diff options
| author | Niko Matsakis <niko@alum.mit.edu> | 2020-05-19 01:09:40 +0000 |
|---|---|---|
| committer | Niko Matsakis <niko@alum.mit.edu> | 2020-06-22 14:33:44 +0000 |
| commit | f2cf9944831f15b1940da85c8fb1b419dec9f074 (patch) | |
| tree | d7f255b1b039ca14ef42a048ede6fc83da60275a /src/test/ui/hrtb | |
| parent | 4199b3ae26007eb9c871b57f56057da586bbd1db (diff) | |
| download | rust-f2cf9944831f15b1940da85c8fb1b419dec9f074.tar.gz rust-f2cf9944831f15b1940da85c8fb1b419dec9f074.zip | |
rewrite leak check to be based on universes
In the new leak check, instead of getting a list of placeholders to track, we look for any placeholder that is part of a universe which was created during the snapshot. We are looking for the following error patterns: * P1: P2, where P1 != P2 * P1: R, where R is in some universe that cannot name P1 This new leak check is more precise than before, in that it accepts this patterns: * R: P1, even if R cannot name P1, because R = 'static is a valid sol'n * R: P1, R: P2, as above Note that this leak check, when running during subtyping, is less efficient than before in some sense because it is going to check and re-check all the universes created since the snapshot. We're going to move when the leak check runs to try and correct that.
Diffstat (limited to 'src/test/ui/hrtb')
| -rw-r--r-- | src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.rs | 5 | ||||
| -rw-r--r-- | src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.stderr | 18 |
2 files changed, 2 insertions, 21 deletions
diff --git a/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.rs b/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.rs index 95b57d6c5bb..f95496a6c3c 100644 --- a/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.rs +++ b/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.rs @@ -2,6 +2,8 @@ // // In particular, we test this pattern in trait solving, where it is not connected // to any part of the source code. +// +// check-pass trait Trait<T> {} @@ -30,9 +32,6 @@ fn main() { // - `?b: ?a` -- solveable if `?b` is inferred to `'static` // - So the subtyping check succeeds, somewhat surprisingly. // This is because we can use `'static`. - // - // NB. *However*, the reinstated leak-check gives an error here. foo::<()>(); - //~^ ERROR not satisfied } diff --git a/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.stderr b/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.stderr deleted file mode 100644 index 1e335f9ee96..00000000000 --- a/src/test/ui/hrtb/hrtb-exists-forall-trait-covariant.stderr +++ /dev/null @@ -1,18 +0,0 @@ -error[E0277]: the trait bound `(): Trait<for<'b> fn(fn(&'b u32))>` is not satisfied - --> $DIR/hrtb-exists-forall-trait-covariant.rs:36:11 - | -LL | fn foo<T>() - | --- required by a bound in this -LL | where -LL | T: Trait<for<'b> fn(fn(&'b u32))>, - | ------------------------------ required by this bound in `foo` -... -LL | foo::<()>(); - | ^^ the trait `Trait<for<'b> fn(fn(&'b u32))>` is not implemented for `()` - | - = help: the following implementations were found: - <() as Trait<fn(fn(&'a u32))>> - -error: aborting due to previous error - -For more information about this error, try `rustc --explain E0277`. |