__GCC_HAVE_SYNC_COMPARE_AND_SWAP_16 not defined on aarch64

[Toebs Douglass]

I wrote:
> I may be utterly, utterly wrong, but I think what's happened makes
> perfect sense at each step of the way but has led to an outcome which
> is crazy.  I publish a lock-free data structure library and I must
> specifically avoid libatomic.

Sorry to be so unconstructive, though.  I don't know enough about GCC
and libatomic to be able to meaningfully make other proposals; I can
only write as an end user.