[Bug fortran/46982] New: SIZE(TRANSFER()) as specification expression

[Bug fortran/46982] New: SIZE(TRANSFER()) as specification expression

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

           Summary: SIZE(TRANSFER()) as specification expression
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Keywords: rejects-valid
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: burnus@gcc.gnu.org


Found at
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A1=ind1012&L=COMP-FORTRAN-90#2

Related: (C_)SIZEOF PR 46641

It seems as if the following might be valid - though none of my compiler
supports it. Steve Lionel wrote: "My reading of F2003 suggests that this
qualifies as a legal specification expression."


Example:

module something_mod
  integer, parameter :: dp = selected_real_kind(12)
  type something_type
      real(dp) :: a = 0.0_dp
      real(dp) :: b = 0.0_dp
      real(dp) :: c = 0.0_dp
  end type something_type
  type(something_type),save :: something
end module something_mod

  use something_mod
  integer :: something_array(size(transfer(something,(/0/))))
end

[Bug fortran/46982] SIZE(TRANSFER()) as specification expression

[kargl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

kargl at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |kargl at gcc dot gnu.org

--- Comment #1 from kargl at gcc dot gnu.org 2010-12-17 03:27:11 UTC ---
(In reply to comment #0)
> Found at
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A1=ind1012&L=COMP-FORTRAN-90#2
> 
> Related: (C_)SIZEOF PR 46641
> 
> It seems as if the following might be valid - though none of my compiler
> supports it. Steve Lionel wrote: "My reading of F2003 suggests that this
> qualifies as a legal specification expression."
> 

I believe James van Buskirk may disagree with Steve Lionel.

See
http://groups.google.com/group/comp.lang.fortran/browse_frm/thread/b36676064392854e#

I haven't looked at the issue farther than noticing James'
clf post.

[Bug fortran/46982] SIZE(TRANSFER()) as specification expression

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-06-23
     Ever confirmed|0                           |1

--- Comment #2 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
Still present at revision 200321. Is the test in comment #0 valid or not?

[Bug fortran/46982] SIZE(TRANSFER()) as specification expression

[anlauf at gmx dot de]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

Harald Anlauf <anlauf at gmx dot de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |anlauf at gmx dot de

--- Comment #3 from Harald Anlauf <anlauf at gmx dot de> ---
(In reply to Dominique d'Humieres from comment #2)
> Still present at revision 200321. Is the test in comment #0 valid or not?

Probably a case where the standard is not intuitive:

  size(transfer(something_type(),(/0/)))

is a constant expression, but

  size(transfer(something,(/0/)))

apparently is not, simply because "something" is not constant.
IMO it also appears not to qualify as a specification expression.

Replacing

  type(something_type),save :: something

by

  type(something_type),parameter :: something = something_type()

would change something into a constant and make the code compile.

So I think that the code in comment #0 is not legal.

[Bug fortran/46982] SIZE(TRANSFER()) as specification expression

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

--- Comment #4 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
> ... So I think that the code in comment #0 is not legal.

Then, should not this PR closed as INVALID?

[Bug fortran/46982] SIZE(TRANSFER()) as specification expression

[anlauf at gmx dot de]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

--- Comment #5 from Harald Anlauf <anlauf at gmx dot de> ---
(In reply to Dominique d'Humieres from comment #4)
> > ... So I think that the code in comment #0 is not legal.
> 
> Then, should not this PR closed as INVALID?

Yes, close as INVALID.

----------------------

For the record: NAG Fortran Compiler Release 5.3.1(928) says

SOMETHING is not permitted in an initialisation expression


and xlf 14.01.0000.0003 complains:

Variable something_array was declared as array of type ADJUSTABLE.  This type
of array is not permitted in a main program.


Only pgi 13.4-0 accepts it, but I believe that it shouldn't.

[Bug fortran/46982] SIZE(TRANSFER()) as specification expression

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46982

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |INVALID

--- Comment #6 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
> Yes, close as INVALID.

Hence closed.