It has been necessary to modify the source code for newer versions of Linux gcc.
In addition math2549.tgz contains new routines for multiple precision arithmetic.
Large integers may be several thousand digits long and all arithmetic works
reasonably with positive and negative numbers.