Bug in minloc and maxloc intrinsic with .false. mask

Tue Mar 7 21:01:00 GMT 2006

On Tue, Mar 07, 2006 at 09:21:06AM +0100, FranÃ§ois-Xavier Coudert wrote:
> > So I infer that gfortran returns a wrong value for these cases, is this
> > right?.
> 
> You're right, and this is know as PR libfortan/25378
> (http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25378).

> Note that this is a F2003 issue, as F95 allowed a processor-dependent
> value to be returned in these cases.

I'll work on this once I have solved all scalar mask issues with
minloc/maxloc/minval/maxval/product etc.

	Thomas