This commit is an implementation of [RFC 1066][rfc] where the conclusion was
that leaking a value is a safe operation in Rust code, so updating the signature
of this function follows suit.

[rfc]: https://github.com/rust-lang/rfcs/blob/master/text/1066-safe-mem-forget.md

Closes #25186

The last one (at least for the moment 😃).

r? @alexcrichton 

- Successful merges: #24576, #24966, #25052, #25131, #25137, #25138, #25139, #25141, #25142, #25144, #25146, #25148, #25154, #25156, #25160, #25173
- Failed merges: 

Nothing uses it anymore.

Remove printing of secret number at final source.

Edit version of #25150.
This commit depends on #25148.

Let me know if I didn't follow correct procedure - seems a trivial change and no one that can be automatically tested.

I noticed this feature added in https://github.com/rust-lang/rust/pull/23289 was missing from the `Search tricks`. Thanks!

r? @steveklabnik

Enhance Google-ability of `.fold()` by mentioning 'reduce' and 'inject' in the docs.

Motivation: [This thread on users.rust-lang.org](https://users.rust-lang.org/t/find-the-shortest-string-in-a-vector/1247)

https://github.com/rust-lang/rust/pull/25080/files#r29634986

r? @huonw 

I know this is the most trivial thing since it's *just* the model lexer, not the real lexer, but [it is simpler to read](https://github.com/rust-lang/rust/issues/15883#issuecomment-72906431) and it'd be great if it was up to date but [it's been rotting](https://github.com/rust-lang/rust/issues/22379) and this is a tiny bit of that.

Thanks!!!!!!

Unsafe destructors were removed, added remarks about needed feature gates.

cc #16676 
r? @steveklabnik 

The error message was misleading, so I adjusted it, and I also added the long diagnostics for this error (resolves one point in #24407).

I was unsure about how to phrase the error message. Is “generic parameter binding” the correct term for this?

For https://github.com/rust-lang/rust/issues/24407

So, I realize this is really late in the game so it's unlikely to be accepted but `FromRawFd`/`FromRawHandle` are necessary for fine grain control over file creation. For example, the current `OpenOptions` does not provide a way to avoid file creation races (there's no way to specify `O_EXCL` or the windows equivalent). Stabilizing these traits and their implementations will give 1.0 users fine-grain control over file creation without committing to any new complex APIs.  Additionally, `AsRawFd`/`AsRawHandle` are already stable so I feel that that stabilizing their inverses is a reasonably small change.

Disclaimer: I'm asking because my crate, tempfile, depends on this feature.

These constants were added in 6f54ce9a and
e8fbd1ce to a consts module that is behind a
gate.

I have not confirmed that these constants do indeed work on either OSX or iOS.

Depends on #24921. r? @alexcrichton 

cc #19986 

r? @Gankro 

The [UnsafeCell documentation says it is undefined behavior](http://doc.rust-lang.org/nightly/std/cell/struct.UnsafeCell.html), so people shouldn't do it.

This happened to catch one case in libstd that was doing this, and I switched that to use an UnsafeCell internally.

Closes #13146 

Closes #20626.

Closes #22463.

Closes #22258.

Closes #22037.

Closes #21701.

Closes #21562.

Closes #21177.

Nothing uses it anymore.

Added documentation of the panicking overflow to the `count` and `position`
methods of `IteratorExt`, also make them more obvious in the code.

Also mention that the `Iterator::next` method of the struct returned by
`IteratorExt::enumerate` can panic due to overflow.

Previously, `try_write` actually only obtained shared read access (but would return a `RwLockWriteGuard` if that access was successful).

Also updates the docs for `try_read` and `try_write`, which were leftover from when those methods returned `Option` instead of `Result`.

Remove printing of secret number at final source.

This commit depends on #25148.

These constants were added in 6f54ce9a and
e8fbd1ce to a consts module that is behind a
gate.

I have not confirmed that these constants do indeed work on either OSX or iOS.

It appears that some of the constants may actually belong in a POSIX module,
but I didn't make these changes here because I don't have access to the POSIX
standard.

Many bounds are currently of the form `T: ?Sized + AsRef<OsStr>` where the
argument is `&T`, but the pattern elsewhere (primarily `std::fs`) has been to
remove the `?Sized` bound and take `T` instead (allowing usage with both
references and owned values). This commit generalizes the possible apis in
`std::env` from `&T` to `T` in this fashion.

The `split_paths` function remains the same as the return value borrows the
input value, so ta borrowed reference is required.