The problem is that you're taking a "negative" number (2's complement of a pointer beyond the 2Gb address) and applying modulo-100. The C++ specs say that the result is implementation-dependent. Your neat x86 processor has different 'div' instructions (div computes both division and modulo) which have different behavior with negative numbers.

To be more precise, there are no negative numbers, they may be interpreted as negative. There are several instructions for a 2's complement operation, div and idiv. With signed operands your compiler uses idiv, and with unsigned operands, div. The result that you're seeing is that you get a negative number. The result is not "way outside the 0..99 range", it's in the range -99..0.

You can read it up in the C++ spec, ANSI ISO IEC 14882 §5.6 sub 4: "The binary / operator yields the quotient, and the binary % operator yields the remainder from the division of the first expression by the second. If the second operand of / or % is zero the behavior is undefined; otherwise (a/b)*b + a % b is equal to a. If both operands are nonnegative then the remainder is nonnegative; if not, the sign of the remainder is implementation-defined."

Your hash function of course solves the problem, but for other pointer operations I'd make sure that you are using unsigneds. Even intermediate results must be unsigned.. so cast the crap if necessary.