Re: storing an integer in a double precision

steve <kargls@xxxxxxxxxxx> writes:

program foo
integer i, j
double precision, volatile :: d
print *, 'Range: ', -huge(i), huge(i)
do i = - huge(i), huge(i)
d = i
j = d
if (i /=3D j) print *, i, j, d
end do
end program foo

Any chance that an optimizer might just replace d=i and j=d
with j=i? I'm guessing no.

.

Relevant Pages