Repository navigation
Conversation
rustbot
added
PG-exploit-mitigations
Collaborator
|
The job Click to see the possible cause of the failure (guessed by this bot) |
adwinwhite
reviewed
Oct 30, 2026
|
|
||
| fn capture_move(x: MD<String>) -> unsafe<> MD<String> { | ||
| let f = || -> unsafe<> MD<String> { unsafe { wrap_binder!(x) } }; | ||
| f() |
Contributor
There was a problem hiding this comment.
We should use x again to verify that it moved?
| //~^ ERROR conflicting implementations of trait `Tr` for type `unsafe<'a> &'a Local` | ||
| impl Tr for &'static Local {} | ||
| impl<'x> Tr for unsafe<> &'x Local {} | ||
| //~^ WARN conflicting implementations of trait `Tr` for type `unsafe<'a> &'a Local` |
Contributor
There was a problem hiding this comment.
This should be ERROR rather than WARN, right?
|
|
||
| // A generic unwrap_binder in a array length position should work (if otherwise allowed). | ||
|
|
||
| #![feature(unsafe_binders, generic_const_exprs)] |
Contributor
There was a problem hiding this comment.
GCE is to be yeeted and we should support GCA instead?
Comment on lines
+15
to
+26
| const _: () = assert!(size_of::<unsafe<'a> &'a u8>() == size_of::<&u8>()); | ||
| const _: () = assert!(size_of::<Option<unsafe<'a> &'a u8>>() == size_of::<&u8>()); | ||
| const _: () = assert!(size_of::<Option<unsafe<> NonZero<u32>>>() == 4); | ||
| const _: () = assert!(align_of::<unsafe<'a> (&'a u8, u16)>() == align_of::<(&u8, u16)>()); | ||
|
|
||
| // The vtable niche survives a binder in the metadata type and in the pointee. | ||
| type BoundMeta = DynMetadata<unsafe<'a> MD<dyn Debug + 'a>>; | ||
| const _: () = assert!(size_of::<Option<BoundMeta>>() == size_of::<BoundMeta>()); | ||
| const _: () = assert!( | ||
| size_of::<Option<*const unsafe<'a> MD<dyn Debug + 'a>>>() | ||
| == size_of::<*const unsafe<'a> MD<dyn Debug + 'a>>() | ||
| ); |
Contributor
There was a problem hiding this comment.
We can assert the layouts of unsafe binders versions are the same as inner types? This currently tests both layout equality and niche optimization.
| } | ||
| // error | ||
| fn wrap_mismatch_bound_vars() { | ||
| let _x: unsafe<'a> &'a u8 = unsafe { wrap_binder!(&0; unsafe<'a, 'b> &'a &'b u8) }; |
Contributor
There was a problem hiding this comment.
We don't have auto-deref into the unsafe binders so the inner types on both sides modulo regions are different, &u8 vs &&u8. This isn't what we want to test, right?
Comment on lines
+18
to
+19
| assert_eq!(type_name::<unsafe<'a> &'a ()>(), "unsafe<> &'_ ()"); | ||
| assert_eq!(type_name::<unsafe<'a> (&'a u8, u8)>(), "unsafe<> (&'_ u8, u8)"); |
Contributor
There was a problem hiding this comment.
Should include bound vars as well?
Comment on lines
+15
to
+18
| fn launder_free<'x>(r: &'x String) -> &'static String { | ||
| let b: unsafe<'a> (&'a u8, &'x String) = unsafe { wrap_binder!((&0, r)) }; | ||
| unsafe { unwrap_binder!(b).1 } | ||
| } |
Contributor
There was a problem hiding this comment.
this is testing the same early param lifetime as the test case above rather than an temporary body region?
| } | ||
|
|
||
| fn h(mut b: unsafe<> u8) -> unsafe<> u8 { | ||
| unsafe { unwrap_binder!(b) = 1 }; |
Contributor
There was a problem hiding this comment.
should comment that we support this?
| //@ edition: 2024 | ||
| //@ check-pass | ||
|
|
||
| // Unsafe binders can be held across `.await`, and the future stays `Send`. |
Contributor
There was a problem hiding this comment.
this is redundant as we test Send in send-sync.rs?
| //@ check-pass | ||
| //@ known-bug: #130516 | ||
|
|
||
| // Casts fwith unsafe binders should not change the metadata kind or the |
Contributor
|
☔ The latest upstream changes (presumably #165234) made this pull request unmergeable. Please resolve the merge conflicts by rebasing. |
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Supersedes #153458, #156373, #154524, #158856
First test adds many tests covering (expected) implementation behavior of unsafe binders. Many of these need further design discussion, but the point here is to figure out some behavior that we can pin an implementation to for testing.
Remaining commits fix broken tests one by one (mostly), and should be fairly independent.