[Bug fortran/54199] New: Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[Bug fortran/54199] New: Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54199

             Bug #: 54199
           Summary: Superfluous diagnostic "is also the name of an
                    intrinsic" for internal procedures
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Keywords: diagnostic
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: burnus@gcc.gnu.org


Compiling the following program with -Wall gives the message:

  function fraction(x)
  1
Warning: 'fraction' declared at (1) is also the name of an intrinsic.  It can
only be called via an explicit interface or if declared EXTERNAL.

I believe this message should not be printed for internal procedures.

(For module procedures, there is a different warning, which can stay:
"'fraction' declared at (1) may shadow the intrinsic of the same name.  In
order to call the intrinsic, explicit INTRINSIC declarations may be required.")


subroutine test()
contains
  function fraction(x)
    real :: x
  end function fraction
end subroutine test


Untested patch:

--- a/gcc/fortran/intrinsic.c
+++ b/gcc/fortran/intrinsic.c
@@ -4510,3 +4518,3 @@ gfc_warn_intrinsic_shadow (const gfc_symbol* sym, bool
in_module, bool func)
                 sym->name, &sym->declared_at);
-  else
+  else if (!sym->ns->proc_name)
     gfc_warning ("'%s' declared at %L is also the name of an intrinsic.  It
can"

[Bug fortran/54199] Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[mikael at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54199

Mikael Morin <mikael at gcc dot gnu.org> changed:

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

--- Comment #1 from Mikael Morin <mikael at gcc dot gnu.org> 2012-08-08 19:55:21 UTC ---
(In reply to comment #0)
> Compiling the following program with -Wall gives the message:
> 
>   function fraction(x)
>   1
> Warning: 'fraction' declared at (1) is also the name of an intrinsic.  It can
> only be called via an explicit interface or if declared EXTERNAL.
> 
> I believe this message should not be printed for internal procedures.

Maybe the second part is confusing, but a warning makes sense IMO.

> 
> (For module procedures, there is a different warning, which can stay:
> "'fraction' declared at (1) may shadow the intrinsic of the same name.  In
> order to call the intrinsic, explicit INTRINSIC declarations may be required.")

That one would be fine for internal procedures, don't you think?

[Bug fortran/54199] Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54199

Tobias Burnus <burnus at gcc dot gnu.org> changed:

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

--- Comment #2 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-08-09 07:52:02 UTC ---
(In reply to comment #1)
> Maybe the second part is confusing, but a warning makes sense IMO.

> > (For module procedures, there is a different warning, which can stay:
> > "'fraction' declared at (1) may shadow the intrinsic of the same name.  In
> > order to call the intrinsic, explicit INTRINSIC declarations may be
> > required.")
> 
> That one would be fine for internal procedures, don't you think?

Regarding a warning: Maybe it makes sense. For independent procedures, it makes
a lot of sense as it is very likely to call the wrong procedure. For modules,
usually the module writer has a good idea about the code (unless the module is
very large) but the main problem are the module users, which inadvertently
use-associating such a procedure.

For internal procedures, the risk is much lower: such code is typically much
smaller and it cannot be accessed from the outside. For the parent procedure,
there is also no way to access the intrinsic procedure - the INTRINSIC only
allows the access for other internal procedures. Thus, the importance of the
warning is single proc > module proc > internal proc (and the question where
one stops warning with -Wall).

In any case, I concur that for internal procedures the warning wording for
module procedures is much better than the current one, whose second part is
plainly wrong.

[Bug fortran/54199] Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54199

Tobias Burnus <burnus at gcc dot gnu.org> changed:

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

--- Comment #4 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-08-09 12:06:57 UTC ---
FIXED on the 4.8 trunk.

[Bug fortran/54199] Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54199

Tobias Burnus <burnus at gcc dot gnu.org> changed:

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

--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-08-09 12:06:36 UTC ---
Author: burnus
Date: Thu Aug  9 12:06:31 2012
New Revision: 190251

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=190251
Log:
2012-08-09  Tobias Burnus  <burnus@net-b.de>

        PR fortran/54199
        * intrinsic.c (gfc_warn_intrinsic_shadow): Better warning
        for internal procedures.

2012-08-09  Tobias Burnus  <burnus@net-b.de>

        PR fortran/54199
        * gfortran.dg/intrinsic_shadow_4.f90: New.


Added:
    trunk/gcc/testsuite/gfortran.dg/intrinsic_shadow_4.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/intrinsic.c
    trunk/gcc/testsuite/ChangeLog

--- Comment #4 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-08-09 12:06:57 UTC ---
FIXED on the 4.8 trunk.

[Bug fortran/54199] Superfluous diagnostic "is also the name of an intrinsic" for internal procedures

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54199

--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-08-09 12:06:36 UTC ---
Author: burnus
Date: Thu Aug  9 12:06:31 2012
New Revision: 190251

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=190251
Log:
2012-08-09  Tobias Burnus  <burnus@net-b.de>

        PR fortran/54199
        * intrinsic.c (gfc_warn_intrinsic_shadow): Better warning
        for internal procedures.

2012-08-09  Tobias Burnus  <burnus@net-b.de>

        PR fortran/54199
        * gfortran.dg/intrinsic_shadow_4.f90: New.


Added:
    trunk/gcc/testsuite/gfortran.dg/intrinsic_shadow_4.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/intrinsic.c
    trunk/gcc/testsuite/ChangeLog