[Bug fortran/55855] New: incorrect warning with type-bound procedure pointer

[Bug fortran/55855] New: incorrect warning with type-bound procedure pointer

[abensonca at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

             Bug #: 55855
           Summary: incorrect warning with type-bound procedure pointer
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: abensonca@gmail.com


The following:

module eventMod
  type, public :: event
     private
    procedure(eventTask ), pointer, public :: task
 end type event
  abstract interface
     logical function eventTask(self)
       import event
       class(event), intent(in) :: self
     end function eventTask
  end interface
end module eventMod
program test
  use eventMod
  implicit none
  logical :: r
  type(event), pointer :: myEvent
  allocate(myEvent)
  r=myEvent%task()
end program test

causes the following warning when compiled (with gfortran 4.8.0):

$ gfortran -v
Using built-in specs.
COLLECT_GCC=gfortran
COLLECT_LTO_WRAPPER=/home/abenson/Galacticus/Tools/libexec/gcc/x86_64-unknown-
linux-gnu/4.8.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-trunk/configure --
prefix=/home/abenson/Galacticus/Tools --enable-languages=c,c++,fortran --
disable-multilib --with-gmp=/home/abenson/Galacticus/Tools
Thread model: posix
gcc version 4.8.0 20121219 (experimental) (GCC)

$ gfortran -o warn.exe warn.F90 -Wall
warn.F90:19.4:

  r=myEvent%task()
    1
Warning: POINTER valued function appears on right-hand side of assignment at
(1)

As far as I can tell, the code runs correctly despite this warning.

the warning is spurious as the "task" function returns a logical, not a
pointer.

[Bug fortran/55855] [OOP] incorrect warning with type-bound procedure on pointer-valued base object

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

janus at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |diagnostic
   Last reconfirmed|                            |2013-01-03
                 CC|                            |janus at gcc dot gnu.org
     Ever Confirmed|0                           |1
            Summary|incorrect warning with      |[OOP] incorrect warning
                   |type-bound procedure        |with type-bound procedure
                   |pointer                     |on pointer-valued base
                   |                            |object

--- Comment #1 from janus at gcc dot gnu.org 2013-01-03 09:15:00 UTC ---
The following draft patch seems to be sufficient to fix it:


Index: gcc/fortran/expr.c
===================================================================
--- gcc/fortran/expr.c    (revision 194763)
+++ gcc/fortran/expr.c    (working copy)
@@ -3151,9 +3151,8 @@ gfc_check_assign (gfc_expr *lvalue, gfc_expr *rval

   /* This is possibly a typo: x = f() instead of x => f().  */
   if (gfc_option.warn_surprising
-      && rvalue->expr_type == EXPR_FUNCTION
-      && rvalue->symtree->n.sym->attr.pointer)
-    gfc_warning ("POINTER valued function appears on right-hand side of "
+      && rvalue->expr_type == EXPR_FUNCTION && gfc_expr_attr (rvalue).pointer)
+    gfc_warning ("POINTER-valued function appears on right-hand side of "
          "assignment at %L", &rvalue->where);

   /* Check size of array assignments.  */

[Bug fortran/55855] [OOP] incorrect warning with procedure pointer component on pointer-valued base object

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

janus at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[OOP] incorrect warning     |[OOP] incorrect warning
                   |with type-bound procedure   |with procedure pointer
                   |on pointer-valued base      |component on pointer-valued
                   |object                      |base object

--- Comment #2 from janus at gcc dot gnu.org 2013-01-03 09:30:11 UTC ---
Adjusting title. On first glance I falsely identified 'task' as a type-bound
procedure (although it is a procedure pointer component).


Further reduced test case:

  implicit none
  type :: event
    procedure(logical), pointer, nopass :: task
  end type event
  logical :: r
  type(event), pointer :: myEvent
  allocate(myEvent)
  r=myEvent%task()
end

[Bug fortran/55855] [OOP] incorrect warning with procedure pointer component on pointer-valued base object

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

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

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

--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> 2013-01-03 11:09:02 UTC ---
(In reply to comment #1)
> The following draft patch seems to be sufficient to fix it:
> +      && rvalue->expr_type == EXPR_FUNCTION
> +      && gfc_expr_attr (rvalue).pointer)

The patch is OK (obvious) with a test case.

(In reply to comment #2)
> Adjusting title. On first glance I falsely identified 'task' as a type-bound
> procedure (although it is a procedure pointer component).

In any case, the patch (I just had tested an identical patch) fixes the issue
and also works with pointer-returning functions and TBP.

[Bug fortran/55855] [OOP] incorrect warning with procedure pointer component on pointer-valued base object

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

janus at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
         AssignedTo|unassigned at gcc dot       |janus at gcc dot gnu.org
                   |gnu.org                     |

--- Comment #4 from janus at gcc dot gnu.org 2013-01-03 15:33:38 UTC ---
(In reply to comment #3)
> (In reply to comment #1)
> > The following draft patch seems to be sufficient to fix it:
> > +      && rvalue->expr_type == EXPR_FUNCTION
> > +      && gfc_expr_attr (rvalue).pointer)
> 
> The patch is OK (obvious) with a test case.

Thanks. I just checked that it is free of regressions, apart from a failure in
assignment_1.f90, due to the changed error message (which I modified for
orthographical reasons):

Index: gcc/testsuite/gfortran.dg/assignment_1.f90
===================================================================
--- gcc/testsuite/gfortran.dg/assignment_1.f90    (revision 194763)
+++ gcc/testsuite/gfortran.dg/assignment_1.f90    (working copy)
@@ -12,7 +12,7 @@ integer, target :: t, s
 t = 1
 p => s
 ! We didn't dereference the pointer in the following line.
-p = f() ! { dg-warning "POINTER valued function" "" }
+p = f() ! { dg-warning "POINTER-valued function" "" }
 p = p+1
 if (p.ne.2) call abort()
 if (p.ne.s) call abort()


Will commit soon.

[Bug fortran/55855] [OOP] incorrect warning with procedure pointer component on pointer-valued base object

[mikael at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

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

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

--- Comment #5 from Mikael Morin <mikael at gcc dot gnu.org> 2013-01-03 16:10:57 UTC ---
(In reply to comment #3)
> In any case, the patch (I just had tested an identical patch) fixes the issue
> and also works with pointer-returning functions and TBP.

If you have more complete testcases, they are pre-approved to go together with
Janus' patch.

[Bug fortran/55855] [OOP] incorrect warning with procedure pointer component on pointer-valued base object

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

--- Comment #6 from janus at gcc dot gnu.org 2013-01-03 16:15:07 UTC ---
Author: janus
Date: Thu Jan  3 16:14:54 2013
New Revision: 194857

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=194857
Log:
2013-01-03  Janus Weil  <janus@gcc.gnu.org>

    PR fortran/55855
    * expr.c (gfc_check_assign): Use 'gfc_expr_attr' to evaluate attributes
    of rvalue. Correct hyphenation in error message.

2013-01-03  Janus Weil  <janus@gcc.gnu.org>

    PR fortran/55855
    * gfortran.dg/assignment_1.f90: Modified.
    * gfortran.dg/assignment_4.f90: New.

Added:
    trunk/gcc/testsuite/gfortran.dg/assignment_4.f90
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/expr.c
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gfortran.dg/assignment_1.f90

[Bug fortran/55855] [OOP] incorrect warning with procedure pointer component on pointer-valued base object

[janus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55855

janus at gcc dot gnu.org changed:

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

--- Comment #7 from janus at gcc dot gnu.org 2013-01-03 20:17:59 UTC ---
(In reply to comment #5)
> (In reply to comment #3)
> > In any case, the patch (I just had tested an identical patch) fixes the issue
> > and also works with pointer-returning functions and TBP.
> 
> If you have more complete testcases, they are pre-approved to go together with
> Janus' patch.

I guess the one test case we have should be enough (the patch should fix all
variants).

r194857 fixes the bug on 4.8 trunk. Closing.

Thanks to Andrew for the report!