[ILUG] 64-int int math library for 32-bit int machines?

[Brian Foster]

 what I am looking for is a set of routines (e.g., a library),
 written in portable pre-1999 C, whose source is available,
 that, using 32-bit maths, implements at least the five basic
 integer operations (+ - * / %) for 64-bit integers, and which
 ideally has both a "signed" and "unsigned" version.

 please note I said using 32-bit maths.  so NO <long long>,
 NO <__int64>, NO <i64>, none of that stuff!  the compilers
 (plural) in question do not have 64-bit integers.

 ideally C, albeit C++ could be Ok, but NO assembly.
 implementations in other languages (e.g., Java) which could
 be used as a model and reference for a C version would be Ok.

 ideally a BSD-ish license, albeit (L)GPL, CPL, or so on
 should be Ok.

 ideally, the "64-bit integers" would be stored as something
 like an array[] or <struct> of two 32-bit integers.  storing
 as ASCII string, or BCD, or so on would be problematical.

 I do not need an arbitrary precision package (such as GNUs
 BigNum), nor rationals, nor fixed point, etc., just signed
 and unsigned 64-bit integers.  and please note I have _not_
 mentioned (time) efficiency, albeit space (size) could be
 a concern.  but must be mathematically sound, and ideally
 also has a comprehensive Test Suite.

 Google™ has not (yet) been of much help here.  (I suppose
 the GCC support library might have something like this?)

 any ideas?  pointers?

cheers!
	-blf-
-- 
«How many surrealists does it take to    |  Brian Foster      Montpellier,
 change a lightbulb?  Three.  One calms  |  blf at utvinternet.ie      FRANCE
 the warthog, and two fill the bathtub   |    Stop E$$o (ExxonMobile)!
 with brightly-colored machine tools.»   |        http://www.stopesso.com