------- Comment #9 from segalemb at usp dot br 2005-10-07 13:57 -------
(In reply to comment #8)
I tried this simple test case:
module param
double precision mutdefc(8,5,7)
data mutdefc(1,1,1) /0.0000d0/
* mutdefc(1,1,2) /0.0000d0/
end module param
and the compilation go straight. Only when all indexes of mutdefc
where equal, the error appears. This indicates that this is not
a duplicate of bug 17737.
> (In reply to comment #7)
> > Note I reduced it using delta so the all of my reduced tescase came exactly
> from the file and nothing
> > else.
>
> This is not true. Even if the constants appearing in the data statements in
> the
> original code are replaced by their values, there will be no line that reads
> data mutdefc(...,7) /.../
> as all initializations are of the form
> data mutdefc(...,1) /.../
> * mutdefc(...,2) /.../
> ...
> So you did more to the testcase.
>
> Given the error location, I'm inclined to believe that it is indeed a
> duplicate,
> but your reduced testcase doesn't prove it.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23884
