<'_, 'tcx> {
) => {
// Prefer these to a global where-clause bound
// (see issue #50825).
DropVictim::drop_if(
is_global(victim_cand) && other.evaluation.must_apply_modulo_regions(),
)
if is_global(victim_cand) && other.evaluation.must_apply_modulo_regions() {
DropVictim::Yes
} else {
DropVictim::No
}
}

(ProjectionCandidate(i, _), ProjectionCandidate(j, _))
| (ObjectCandidate(i), ObjectCandidate(j)) => {
// Arbitrarily pick the lower numbered candidate for backwards
// compatibility reasons. Don't let this affect inference.
DropVictim::drop_if(i < j && !needs_infer)
if i < j && !needs_infer { DropVictim::Yes } else { DropVictim::No }
}
(ObjectCandidate(_), ProjectionCandidate(..))
| (ProjectionCandidate(..), ObjectCandidate(_)) => {
Expand Down Expand Up @@ -2015,65 +2017,55 @@ impl<'tcx> SelectionContext<'_, 'tcx> {
}
}

match tcx.impls_are_allowed_to_overlap(other_def, victim_def) {
// For #33140 the impl headers must be exactly equal, the trait must not have
// any associated items and there are no where-clauses.
//
// We can just arbitrarily drop one of the impls.
Some(ty::ImplOverlapKind::Issue33140) => {
assert_eq!(other.evaluation, victim.evaluation);
DropVictim::Yes
}
// For candidates which already reference errors it doesn't really
// matter what we do 🤷
Some(ty::ImplOverlapKind::Permitted { marker: false }) => {
DropVictim::drop_if(other.evaluation.must_apply_considering_regions())
}
Some(ty::ImplOverlapKind::Permitted { marker: true }) => {
// Subtle: If the predicate we are evaluating has inference
// variables, do *not* allow discarding candidates due to
// marker trait impls.
//
// Without this restriction, we could end up accidentally
// constraining inference variables based on an arbitrarily
// chosen trait impl.
//
// Imagine we have the following code:
//
// ```rust
// #[marker] trait MyTrait {}
// impl MyTrait for u8 {}
// impl MyTrait for bool {}
// ```
//
// And we are evaluating the predicate `<_#0t as MyTrait>`.
//
// During selection, we will end up with one candidate for each
// impl of `MyTrait`. If we were to discard one impl in favor
// of the other, we would be left with one candidate, causing
// us to "successfully" select the predicate, unifying
// _#0t with (for example) `u8`.
//
// However, we have no reason to believe that this unification
// is correct - we've essentially just picked an arbitrary
// *possibility* for _#0t, and required that this be the *only*
// possibility.
//
// Eventually, we will either:
// 1) Unify all inference variables in the predicate through
// some other means (e.g. type-checking of a function). We will
// then be in a position to drop marker trait candidates
// without constraining inference variables (since there are
// none left to constrain)
// 2) Be left with some unconstrained inference variables. We
// will then correctly report an inference error, since the
// existence of multiple marker trait impls tells us nothing
// about which one should actually apply.
DropVictim::drop_if(
!needs_infer && other.evaluation.must_apply_considering_regions(),
)
if other.evaluation.must_apply_considering_regions() {
match tcx.impls_are_allowed_to_overlap(other_def, victim_def) {
Some(ty::ImplOverlapKind::Permitted { marker: true }) => {
// Subtle: If the predicate we are evaluating has inference
// variables, do *not* allow discarding candidates due to
// marker trait impls.
//
// Without this restriction, we could end up accidentally
// constraining inference variables based on an arbitrarily
// chosen trait impl.
//
// Imagine we have the following code:
//
// ```rust
// #[marker] trait MyTrait {}
// impl MyTrait for u8 {}
// impl MyTrait for bool {}
// ```
//
// And we are evaluating the predicate `<_#0t as MyTrait>`.
//
// During selection, we will end up with one candidate for each
// impl of `MyTrait`. If we were to discard one impl in favor
// of the other, we would be left with one candidate, causing
// us to "successfully" select the predicate, unifying
// _#0t with (for example) `u8`.
//
// However, we have no reason to believe that this unification
// is correct - we've essentially just picked an arbitrary
// *possibility* for _#0t, and required that this be the *only*
// possibility.
//
// Eventually, we will either:
// 1) Unify all inference variables in the predicate through
// some other means (e.g. type-checking of a function). We will
// then be in a position to drop marker trait candidates
// without constraining inference variables (since there are
// none left to constrain)
// 2) Be left with some unconstrained inference variables. We
// will then correctly report an inference error, since the
// existence of multiple marker trait impls tells us nothing
// about which one should actually apply.
if needs_infer { DropVictim::No } else { DropVictim::Yes }
}
Some(_) => DropVictim::Yes,
None => DropVictim::No,
}
None => DropVictim::No,
} else {
DropVictim::No
}
}

Expand Down
6 changes: 1 addition & 5 deletions tests/ui/marker_trait_attr/overlap-marker-trait-with-static-lifetime.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,8 +1,4 @@
// known-bug: #89515
//
// The trait solver cannot deal with ambiguous marker trait impls
// if there are lifetimes involved. As we must not special-case any
// regions this does not work, even with 'static
// check-pass
#![feature(marker_trait_attr)]

#[marker]
Expand Down
31 changes: 0 additions & 31 deletions tests/ui/marker_trait_attr/overlap-marker-trait-with-static-lifetime.stderr
View file
Open in desktop

This file was deleted.

14 changes: 3 additions & 11 deletions tests/ui/marker_trait_attr/overlapping-impl-1-modulo-regions.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,17 +1,9 @@
// known-bug: #109481
//
// While the `T: Copy` is always applicable when checking
// that the impl `impl<T: Copy> F for T {}` is well formed,
// the old trait solver can only approximate this by checking
// that there are no inference variables in the obligation and
// no region constraints in the evaluation result.
//
// Because of this we end up with ambiguity here.
// check-pass
#![feature(marker_trait_attr)]

#[marker]
pub trait F {}
impl<T: Copy> F for T {}
impl<T: 'static> F for T {}
impl<T> F for T where T: Copy {}
impl<T> F for T where T: 'static {}

fn main() {}
14 changes: 0 additions & 14 deletions tests/ui/marker_trait_attr/overlapping-impl-1-modulo-regions.stderr
View file
Open in desktop

This file was deleted.