On Tue, Mar 06, 2018 at 10:40:14PM +0100, Thomas Koenig wrote:
> the attached patch fixes a bug, partly an 8 regression, for
> simplifying an expression containing minloc or maxloc.
> The underlying problem was that
> integer, dimension(0), parameter :: z=0
> ended up as EXPR_CONSTANT even though the rank was one, which
> was then passed to the simplification routines, which either
> ICEd or gave wrong results.
> In doing this, I had to change the logic of the is_size_zero_array
> function. Trying to call it from within the simplification rountines
> led to the simplification routines to be called, and so on... until
> the stack ran out.
> As soon as this is committed, I'll also look if there is anything left
> in PR66128, and close that bug if appropriate
I don't think there is anything of value left unless you want to
investigate the old g77 compatibility functions.
> Regression-tested. OK for trunk?