With the current master branch and the values of
n = 16000000 and
p = 3.1444753148558566e-10
for the Binomial distribution

it got stuck in an infinite loop in the BINV algorithm

I know its not reproducible because you need to get an unlucky random number for it to happen. But it is described in the GSL source code https://github.com/ampl/gsl/blob/47862078af4a204c4220f028656a19ddd3922144/randist/binomial_tpe.c#L100