This reverts commit b15785b6.

Auto merge of #53588 - tristanburgess:52985_diagnostics_no_concrete_type_behind_existential_type, r=oli-obk

52985 diagnostics no concrete type behind existential type

@oli-obk FYI. See below for new cycle error generated.

```rust
error[E0391]: cycle detected when processing `Foo`
 --> /dev/staging/existential_type_no_concrete_type_nouse_potential.rs:3:1
  |
3 | existential type Foo: Copy;
  | ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  |
note: ...which requires processing `bar`...
 --> /dev/staging/existential_type_no_concrete_type_nouse_potential.rs:6:23
  |
6 |   fn bar(x: Foo) -> Foo {
  |  _______________________^
7 | |     x
8 | | }
  | |_^
  = note: ...which again requires processing `Foo`, completing the cycle

error: aborting due to previous error

For more information about this error, try `rustc --explain E0391`.
```

Fix unsoundness for VecDeque

 See individual commit for more details.

r? @RalfJung.

Fixes https://github.com/rust-lang/rust/issues/53566, fixes https://github.com/rust-lang/rust/issues/53529

Use optimized SmallVec implementation

This PR replaces current SmallVec implementation with the one from the Servo project.

Closes https://github.com/rust-lang/rust/issues/51640

r? @Mark-Simulacrum

Implement try block expressions

I noticed that `try` wasn't a keyword yet in Rust 2018, so...

~~Fix​es https://github.com/rust-lang/rust/issues/52604~~ That was fixed by PR https://github.com/rust-lang/rust/pull/53135
cc https://github.com/rust-lang/rust/issues/31436 https://github.com/rust-lang/rust/issues/50412

Stabilize a few secondary macro features

- `tool_attributes` - closes https://github.com/rust-lang/rust/issues/44690
- `proc_macro_path_invoc` - this feature was created due to issues with tool attributes (https://github.com/rust-lang/rust/issues/51277), those issues are now fixed (https://github.com/rust-lang/rust/pull/52841)
- partially `proc_macro_gen` - this feature was created due to issue https://github.com/rust-lang/rust/issues/50504, the issue is now fixed (https://github.com/rust-lang/rust/pull/51952), so proc macros can generate modules. They still can't generate `macro_rules` items though due to unclear hygiene interactions.

Feature gate where clauses on associated type impls

Fixes #52913. This doesn't address the core problem, which is tracked by https://github.com/rust-lang/rust/issues/47206. However, it fixes the stable-to-stable regression: you now have to enable `#![feature(generic_associated_types)]` to trigger the weird behaviour.

Merge `IdxSet` and `IdxSetBuf`

Because it simplifies things.

@r? nikomatsakis

  - Original cycle error diagnostics PR'd against this issue caught
panic-causing error while resolving std::mem::transmute calls
  - Now, catch invalid use case of not providing a concrete sized type
behind existential type in definining use case.
  - Update relevant test to reflect this new error

52985: revert normalize query changes
      - PR 53588 invalidates 53316, causing a correct cycle error to occur
    with a good span.
      - Don't need to revert the whole merge as the test files are
    still fine, just need to revert the normalize query changes.
      - It should now be correct that infinite recursion detected during
    normalize query type folding is a bug, should have been caught earlier
    (when resolving the existential type's defining use cases).

52985: code review impl
  - Only cause cycle error if anonymous type resolves to anonymous type
that has the same def id (is the same type) as the original (parent)
type.
  - Add test case to cover this case for existential types.

52985: remove Ty prefix from TyAnon
  - To align with changes per commit 6f637da5

`tool_attributes`, `proc_macro_path_invoc`, partially `proc_macro_gen`

Allow panicking with string literal messages inside constants

r? @EddyB

cc https://github.com/rust-lang/rust/issues/51999

we can't implement things like `panic!("foo: {}", x)` right now because we can't call trait methods (most notably `Display::fmt`) inside constants. Also most of these impls probably have loops and conditions, so it's messy anyway.

But hey `panic!("foo")` works at least.

cc @japaric got any test ideas for `#![no_std]`?

Rollup of 10 pull requests

Successful merges:

 - #53418 (Mark some suggestions as MachineApplicable)
 - #53431 (Moved some feature gate ui tests to correct location)
 - #53442 (Update version of rls-data used with save-analysis)
 - #53504 (Set applicability for more suggestions.)
 - #53541 (Fix missing impl trait display as ret type)
 - #53544 (Point at the trait argument when using unboxed closure)
 - #53558 (Normalize source line and column numbers.)
 - #53562 (Lament the invincibility of the Turbofish)
 - #53574 (Suggest direct raw-pointer dereference)
 - #53585 (Remove super old comment on function that parses items)

Failed merges:

 - #53472 (Use FxHash{Map,Set} instead of the default Hash{Map,Set} everywhere in rustc.)
 - #53563 (use String::new() instead of String::from(""), "".to_string(), "".to_owned() or "".into())

r? @ghost

Rename TyVariants and variants

- Rename `TypeVariants` to `TyKind`.
- Remove the `Ty` prefix from each one of its variants (plus the identically-named variants of `PrimTy`).
- Rename `ty::Slice` to `ty::List`.

The new names look cleaner.

r? @EddyB

Remove super old comment on function that parses items

This comment was added more than 5 years ago in ab03c1e4. As far as anyone reading this comment today needs to know, the function has never parsed items from inside an extern crate.

Suggest direct raw-pointer dereference

People often come looking for some kind of `as_ref_unchecked` method on
raw pointers that would give them `&T` and not `Option<&T>` when they
are sure the pointer is not NULL.

There's no such method, but taking a reference of the dereferenced
pointer accomplishes the same thing. Therefore, suggest using that, at
the `as_ref` site ‒ it's a place people are likely going to look into.

Lament the invincibility of the Turbofish

Here a test case is added to ensure that any others attempting to drive the Turbofish to extinction have second thoughts. Previously the [entire test suite would succeed](https://github.com/rust-lang/rust/pull/53511) if generic arguments were accepted without disambiguation, making for [confusing and heartbreaking circumstances](https://github.com/rust-lang/rfcs/pull/2527).

Normalize source line and column numbers.

Fixes #53547.

r? @EddyB

Point at the trait argument when using unboxed closure

Fix #53534.

r? @varkor

Fix missing impl trait display as ret type

I need to convert a `TraitPredicate` into a `TraitBound` to get the returned impl trait. So far, didn't find how or even if it was the good way to do it.

cc @EddyB @oli-obk (since you're the one behind the change apparently 😉)

Set applicability for more suggestions.

Converts a couple more calls to `span_suggestion_with_applicability`  (#50723). To be on the safe side, I marked suggestions that depend on the intent of the user or that are potentially lossy conversions as MaybeIncorrect.

r? @estebank

Update version of rls-data used with save-analysis

This part 1/3 for fixing rust-lang/rust#53440.

Moved some feature gate ui tests to correct location

None

Mark some suggestions as MachineApplicable

I think the following suggestions should be safe to mark as `MachineApplicable`.

r? @estebank

fix array drop glue: properly turn raw ptr into reference

Discovered while working on https://github.com/rust-lang/rust/pull/53424: The generated drop glue uses an assignment `ptr = cur` where `ptr` is a reference and `cur` a raw pointer. This is not well-formed MIR.

Do we have MIR sanity checks that run on the drop glue and should have caught this?

r? @EddyB

Remove Ty prefix from Ty{Adt|Array|Slice|RawPtr|Ref|FnDef|FnPtr|Dynamic|Closure|Generator|GeneratorWitness|Never|Tuple|Projection|Anon|Infer|Error}

Buffer LLVM's object output stream

In some profiling on OSX I saw the `write` syscall as quite high up on
the profiling graph, which is definitely not good! It looks like we're
setting the output stream of an object file as directly to a file
descriptor which means that we run the risk of doing lots of little
writes rather than a few large writes.

This commit fixes this issue by adding a buffered stream on the output,
causing the `write` syscall to disappear from the profiles on OSX.