This was just my attempt at being smart, I didn't realize
`compare_exchange` returned the old value (I'm dumb!), so I thought
that if the value was 1 then it must have become 2. Normally with
small errors like these you should leave a comment explaining why, but
really, compare and *exchange* is pretty obvious. My bad.

See #151

Use cbitset crate instead of custom bitset implementation in select

See merge request redox-os/relibc!221

Well I'm dumb... This test was already ran in a way that wasn't
verifying output, because this test outputs things that can
differ. Excuse me for not noticing!

This reverts commit 0af78b1e.

Rename __assert to __assert_fail

See merge request !212

Implement remaining LCG functions

See merge request !219

Implement posix_memalign()

See merge request !220

Fix out-of-bounds error in strsignal

See merge request !222

Make valloc() get page size through sysconf(), add tests

See merge request !216

Implement LCG pseudorandom number functions

See merge request !213

Add memrchr()

See merge request !214

Spring cleanup

See merge request !217

This makes relibc more compatible with other libc implementations

The race condition didn't trigger when I tested it, so I assumed I had
solved it...

Some recipes require the headers, but they don't explicitly say
so. Parallel make (-j`nproc`) might start compiling the libs when the
headers aren't done yet.