… idea.

…r the mutex holding it

Co-authored-by: LeSeulArtichaut <leseulartichaut@gmail.com>

These lead to inference regressions (mostly in tests) in code that looks
like:

    let socket = std::net::SocketAddrV4::new(std::net::Ipv4Addr::new(127, 0, 0, 1), 8080);
    assert_eq!(socket, "127.0.0.1:8080".parse().unwrap());

That compiles as of stable 1.44.0 but fails in beta with:

    error[E0284]: type annotations needed
     --> src/main.rs:3:41
      |
    3 |     assert_eq!(socket, "127.0.0.1:8080".parse().unwrap());
      |                                         ^^^^^ cannot infer type for type parameter `F` declared on the associated function `parse`
      |
      = note: cannot satisfy `<_ as std::str::FromStr>::Err == _`
    help: consider specifying the type argument in the method call
      |
    3 |     assert_eq!(socket, "127.0.0.1:8080".parse::<F>().unwrap());
      |

This commit normalizes projections which contain opaque types (opaque types
are otherwise linted against, which is would have previously made the
test cases added in this commit fail).

Signed-off-by: David Wood <david@davidtw.co>

Co-authored-by: David Tolnay <dtolnay@gmail.com>

Based on the review made by dtolnay.

…p-elab, r=oli-obk

Check for live drops in constants after drop elaboration

Resolves rust-lang#66753.

This PR splits the MIR "optimization" pass series in two and introduces a query–`mir_drops_elaborated_and_const_checked`–that holds the result of the `post_borrowck_cleanup` analyses and checks for live drops. This query is invoked in `rustc_interface` for all items requiring const-checking, which means we now do `post_borrowck_cleanup` for items even if they are unused in the crate.

As a result, we are now more precise about when drops are live. This is because drop elaboration can e.g. eliminate drops of a local when all its fields are moved from. This does not mean we are doing value-based analysis on move paths, however; Storing a `Some(CustomDropImpl)` into a field of a local will still set the qualifs for that entire local.

r? @oli-obk

…atthewjasper

On recursive ADT, provide indirection structured suggestion

…ample, r=dtolnay

Example about explicit mutex dropping

Fixes rust-lang#67457.

Following the remarks made in rust-lang#73074, I added an example on the main `Mutex` type, with a situation where there is mutable data and a computation result.

In my testing it is effectively needed to explicitly drop the lock, else it deadlocks.

r? @dtolnay because you were the one to review the previous PR.

…in-projections, r=estebank

lint: normalize projections using opaque types

Fixes rust-lang#73251.

This PR normalizes projections which use opaque types (opaque types are otherwise linted against, which is would have previously made the test cases added in this PR fail).

Revert heterogeneous SocketAddr PartialEq impls

Originally added in rust-lang#72239.

These lead to inference regressions (mostly in tests) in code that looks like:

```rust
let socket = SocketAddrV4::new(Ipv4Addr::new(127, 0, 0, 1), 8080);
assert_eq!(socket, "127.0.0.1:8080".parse().unwrap());
```

That compiles as of stable 1.44.0 but fails in beta with:

```console
error[E0284]: type annotations needed
 --> src/main.rs:3:41
  |
3 |     assert_eq!(socket, "127.0.0.1:8080".parse().unwrap());
  |                                         ^^^^^ cannot infer type for type parameter `F` declared on the associated function `parse`
  |
  = note: cannot satisfy `<_ as std::str::FromStr>::Err == _`
help: consider specifying the type argument in the method call
  |
3 |     assert_eq!(socket, "127.0.0.1:8080".parse::<F>().unwrap());
  |
```

Closes rust-lang#73242.