The integer type tracking borrow count has a typedef called `BorrowFlag`. This type should be used instead of explicit `isize`.

Usage of `Pin::new_unchecked(&mut …)` is dangerous with `poll_fn`, even
though the `!Unpin`-infectiousness has made things smoother.
Nonetheless, there are easy ways to avoid the need for any `unsafe`
altogether, be it through `Box::pin`ning, or the `pin!` macro. Since the
latter only works within an `async` context, showing an example
artifically introducing one ought to help people navigate this subtlety
with safety and confidence.

…d_macro_suggestions

…ve(Default)

…n-clarifications, r=thomcc

[doc] `poll_fn`: explain how to `pin` captured state safely

Usage of `Pin::new_unchecked(&mut …)` is dangerous with `poll_fn`, even though the `!Unpin`-infectiousness has made things smoother. Nonetheless, there are easy ways to avoid the need for any `unsafe` altogether, be it through `Box::pin`ning, or the `pin!` macro. Since the latter only works within an `async` context, showing an example artificially introducing one ought to help people navigate this subtlety with safety and confidence.

## Preview

https://user-images.githubusercontent.com/9920355/230092494-da22fdcb-0b8f-4ff4-a2ac-aa7d9ead077a.mov

``@rustbot`` label +A-docs

…e_impl, r=cjgillot

Simplify `Span::source_callee` impl

Imo the iterator impl is easier to grasp.

Use BorrowFlag instead of explicit isize

The integer type tracking borrow count has a typedef called `BorrowFlag`. This type should be used instead of explicit `isize`.

…ve1, r=WaffleLapkin

Rewrite various resolve/diagnostics errors as translatable diagnostics

additional question:

For trivial strings is it ever accepted to use `fluent_generated::foo` in a `label` for example? Or is an empty struct `Diagnostic` preferred?

…e, r=oli-obk

Continue folding in query normalizer on weak aliases

Fixes rust-lang#112752
Fixes rust-lang#112731 (same root cause, so didn't make a test for it)
fixes rust-lang#112776

r? ``@oli-obk``

Treat TAIT equation as always ambiguous in coherence

Not sure why we weren't treating all TAIT equality as ambiguous -- this behavior combined with `DefineOpaqueTypes::No` leads to coherence overlap failures, since we incorrectly consider impls as not overlapping because the obligation `T: From<Foo>` doesn't hold.

Fixes rust-lang#112765

…ce, r=fee1-dead

Don't ICE on bound var in `reject_fn_ptr_impls`

We may try to use an impl like `impl<T: FnPtr> PartialEq {}` to satisfy a predicate like `for<T> T: PartialEq` -- don't ICE in that case.

Fixes rust-lang#112735

…-obk

Fix linker failures when #[global_allocator] is used in a dependency

Fixes rust-lang#112715