[Fortran] Improved dependency analysis of pointer variables

[Tobias Schlüter]

Quoting Tobias Schlüter <Tobias.Schlueter@Physik.Uni-Muenchen.DE>:
> I marveled at that line earlier myself.  I think it's wrong.  If offset is
> the
> number of storage units / bytes / whatever that the object is into the common
> block, then the inequality would not be correct for a dependency where the
> stride is negative (i.e. a(5:1:-1)), so I think the right course of action is
> to remove that check, and only do smart stuff if there are no equivalences
> involved.

Forgot to say: if we decide to keep it, the comment in front of the function
should be enhanced, I couldn't tell from it which argument represents the
source, and which the destination.

- Tobi