A PARAMETER value is not allowed in a DATA statement, similar to an 
EQUIVALENCE.

The check for this was in gfc_assign_data_value() in data.cc which turns 
out to be too late when trying to assign a zero sized array.

To correct this, the check is moved to match_variable() in primary.cc 
where  a similar check for EQUIVALENCE is already being performed.

Regression tested on x86_64-linux-gnu.  I will create a test case from 
the case presented in the PR which is trivial.  There are already two 
other tests in the test suite that exercise this check.

OK for trunk?

Regards,

Jerry