This adds support for multiple-precision floating-point computation PR: ports/180004 Submitted by: Dan Lukes <dan@obluda.cz>