[PATCH v2] handle sem_t with ILP32 and __HAVE_64B_ATOMICS

[Torvald Riegel]

On Mon, 2015-01-26 at 18:35 -0500, Chris Metcalf wrote:
> This version reflects Torvald's recent comments.

I think this looks good, but we'll have to find consensus with Carlos
first.  I also wouldn't be opposed to at first just enable the
64b-algorithm if __HAVE_64B_ATOMICS and _LP64 are defined (if the latter
is the right check...).