Rob Browning <rlb@defaultvalue.org> writes:
>
> There may well be bugs

A small one in (gcd n 0), test and possible fix below.  The result
should of course be abs(n) but mpz_gcd_ui can't return that if it
doesn't fit a ulong.

> where
> special casing with GMP operations could help, but it'll probably take
> some good benchmarking to know for sure.

ash (mentioned in the comments) would definitely benefit from the mpz
2exp functions.

In gcd, it'd be nice for the inum/bignum case to share code with the
bignum/inum case, like lcm does.