diff options
| author | bors <bors@rust-lang.org> | 2022-04-30 13:52:29 +0000 |
|---|---|---|
| committer | bors <bors@rust-lang.org> | 2022-04-30 13:52:29 +0000 |
| commit | d201c812d40932509b2b5307c0b20c1ce78d21da (patch) | |
| tree | febd00751fa6caa1f704cdcabe5249a7643643e7 /compiler/rustc_mir_transform/src/coverage/debug.rs | |
| parent | 76d4862fdd131b6f79dc0a31857f888d26bcdb27 (diff) | |
| parent | b4e3e62c9cde00637fc2e69b24581ba77494035b (diff) | |
| download | rust-d201c812d40932509b2b5307c0b20c1ce78d21da.tar.gz rust-d201c812d40932509b2b5307c0b20c1ce78d21da.zip | |
Auto merge of #93803 - cjgillot:unify-bounds, r=oli-obk
Handle generic bounds in a uniform way in HIR Generic bounds in HIR used to be split between bounds in the parameter definition and bounds in a where clause. This PR attempts to store all of those as where predicates. This effectively desugars ```rust fn foo<T: Default, U>(x: impl Copy) where U: Clone ``` into ```rust fn foo<T, U, _V>(x: _V) where T: Default, U: Clone, _V: Copy ``` (where _V is actually hidden and called "impl Copy"). I managed to make compiler warnings more uniform. About rustdoc: is making this desugaring user-visible acceptable? About clippy: I don't understand the subtle logic in the `needless-lifetimes` lint. r? `@estebank`
Diffstat (limited to 'compiler/rustc_mir_transform/src/coverage/debug.rs')
0 files changed, 0 insertions, 0 deletions