Wanted: arbitrary precision integer arithmetic

[Robert D. Silverman]

I have a set of routines that beat the UNIX routines by over an order
of magnitude. These include not only +,-,*,/, but also things like
GCD's modular inverses, Jacobi symbol, modular exponentiation, etc. etc.
They are public domain and I use them in my factoring and prime testing
work. They run on SUN-3's and VAXEN and are 99% portable. They do use,
however 3 routines ~5 lines each to get at the ediv and emul instructions
and are written in assembler.

Bob Silverman