[Bug fortran/51512] New: RFC: Bogus "Return value of function"/"INTENT(OUT) was not set" with allocatable results/dummies

[Bug fortran/51512] New: RFC: Bogus "Return value of function"/"INTENT(OUT) was not set" with allocatable results/dummies

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51512

             Bug #: 51512
           Summary: RFC: Bogus "Return value of function"/"INTENT(OUT) was
                    not set" with allocatable results/dummies
    Classification: Unclassified
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Keywords: diagnostic
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: burnus@gcc.gnu.org


gfortran -Wall prints for the following program:

  Warning: Return value of function 'f' at (1) not set

However, as "f" is allocatable, the program is valid and the function result is
an unallocated array.

RFC: Should we really print a warning? The program is valid, though most of the
time this indicates a programming error.

In GCC 4.3, gfortran did not warn; 4.4 to 4.7 do.


function f()
  integer, allocatable :: f(:)
end function f


Related:

  integer function f(x)
    integer, allocatable, intent(out) :: x(:)
    f = 5
  end function f

Again, gcc 4.5 (!) to 4.7 print a warning:
   Warning: Dummy argument 'x' at (1) was declared INTENT(OUT) but was not set

But again, the code is questionable but perfectly standard conform and in some
cases even useful.

[Bug fortran/51512] RFC: Bogus "Return value of function"/"INTENT(OUT) was not set" with allocatable results/dummies

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51512

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-06-12
     Ever confirmed|0                           |1

--- Comment #1 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
Confirmed at revision 199988.

> RFC: Should we really print a warning? The program is valid, 
> though most of the time this indicates a programming error.

IMO this is exactly what warnings are for. I suggest to close this PR as
WONTFIX.