[Bug fortran/30239] New: duplicate data type assignment not detected

[Bug fortran/30239] New: duplicate data type assignment not detected

[dfranke at gcc dot gnu dot org]

The Intel Fortan Compiler issues an error on the code below

INTEGER FUNCTION foo()
  INTEGER :: foo
  foo = 0
END FUNCTION

gfortran takes it without further notice. Letting "INTEGER :: foo" be "REAL ::
foo" gives "Error: Symbol 'foo' at (1) already has basic type of INTEGER" on
gfortran as well.

$> gfortran-svn -v
gcc version 4.3.0 20061128 (experimental)


-- 
           Summary: duplicate data type assignment not detected
           Product: gcc
           Version: 4.3.0
            Status: UNCONFIRMED
          Keywords: diagnostic
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: dfranke at gcc dot gnu dot org
  GCC host triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[burnus at gcc dot gnu dot org]

------- Comment #1 from burnus at gcc dot gnu dot org  2006-12-17 20:15 -------
This is because in symbol.c's gfc_add_type there is

      const char *msg = "Symbol '%s' at %L already has basic type of %s";
      if (!(sym->ts.type == ts->typesym->ts.type == ts->type
             && (sym->attr.flavor == FL_PROCEDURE || sym->attr.result))
           || gfc_notification_std (GFC_STD_GNU) == ERROR
           || pedantic)
        {
          gfc_error (msg, sym->name, where, gfc_basic_typename (sym->ts.type));
          return FAILURE;
        }

The !(sym->ts.type == ts->type) prevents the error message. Removing this check
should all what is needed for this bug.
But this might break other cases were it is allowed. (I didn't check/though
about those, however).


-- 

burnus at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |burnus at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2006-12-17 20:15:55
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[pault at gcc dot gnu dot org]

------- Comment #2 from pault at gcc dot gnu dot org  2006-12-18 06:01 -------
This was a fix for a PR about a year ago - the std=gnu is meant, obviously, to
enforce all versions of the standard on this.  However, a number of other
compilers did/do permit this wrinkle on the standard and it even appears in one
of Meissner's examples.

> should all what is needed for this bug.
> But this might break other cases were it is allowed. (I didn't check/though
> about those, however).
> 

Paul


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[burnus at gcc dot gnu dot org]

------- Comment #3 from burnus at gcc dot gnu dot org  2006-12-18 07:49 -------
> This was a fix for a PR about a year ago - the std=gnu is meant, obviously, to
> enforce all versions of the standard on this.  However, a number of other
> compilers did/do permit this wrinkle on the standard and it even appears in one
> of Meissner's examples.

Thus: Should we warn? Give an error (as ifort, NAG f95 and g95 do)? Or keep the
status quo where an error is given only for -pedantic and -std=f95/f2003 ?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[dfranke at gcc dot gnu dot org]

------- Comment #4 from dfranke at gcc dot gnu dot org  2006-12-18 08:20 -------
Ups, I didn't check with -pedantic or the -std options.
Since others treat it as an error, I think, a warning in -std=gnu should be the
very least. 


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[domob at gcc dot gnu dot org]

-- 

domob at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |domob at gcc dot gnu dot org
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2006-12-17 20:15:55         |2008-08-22 07:18:27
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[domob at gcc dot gnu dot org]

------- Comment #5 from domob at gcc dot gnu dot org  2008-08-22 11:40 -------
What's with this PR, do you have any decision?  I'm not sure about a `fixed'
warning, as this seems to be not common with gfortran, right?

But I'd suggest maybe a -Wsurprising warning, so that at least -Wall will
notify about this thing together with -pedantic or -std=*.  What do you think
about this?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[dfranke at gcc dot gnu dot org]

------- Comment #6 from dfranke at gcc dot gnu dot org  2008-08-22 12:46 -------
FUNCTION foo()
  INTEGER :: foo
  INTEGER :: foo
  foo = 42
END FUNCTION

This is the same and I'd expect (at least) a warning here -- without -pedantic.
IMO, one should be warned (-Wsurpring/-Wall, no -pedantic, in any standard) if
the type is confirmed as shown above.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[domob at gcc dot gnu dot org]

------- Comment #7 from domob at gcc dot gnu dot org  2008-08-22 20:37 -------
Subject: Bug 30239

Author: domob
Date: Fri Aug 22 20:36:12 2008
New Revision: 139499

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=139499
Log:
2008-08-22  Daniel Kraft  <d@domob.eu>

        PR fortran/30239
        * symbol.c (gfc_add_type): Warn on -Wsurprising if a function-result
        type is re-declared but neither -pedantic nor -std=f* is given and so
        this is no error.
        * invoke.texi (-Wsurprising): Document this new behaviour.


Added:
    trunk/gcc/testsuite/gfortran.dg/duplicate_type_1.f90
    trunk/gcc/testsuite/gfortran.dg/duplicate_type_2.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/invoke.texi
    trunk/gcc/fortran/symbol.c
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239

[Bug fortran/30239] duplicate data type assignment not detected

[domob at gcc dot gnu dot org]

------- Comment #8 from domob at gcc dot gnu dot org  2008-08-22 20:38 -------
I think we can fix this now, added a -Wsurprising warning.


-- 

domob at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30239