Bug 44863 - [OOP] Fortran runtime error: internal error: bad hash value in dynamic dispatch
Summary: [OOP] Fortran runtime error: internal error: bad hash value in dynamic dispatch
Status: RESOLVED FIXED
Alias: None
Product: gcc
Classification: Unclassified
Component: fortran (show other bugs)
Version: 4.5.1
: P3 normal
Target Milestone: 4.6.0
Assignee: Not yet assigned to anyone
URL:
Keywords: wrong-code
Depends on:
Blocks:
 
Reported: 2010-07-08 02:09 UTC by David Car
Modified: 2016-11-16 17:46 UTC (History)
2 users (show)

See Also:
Host:
Target:
Build:
Known to work: 4.6.0
Known to fail: 4.5.1
Last reconfirmed:


Attachments
test case described in post (771 bytes, text/plain)
2010-07-08 02:11 UTC, David Car
Details

Note You need to log in before you can comment on or make changes to this bug.
Description David Car 2010-07-08 02:09:17 UTC
The following code compiles fine but produces a runtime error when trying to perform a dynamic dispatch:

!================================================================
module BaseStrategy

  type, public, abstract :: Strategy
   contains
     procedure(strategy_update), pass( this ), deferred :: update
     procedure(strategy_pre_update), pass( this ), deferred :: preUpdate
     procedure(strategy_post_update), pass( this ), deferred :: postUpdate
  end type Strategy

  abstract interface
     subroutine strategy_update( this )
       import Strategy
       class (Strategy), target, intent(in) :: this
     end subroutine strategy_update
  end interface

  abstract interface
     subroutine strategy_pre_update( this )
       import Strategy
       class (Strategy), target, intent(in) :: this
     end subroutine strategy_pre_update
  end interface

  abstract interface
     subroutine strategy_post_update( this )
       import Strategy
       class (Strategy), target, intent(in) :: this
     end subroutine strategy_post_update
  end interface

end module BaseStrategy

!==============================================================
module LaxWendroffStrategy

  use BaseStrategy

  private :: update, preUpdate, postUpdate

  type, public, extends( Strategy ) :: LaxWendroff
     class (Strategy), pointer :: child => null()
     contains
       procedure, pass( this ) :: update
       procedure, pass( this ) :: preUpdate
       procedure, pass( this ) :: postUpdate
  end type LaxWendroff

contains

  subroutine update( this )
    class (LaxWendroff), target, intent(in) :: this

    print *, 'Calling LaxWendroff update'
  end subroutine update

  subroutine preUpdate( this )
    class (LaxWendroff), target, intent(in) :: this

    print *, 'Calling LaxWendroff preUpdate'
  end subroutine preUpdate

  subroutine postUpdate( this )
    class (LaxWendroff), target, intent(in) :: this

    print *, 'Calling LaxWendroff postUpdate'
  end subroutine postUpdate

end module LaxWendroffStrategy
!===============================================================
module KEStrategy

  use BaseStrategy
  ! Uncomment the line below and it runs fine                                                    
  ! use LaxWendroffStrategy                                                                      

  private :: update, preUpdate, postUpdate

  type, public, extends( Strategy ) :: KE
     class (Strategy), pointer :: child => null()
     contains
       procedure, pass( this ) :: update
       procedure, pass( this ) :: preUpdate
       procedure, pass( this ) :: postUpdate
  end type KE

contains

  subroutine init( this, other )
    class (KE), intent(inout) :: this
    class (Strategy), target, intent(in) :: other

    this % child => other
  end subroutine init

  subroutine update( this )
    class (KE), target, intent(in) :: this

    if ( associated( this % child ) ) then
       call this % child % update()
    end if

    print *, 'Calling KE update'
  end subroutine update

 subroutine preUpdate( this )
    class (KE), target, intent(in) :: this

    if ( associated( this % child ) ) then
       call this % child % preUpdate()
    end if

    print *, 'Calling KE preUpdate'
  end subroutine preUpdate

  subroutine postUpdate( this )
    class (KE), target, intent(in) :: this

    if ( associated( this % child ) ) then
       call this % child % postUpdate()
    end if

    print *, 'Calling KE postUpdate'
  end subroutine postUpdate

end module KEStrategy

!=============================================================                
program main

  use LaxWendroffStrategy
  use KEStrategy

  type :: StratSeq
     class (Strategy), pointer :: strat => null()
  end type StratSeq

  type (LaxWendroff), target :: lw_strat
  type (KE), target :: ke_strat

  allocate( seq(10) )

  call init( ke_strat, lw_strat )
  call ke_strat % preUpdate()
  call ke_strat % update()
  call ke_strat % postUpdate()                 

end program main

The specific runtime error is:

At line 111 of file test.f90
Fortran runtime error: internal error: bad hash value in dynamic dispatch
 
Line 111 above is the line:
       call this % child % preUpdate()

If I uncomment the line the comment "! Uncomment the line below and it runs fine" the code builds and runs without an error.
Comment 1 David Car 2010-07-08 02:11:17 UTC
Created attachment 21136 [details]
test case described in post
Comment 2 Tobias Burnus 2010-07-08 07:22:51 UTC
Can you update to GCC 4.6? I get the same error with 4.5, but with 4.6 I get the same result as with NAG or Intel. The reason is that gfortran now uses a proper vtable. Or in the words of http://gcc.gnu.org/gcc-4.6/changes.html :

* Improved support for polymorphism between libraries and programs and for complicated inheritance patterns (cf. http://gcc.gnu.org/wiki/OOP)

Newer binaries can be found at http://gcc.gnu.org/wiki/GFortranBinaries
Comment 3 janus 2010-07-10 08:35:44 UTC
Closing as fixed.
Comment 4 janus 2010-08-21 14:51:19 UTC
Subject: Bug 44863

Author: janus
Date: Sat Aug 21 14:50:57 2010
New Revision: 163445

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=163445
Log:
2010-08-21  Janus Weil  <janus@gcc.gnu.org>

	PR fortran/45271
	PR fortran/45290
	* class.c (add_proc_comp): Add static initializer for PPCs.
	(add_procs_to_declared_vtab): Modified comment.
	* module.c (mio_component): Add argument 'vtype'. Don't read/write the
	initializer if the component is part of a vtype.
	(mio_component_list): Add argument 'vtype', pass it on to
	'mio_component'.
	(mio_symbol): Modified call to 'mio_component_list'.
	* trans.h (gfc_conv_initializer): Modified prototype.
	(gfc_trans_assign_vtab_procs): Removed.
	* trans-common.c (create_common): Modified call to
	'gfc_conv_initializer'.
	* trans-decl.c (gfc_get_symbol_decl,get_proc_pointer_decl,
	gfc_emit_parameter_debug_info): Modified call to
	'gfc_conv_initializer'.
	(build_function_decl): Remove assertion.
	* trans-expr.c (gfc_conv_derived_to_class,gfc_trans_class_assign):
	Removed call to 'gfc_trans_assign_vtab_procs'.
	(gfc_conv_initializer): Add argument 'procptr'.
	(gfc_conv_structure): Modified call to 'gfc_conv_initializer'.
	(gfc_trans_assign_vtab_procs): Removed.
	* trans-stmt.c (gfc_trans_allocate): Removed call to
	'gfc_trans_assign_vtab_procs'.


2010-08-21  Janus Weil  <janus@gcc.gnu.org>

	PR fortran/44863
	PR fortran/45271
	PR fortran/45290
	* gfortran.dg/dynamic_dispatch_10.f03: New (PR 44863 comment #1).
	* gfortran.dg/pointer_init_5.f90: New (PR 45290 comment #6).
	* gfortran.dg/typebound_call_18.f03: New (PR 45271 comment #3).

Added:
    trunk/gcc/testsuite/gfortran.dg/dynamic_dispatch_10.f03
    trunk/gcc/testsuite/gfortran.dg/pointer_init_5.f90
    trunk/gcc/testsuite/gfortran.dg/typebound_call_18.f03
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/class.c
    trunk/gcc/fortran/module.c
    trunk/gcc/fortran/trans-common.c
    trunk/gcc/fortran/trans-decl.c
    trunk/gcc/fortran/trans-expr.c
    trunk/gcc/fortran/trans-stmt.c
    trunk/gcc/fortran/trans.h
    trunk/gcc/testsuite/ChangeLog

Comment 5 janus 2010-08-21 14:59:47 UTC
r163445 adds comment #1 to the testsuite, because it uncovered a regression in my recent patch for PR 45271 (thanks to Dominique for noticing).