try this benchmark after we get flags passed through to isPrime

> I've tested against Morain's ecpp-6.4.5 and got other people to test against primo on
> Windows; I suspect part of the issue is that those implementations are distributed as
> 32-bit executables so don't get the benefit of 64-bit libgmp arithmetic.  There is no
> open-source ECPP implementation that I'm aware of, unless Franke's fastECPP happens to be
> open source and he didn't mention this in the paper.
> 
> On the same hardware, I find that ecpp for 10^500+961 takes 428.2 seconds and
> isprime(10^500+961,2) under pari takes 93.8.  For 10^800+1537 the timings are 1537 seconds
> for ECPP and 508.3 for pari.
> 
