std: Implement TLS for wasm32-unknown-unknown

This adds an implementation of thread local storage for the
`wasm32-unknown-unknown` target when the `atomics` feature is
implemented. This, however, comes with a notable caveat of that it
requires a new feature of the standard library, `wasm-bindgen-threads`,
to be enabled.

Thread local storage for wasm (when `atomics` are enabled and there's
actually more than one thread) is powered by the assumption that an
external entity can fill in some information for us. It's not currently
clear who will fill in this information nor whose responsibility it
should be long-term. In the meantime there's a strategy being gamed out
in the `wasm-bindgen` project specifically, and the hope is that we can
continue to test and iterate on the standard library without committing
to a particular strategy yet.

As to the details of `wasm-bindgen`'s strategy, LLVM doesn't currently
have the ability to emit custom `global` values (thread locals in a
`WebAssembly.Module`) so we leverage the `wasm-bindgen` CLI tool to do
it for us. To that end we have a few intrinsics, assuming two global values:

* `__wbindgen_current_id` - gets the current thread id as a 32-bit
  integer. It's `wasm-bindgen`'s responsibility to initialize this
  per-thread and then inform libstd of the id. Currently `wasm-bindgen`
  performs this initialization as part of the `start` function.
* `__wbindgen_tcb_{get,set}` - in addition to a thread id it's assumed
  that there's a global available for simply storing a pointer's worth
  of information (a thread control block, which currently only contains
  thread local storage). This would ideally be a native `global`
  injected by LLVM, but we don't have a great way to support that right
  now.

To reiterate, this is all intended to be unstable and purely intended
for testing out Rust on the web with threads. The story is very likely
to change in the future and we want to make sure that we're able to do
that!

The #[panic_handler] attribute can be applied to non-functions

Fixes #54896.

This commit extends the existing lang items functionality to assert
that the `#[lang_item]` attribute is only found on the appropriate item
for any given lang item. That is, language items representing traits
must only ever have their corresponding attribute placed on a trait, for
example.

r? @nagisa

miri engine: Fix run-time validation

This fixes all false positives that came up when actually enabling this in miri.

r? @oli-obk

miri needs this extra flexibility

structured suggestion for E0223 ambiguous associated type

(routine (and when are we going to be done finding these, anyway?) but something that stuck out to me while glancing at #54970)

r? @estebank

Point to variable in `asm!` macro when failing borrowck

Fix #34940.

[NLL] Check user types are well-formed

Also contains a change of span for AscribeUserType.
I'm not quite sure if this was what @nikomatsakis was thinking.

Closes #54620

r? @nikomatsakis

Rollup of 16 pull requests

Successful merges:

 - #54755 (Documents reference equality by address (#54197))
 - #54811 (During rustc bootstrap, make default for `optimize` independent of `debug`)
 - #54825 (NLL says "borrowed content" instead of more precise "dereference of raw pointer")
 - #54860 (Add doc comments about safest way to initialize a vector of zeros)
 - #54869 (Fix mobile docs)
 - #54891 (Fix tracking issue for Once::is_completed)
 - #54913 (doc fix: it's auto traits that make for automatic implementations)
 - #54920 (Fix handling of #[must_use] on unit and uninhabited types)
 - #54932 (A handful of random string-related improvements)
 - #54936 (impl Eq+Hash for TyLayout)
 - #54950 (std: Synchronize global allocator on wasm32)
 - #54956 ("(using ..." doesn't have the matching ")")
 - #54958 (add a macro for static (compile-time) assertions)
 - #54967 (Remove incorrect span for second label inner macro invocation)
 - #54983 (Fix slice's benchmarks)
 - #54989 (Fix spelling in the documentation to htmldocck.py)

Failed merges:

r? @ghost

Fix spelling in the documentation to htmldocck.py

I was reading through htmldocck.py, and decided to attempt to clean it up a little bit. Let me know if you disagree with my edits.

Fix slice's benchmarks

Fixes #54013.

Remove incorrect span for second label inner macro invocation

A fix for issue #54841

add a macro for static (compile-time) assertions

Cc @oli-obk

"(using ..." doesn't have the matching ")"

Fixes #54948.

std: Synchronize global allocator on wasm32

We originally didn't have threads, and now we're starting to add them!
Make sure we properly synchronize access to dlmalloc when the `atomics`
feature is enabled for `wasm32-unknown-unknown`.

impl Eq+Hash for TyLayout

As proposed by @EddyB at https://github.com/rust-lang/rust/pull/53671#pullrequestreview-159761136.

I have an upcoming PR that would also significantly benefit from this.

A handful of random string-related improvements

- remove a few allocations in `errors/diagnostic`
- make `build_helper::gnu_target` return `Cow<str>`

Fix handling of #[must_use] on unit and uninhabited types

Fixes https://github.com/rust-lang/rust/issues/54828.

doc fix: it's auto traits that make for automatic implementations

Being a marker trait is not good enough (that just means "no items in the trait").

r? @alexcrichton who [originally wrote these docs](https://github.com/RalfJung/rust/commit/0a13f1abafe70cddf34bf2b2ba3946c418ed6241).

Fix tracking issue for Once::is_completed

https://github.com/rust-lang/rust/pull/53027 was merged without a tracking issue. I just filed https://github.com/rust-lang/rust/issues/54890. CC @matklad

Fix mobile docs

Fixes #54836.

<img width="1440" alt="screen shot 2018-10-06 at 18 53 19" src="https://user-images.githubusercontent.com/3050060/46573683-1b4cd700-c999-11e8-9e6b-86a23b332e22.png">

r? @QuietMisdreavus

Add doc comments about safest way to initialize a vector of zeros

This adds more information about the vec! macro as discussed in #54628. I think this is a good starting point, but I think additional detail is needed so that we can explain why vec! is safer than the alternatives.

NLL says "borrowed content" instead of more precise "dereference of raw pointer"

Part of #52663.

Previously, move errors involving the dereference of a raw pointer would
say "borrowed content". This commit changes it to say "dereference of
raw pointer".

r? @nikomatsakis
cc @pnkfelix

Rollup merge of #54811 - pnkfelix:issue-24840-separate-bootstrap-default-for-optimize-from-debug-setting, r=nikomatsakis

During rustc bootstrap, make default for `optimize` independent of `debug`

It may have taken me three and a half years, but I'm following through on my ["threat"](https://github.com/rust-lang/rust/issues/24840#issuecomment-97911700)

Fix #24840

Documents reference equality by address (#54197)

Clarification of the use of `ptr::eq` to test equality of references via address by pointer coercion,  regarding issue #54197 .

The same example as in `ptr::eq` docs is shown here to clarify that
`PartialEq` compares values pointed-to instead of via address (which can be desired in some cases)

Fix #54224 (const promotion regression)

r? @EddyB

Improve error display for codeblocks in rustdoc

Part of #53919.

r? @QuietMisdreavus

Simplify implementation of align_offset slightly