Simplifying the code of methods: `nth`, `fold`, `rposition`, and iterators: `Filter`, `FilterMap`, `SkipWhile`.

```
before
test iter::bench_multiple_take      ... bench:        15 ns/iter (+/- 0)
test iter::bench_rposition          ... bench:       349 ns/iter (+/- 94)
test iter::bench_skip_while         ... bench:       158 ns/iter (+/- 6)

after
test iter::bench_multiple_take      ... bench:        15 ns/iter (+/- 0)
test iter::bench_rposition          ... bench:       314 ns/iter (+/- 2)
test iter::bench_skip_while         ... bench:       107 ns/iter (+/- 0)
```
@koalazen has the code for `Skip`.

Once #16011 is fixed, `min_max` could use a for loop.

 Closes #6334

Previously the implementation detection regex would detect
`impl fmt::Show for MyStruct` as `fmt`. Now it will be detected as
`fmt::Show for MyStruct`. Implementations such as `impl MyStruct` will
still be detected as `MyStruct`.

Simplifying the code of methods: nth, fold, rposition
and iterators: Filter, FilterMap, SkipWhile
Adding basic benchmarks

… to separate text from the edge of the newly added background.

This commit stabilizes the `std::sync::atomics` module, renaming it to
`std::sync::atomic` to match library precedent elsewhere, and tightening
up behavior around incorrect memory ordering annotations.

The vast majority of the module is now `stable`. However, the
`AtomicOption` type has been deprecated, since it is essentially unused
and is not truly a primitive atomic type. It will eventually be replaced
by a higher-level abstraction like MVars.

Due to deprecations, this is a:

[breaking-change]

This teach rustdoc to add `<meta name="description">` and `<meta name="keywords">` tags to crate docs. Description is important for search engines because they display it as the page description. Keywords are less useful but still generally recommended.

This also changes the "stability dashboard" link to just say "stability", because the current link takes up a lot of space.

cc https://github.com/rust-lang/rust/issues/12466

For a good measure, implemented target_record_stack_bounds for 32-bit Windows as well.

There's a good long comment explaining why. The tl;dr; is that I have no idea
why this is necessary, but it gets --test to work on windows which is something,
right?

cc #13259
cc #16275
cc rust-lang/cargo#302

I found the current docs to be a little confusing. I believe this makes them straight to the point and less confusing.

This ended up passing through the lexer but dying later on in parsing when it
wasn't handled. The strategy taken was to copy the `str_lit` funciton, but adapt
it for bytes.

Closes #16278

The 30-minute intro had a comment mentioning that `+=` will be included in the future. It's already included, and this fixes it to use `+=`.

Fixes missing overflow lint for i64 #14269

The `type_overflow` lint, doesn't catch the overflow for `i64` because the overflow happens earlier in the parse phase when the `u64` as biggest possible int gets casted to `i64` , without checking the for
overflows.
We can't lint in the parse phase, so we emit a compiler error, as we do for overflowing `u64`

Perhaps a consistent behaviour would be to emit a parse error for *all*  overflowing integer types.

See #14269

Apparently the units are in milliseconds, not in seconds!

Follow-up to #16215

Gedit currently lacks syntax highlighting for macros.

The reference manual said that code is interpreted as UTF-8 text and a implementation will normalize it to NFKC. However, rustc doesn't do any normalization now.

We may want to do any normalization for symbols, but normalizing whole text seems harmful because doing so loses some sort of information even if we choose a non-K variant of normalization.

I'd suggest removing "normalized to Unicode normalization form NFKC" phrase for the present so that the manual represents the current state properly. When we address the problem (with a RFC?), then the manual should be updated.

Closes #12388.

Reference: https://github.com/rust-lang/rust/issues/2253

The `type_overflow` lint, doesn't catch the overflow for `i64` because
the overflow happens earlier in the parse phase when the `u64` as biggest
possible int gets casted to `i64` , without checking the for overflows.
We can't lint in the parse phase, so a refactoring of the `LitInt` type
was necessary.

The types `LitInt`, `LitUint` and `LitIntUnsuffixed` where merged to one
type `LitInt` which stores it's value as `u64`. An additional parameter was
added which indicate the signedness of the type and the sign of the value.

This operation seems common enough that it would be convenient for it to be a standard method.

Fixes #15830.

The original trick used to trigger unwinds would not work with GCC's implementation of SEH, so I had to invent a new one: rust_try now consists of two routines: the outer one, whose handler triggers unwinds, and the inner one, that stops unwinds by having a landing pad that swallows exceptions and passes them on to the outer routine via a normal return.