Bug#122114: marked as done (gcc-3.0: Weird SegFault on exit())
Your message dated Thu, 30 May 2002 00:13:09 +0200
with message-id <15605.21109.31516.104034@gargle.gargle.HOWL>
and subject line Bug#122114: feedback wanted on your Debian report #122114
has caused the attached Bug report to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what I am
talking about this indicates a serious mail system misconfiguration
somewhere. Please contact me immediately.)
Debian bug tracking system administrator
(administrator, Debian Bugs database)
--------------------------------------
Received: (at submit) by bugs.debian.org; 2 Dec 2001 15:27:07 +0000
>From jlgg@operamail.com Sun Dec 02 09:27:07 2001
Return-path: <jlgg@operamail.com>
Received: from 172-madr-x47.libre.retevision.es (caballero.localnet) [62.83.24.172]
by master.debian.org with esmtp (Exim 3.12 1 (Debian))
id 16AYWI-0001tm-00; Sun, 02 Dec 2001 09:27:07 -0600
Received: from jl by caballero.localnet with local (Exim 3.32 #1 (Debian))
id 16AYW4-0000FF-00; Sun, 02 Dec 2001 16:26:52 +0100
From: José Luis González <jlgg@operamail.com>
Subject: gcc-3.0: Weird SegFault on exit()
To: submit@bugs.debian.org
X-Mailer: bug 3.3.10
Message-Id: <E16AYW4-0000FF-00@caballero.localnet>
Sender: =?iso-8859-1?Q?Jos=E9_Luis_Gonz=E1lez_Gonz=E1lez?= <jlgg@operamail.com>
Date: Sun, 02 Dec 2001 16:26:52 +0100
Delivered-To: submit@bugs.debian.org
Package: gcc-3.0
Version: 1:3.0.2-3
Severity: normal
Weird reproducible bug for gcc-3.0 on i386 (I have not checked other
architectures).
The program runs correctly, but when exiting, it segfaults. Doing a
backtrace reveals the segfault was out of the program, on system
libraries. This seems to be a libc bug, but it does not happen with
gcc 2.95.4.
I have placed a program that reproduces this bug on
http://garfield.asoc.euitt.upm.es/~jlg/debfiles.c
Removing function "position" the segfault does not happen, which
makes this bug even more obscure because "position" does not
touch the C library nor modifies external data (it's just a binary
search on an array of strings. The same program without "position"
(it is replaced with a macro function returning always -1 (not found):
http://garfield.asoc.euitt.upm.es/~jlg/debfiles2.c
Running the program without arguments is enough to get the bug.
The backtrace from debfiles.c:
(gdb) run
Starting program: /home/jl/debfiles
Program received signal SIGSEGV, Segmentation fault.
0x40117922 in __deregister_frame_info_bases () from /lib/libc.so.6
(gdb) where
#0 0x40117922 in __deregister_frame_info_bases () from /lib/libc.so.6
#1 0x40119b55 in __umoddi3 () from /lib/libc.so.6
#2 0x4000b4c6 in _dl_fini () from /lib/ld-linux.so.2
#3 0x4004afa3 in exit () from /lib/libc.so.6
#4 0x40038665 in __libc_start_main () from /lib/libc.so.6
-- System Information
Debian Release: testing/unstable
Kernel Version: Linux caballero 2.4.16 #2 mié nov 28 20:28:55 CET 2001 i586 unknown
Versions of the packages gcc-3.0 depends on:
ii binutils 2.11.92.0.10-4 The GNU assembler, linker and binary utiliti
ii cpp-3.0 3.0.2-3 The GNU C preprocessor.
ii gcc-3.0-base 3.0.2-3 The GNU Compiler Collection (base package).
ii libc6 2.2.4-5 GNU C Library: Shared libraries and Timezone
ii libgcc1 3.0.2-3 GCC support library.
---------------------------------------
Received: (at 122114-done) by bugs.debian.org; 29 May 2002 22:15:18 +0000
>From doko@cs.tu-berlin.de Wed May 29 17:15:18 2002
Return-path: <doko@cs.tu-berlin.de>
Received: from mail.cs.tu-berlin.de [130.149.17.13] (root)
by master.debian.org with esmtp (Exim 3.12 1 (Debian))
id 17DBiv-0005X4-00; Wed, 29 May 2002 17:15:17 -0500
Received: from bolero.cs.tu-berlin.de (daemon@bolero.cs.tu-berlin.de [130.149.19.1])
by mail.cs.tu-berlin.de (8.9.3/8.9.3) with ESMTP id AAA03760;
Thu, 30 May 2002 00:13:09 +0200 (MET DST)
Received: (from doko@localhost)
by bolero.cs.tu-berlin.de (8.11.6+Sun/8.9.3) id g4TMD9d18856;
Thu, 30 May 2002 00:13:09 +0200 (MEST)
From: Matthias Klose <doko@cs.tu-berlin.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
Message-ID: <15605.21109.31516.104034@gargle.gargle.HOWL>
Date: Thu, 30 May 2002 00:13:09 +0200
To: =?iso-8859-1?Q?Jos=E9_Luis_Gonz=E1lez_Gonz=E1lez?= <jlg80@mi.madritel.es>,
122114-done@bugs.debian.org
Subject: Re: Bug#122114: feedback wanted on your Debian report #122114
In-Reply-To: <[🔎] 20020529140137.GB335@caballero.dyndns.org>
References: <[🔎] 20020529140137.GB335@caballero.dyndns.org>
X-Mailer: VM 7.03 under 21.4 (patch 6) "Common Lisp" XEmacs Lucid
Delivered-To: 122114-done@bugs.debian.org
Jos=E9 Luis Gonz=E1lez Gonz=E1lez writes:
> > Please could you recheck with gcc-3.0.4 and gcc-3.1 (in unstable) a=
nd
> > send the arguments you pass to reproduce the segfault? which comman=
d
> > line do you use for compiling?
>=20
> I have checked with gcc 3.0.4 and it does not trigger the bug. Moreov=
er,
> I have checked with the binary that gave me the segfault (compiled wi=
th
> 3.0.2 or 3.0.3) and it doesn't segfault anymore.
>=20
> I'm sorry I can't test with 3.1. Anyway, it seems the bug was elsewhe=
re,
> probably on glibc, as the same binary that segfaulted on two machines=
> now works fine, and also freshly compiled binaries don't reproduce it=
.
> My belief is also reinforced by the fact that I tested 3.0.4 as soon =
as
> it entered woody and, by then, the binaries produced by it segfaulted=
> while those from 2.95.4 didn't. Weird was the fact that the program
> segfaulted AFTER exiting (on _exit()) and I didn't play with atexit()=
or
> similar functions; actually the program didn't touch data outside its=
> own local data, so the segfault on _exit() seemed to be a bug on glib=
c,
> but I filed the report for gcc-3.0 because gcc 2.95.4 proced a binary=
> that did not segfault.
>=20
> As I no longer suffer this bug, I would not mind if it were closed.
Closed.
--
To UNSUBSCRIBE, email to debian-gcc-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: