…nsion, r=<try>

expand free alias types in the auto-trait orphan check

…_alias_expansion, r=fmease

expand free alias types in the auto-trait orphan check

fixes rust-lang#157756.

auto-trait impls are only allowed on nominal types. the check for that looked at the self type as written, without expanding it, so a lazy type alias got rejected even when it pointed straight at a type that's fine on its own:

```rust
#![feature(lazy_type_alias)]
struct Local;
type Alias = Local;
unsafe impl Sync for Alias {} // used to be E0321
```

the locality check right above it already normalizes the self type, so the two were looking at different things. imo the cleanest fix is to expand free aliases first, with the same `expand_free_alias_tys` coherence already calls in a couple of other spots. anything that should still be rejected (trait objects, opaque types, bare type params) survives the expansion and errors like before.

one thing worth a second look: `impl_for_weak_alias.rs` used `type Alias = (impl Sized, u8)` and expected an error. idk if that was deliberate, but the alias expands to a tuple, and a tuple is fine for a local auto trait if you write it by hand, so it passes now and i made it check-pass. lmk if i'm missing why it should still be rejected.

Rollup of 18 pull requests

Successful merges:

 - #158692 (Add release notes for 1.96.1)
 - #134021 (Implement `IntoIterator` for `[&[mut]] Box<[T; N], A>`)
 - #152860 (Port the `without_debuginfo` test from `backtrace-rs` to the testsuite)
 - #155932 (MIR Call terminator: evaluate destination place before arguments)
 - #156777 (Add -Zautodiff_post_passes flag to limit which llvm passes to run after enzyme to make autodiff tests more robust)
 - #157151 (JSON target specs: remove 'x86-softfloat' compatibility alias)
 - #157835 (expand free alias types in the auto-trait orphan check)
 - #158377 (add `-Zforce-intrinsic-fallback` flag)
 - #158434 (delegation: refactor AST -> HIR lowering)
 - #158552 (make some tidy errors around python easier to understand)
 - #158624 (borrowck: Introduce BlameConstraint::to_obligation_cause_from_path())
 - #158704 (Optimize `ArrayChunks::try_rfold` with `DoubleEndedIterator::next_chunk_back`)
 - #158711 (library: Comment on libtest's dicey internal soundness)
 - #158539 (Move `SizeHint` and `IoHandle` to `core::io`)
 - #158659 (refactor the normalization in `coerce_shared_info`)
 - #158689 (resolver: don't use `Finalize` when resolving visibilities during AST expansion)
 - #158698 (Update TypeVisitable implementation)
 - #158706 (Tweaks to MIR building scope API)

…_alias_expansion, r=fmease

expand free alias types in the auto-trait orphan check

fixes rust-lang#157756.

auto-trait impls are only allowed on nominal types. the check for that looked at the self type as written, without expanding it, so a lazy type alias got rejected even when it pointed straight at a type that's fine on its own:

```rust
#![feature(lazy_type_alias)]
struct Local;
type Alias = Local;
unsafe impl Sync for Alias {} // used to be E0321
```

the locality check right above it already normalizes the self type, so the two were looking at different things. imo the cleanest fix is to expand free aliases first, with the same `expand_free_alias_tys` coherence already calls in a couple of other spots. anything that should still be rejected (trait objects, opaque types, bare type params) survives the expansion and errors like before.

one thing worth a second look: `impl_for_weak_alias.rs` used `type Alias = (impl Sized, u8)` and expected an error. idk if that was deliberate, but the alias expands to a tuple, and a tuple is fine for a local auto trait if you write it by hand, so it passes now and i made it check-pass. lmk if i'm missing why it should still be rejected.

…uwer

Rollup of 20 pull requests

Successful merges:

 - #158692 (Add release notes for 1.96.1)
 - #134021 (Implement `IntoIterator` for `[&[mut]] Box<[T; N], A>`)
 - #155932 (MIR Call terminator: evaluate destination place before arguments)
 - #155989 (Update `transmute_copy` to ub_checks and `?Sized`)
 - #156777 (Add -Zautodiff_post_passes flag to limit which llvm passes to run after enzyme to make autodiff tests more robust)
 - #157151 (JSON target specs: remove 'x86-softfloat' compatibility alias)
 - #157835 (expand free alias types in the auto-trait orphan check)
 - #157857 (Stabilize `#[my_macro] mod foo;` (part of `proc_macro_hygiene`))
 - #158377 (add `-Zforce-intrinsic-fallback` flag)
 - #158434 (delegation: refactor AST -> HIR lowering)
 - #158552 (make some tidy errors around python easier to understand)
 - #158624 (borrowck: Introduce BlameConstraint::to_obligation_cause_from_path())
 - #158704 (Optimize `ArrayChunks::try_rfold` with `DoubleEndedIterator::next_chunk_back`)
 - #158711 (library: Comment on libtest's dicey internal soundness)
 - #158751 (rustdoc: Fix crash when trying to inline foreign item which cannot have attributes)
 - #158539 (Move `SizeHint` and `IoHandle` to `core::io`)
 - #158659 (refactor the normalization in `coerce_shared_info`)
 - #158689 (resolver: don't use `Finalize` when resolving visibilities during AST expansion)
 - #158698 (Update TypeVisitable implementation)
 - #158706 (Tweaks to MIR building scope API)

…uwer

Rollup of 19 pull requests

Successful merges:

 - #158692 (Add release notes for 1.96.1)
 - #134021 (Implement `IntoIterator` for `[&[mut]] Box<[T; N], A>`)
 - #155932 (MIR Call terminator: evaluate destination place before arguments)
 - #155989 (Update `transmute_copy` to ub_checks and `?Sized`)
 - #156777 (Add -Zautodiff_post_passes flag to limit which llvm passes to run after enzyme to make autodiff tests more robust)
 - #157151 (JSON target specs: remove 'x86-softfloat' compatibility alias)
 - #157835 (expand free alias types in the auto-trait orphan check)
 - #157857 (Stabilize `#[my_macro] mod foo;` (part of `proc_macro_hygiene`))
 - #158434 (delegation: refactor AST -> HIR lowering)
 - #158552 (make some tidy errors around python easier to understand)
 - #158624 (borrowck: Introduce BlameConstraint::to_obligation_cause_from_path())
 - #158704 (Optimize `ArrayChunks::try_rfold` with `DoubleEndedIterator::next_chunk_back`)
 - #158711 (library: Comment on libtest's dicey internal soundness)
 - #158751 (rustdoc: Fix crash when trying to inline foreign item which cannot have attributes)
 - #158539 (Move `SizeHint` and `IoHandle` to `core::io`)
 - #158659 (refactor the normalization in `coerce_shared_info`)
 - #158689 (resolver: don't use `Finalize` when resolving visibilities during AST expansion)
 - #158698 (Update TypeVisitable implementation)
 - #158706 (Tweaks to MIR building scope API)

Rollup merge of #157835 - Dnreikronos:orphan_auto_trait_free_alias_expansion, r=fmease

expand free alias types in the auto-trait orphan check

fixes #157756.

auto-trait impls are only allowed on nominal types. the check for that looked at the self type as written, without expanding it, so a lazy type alias got rejected even when it pointed straight at a type that's fine on its own:

```rust
#![feature(lazy_type_alias)]
struct Local;
type Alias = Local;
unsafe impl Sync for Alias {} // used to be E0321
```

the locality check right above it already normalizes the self type, so the two were looking at different things. imo the cleanest fix is to expand free aliases first, with the same `expand_free_alias_tys` coherence already calls in a couple of other spots. anything that should still be rejected (trait objects, opaque types, bare type params) survives the expansion and errors like before.

one thing worth a second look: `impl_for_weak_alias.rs` used `type Alias = (impl Sized, u8)` and expected an error. idk if that was deliberate, but the alias expands to a tuple, and a tuple is fine for a local auto trait if you write it by hand, so it passes now and i made it check-pass. lmk if i'm missing why it should still be rejected.