[Bug fortran/55469] New: memory leak on read with istat.ne.0

[Bug fortran/55469] New: memory leak on read with istat.ne.0

[Joost.VandeVondele at mat dot ethz.ch]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

             Bug #: 55469
           Summary: memory leak on read with istat.ne.0
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: Joost.VandeVondele@mat.ethz.ch


The following testcase leads to memory leaks with gfortran 4.5/4.6/4.7/4.8 (as
found by valgrind) 4.1 seems not to leak (but has a couple of warnings).

REAL :: z
INTEGER :: istat
CHARACTER(LEN=3) :: t
t="NVE"
READ (UNIT=t,FMT=*,IOSTAT=istat) z
END

note that istat.NE.0 in this case.

==37422== 300 bytes in 1 blocks are definitely lost in loss record 1 of 1
==37422==    at 0x4A057F4: calloc (vg_replace_malloc.c:593)
==37422==    by 0x4C298FF: _gfortrani_xcalloc (memory.c:56)
==37422==    by 0x4CE2A82: l_push_char.isra.2 (list_read.c:641)
==37422==    by 0x4CE3223: read_real (list_read.c:1634)
==37422==    by 0x4CE504E: _gfortrani_list_formatted_read (list_read.c:1895)

[Bug fortran/55469] memory leak on read with istat.ne.0

[Joost.VandeVondele at mat dot ethz.ch]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |Joost.VandeVondele at mat
                   |                            |dot ethz.ch

--- Comment #1 from Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> 2012-11-26 08:01:27 UTC ---
BTW, would there be a simple workaround ?

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

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> 2012-11-26 09:26:46 UTC ---
Untested:

--- a/libgfortran/io/list_read.c
+++ b/libgfortran/io/list_read.c
@@ -1966,8 +1966,8 @@ list_formatted_read_scalar (st_parameter_dt *dtp, bt
type, void *p,
     }

+cleanup:
   if (--dtp->u.p.repeat_count <= 0)
     free_saved (dtp);

-cleanup:
   if (err == LIBERROR_END)
     hit_eof (dtp);

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-11-26 10:23:54 UTC ---
Didn't help. The following should work. The crucial part is "free_line". At a
glance free_saved(dtp) (here and in comment 2) seems also to be sensible, but
one should read through the file to check that it is indeed correct - and to
find other places where free_line is missing.

--- a/libgfortran/io/list_read.c
+++ b/libgfortran/io/list_read.c
@@ -1757,8 +1757,9 @@ read_real (st_parameter_dt *dtp, void * dest, int length)
  bad_real:

+  free_line (dtp);
+  free_saved (dtp);
   if (nml_bad_return (dtp, c))
     return;

-  free_saved (dtp);
   if (c == EOF)
     {

[Bug fortran/55469] memory leak on read with istat.ne.0

[matthias.krack at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

Matthias Krack <matthias.krack at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |matthias.krack at gmail dot
                   |                            |com

--- Comment #4 from Matthias Krack <matthias.krack at gmail dot com> 2012-11-28 14:58:23 UTC ---
Thanks Tobias, the fix you propose in comment 3 seems indeed to solve the
problem for the test case of Joost as well as for the valgrind based regtest of
the current CP2K development version in which this memory leak originally
showed up.

[Bug fortran/55469] memory leak on read with istat.ne.0

[matthias.krack at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #5 from Matthias Krack <matthias.krack at gmail dot com> 2012-11-29 10:16:06 UTC ---
Memory leaks are still present for other choices of the string t.
For instance t="." or t="./" is still causing memory leaks for the test case in
comment 1 as detected by valgrind 3.8.1 for a patched (see comment 3) gcc
version 4.7.3 20121129 (prerelease) (GCC)

==25497== 300 bytes in 1 blocks are definitely lost in loss record 1 of 1
==25497==    at 0x4A0765C: malloc (vg_replace_malloc.c:270)
==25497==    by 0x4C29208: _gfortrani_get_mem (memory.c:43)
==25497==    by 0x4CDD799: push_char (list_read.c:78)
==25497==    by 0x4CDE892: read_real (list_read.c:1408)
==25497==    by 0x4CE0F86: _gfortrani_list_formatted_read (list_read.c:1876)

It seems that the logic for the memory cleanup is still incomplete.

Note, that the memory leaks appear only if the IOSTAT parameter is present in
the READ statement.

[Bug fortran/55469] memory leak on read with istat.ne.0

[Joost.VandeVondele at mat dot ethz.ch]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #6 from Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> 2012-11-29 10:23:13 UTC ---
Is that for the more complete patch posted here:

http://gcc.gnu.org/ml/fortran/2012-11/msg00083.html

BTW, wrong PR number in that message.

[Bug fortran/55469] memory leak on read with istat.ne.0

[matthias.krack at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #7 from Matthias Krack <matthias.krack at gmail dot com> 2012-11-29 11:12:56 UTC ---
I also checked the more complete patch PR5469, but it shows still the memory
leaks as described in comment 5 for the 4_7-branch. The same is true for the
gcc trunk version.

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2012-11-29
         AssignedTo|unassigned at gcc dot       |burnus at gcc dot gnu.org
                   |gnu.org                     |
     Ever Confirmed|0                           |1

--- Comment #8 from Tobias Burnus <burnus at gcc dot gnu.org> 2012-11-29 11:39:40 UTC ---
(In reply to comment #5)
> Memory leaks are still present for other choices of the string t.
> For instance t="." or t="./" is still causing memory leaks

Missed a failure as it was hidden behind a function call (convert_real, in a
different file).

(In reply to comment #6)
> BTW, wrong PR number in that message.

Both issues have been fixed in the new patch at:
  http://gcc.gnu.org/ml/fortran/2012-11/msg00092.html

[Bug fortran/55469] memory leak on read with istat.ne.0

[Joost.VandeVondele at mat dot ethz.ch]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #9 from Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> 2013-02-07 05:57:43 UTC ---
This 

http://gcc.gnu.org/ml/gcc/2013-02/msg00068.html

seems the same/similar issue. Was there consensus about the patch ?

[Bug fortran/55469] memory leak on read with istat.ne.0

[Joost.VandeVondele at mat dot ethz.ch]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #10 from Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> ---
This bug is a bit of showstopper for doing automatic leak testing with
gperftools (essentially all our regtests have a non-zero error code with this
leak being detected). Any chance to revive one of the patches posted ?

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #11 from Tobias Burnus <burnus at gcc dot gnu.org> ---
Author: burnus
Date: Tue Oct  1 20:52:49 2013
New Revision: 203086

URL: http://gcc.gnu.org/viewcvs?rev=203086&root=gcc&view=rev
Log:
2013-10-01  Tobias Burnus  <burnus@net-b.de>

        PR fortran/55469
        * io/list_read (parse_repeat, read_integer, read_character,
        parse_real, read_real, check_type, list_formatted_read_scalar,
        finish_list_read): Call list_free.


Modified:
    trunk/libgfortran/ChangeLog
    trunk/libgfortran/io/list_read.c

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #12 from Tobias Burnus <burnus at gcc dot gnu.org> ---
Author: burnus
Date: Fri Oct  4 10:06:23 2013
New Revision: 203201

URL: http://gcc.gnu.org/viewcvs?rev=203201&root=gcc&view=rev
Log:
2013-10-04  Tobias Burnus  <burnus@net-b.de>

        PR fortran/55469
        * io/list_read (parse_repeat, read_integer, read_character,
        parse_real, read_real, check_type, list_formatted_read_scalar,
        finish_list_read): Call list_free.


Modified:
    branches/gcc-4_8-branch/libgfortran/io/list_read.c

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

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

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

--- Comment #14 from Tobias Burnus <burnus at gcc dot gnu.org> ---
FIXED on the trunk and the 4.8 branch.

Thanks for the report – and sorry for taking that long to get it fixed.
>From gcc-bugs-return-431071-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Fri Oct 04 10:26:46 2013
Return-Path: <gcc-bugs-return-431071-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 31760 invoked by alias); 4 Oct 2013 10:26:46 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 31720 invoked by uid 48); 4 Oct 2013 10:26:43 -0000
From: "paolo.carlini at oracle dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/58560] [4.7/4.8/4.9 Regression] [c++11] ICE with auto in typedef
Date: Fri, 04 Oct 2013 10:26:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c++
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords: error-recovery, ice-on-invalid-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: paolo.carlini at oracle dot com
X-Bugzilla-Status: ASSIGNED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 4.9.0
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: bug_status
Message-ID: <bug-58560-4-FbpoPMv6u4@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-58560-4@http.gcc.gnu.org/bugzilla/>
References: <bug-58560-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-10/txt/msg00215.txt.bz2
Content-length: 379

http://gcc.gnu.org/bugzilla/show_bug.cgi?idX560

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Mine.

[Bug fortran/55469] memory leak on read with istat.ne.0

[burnus at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55469

--- Comment #13 from Tobias Burnus <burnus at gcc dot gnu.org> ---
Author: burnus
Date: Fri Oct  4 10:08:57 2013
New Revision: 203202

URL: http://gcc.gnu.org/viewcvs?rev=203202&root=gcc&view=rev
Log:
Add missing changelog entry for Rev. 203201:
2013-10-04  Tobias Burnus  <burnus@net-b.de>

        PR fortran/55469
        * io/list_read (parse_repeat, read_integer, read_character,
        parse_real, read_real, check_type, list_formatted_read_scalar,
        finish_list_read): Call list_free.

Modified:
    branches/gcc-4_8-branch/libgfortran/ChangeLog