* environment
* nonexistent
* potential
* prefixes
* stabilization
* suffixes
* use

Signed-off-by: Josh Soref <[email protected]>

Adds support for LoongArch.

Negating a non-zero integer currently requires unpacking to a
primitive and re-wrapping. Since negation of non-zero signed
integers always produces a non-zero result, it is safe to
implement `Neg` for `NonZeroI{N}`.

The new `impl` is marked as stable because trait implementations
for two stable types can't be marked unstable.

Implement `Neg` for signed non-zero integers.

Negating a non-zero integer currently requires unpacking to a primitive and re-wrapping. Since negation of non-zero signed integers always produces a non-zero result, it is safe to implement `Neg` for `NonZeroI{N}`.

The new `impl` is marked as stable because trait impls for two stable types can't be marked unstable.

See discussion on rust-lang/libs-team#105 for additional context.

…acrum

Spelling misc

These two files seem to be fairly distinct from everything else.

That said, if this project doesn't like changing changelogs, I'm happy to drop the changes to `RELEASES.md`

cmp doc examples improvements

Most changes are for stylistic consistency, with some changes to provide more clarity.

…x-raw-sys, r=Mark-Simulacrum

bootstrap: Update linux-raw-sys to 0.3.2

Adds support for LoongArch.

Make `impl Debug for Span` not panic on not having session globals.

I hit the panic that this patch avoids while messing with the early lints in `rustc_session::config::build_session_options()`. The rest of that project is not finished, but this seemed like a self-contained improvement.

(Should changes like this add tests? I don't see similar unit tests.)

…_unsafe_op_in_unsafe_fn, r=Nilstrieb

`deny(unsafe_op_in_unsafe_fn)` in `rustc_data_structures`

r? `@Nilstrieb`

I couldn't bring myself to document the safety in big `unsafe` functions but ehh

…compiler-errors

fix lint regression in `non_upper_case_globals`

Fixes rust-lang#110573

The issue also exists for inherent associated types (where I copied my impl from). `EarlyContext` is more involved to fix in this way, so I'll leave it be for now (note it's unstable so that's not urgent).

r? `@compiler-errors`