This contains several fixes, including the FD_ISSET signature change
that the previous patch in this PR needs.

dirent contains a flexible array member, so don't test its sizeof, don't
allow it to be copied, and don't represent it with an artificial size.

Also, locale_t is a typedef for a pointer to an opaque struct, so
represent it that way explicitly.

Rustc is changing it to be target_os; for now, recognize both.

Bump to 0.2.52

This will allow the crt-static feature to be utilized correctly on Redox, #1322

This fixes issue #1318 - on Solarish systems __posix_readdir_r is only

defined for 32-bit Solarish targets.  Since rust doesn't currently support 32-bit Solarish targets, the line is both unnecessary and wrong.

As found with the latest fd binary, removing the line fixed the fd build (and fd then works as expected).

redox: support crt-static

This reuses the code from musl to add support to Redox for crt-static.

Linking to `m` is also unnecessary as it is included in `c`

defined for 32-bit Solarish targets.  Since rust doesn't currently
support 32-bit Solarish targets, the line is both unnecessary and wrong.

Remove duplicate target from ci/build.sh

I'm almost certain this is not intentional

I'm almost certain this is not intentional

 Add binding for new `__wasilibc_find_relpath` API

This'll be used in libstd upstream, and this also enables wasi on CI!

Added recently and will be used in libstd!

add login_tty api from libutil

Add an issue template

Now that wasi is in nightlies, we can run it on PRs!

Bump to 0.2.51

Pulls in the new wasi target support!

Add mq_getfd_np() on FreeBSD and fix mqd_t on DragonFlyBSD

[`mq_getfd_np()` was added in FreeBSD 11](https://svnweb.freebsd.org/base/stable/11/include/mqueue.h?revision=306905&view=markup). I'm already using it in my [posixmq crate](https://github.com/tormol/posixmq) for mio/kqueue integration, and I've tested it both in virtualbox and on Cirrus-CI.

[`mqd_t` in an `int` on DragonFlyBSD](https://github.com/DragonFlyBSD/DragonFlyBSD/blob/e7ab884bd49753f8884eb597d10d6569a08fa0df/sys/sys/types.h#L139) even though it's a pointer on FreeBSD, because [DragonflyBSD's implementation is based on NetBSD](https://github.com/DragonFlyBSD/DragonFlyBSD/commit/f2df0f7ca6de3f944dcbfb9923750d7c88984535). The definitions for `mq_attr` are already separate and correct.
Does fixing this count as a breaking change? I think the current definition will work in most cases, because IIRC the calling convention means that `mqd_t` is always passed via registers, the upper 32 bits might just contain garbage.
I've *not* tested this change on DragonFlyBSD.

I want to add mq symbols for solarish, but is the CDDL license compatible with Apache-2.0 and MIT?

Add intial support for wasm32-unknown-wasi

This target is [being proposed][LINK] int he rust-lang/rust repository
and this is intended to get coupled with that proposal. The definitions
here all match the upstream reference-sysroot definitions and the
functions all match the reference sysroot as well. The linkage here is
described more in detail on the Rust PR itself, but in general it's
similar to musl.

Automatic verification has been implemented in the same manner as other
targets, and it's been used locally to develop this PR and catch errors
in the bindings already written (also to help match the evolving sysroot
of wasi). The verification isn't hooked up to CI yet though because
there is no wasi target distributed via rustup just yet, but once that's
done I'll file a follow-up PR to execute verification on CI.

[LINK]: https://github.com/rust-lang/rust/pull/59464

This target is [being proposed][LINK] int he rust-lang/rust repository
and this is intended to get coupled with that proposal. The definitions
here all match the upstream reference-sysroot definitions and the
functions all match the reference sysroot as well. The linkage here is
described more in detail on the Rust PR itself, but in general it's
similar to musl.

Automatic verification has been implemented in the same manner as other
targets, and it's been used locally to develop this PR and catch errors
in the bindings already written (also to help match the evolving sysroot
of wasi). The verification isn't hooked up to CI yet though because
there is no wasi target distributed via rustup just yet, but once that's
done I'll file a follow-up PR to execute verification on CI.

[LINK]:

Don't fix asmjs and wasm32 build jobs

This doesn't fix the jobs, but at least now libc-test reports no errors, even though they still fail to link.