Moving integer division to floating-point is trivial

Shows that floating point trunc(x/y) equals integer x/y provided the integers fit in the floating point format.