PR fortran/18899: ubound() failing for assumed-shape arrays

[Steven Bosscher]

On Tuesday 13 September 2005 09:33, Richard Sandiford wrote:
> Steven Bosscher <stevenb@suse.de> writes:
> > On Monday 12 September 2005 11:46, Richard Sandiford wrote:
> >> I think the two cases where we need a lower bound of 1 are:
> >>
> >>     - when referring to a section of b (because lbound() is then
> >>       defined to be 1 for every dimension)
> >>
> >>     - when the array will be used as a pointer target (because the
> >>       pointer code assumes this, and because applying lbound() to
> >>       the pointer should also give 1 for every dimension)
> >
> > Could you perhaps add test cases for that too, then?
>
> OK, the testcase now checks pointer bounds too.  The patch is
> otherwise the same as before.
>
> Bootstrapped & regression-tested on i686-pc-linux-gnu.  OK to install?

It looks OK to me.

Gr.
Steven