the kernel currently enforce that a stack is immutable. calling mmap(2) or 
mprotect(2) to change it will result in EPERM, which generate a panic!().

so just do like for Linux, and trust the kernel to do the right thing.

…ction

…piler-errors

Suggest `==` to the first expr which has `ExprKind::Assign` kind

follow-up to rust-lang#102708

[playground](https://play.rust-lang.org/?version=nightly&mode=debug&edition=2021&gist=4241dc33ed8af02e1ef530d6b14903fd)

…m-ou-se

openbsd: don't reallocate a guard page on the stack.

the kernel currently enforce that a stack is immutable. calling mmap(2) or  mprotect(2) to change it will result in EPERM, which generate a panic!().

so just do like for Linux, and trust the kernel to do the right thing.

…r=cjgillot

Print return-position `impl Trait` in trait verbosely if `-Zverbose`

Makes the behavior a bit closer to regular `impl Trait` printing

…istentials, r=cjgillot

Sort elaborated existential predicates in `object_ty_for_trait`

r? `@cjgillot`

I think that rust-lang#102845 caused rust-lang#102933. Depending on the order that we elaborate these existential projection predicates, there's no guarantee that they'll be sorted by def id, which is what is failing the assertion in the issue.

Fixes rust-lang#102933
Fixes rust-lang#102973

Use `full_res` instead of `expect_full_res`

Fixes rust-lang#102946
Fixes rust-lang#102978

…e1-dead

Delay `is_intrinsic` query until after we've determined the callee is a function

Fixes rust-lang#102985