Instead of instaling OCaml bindings in a location where installation
will not fail, don't build them in the first place.

This wins 3% on `unicode_normalization`.

The commit also removes the debug statement, because they annoyed me.
This change wins another 1% on `unicode_normalization`, at least partly
because it no longer needs to increment `iteration`.

Building Servo with a recent Nightly produces:

```rust
warning: use of deprecated attribute `plugin`: compiler plugins are deprecated. See rust-lang#29597
  --> components/script/lib.rs:14:1
   |
14 | #![plugin(script_plugins)]
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^ help: remove this attribute
   |
   = note: `#[warn(deprecated)]` on by default
```

First, linking to rust-lang#29597 is not ideal
since there is pretty much no discussion there of the deprecation
and what can be used instead.
This PR changes the link to the deprecation PR which does have more discussion.

Second, the “remove this attribute” suggestion is rather unhelpful.
Just because a feature is deprecated doesn’t mean that simply removing its use
without a replacement is acceptable.

In the case of custom lint, there is no replacement available.
Prefixing a message with “help:” when telling users that they’re screwed
honestly feels disrespectful.

This PR also changes the message to be more factual.

Co-Authored-By: Mark Rousskov <[email protected]>

The current code has this impl:
```
impl<T: Into<TokenStream>> iter::FromIterator<T> for TokenStream
```
If given an `IntoIterator<Item = TokenTree>`, it will convert each individual
`TokenTree` to a `TokenStream` (at the cost of two allocations: a `Vec`
and an `Lrc`). It will then merge those `TokenStream`s into a single
`TokenStream`. This is inefficient.

This commit changes the impl to this less general one:
```
impl iter::FromIterator<TokenTree> for TokenStream
```
It collects the `TokenTree`s into a single `Vec` first and then converts that
to a `TokenStream` by wrapping it in a single `Lrc`. The previous generality
was unnecessary; no other code needs changing.

This change speeds up several benchmarks by up to 4%.

Because most of the call sites have an easier time working with a
`TokenTree` instead of a `TokenStream`.

Likewise for `NestedMetaItem::tokens()`. Also, add
`MetaItemKind::token_trees_and_joints()`, which `MetaItemKind::tokens()`
now calls.

This avoids some unnecessary `TokenTree` to `TokenStream` conversions,
and removes the need for the clumsy
`TokenStream::append_to_tree_and_joint_vec()`.

Document JSON message output.

This documents the JSON messages in the rustc book.

Disable Go and OCaml bindings when building LLVM

Instead of instaling OCaml bindings in a location where installation
will not fail, don't build them in the first place.

…-E0575, r=kinnison

Add long error explanation for E0575

Part of rust-lang#61137.

…komatsakis

Add more coherence tests

I've wrote the missing test cases listed in [this google doc](https://docs.google.com/spreadsheets/d/1WlroTEXE6qxxGvEOhICkUpqguYZP9YOZEvnmEtSNtM0/edit#gid=0)

> The other thing that might be useful is to rename the existing tests so they all fit the new naming scheme we were using.

I'm not entirely sure how to do this. If everything from the google sheet is covered could I just remove the remaining tests in `src/test/ui/coherence` or is there something in there that should remain?

cc rust-lang#63599

r? @nikomatsakis

…Tree-to-TokenStream-conversions, r=petrochenkov

Avoid unnecessary `TokenTree` to `TokenStream` conversions

A `TokenStream` contains any number of `TokenTrees`. Therefore, a single `TokenTree` can be promoted to a `TokenStream`. But doing so costs two allocations: one for the single-element `Vec`, and one for the `Lrc`. (An `IsJoint` value also must be added; the default is `NonJoint`.)

The current code converts `TokenTree`s to `TokenStream`s unnecessarily in a few places. This PR removes some of these unnecessary conversions, both simplifying the code and speeding it up.

r? @petrochenkov

…sakis

Use a sharded dep node to dep node index map

Split out from rust-lang#61845 and based on rust-lang#63756.

r? @nikomatsakis

…size, r=nikomatsakis

Speed up `LexicalResolve::expansion()`

A couple of improvements that speed up `unicode_normalization` by about 4%. The first commit was enabled by the improvements to `BitSet` iteration in rust-lang#65425.

r? @nikomatsakis

…-E0584, r=kinnison

Add long error explanation for E0584

Part of rust-lang#61137.

r? @kinnison

properly document panics in div_euclid and rem_euclid

For signed numbers, document that `div_euclid` and `rem_euclid` panic not just when `rhs` is 0, but also when the division overflows.

For unsigned numbers, document that `div_euclid` and `rem_euclid` panic when `rhs` is 0.

Plugins deprecation: don’t suggest simply removing the attribute

Building Servo with a recent Nightly produces:

```rust
warning: use of deprecated attribute `plugin`: compiler plugins are deprecated. See rust-lang#29597
  --> components/script/lib.rs:14:1
   |
14 | #![plugin(script_plugins)]
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^ help: remove this attribute
   |
   = note: `#[warn(deprecated)]` on by default
```

First, linking to rust-lang#29597 is not ideal since there is pretty much no discussion there of the deprecation and what can be used instead. This PR changes the link to the deprecation PR which does have more discussion.

Second, the “remove this attribute” suggestion is rather unhelpful. Just because a feature is deprecated doesn’t mean that simply removing its use without a replacement is acceptable.

In the case of custom lint, there is no replacement available. Prefixing a message with “help:” when telling users that they’re screwed honestly feels disrespectful.

This PR also changes the message to be more factual.

…r=simulacrum

add option to ping llvm ice-breakers to triagebot

reorder fmt docs for more clarity

I adjusted these docs in rust-lang#65332 but wasn't happy with the result when seeing it in rustdoc. So this reorders the subsections in the "Formatting Parameters" section to be more logical (subsections that reference `width` come after the `width` subsection) and they also all have examples now.

Update backtrace to 0.3.40

Diff: rust-lang/backtrace-rs@0.3.37...b5cc5b1

Pretty low risk, considering the only changes are in low-tier targets.

r? @cramertj
cc @alexcrichton