[Bug fortran/35959] New: Recursive function with allocatable array

[Bug fortran/35959] New: Recursive function with allocatable array

[michael dot baudin at gmail dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 3388 bytes --]

Problem with gfortran while developing a recursive function
which allocates an array stored in a derived-type.
This is the sample test :

module testmod

#define _DYNAMIC_TYPE allocatable
!#define _DYNAMIC_TYPE pointer

 type, public :: t_type
    integer, dimension(:), _DYNAMIC_TYPE :: chars
 end type t_type
 integer, save :: callnb = 0
contains
 recursive function recursivefunc ( this ) result ( match )
   type(t_type), intent(in) :: this
   type(t_type) :: subpattern
   logical :: thisalloc
   integer :: thislength
   logical :: match
   write ( * , * ) "recursivefunc [" , callnb , "]"
   thislength = size ( this % chars )
   write ( * , * ) "length :", thislength
   callnb = callnb + 1
   thisalloc = allocated ( this % chars )
!    thisalloc = associated ( this % chars )
   if ( .NOT. thisalloc ) then
      write ( 6 , * ) "STOP with error !"
      stop
   endif
   if ( thislength == 0 ) then
      match = .true.
      return
   endif
   allocate ( subpattern % chars ( thislength - 1 ) )
   match = recursivefunc ( subpattern )
 end function recursivefunc
end module testmod

program testprog
 use testmod
 implicit none
 type(t_type) :: this
 logical :: match
 allocate ( this % chars ( 10 ))
 match = recursivefunc ( this )
 print * , "match :", match
end program testprog

Compile it with :

gfortran -x f95-cpp-input testZeroSizeRecursive2.f90 -o
testZeroSizeRecursive2.exe
(flibs-workbench) 60 % testZeroSizeRecursive2.exe
 recursivefunc [           0 ]
 length :          10
 recursivefunc [           1 ]
 length :           9
 STOP with error !

This is because the allocated statement return .false., even
 if the array is really allocated.
Trying the "pointer" version by modifying the #define preprocessing statement
and the "allocated" to associated" make the problem disappear :

 recursivefunc [           0 ]
 length :          10
 recursivefunc [           1 ]
 length :           9
 recursivefunc [           2 ]
 length :           8
 recursivefunc [           3 ]
 length :           7
 recursivefunc [           4 ]
 length :           6
 recursivefunc [           5 ]
 length :           5
 recursivefunc [           6 ]
 length :           4
 recursivefunc [           7 ]
 length :           3
 recursivefunc [           8 ]
 length :           2
 recursivefunc [           9 ]
 length :           1
 recursivefunc [          10 ]
 length :           0
 match : T

This is the version of gfortran I currently use ( I downloaded it today
and I think that it is up-todate ) :

(flibs-workbench) 61 % gfortran -v
Using built-in specs.
Target: i586-pc-mingw32
Configured with: ../trunk/configure --prefix=/mingw
--enable-languages=c,fortran --with-gmp=/home/FX/local
--with-ld=/mingw/bin/ld --with-as=/mingw/bin/as --disable-werror
--enable-bootstrap --enable-threads --disable-nls
--build=i586-pc-mingw32 --enable-libgomp --disable-shared
Thread model: win32
gcc version 4.4.0 20080312 (experimental) [trunk revision 133139] (GCC)

Regards,

Michaël Baudin


-- 
           Summary: Recursive function with allocatable array
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: michael dot baudin at gmail dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[pault at gcc dot gnu dot org]

------- Comment #1 from pault at gcc dot gnu dot org  2008-04-17 05:59 -------
Here is a further reduced version of the testcase:

program testprog
  type :: t_type
    integer, dimension(:), allocatable :: chars
  end type t_type
  integer, save :: callnb = 0
  type(t_type) :: this
  allocate ( this % chars ( 4))
  print * , "match :", recursivefunc ( this ), callnb
contains
  recursive function recursivefunc ( this ) result ( match )
    type(t_type), intent(in) :: this
    type(t_type) :: subpattern

    logical :: match

    callnb = callnb + 1
    match = (callnb == 10)

    if ((.NOT. allocated (this % chars)) .OR. match) return
    allocate ( subpattern % chars ( 4 ) )
    match = recursivefunc ( subpattern )
  end function recursivefunc
end program testprog

The code that the function produces is:

recursivefunc (struct t_type & this)
{
  logical(kind=4) match;
  static struct t_type subpattern = {};

  subpattern.chars.data = 0B;
  callnb = callnb + 1;
  match = callnb == 10;
  if ((integer(kind=4)[0:] *) this->chars.data == 0B || match) goto
__return_recursivefunc;
  {
    void * D.998;

    subpattern.chars.dtype = 265;
    subpattern.chars.dim[0].lbound = 1;
    subpattern.chars.dim[0].ubound = 4;
    subpattern.chars.dim[0].stride = 1;
    if (subpattern.chars.data == 0B)
      {
        {
          void * D.999;

          D.999 = __builtin_malloc (16);
          if (D.999 == 0B)
            {
              _gfortran_os_error (&"Out of memory"[1]{lb: 1 sz: 1});
            }
          D.998 = D.999;
        }
      }
    else
      {
        _gfortran_runtime_error (&"Attempting to allocate already allocated
array"[1]{lb: 1 sz: 1});
      }
    subpattern.chars.data = D.998;
    subpattern.chars.offset = -1;
  }
  match = recursivefunc (&subpattern);
  __return_recursivefunc:;
  if (subpattern.chars.data != 0B)
    {
      __builtin_free (subpattern.chars.data);
    }
  subpattern.chars.data = 0B;
  return match;
}

Being a local variable in a recursive procedure, it is declared to be static. 
Having allocatable components, it is nullified on entry to the procedure.  This
explains the observed behaviour.

I believe that it is a bug but will have to take a close look at the standard
to confirm it - in fact, I have a suspicion that it might not be covered.  If
the latter is the case, I would be inclined to modify gfortran because the
unexpectd behaviour.

Confirmed for now

Paul


-- 

pault at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pault at gcc dot gnu dot org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2008-04-17 05:59:41
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[pault at gcc dot gnu dot org]

------- Comment #2 from pault at gcc dot gnu dot org  2008-04-17 20:37 -------
This is indeed a bug.  A fix will be posted in half an hour or so.

Paul


-- 

pault at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |pault at gcc dot gnu dot org
                   |dot org                     |
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2008-04-17 05:59:41         |2008-04-17 20:37:53
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[michael dot baudin at gmail dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1637 bytes --]



------- Comment #3 from michael dot baudin at gmail dot com  2008-04-18 08:13 -------
Subject: Re:  Recursive function with allocatable array

Hi Paul,

The generic source code that you sent to me is interesting
and you analysis is very clear. Thank you for taking the time
to fix this.

But I wonder how you produced this source (which have many in common
with C), because I looked in the  gfortran manual, and I obviously
did not find any option to generate the generic source code,
which is an internal step. Is that available only with the middle-end
of gfortran and not available for end-users ?

Best regards,

Michaël


On Thu, Apr 17, 2008 at 10:37 PM, pault at gcc dot gnu dot org
<gcc-bugzilla@gcc.gnu.org> wrote:
>
>
>  ------- Comment #2 from pault at gcc dot gnu dot org  2008-04-17 20:37 -------
>  This is indeed a bug.  A fix will be posted in half an hour or so.
>
>
>  Paul
>
>
>  --
>
>  pault at gcc dot gnu dot org changed:
>
>            What    |Removed                     |Added
>  ----------------------------------------------------------------------------
>          AssignedTo|unassigned at gcc dot gnu   |pault at gcc dot gnu dot org
>                    |dot org                     |
>              Status|NEW                         |ASSIGNED
>    Last reconfirmed|2008-04-17 05:59:41         |2008-04-17 20:37:53
>
>
>                date|                            |
>
>
>  http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959
>
>  ------- You are receiving this mail because: -------
>  You reported the bug, or are watching the reporter.
>


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[jvdelisle at gcc dot gnu dot org]

------- Comment #4 from jvdelisle at gcc dot gnu dot org  2008-04-18 14:04 -------
Reply to: But I wonder how you produced this source (which have many in common
with C),

The intermediate code can be listed using the -fdump-tree-original compiler
flag. It will be placed in a file with the word original appended to the name.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[michael dot baudin at gmail dot com]

------- Comment #5 from michael dot baudin at gmail dot com  2008-04-18 14:11 -------
Subject: Re:  Recursive function with allocatable array

Thank you very much.
I will try immediately.

On Fri, Apr 18, 2008 at 4:04 PM, jvdelisle at gcc dot gnu dot org
<gcc-bugzilla@gcc.gnu.org> wrote:
>  ------- Comment #4 from jvdelisle at gcc dot gnu dot org  2008-04-18 14:04 -------
>  Reply to: But I wonder how you produced this source (which have many in common
>  with C),
>
>  The intermediate code can be listed using the -fdump-tree-original compiler
>  flag. It will be placed in a file with the word original appended to the name.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[michael dot baudin at gmail dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1259 bytes --]



------- Comment #6 from michael dot baudin at gmail dot com  2008-04-18 15:07 -------
Subject: Re:  Recursive function with allocatable array

I used the -fdump-tree-original option to compare the version produced on this
source code with allocatable and pointer.
Only the allocatable contains the following statements :

  static struct t_type subpattern = {};
  subpattern.chars.data = 0B;

I understand that the statement "  subpattern.chars.data = 0B;" allows
to free the array when it goes out of scope, as the standard says about
allocatable arrays (which explains why it is not in the pointer version) :

  if (subpattern.chars.data != 0B)
    {
      __builtin_free (subpattern.chars.data);
    }
  subpattern.chars.data = 0B;

The statement "static struct t_type subpattern = {};", which is not in the
pointer version, is the cause of the problem so that the
solution is to remove it : at second call, the array has been
initialiazed to null, and because it is static, it memory is still deallocated
on recursive calls.
An allocatable array in a recursive subroutine should not be static
(in fortran,
that means having the SAVE attribute), isn't it ?
Am I right ?

Best regards,

Michaël


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[pault at gcc dot gnu dot org]

------- Comment #7 from pault at gcc dot gnu dot org  2008-04-19 21:56 -------
Subject: Bug 35959

Author: pault
Date: Sat Apr 19 21:55:24 2008
New Revision: 134472

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=134472
Log:
2008-04-19  Paul Thomas  <pault@gcc.gnu.org>

        PR fortran/35944
        PR fortran/35946
        PR fortran/35947
        * trans_array.c (gfc_trans_array_constructor): Temporarily
        realign loop, if loop->from is not zero, before creating
        the temporary array and provide an offset.

        PR fortran/35959
        * trans-decl.c (gfc_init_default_dt): Add gfc_ prefix to name
        and allow for NULL body.  Change all references from
        init_default_dt to gfc_init_default_dt.
        * trans.h : Add prototype for gfc_init_default_dt.
        * trans-array.c (gfc_trans_deferred_vars): After nullification
        call gfc_init_default_dt for derived types with allocatable
        components.

2008-04-19  Paul Thomas  <pault@gcc.gnu.org>

        PR fortran/35944
        PR fortran/35946
        PR fortran/35947
        * gfortran.dg/array_constructor_23.f: New test.

        PR fortran/35959
        * gfortran.dg/alloc_comp_default_init_2.f90: New test.
        * gfortran.dg/alloc_comp_basics_1.f90: Change occurrences of
        "builtin_free" to 27.
        * gfortran.dg/alloc_comp_constructor_1.f90: Change occurrences
        of "builtin_free" to 21.

Added:
    trunk/gcc/testsuite/gfortran.dg/alloc_comp_default_init_2.f90
    trunk/gcc/testsuite/gfortran.dg/array_constructor_23.f
Modified:
    trunk/gcc/fortran/ChangeLog
    trunk/gcc/fortran/trans-array.c
    trunk/gcc/fortran/trans-decl.c
    trunk/gcc/fortran/trans.h
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gfortran.dg/alloc_comp_basics_1.f90
    trunk/gcc/testsuite/gfortran.dg/alloc_comp_constructor_1.f90


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[pault at gcc dot gnu dot org]

------- Comment #8 from pault at gcc dot gnu dot org  2008-04-19 22:31 -------
Subject: Bug 35959

Author: pault
Date: Sat Apr 19 22:30:03 2008
New Revision: 134475

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=134475
Log:
2008-04-19  Paul Thomas  <pault@gcc.gnu.org>

        PR fortran/35944
        PR fortran/35946
        PR fortran/35947
        * trans_array.c (gfc_trans_array_constructor): Temporarily
        realign loop, if loop->from is not zero, before creating
        the temporary array and provide an offset.

        PR fortran/35959
        * trans-decl.c (gfc_init_default_dt): Add gfc_ prefix to name
        and allow for NULL body.  Change all references from
        init_default_dt to gfc_init_default_dt.
        * trans.h : Add prototype for gfc_init_default_dt.
        * trans-array.c (gfc_trans_deferred_vars): After nullification
        call gfc_init_default_dt for derived types with allocatable
        components.

2008-04-19  Paul Thomas  <pault@gcc.gnu.org>

        PR fortran/35944
        PR fortran/35946
        PR fortran/35947
        * gfortran.dg/array_constructor_23.f: New test.

        PR fortran/35959
        * gfortran.dg/alloc_comp_default_init_2.f90: New test.
        * gfortran.dg/alloc_comp_basics_1.f90: Change occurrences of
        "builtin_free" to 27.
        * gfortran.dg/alloc_comp_constructor_1.f90: Change occurrences
        of "builtin_free" to 21.

Added:
   
branches/gcc-4_3-branch/gcc/testsuite/gfortran.dg/alloc_comp_default_init_2.f90
    branches/gcc-4_3-branch/gcc/testsuite/gfortran.dg/array_constructor_23.f
Modified:
    branches/gcc-4_3-branch/gcc/fortran/ChangeLog
    branches/gcc-4_3-branch/gcc/fortran/trans-array.c
    branches/gcc-4_3-branch/gcc/fortran/trans-decl.c
    branches/gcc-4_3-branch/gcc/fortran/trans.h
    branches/gcc-4_3-branch/gcc/testsuite/ChangeLog
    branches/gcc-4_3-branch/gcc/testsuite/gfortran.dg/alloc_comp_basics_1.f90
   
branches/gcc-4_3-branch/gcc/testsuite/gfortran.dg/alloc_comp_constructor_1.f90


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[pault at gcc dot gnu dot org]

------- Comment #9 from pault at gcc dot gnu dot org  2008-04-19 22:32 -------
Fixed on trunk and 4.3.

Thanks for the report.

Paul


-- 

pault at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959

[Bug fortran/35959] Recursive function with allocatable array

[michael dot baudin at gmail dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 494 bytes --]



------- Comment #10 from michael dot baudin at gmail dot com  2008-04-24 19:55 -------
Subject: Re:  Recursive function with allocatable array

Thank you for fixing the bug.

Michaël

On Sun, Apr 20, 2008 at 12:32 AM, pault at gcc dot gnu dot org
<gcc-bugzilla@gcc.gnu.org> wrote:
>
>
>  ------- Comment #9 from pault at gcc dot gnu dot org  2008-04-19 22:32 -------
>  Fixed on trunk and 4.3.
>
>  Thanks for the report.
>
>  Paul


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35959