This commit brings the `Error` trait in line with the [Error interoperation
RFC](https://github.com/rust-lang/rfcs/pull/201) by adding downcasting,
which has long been intended. This change means that for any `Error`
trait objects that are `'static`, you can downcast to concrete error
types.

To make this work, it is necessary for `Error` to inherit from
`Reflect` (which is currently used to mark concrete types as "permitted
for reflection, aka downcasting"). This is a breaking change: it means
that impls like

```rust
impl<T> Error for MyErrorType<T> { ... }
```

must change to something like

```rust
impl<T: Reflect> Error for MyErrorType<T> { ... }
```

except that `Reflect` is currently unstable (and should remain so for
the time being). For now, code can instead bound by `Any`:

```rust
impl<T: Any> Error for MyErrorType<T> { ... }
```

which *is* stable and has `Reflect` as a super trait. The downside is
that this imposes a `'static` constraint, but that only
constrains *when* `Error` is implemented -- it does not actually
constrain the types that can implement `Error`.

[breaking-change]

Associated constants were now showing up for traits and would panic if they were
found on an inherent impl. This commit unblocks the nighly builders.

Closes #24446.

Diagnostic errors are now checked for uniqueness across the compiler and
error metadata is written to JSON files.

Make whitespace consistent

Addresses part of #10381.

Closes #9116.

The problem here is that this test doesn't even compile for Android.

See #24958.

This adds some missing punctuation and converts uses of "Gets" to
"Returns". This sounds better to my ear, but more importantly is
more consistent with the documentation from other files.

THis includes tests for struct and enum. (I suspect the closure case
is actually unreachable, but i see no harm in including it.)

This adds some missing punctuation, adds a missing word, and
corrects a bug in the description of `send_to`, which actually
returns the number of bytes written on success.

Fixes #24925.

This is OK to do given:
  - PIE is supported on Android starting with API 16.
  - The bots are running API 18.
  - API < 16 now has a 12.5% market share[0] as of 2015-04-29.

Unfortunately, this breaks backtrace support. See #17520.

Closes #17437.

[0] https://developer.android.com/about/dashboards/index.html

Note it is safe, with respect to autobuilds, to land before #24945.

(In other words, landing this sooner won't break things for anyone any
worse than they were already broken, since there are *other* tests
that also add `-g` to their flags via `compile-flags: -g`.)

Currently if a standard I/O handle is set to inherited on Windows, no action is
taken and the slot in the process information description is set to
`INVALID_HANDLE_VALUE`. Due to our passing of `STARTF_USESTDHANDLES`, however,
this means that the handle is actually set to nothing and if a child tries to
print it will generate an error.

This commit fixes this behavior by explicitly creating stdio handles to be
placed in these slots by duplicating the current process's I/O handles. This is
presumably what previously happened silently by using a file-descriptor-based
implementation instead of a `HANDLE`-centric implementation.

Along the way this cleans up a lot of code in `Process::spawn` for Windows by
ensuring destructors are always run, using more RAII, and limiting the scope of
`unsafe` wherever possible.

```asm
    movzbl      %dil, %eax
    bsfl        %eax, %eax
    movl        $32, %ecx
    cmovnel     %eax, %ecx
    cmpl        $32, %ecx
    movl        $8, %eax
    cmovnel     %ecx, %eax
```

which has some unnecessary overhead, having two conditional moves.

To improve the codegen, we can zero extend the 8 bit integer, then set
bit 8 and perform a cttz operation on the extended value. That way
there's no conditional operation involved at all.

compiler plugins are hosted unter "compiler-plugins.html" not "plugins.html"

use visible characters for the multibyte character filler.

Since #24783, the style guidelines recommend that unit tests should live in a submodule `tests` rather than `test` to not clash with the possible use of libtest. This is especially important for benchmark tests as they require libtest. Fixes #24923.

Fix #24687