Debian Bug report logs - #493045
Segmentation Fault in cron console PHP Job

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Timo Kirst <tkirst@hot-chilli.net>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: Segmentation Fault in cron console PHP Job

Date: Thu, 31 Jul 2008 01:24:10 +0200

Package: php5
Version: 5.2.6-2

Hi there,

exactly since we upgraded to this package we get intermittent
Segmentation Faults from two cron jobs each running a php skript every 5 minutes.

During the last 7 days this happened 20 times. The scripts have not
changed for months now and they ran without any problem at all.

One example (content of the cron status mail):

/bin/sh: line 1: 13954 Segmentation fault      /sbin/mailbox_watcher >>/var/log/mailbox_watcher.log

This mailbox watcher does some checks on a mysql DB and synchronizes
possible changes with the filesystem on the harddrive. Nothing special
and it seems that most of the time (99%) everything runs fine. What is
also confirmed is that it does not depend on the code cases that are
executed, this means it does not depend whether the check skript
returns without having to synchronize or after doing so.

I am simply not able to reproduce it. Although it keeps happening.

Any thoughts on that? If you see any procedure, logs, etc I could try to narrow
the problem down with, pls let me know.

One word about the complexity of the two skripts. They are totally
different. The one is just a "dirty" procedural code, the other is
more object oriented with a class defined.

See the first lines of this skript below:

> #!/usr/bin/php5
> <?php
>
> define('_UNDEFINED',0);
> define('_FATAL',1);
> define('_CRITICAL',2);
> define('_WARNING',3);
> define('_INFO',4);
>
> define('_EXITTHRESHOLD',2);
>
> define('_CRLF',"\n");
>
> class checkAuthDB {
>
>     var $ident;
>     var $conf;
>     var $dblinks;
>     var $lastrun;
>     var $user;
>     var $logIsOpen;
>     var $logfp;
>
>     function __construct($ident) {
>
> (...)

Ok, and here the first lines of the mailbox_watcher

> #!/usr/bin/php5
> <?php
>
> // initialising some variables...
>
>         // mysql connect information (if needed)
>         $db_host="...";
>         $db_user="...";
>         $db_password="...";
>         $db_name="...";
>
>         $output_init=false;
>
>         // registering needed functions...
>
>         function insdate() {
>
> (...)


Thanks in advance for all your efforts!!

-- 
Regards,

Timo Kirst

Message #8 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Raphael Geissert <atomo64@gmail.com>

To: Timo Kirst <tkirst@hot-chilli.net>, 493045@bugs.debian.org

Cc: control@bugs.debian.org

Subject: Re: [php-maint] Bug#493045: Segmentation Fault in cron console PHP Job

Date: Wed, 30 Jul 2008 21:56:30 -0500

[Message part 1 (text/plain, inline)]

tag 493045 moreinfo
thanks

On Wednesday 30 July 2008, Timo Kirst wrote:
> Package: php5
> Version: 5.2.6-2
>
> Hi there,
>
> exactly since we upgraded to this package we get intermittent
> Segmentation Faults from two cron jobs each running a php skript every 5
> minutes.
>
> During the last 7 days this happened 20 times. The scripts have not
> changed for months now and they ran without any problem at all.
>
> One example (content of the cron status mail):
>
> /bin/sh: line 1: 13954 Segmentation fault      /sbin/mailbox_watcher
> >>/var/log/mailbox_watcher.log
>
...
>
> I am simply not able to reproduce it. Although it keeps happening.

What do you mean by that? you can not reproduce it in a separate environment, 
or when manually running the cronjob, or what?

Could you please provide any or both scripts? if you can't: do they use 
time/date functions? what version of the libmysqlclient15off package do you 
have? 

Could you please install php5-dbg and run the script under gdb.

>
> Any thoughts on that? If you see any procedure, logs, etc I could try to
> narrow the problem down with, pls let me know.
>
> One word about the complexity of the two skripts. They are totally
> different. The one is just a "dirty" procedural code, the other is
> more object oriented with a class defined.
>
> See the first lines of this skript below:
...
>
> Ok, and here the first lines of the mailbox_watcher
>
...

Those lines are actually useless.

>
> Thanks in advance for all your efforts!!

Regards,
-- 
Atomo64 - Raphael

Please avoid sending me Word, PowerPoint or Excel attachments.
See http://www.gnu.org/philosophy/no-word-attachments.html

[signature.asc (application/pgp-signature, inline)]

Message #15 received at submit@bugs.debian.org (full text, mbox, reply):

From: Martijn Grendelman <martijn@pocos.nl>

To: Timo Kirst <tkirst@hot-chilli.net>, 493045@bugs.debian.org

Cc: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: Re: Bug#493045: Segmentation Fault in cron console PHP Job

Date: Thu, 31 Jul 2008 09:04:17 +0200

[Message part 1 (text/plain, inline)]

> Package: php5
> Version: 5.2.6-2
> 
> Hi there,
> 
> exactly since we upgraded to this package we get intermittent
> Segmentation Faults from two cron jobs each running a php skript every 5 minutes.

I experienced the same behaviour with my own backported PHP 5.2.6 on
Etch. IIRC it was only with the CLI version and not with the Apache module.

I didn't file a bug report, because I don't use testing or unstable.
However, the problem went away when I rebuilt PHP, linking against the
normal libmysqlclient, instead of libmysqlclient_r.

> During the last 7 days this happened 20 times. The scripts have not
> changed for months now and they ran without any problem at all.
> 
> One example (content of the cron status mail):
> 
> /bin/sh: line 1: 13954 Segmentation fault      /sbin/mailbox_watcher >>/var/log/mailbox_watcher.log
> 
> This mailbox watcher does some checks on a mysql DB and synchronizes
> possible changes with the filesystem on the harddrive. Nothing special
> and it seems that most of the time (99%) everything runs fine. What is
> also confirmed is that it does not depend on the code cases that are
> executed, this means it does not depend whether the check skript
> returns without having to synchronize or after doing so.
> 
> I am simply not able to reproduce it. Although it keeps happening.
> 
> Any thoughts on that? If you see any procedure, logs, etc I could try to narrow
> the problem down with, pls let me know.
> 
> One word about the complexity of the two skripts. They are totally
> different. The one is just a "dirty" procedural code, the other is
> more object oriented with a class defined.
> 
> See the first lines of this skript below:
> 
>> #!/usr/bin/php5
>> <?php
>>
>> define('_UNDEFINED',0);
>> define('_FATAL',1);
>> define('_CRITICAL',2);
>> define('_WARNING',3);
>> define('_INFO',4);
>>
>> define('_EXITTHRESHOLD',2);
>>
>> define('_CRLF',"\n");
>>
>> class checkAuthDB {
>>
>>     var $ident;
>>     var $conf;
>>     var $dblinks;
>>     var $lastrun;
>>     var $user;
>>     var $logIsOpen;
>>     var $logfp;
>>
>>     function __construct($ident) {
>>
>> (...)
> 
> Ok, and here the first lines of the mailbox_watcher
> 
>> #!/usr/bin/php5
>> <?php
>>
>> // initialising some variables...
>>
>>         // mysql connect information (if needed)
>>         $db_host="...";
>>         $db_user="...";
>>         $db_password="...";
>>         $db_name="...";
>>
>>         $output_init=false;
>>
>>         // registering needed functions...
>>
>>         function insdate() {
>>
>> (...)
> 
> 
> Thanks in advance for all your efforts!!


Best regards,

Martijn Grendelman

[smime.p7s (application/x-pkcs7-signature, attachment)]

Message #25 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Timo Kirst <tkirst@hot-chilli.net>

To: Raphael Geissert <atomo64@gmail.com>, 493045@bugs.debian.org

Cc: control@bugs.debian.org

Subject: Re: [php-maint] Bug#493045: Segmentation Fault in cron console PHP Job

Date: Sat, 2 Aug 2008 13:28:48 +0200

[Message part 1 (text/plain, inline)]

Hello Raphael,

thx for the quick reply and sorry for the late response.

I attached one of the modules which causes the problem. It uses the
mktime and the time function. The checkAuthDB.conf file which gets included
just carries the config array with the localhost mysql connect
information.

Version of libmysqlclient15off is 5.0.51a-11

See my other comments below.


On Thursday 31 July 2008, Raphael Geissert wrote:

> tag 493045 moreinfo
> thanks

> On Wednesday 30 July 2008, Timo Kirst wrote:
>> Package: php5
>> Version: 5.2.6-2
>>
>> Hi there,
>>
>> exactly since we upgraded to this package we get intermittent
>> Segmentation Faults from two cron jobs each running a php skript every 5
>> minutes.
>>
>> During the last 7 days this happened 20 times. The scripts have not
>> changed for months now and they ran without any problem at all.
>>
>> One example (content of the cron status mail):
>>
>> /bin/sh: line 1: 13954 Segmentation fault      /sbin/mailbox_watcher
>> >>/var/log/mailbox_watcher.log
>>
> ...
>>
>> I am simply not able to reproduce it. Although it keeps happening.
> What do you mean by that? you can not reproduce it in a separate environment,
> or when manually running the cronjob, or what?

I can not reproduce it means that I did not find a way where the error
appears every single time.

> Could you please provide any or both scripts? if you can't: do they use
> time/date functions? what version of the libmysqlclient15off package do you
> have?

The answers to these questions are stated above.

> Could you please install php5-dbg and run the script under gdb.

I will do that this weekend and report then.

>>
>> Any thoughts on that? If you see any procedure, logs, etc I could try to
>> narrow the problem down with, pls let me know.
>>
>> One word about the complexity of the two skripts. They are totally
>> different. The one is just a "dirty" procedural code, the other is
>> more object oriented with a class defined.
>>
>> See the first lines of this skript below:
> ...
>>
>> Ok, and here the first lines of the mailbox_watcher
>>
> ...

> Those lines are actually useless.

ok.

-- 
Regards, Timo

[checkAuthDB.php (application/octet-stream, attachment)]

Message #34 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Helmut <helmutjd@gmail.com>

To: 493045@bugs.debian.org

Subject: Segmentation Fault in cron console PHP Job

Date: Mon, 15 Sep 2008 15:23:55 -0700

Hello all,

I'm experiencing this same issue and hopefully I can address some of the 
unanswered questions above.  There was some talk about this being 
related to 495575 but I'm using php5-mysql 5.2.6-3 and 
libmysqlclient15off 5.0.51a-12 and I'm still seeing this behavior.

To answer some earlier questions from this report:
- it's not dependent on any particular PHP script; even linting will 
intermittently cause a segfault (i.e. echo test>test.php; php -l test.php)
- the segmentation faults are intermittent; if you run a simple BASH for 
loop calling "php -q test.php" a few dozen times, it'll segfault on 
roughly 30% of the loop iterations (see below)
- it's somehow directly related to the MySQL extension; removing 
/etc/php5/conf.d/mysql*.ini resolves the problem
- it doesn't appear to be a conflict or load-order issue with any other 
extension; I've tried removing all but the MySQL extension and it still 
segfaults; as soon as you remove the MySQL extension all is well

my setup:

root@titanium:/tmp/bt# cat /etc/debian_version
lenny/sid

root@titanium:/tmp/bt# dpkg -l | grep mysql | awk '{print $2 " " $3}'
libdbd-mysql-perl 4.007-1
libmysqlclient15-dev 5.0.51a-12
libmysqlclient15off 5.0.51a-12
mysql-client 5.0.51a-12
mysql-client-5.0 5.0.51a-12
mysql-common 5.0.51a-12
mysql-server-5.0 5.0.51a-12
php4-mysql 6:4.4.4-9+lenny1
php5-mysql 5.2.6-3

root@titanium:/tmp/bt# dpkg -l | grep php5- | awk '{print $2 " " $3}'
php5-cli 5.2.6-3
php5-common 5.2.6-3
php5-curl 5.2.6-3
php5-dbg 5.2.6-3
php5-dev 5.2.6-3
php5-gd 5.2.6-3
php5-imap 5.2.6-3
php5-interbase 5.2.6-3
php5-mcrypt 5.2.6-3
php5-mysql 5.2.6-3


backtrace:

root@titanium:/tmp/bt# echo test > test.php

root@titanium:/tmp/bt# ulimit -c unlimited; for f in `seq 0 10`; do php 
-l test.php; done
No syntax errors detected in test.php
No syntax errors detected in test.php
No syntax errors detected in test.php
No syntax errors detected in test.php
No syntax errors detected in test.php
No syntax errors detected in test.php
No syntax errors detected in test.php
No syntax errors detected in test.php
Segmentation fault (core dumped)
No syntax errors detected in test.php
No syntax errors detected in test.php
Segmentation fault (core dumped)
No syntax errors detected in test.php

root@titanium:/tmp/bt# gdb /usr/bin/php /tmp/bt/core.30230
GNU gdb 6.8-debian
Copyright (C) 2008 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later 
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i486-linux-gnu"...

warning: Can't read pathname for load map: Input/output error.
Reading symbols from /lib/i686/nosegneg/libcrypt.so.1...done.
Loaded symbols for /lib/i686/nosegneg/libcrypt.so.1
Reading symbols from /usr/lib/libz.so.1...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /usr/lib/libpanel.so.5...done.
Loaded symbols for /usr/lib/libpanel.so.5
Reading symbols from /lib/libncurses.so.5...done.
Loaded symbols for /lib/libncurses.so.5
Reading symbols from /usr/lib/i686/cmov/libssl.so.0.9.8...done.
Loaded symbols for /usr/lib/i686/cmov/libssl.so.0.9.8
Reading symbols from /usr/lib/libdb-4.6.so...done.
Loaded symbols for /usr/lib/libdb-4.6.so
Reading symbols from /lib/libbz2.so.1.0...done.
Loaded symbols for /lib/libbz2.so.1.0
Reading symbols from /usr/lib/libpcre.so.3...done.
Loaded symbols for /usr/lib/libpcre.so.3
Reading symbols from /lib/i686/nosegneg/libresolv.so.2...done.
Loaded symbols for /lib/i686/nosegneg/libresolv.so.2
Reading symbols from /lib/i686/nosegneg/libm.so.6...done.
Loaded symbols for /lib/i686/nosegneg/libm.so.6
Reading symbols from /lib/i686/nosegneg/libdl.so.2...done.
Loaded symbols for /lib/i686/nosegneg/libdl.so.2
Reading symbols from /lib/i686/nosegneg/libnsl.so.1...done.
Loaded symbols for /lib/i686/nosegneg/libnsl.so.1
Reading symbols from /usr/lib/libgssapi_krb5.so.2...done.
Loaded symbols for /usr/lib/libgssapi_krb5.so.2
Reading symbols from /usr/lib/libkrb5.so.3...done.
Loaded symbols for /usr/lib/libkrb5.so.3
Reading symbols from /usr/lib/libk5crypto.so.3...done.
Loaded symbols for /usr/lib/libk5crypto.so.3
Reading symbols from /lib/libcom_err.so.2...done.
Loaded symbols for /lib/libcom_err.so.2
Reading symbols from /usr/lib/libxml2.so.2...done.
Loaded symbols for /usr/lib/libxml2.so.2
Reading symbols from /lib/i686/nosegneg/libc.so.6...done.
Loaded symbols for /lib/i686/nosegneg/libc.so.6
Reading symbols from /usr/lib/i686/cmov/libcrypto.so.0.9.8...done.
Loaded symbols for /usr/lib/i686/cmov/libcrypto.so.0.9.8
Reading symbols from /lib/i686/nosegneg/libpthread.so.0...done.
Loaded symbols for /lib/i686/nosegneg/libpthread.so.0
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /usr/lib/libkrb5support.so.0...done.
Loaded symbols for /usr/lib/libkrb5support.so.0
Reading symbols from /lib/libkeyutils.so.1...done.
Loaded symbols for /lib/libkeyutils.so.1
Reading symbols from 
/usr/lib/php5/20060613+lfs/ioncube_loader_lin_5.2.so...done.
Loaded symbols for /usr/lib/php5/20060613+lfs/ioncube_loader_lin_5.2.so
Reading symbols from /lib/i686/nosegneg/librt.so.1...done.
Loaded symbols for /lib/i686/nosegneg/librt.so.1
Reading symbols from /lib/i686/nosegneg/libnss_files.so.2...done.
Loaded symbols for /lib/i686/nosegneg/libnss_files.so.2
Failed to read a valid object file image from memory.
Core was generated by `php -l test.php'.
Program terminated with signal 11, Segmentation fault.
[New process 30231]
[New process 30230]
#0  0xb6bdfe60 in ?? ()
(gdb) bt
#0  0xb6bdfe60 in ?? ()
#1  0xb781f506 in start_thread () from /lib/i686/nosegneg/libpthread.so.0
#2  0xb7a685be in clone () from /lib/i686/nosegneg/libc.so.6
(gdb) quit


removing the mysql extension:

root@titanium:/tmp/bt# mkdir mysql && mv /etc/php5/conf.d/mysql* mysql
root@titanium:/tmp/bt# for f in `seq 0 100`; do php -l test.php; done | 
grep -c 'No syntax'
101

If I can provide any further info, just let me know.

Thanks,
Steve

Message #39 received at 493045@bugs.debian.org (full text, mbox, reply):

From: "Frank van Viegen" <frank@paiq.nl>

To: 493045@bugs.debian.org

Subject: backtrace

Date: Sat, 25 Oct 2008 17:13:40 +0200

Same problem here. 'php --help'  exits with a segfault about 30% of
the times. The problem only occurs when mysql or mysqli is loaded;
default settings are used. As the problem seems to be thread-related,
it may be relevant to know that this is on a dual processor box,
running Linux version 2.6.22-3-686 (Debian 2.6.22-6).

Package versions:

# dpkg -l | grep mysql | awk '{print $1" "$2" "$3}'
ii libdbd-mysql-perl 4.007-1
ii libdbd-mysql-ruby 0.2.2-1
ii libdbd-mysql-ruby1.8 0.2.2-1
ii libmysql-ruby1.8 2.7.4-1
ii libmysqlclient15off 5.0.51a-15
ii mysql-client-5.0 5.0.51a-15
ii mysql-common 5.0.51a-15
ii mysql-server 5.0.51a-15
ii mysql-server-5.0 5.0.51a-15
ii php5-mysql 5.2.6-5

And a backtrace. Quite a few missing symbols though. If required,
please hint me on how to resolve that.

# gdb `which php`
GNU gdb 6.8-debian
Copyright (C) 2008 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "i486-linux-gnu"...
(no debugging symbols found)
(gdb) r --help
Starting program: /usr/bin/php --help
(no debugging symbols found)
(no debugging symbols found)
[.....]
(no debugging symbols found)
[Thread debugging using libthread_db enabled]
(no debugging symbols found)
(no debugging symbols found)
(no debugging symbols found)

[New Thread 0xb77ef6d0 (LWP 15714)]
[New Thread 0xaec47b90 (LWP 15715)]
Usage: php [options] [-f] <file> [--] [args...]
       php [options] -r <code> [--] [args...]
       php [options] [-B <begin_code>] -R <code> [-E <end_code>] [--] [args...]
       php [options] [-B <begin_code>] -F <file> [-E <end_code>] [--] [args...]
       php [options] -- [args...]
       php [options] -a

  -a               Run interactively
  -c <path>|<file> Look for php.ini file in this directory
  -n               No php.ini file will be used
  -d foo[=bar]     Define INI entry foo with value 'bar'
  -e               Generate extended information for debugger/profiler
  -f <file>        Parse and execute <file>.
  -h               This help
  -i               PHP information
  -l               Syntax check only (lint)
  -m               Show compiled in modules
  -r <code>        Run PHP <code> without using script tags <?..?>
  -B <begin_code>  Run PHP <begin_code> before processing input lines
  -R <code>        Run PHP <code> for every input line
  -F <file>        Parse and execute <file> for every input line
  -E <end_code>    Run PHP <end_code> after processing all input lines
  -H               Hide any passed arguments from external tools.
  -s               Display colour syntax highlighted source.
  -v               Version number
  -w               Display source with stripped comments and whitespace.
  -z <file>        Load Zend extension <file>.

  args...          Arguments passed to script. Use -- args when first argument
                   starts with - or script is read from stdin

  --ini            Show configuration file names

  --rf <name>      Show information about function <name>.
  --rc <name>      Show information about class <name>.
  --re <name>      Show information about extension <name>.
  --ri <name>      Show configuration for extension <name>.

(no debugging symbols found)

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0xaec47b90 (LWP 15715)]
0xb77e93cf in ?? () from /lib/libgcc_s.so.1
(gdb) bt
#0  0xb77e93cf in ?? () from /lib/libgcc_s.so.1
#1  0xb77e9c63 in ?? () from /lib/libgcc_s.so.1
#2  0xb77e9d8b in _Unwind_ForcedUnwind () from /lib/libgcc_s.so.1
#3  0xb780aee6 in _Unwind_ForcedUnwind () from /lib/i686/cmov/libpthread.so.0
#4  0xb7808901 in __pthread_unwind () from /lib/i686/cmov/libpthread.so.0
#5  0xb7802a20 in pthread_exit () from /lib/i686/cmov/libpthread.so.0
#6  0xb7a57a74 in pthread_exit () from /lib/i686/cmov/libc.so.6
#7  0xb72a5e7e in ?? ()
#8  0x00000000 in ?? ()
(gdb)

Message #44 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Christian Weiske <cweiske@cweiske.de>

To: 493045@bugs.debian.org

Subject: same issue here

Date: Sat, 15 Nov 2008 22:11:32 +0100

[Message part 1 (text/plain, inline)]

Hello,


We are experiencing the same issue but in about 98% of all php runs on
cli.

$ php -v
PHP 5.2.6-5 with Suhosin-Patch 0.9.6.2 (cli) (built: Oct  5 2008
13:07:13) Copyright (c) 1997-2008 The PHP Group
Zend Engine v2.2.0, Copyright (c) 1998-2008 Zend Technologies
Speicherzugriffsfehler

The "Speicherzugriffsfehler" is the segfault. It happens whenever
mysql.ini or mysqli.ini is activated.


The backtrace with gdb is:

Could not open /proc/12203/status
(gdb) bt
#0  0xb7894771 in __nptl_create_event () from /lib/libpthread.so.0
#1  0xb7895b7b in pthread_create@@GLIBC_2.1 () from /lib/libpthread.so.0
#2  0xb6e84e54 in my_thread_global_init ()
from /usr/lib/libmysqlclient_r.so.15 #3  0xb6e7fcd1 in my_init ()
from /usr/lib/libmysqlclient_r.so.15 #4  0xb6e7e73b in
mysql_server_init () from /usr/lib/libmysqlclient_r.so.15 #5
0xb702efe4 in zm_startup_mysql ()
from /usr/lib/php5/20060613+lfs/mysql.so #6  0x082d8c1e in
zend_startup_module_ex () #7  0x082ddbfd in zend_hash_apply () #8
0x082d7441 in zend_startup_modules () #9  0x0828a410 in
php_module_startup () #10 0x083547bd in php_signal ()
#11 0x08355017 in main ()


So it seems indeed to be coming from libmysqlclient, here at
version 5.0.51a-17.

-- 
Regards/Mit freundlichen Grüßen
Christian Weiske

-= Geeking around in the name of science since 1982 =-

[signature.asc (application/pgp-signature, attachment)]

Message #53 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Berenyi Mihaly <huf@parawag.net>

To: 493045@bugs.debian.org

Subject: same issue here too

Date: Fri, 9 Jan 2009 18:27:52 +0100

hi,
we're experiencing this same issue on a debian lenny with kernel
2.6.18-5-vserver-686.
the machine is x86 32bit 
enabling either mysql or mysqli (i tried disabling/enabling other
modules but they seem to have zero bearing on this issue) 
causes php to segfault quite often.
i tried getting a coredump and looking at the backtrace, but got the
same info as message #34 in this thread.

installed versions are:
php5-cli 5.2.6.dfsg.1-0.1~lenny1
php5-common 5.2.6.dfsg.1-0.1~lenny1
php5-dbg 5.2.6.dfsg.1-0.1~lenny1
php5-mysql 5.2.6.dfsg.1-0.1~lenny1
libmysqlclient15off 5.0.51a-20

hope this helps, 
	mihaly
-- 
huf (Berenyi Mihaly) - huf@keszen.hu - http://parawag.net
..........it's life jim, but not as we know it...........

Message #58 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Mezei Áron <aron.mezei@nws.hu>

To: 493045@bugs.debian.org

Subject: same issue at php 5.2.6.dfsg.1-1+lenny2 and mysql 5.0.51a-24

Date: Wed, 11 Feb 2009 21:23:49 +0100

[Message part 1 (text/plain, inline)]

Hi

I'm experiencing the same issue as #53, and upgrading mysql and it's 
libs from 5.0.51a-21 to 5.0.51a-24 didn't help (php packages' version: 
5.2.6.dfsg.1-1+lenny2).

Regards,

Aron MEZEI

[Message part 2 (text/html, inline)]

Message #63 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Martin Koegler <mkoegler@auto.tuwien.ac.at>, 515143@bugs.debian.org

Cc: 513204@bugs.debian.org, 493045@bugs.debian.org, 513204-submitter@bugs.debian.org, 493045-submitter@bugs.debian.org

Subject: Re: Bug#515143: Race condition in libmysqlclient15_r

Date: Sat, 14 Feb 2009 00:50:49 +0100

[Message part 1 (text/plain, inline)]

hi martin,

thanks for digging this up!  these crashes have been really difficult to
investigate on the php-side of things.  

we are so ridiculously close to releasing that i don't think it's possible
to get a fix for this into lenny before that happens, but something like
this should definitely be fixed in lenny-proposed-updates and propagated
into the next release update.

i'm cc'ing the other bugs that you've referenced, which i'll keep seperate
for now until you can confirm the problem and fix.


	sean

On Fri, Feb 13, 2009 at 11:11:17PM +0100, Martin Koegler wrote:
> Package: libmysqlclient15off
> Version: 5.0.51a-24
> 
> There is a race condition in libmysqlclient15off:
> http://bugs.mysql.com/bug.php?id=42850
> 
> On one vmware system, I can trigger this race condition with
> php5-cli+php5-mysql.
> 
> Message 10 of #513204 and message 53 of #493045 look similar to resulting
> segfault on my system.
> 
> The easiest solution would be to drop the workaround, which introduced
> this race condition:
> http://lists.mysql.com/commits/16642
> 
> I'm currently testing, if it is still needed for libc6 2.7-18.
> 
> mfg Martin Kögler
> 
> 
> 

-- 

[signature.asc (application/pgp-signature, inline)]

Message #71 received at 493045@bugs.debian.org (full text, mbox, reply):

From: mkoegler@auto.tuwien.ac.at (Martin Koegler)

To: 515143@bugs.debian.org

Cc: 513204@bugs.debian.org, 493045@bugs.debian.org, 513204-submitter@bugs.debian.org, 493045-submitter@bugs.debian.org, sean finney <seanius@debian.org>

Subject: Bug#515143: [PATCH] remove glibc workaround

Date: Sat, 14 Feb 2009 16:11:11 +0100

[Message part 1 (text/plain, inline)]

The attaced patch removes the startup of the dummy thread.

With this patch, I can't trigger a segfault in php-cli anymore.
strace confirms, that that php does not spawn a thread any more.
The patch passes the mysql server test suite in the build.

So I assume, that the workround is not neccesary in lenny any more.

mfg Martin Kögler

[mysql-race.patch (text/x-diff, attachment)]

Message #79 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Norbert Tretkowski <norbert@tretkowski.de>

To: Martin Koegler <mkoegler@auto.tuwien.ac.at>, 515143@bugs.debian.org

Cc: 513204@bugs.debian.org, 493045-submitter@bugs.debian.org, 513204-submitter@bugs.debian.org, 493045@bugs.debian.org

Subject: Re: [debian-mysql] Bug#515143: [PATCH] remove glibc workaround

Date: Sun, 15 Feb 2009 12:58:44 +0100

Am Samstag, den 14.02.2009, 16:11 +0100 schrieb Martin Koegler:
> The attaced patch removes the startup of the dummy thread.
> 
> With this patch, I can't trigger a segfault in php-cli anymore.
> strace confirms, that that php does not spawn a thread any more.
> The patch passes the mysql server test suite in the build.
> 
> So I assume, that the workround is not neccesary in lenny any more.

Thanks, I'll discuss this with the MySQL developers, and put it on my
TODO list for 5.0r1.

	Norbert

Message #87 received at 493045@bugs.debian.org (full text, mbox, reply):

From: pbrufal <pbrufal@servitux.es>

To: Debian Bug Tracking System <493045@bugs.debian.org>

Subject: php5-cli: temporary solution for run php5 without segfaults

Date: Mon, 09 Mar 2009 18:34:22 +0100

Package: php5-cli
Version: 5.2.6.dfsg.1-1+lenny2
Followup-For: Bug #493045


        This is a temporary solution

1) Install php5-dev, libmysqlclient15-dev and the source of php5 (apt-get source php5)
2) cd php5-5.2.6.dfsg.1/ext/mysql/
3) run "phpize" in this directory
4) ./configure && make && make install

        Now you can use php in the command line without segfaults.

-- System Information:
Debian Release: 5.0
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 2.6.22.6 (SMP w/1 CPU core)
Locale: LANG=es_ES@euro, LC_CTYPE=es_ES@euro (charmap=ISO-8859-15)
Shell: /bin/sh linked to /bin/bash

Versions of packages php5-cli depends on:
ii  libbz2-1.0         1.0.5-1               high-quality block-sorting file co
ii  libc6              2.7-18                GNU C Library: Shared libraries
ii  libcomerr2         1.41.3-1              common error description library
ii  libdb4.6           4.6.21-11             Berkeley v4.6 Database Libraries [
ii  libkrb53           1.6.dfsg.4~beta1-5    MIT Kerberos runtime libraries
ii  libmagic1          4.26-1                File type determination library us
ii  libncurses5        5.7+20081213-1        shared libraries for terminal hand
ii  libpcre3           7.6-2.1               Perl 5 Compatible Regular Expressi
ii  libssl0.9.8        0.9.8g-15             SSL shared libraries
ii  libxml2            2.6.32.dfsg-5         GNOME XML library
ii  mime-support       3.44-1                MIME files 'mime.types' & 'mailcap
ii  php5-common        5.2.6.dfsg.1-1+lenny2 Common files for packages built fr
ii  tzdata             2008h-2               time zone and daylight-saving time
ii  ucf                3.0016                Update Configuration File: preserv
ii  zlib1g             1:1.2.3.3.dfsg-12     compression library - runtime

php5-cli recommends no packages.

Versions of packages php5-cli suggests:
ii  php-pear           5.2.6.dfsg.1-1+lenny2 PEAR - PHP Extension and Applicati

-- no debconf information

Message #92 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Raphael Mazelier <raph@b3g.fr>

To: 493045@bugs.debian.org

Subject: Segmentation Fault in cron console PHP Job

Date: Fri, 20 Mar 2009 18:15:23 +0100

Package: php5-cli
Version: 5.2.6.dfsg.1-1+lenny2
Followup-For: Bug #493045


1) Install php5-dev, libmysqlclient15-dev and the source of php5 (apt-get source php5)
2) cd php5-5.2.6.dfsg.1/ext/mysql/
3) run "phpize" in this directory
4) ./configure&&  make&&  make install


I try the above procedure with no luck  I still have some random segfault :
for example :  segfault at b71bee90 ip b71bee90 sp b71433ac error 14 in 
librt-2.7.so[b76e9000+7000]
with lenny kernel  2.6.26-1-686-bigmem #1 SMP

Two things to notice :
- this server use very intensively php-cli mysql scripts.
- when I revert to an old etch kernel (2.6.18-6-686-bigmem #1 SMP) the 
php scripts run just fine.

So I'm convinced that's the problem may be related to a kernel change.

-- 

 Raphael Mazelier
 Tel: +33 175 00 85 23
 Direction Technique - Responsable Systemes
 B3G Telecom - 15-17 rue Auber 75009 Paris
 

Message #97 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Norbert Tretkowski <norbert@tretkowski.de>, 515143@bugs.debian.org

Cc: 513204@bugs.debian.org, 493045@bugs.debian.org, 524366@bugs.debian.org

Subject: Re: [debian-mysql] Bug#515143: Bug#515143: [PATCH] remove glibc workaround

Date: Thu, 16 Apr 2009 23:17:19 +0200

[Message part 1 (text/plain, inline)]

hi norbert,

On Sun, Feb 15, 2009 at 12:58:44PM +0100, Norbert Tretkowski wrote:
> Am Samstag, den 14.02.2009, 16:11 +0100 schrieb Martin Koegler:
> > The attaced patch removes the startup of the dummy thread.
> > 
> > With this patch, I can't trigger a segfault in php-cli anymore.
> > strace confirms, that that php does not spawn a thread any more.
> > The patch passes the mysql server test suite in the build.
> > 
> > So I assume, that the workround is not neccesary in lenny any more.
> 
> Thanks, I'll discuss this with the MySQL developers, and put it on my
> TODO list for 5.0r1.

ping :)

the users are getting restless (and sarcastic) in php land...  do you
need help preparing something for s-p-u?


	sean

[signature.asc (application/pgp-signature, inline)]

Message #102 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Norbert Tretkowski <norbert@tretkowski.de>

To: sean finney <seanius@debian.org>

Cc: 515143@bugs.debian.org, 513204@bugs.debian.org, 493045@bugs.debian.org, 524366@bugs.debian.org

Subject: Re: [debian-mysql] Bug#515143: Bug#515143: [PATCH] remove glibc workaround

Date: Fri, 17 Apr 2009 10:49:21 +0200

Am Donnerstag, den 16.04.2009, 23:17 +0200 schrieb sean finney:
> the users are getting restless (and sarcastic) in php land...

I don't care about sarcasm in php land...

> do you need help preparing something for s-p-u?

Yes. I'm currently the only person working on the MySQL packages in
Debian, and could really need some help.

	Norbert

Message #107 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Michael Bonfils <murlock42@gmail.com>, 513204@bugs.debian.org, Raphael Geissert <atomo64+debian@gmail.com>, Todd Courtnage <todd@cambrianhouse.com>, Martin Koegler <mkoegler@auto.tuwien.ac.at>, 515143@bugs.debian.org, Norbert Tretkowski <norbert@tretkowski.de>, "Raoul Bhatia [IPAX]" <r.bhatia@ipax.at>, Lawrence Ong <long.debian@gmail.com>, Simon Waters <simonw@zynet.net>

Cc: 493045-submitter@bugs.debian.org, 513204-submitter@bugs.debian.org, 493045@bugs.debian.org, 524366@bugs.debian.org, control@bugs.debian.org, Debian PHP Maintainers <pkg-php-maint@lists.alioth.debian.org>

Subject: request for testing fix for #513204

Date: Sat, 18 Apr 2009 17:55:31 +0200

[Message part 1 (text/plain, inline)]

tags 513204 moreinfo
thanks

hi everyone,

after speaking with the mysql maintainer i've prepared a proposed update
for the stable/lenny version of mysql containing the fix from Martin Koegler.
for those of you who are affected, can you please try the following in
your sources.list:

deb http://people.debian.org/~seanius/mysql/513204/amd64 ./

or

deb http://people.debian.org/~seanius/mysql/513204/i386 ./

depending on your architecture, and apt-get update/upgrade.
the new version of libmysqlclient15off containing the fix should be
5.0.51a-24lenny1.  you probably don't need any more than the updated
libmysqlclient15off package (so you could also just manually download
and install it).

please report back with success (or failure) reports, and if it sounds
like this fixes the problems with no ill effects then i'll send this
to stable-proposed-updates and it will be hopefully included in the
next stable point release.


	sean

-- 

[signature.asc (application/pgp-signature, inline)]

Message #115 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Michael Bonfils <murlock42@gmail.com>, 513204@bugs.debian.org, Raphael Geissert <atomo64+debian@gmail.com>, Todd Courtnage <todd@cambrianhouse.com>, Martin Koegler <mkoegler@auto.tuwien.ac.at>, 515143@bugs.debian.org, Norbert Tretkowski <norbert@tretkowski.de>, "Raoul Bhatia [IPAX]" <r.bhatia@ipax.at>, Lawrence Ong <long.debian@gmail.com>, Simon Waters <simonw@zynet.net>

Cc: 493045-submitter@bugs.debian.org, 513204-submitter@bugs.debian.org, 493045@bugs.debian.org, 524366@bugs.debian.org

Subject: Re: request for testing fix for #513204

Date: Sat, 18 Apr 2009 18:22:35 +0200

[Message part 1 (text/plain, inline)]

sorry for the mail flood..

On Sat, Apr 18, 2009 at 05:55:31PM +0200, sean finney wrote:
> depending on your architecture, and apt-get update/upgrade.
> the new version of libmysqlclient15off containing the fix should be
> 5.0.51a-24lenny1.  you probably don't need any more than the updated
> libmysqlclient15off package (so you could also just manually download
> and install it).

oh, and maybe it goes without saying but just to be totally clear: you'll
also need to stop / start apache2 as well.


	sean

[signature.asc (application/pgp-signature, inline)]

Message #123 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Gurr <gurr@mellesleg.hu>

To: 493045@bugs.debian.org

Subject: Segmentation Fault in cron console PHP Job

Date: Thu, 14 May 2009 17:19:16 +0200

Hi!

I don't know how others stand with this, but i still experience
segfaults. I have two computers running debian lenny, with
apache+mysql+php5 installed from package. My cron php's are random
segfaulting, altough the mysql stuffs have been upgraded:

anubis:~# dpkg -l | grep mysql | awk '{print $2 " " $3}'
libdbd-mysql-perl 4.007-1
libmysqlclient15off 5.0.51a-24+lenny1
mysql-client-5.0 5.0.51a-24+lenny1
mysql-common 5.0.51a-24+lenny1
mysql-server-5.0 5.0.51a-24+lenny1
php5-mysql 5.2.6.dfsg.1-1+lenny3
zabbix-server-mysql 1:1.4.6-1
anubis:~#

One of the computers is a dual core s775 intel, the other is a older
s468 intel P4 (single core, hyperthreading enabled).

Just drop me a letter if you need more info.

-- 
Best regards:
 Zsolt Gurmai                          mailto:gurr@mellesleg.hu

Message #128 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Patrick Li <patrick@papaq.org>

To: 493045@bugs.debian.org

Subject: Re: request for testing fix for #513204

Date: Wed, 17 Jun 2009 15:06:52 +1200

Hi Sean,

I grabbed the latest svn code from
svn://svn.debian.org/pkg-mysql/mysql-dfsg-5.0/branches/lenny-proposed/debian
and complied against mysql-dfsg-5.0_5.0.51a.orig.tar.gz.

I need to revert 2 hunks from
patches/65_fix_gis_functions_crash.dpatch in order for it to patch
cleanly.

I cannot reproduce the segfaults on the new libmysqlclient15off. I was
able to repo this quite easily before the upgrade.

So I can confirm this is fixed.

Patrick Li

Message #133 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Patrick Li <patrick@papaq.org>, 493045@bugs.debian.org, Norbert Tretkowski <nobse@debian.org>

Subject: Re: [php-maint] Bug#493045: request for testing fix for #513204

Date: Wed, 17 Jun 2009 10:58:21 +0200

[Message part 1 (text/plain, inline)]

hi patrick,

On Wed, Jun 17, 2009 at 03:06:52PM +1200, Patrick Li wrote:
> I grabbed the latest svn code from
> svn://svn.debian.org/pkg-mysql/mysql-dfsg-5.0/branches/lenny-proposed/debian
> and complied against mysql-dfsg-5.0_5.0.51a.orig.tar.gz.
> 
> I need to revert 2 hunks from
> patches/65_fix_gis_functions_crash.dpatch in order for it to patch
> cleanly.

hrm... it looks like this backported patch doesn't quite work.  i took
a quick look and it seems that most (not all) of the changes are already
present in this version.  if i try to do a reverse patch only a few
hunks don't apply, all of which are in the test directory (see below).
i'm not familiar with this particular problem fix... norbert could you
check to see if this patch is actually needed?

at this point i've heard back from a number of folks (both publically and
privately), so i feel a bit more comfortable about pushing something towards
s-p-u so i can prepare an upload as soon as we clear the air on this issue.


	sean

(reverse patch output)

patch -R --dry-run -p1 < debian/patches/65_fix_gis_functions_crash.dpatch                :)
patching file mysql-test/r/gis-rtree.result
Hunk #4 FAILED at 1426.
1 out of 4 hunks FAILED -- saving rejects to file mysql-test/r/gis-rtree.result.rej
patching file mysql-test/r/gis.result
patching file mysql-test/t/gis-rtree.test
Hunk #4 FAILED at 802.
1 out of 4 hunks FAILED -- saving rejects to file mysql-test/t/gis-rtree.test.rej
patching file mysql-test/t/gis.test
patching file sql/item_geofunc.cc
Hunk #1 succeeded at 74 (offset 4 lines).
Hunk #2 succeeded at 87 (offset 4 lines).
Hunk #3 succeeded at 341 (offset 4 lines).
Hunk #4 succeeded at 372 (offset 4 lines).
Hunk #5 succeeded at 395 (offset 4 lines).
Hunk #6 succeeded at 410 (offset 4 lines).

[signature.asc (application/pgp-signature, inline)]

Message #138 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Gurr <gurr@mellesleg.hu>

To: 493045@bugs.debian.org

Subject: the segfaults are back, and are at large

Date: Thu, 1 Oct 2009 12:35:09 +0200

A short piece of log:

Oct  1 06:26:01 anubis /USR/SBIN/CRON[3818]: (root) CMD (/root/minuter)
Oct  1 06:27:01 anubis /USR/SBIN/CRON[3920]: (root) CMD (/root/minuter)
Oct  1 06:27:01 anubis kernel: php[3925]: segfault at 7f883c743edb ip 00007f883c2ba1af sp 0000000042341c90 error 4 in libgcc_s.so.1[7f883c2aa000+16000]
Oct  1 06:27:01 anubis kernel: php[3944]: segfault at 7fdec52f0edb ip 00007fdec50841af sp 00000000425e8c90 error 4 in libgcc_s.so.1[7fdec5074000+16000]
Oct  1 06:28:01 anubis /USR/SBIN/CRON[3974]: (root) CMD (/root/minuter)
Oct  1 06:28:01 anubis kernel: php[3979]: segfault at 7f0a7f33cedb ip 00007f0a7f6eb1af sp 0000000040e3fc90 error 4 in libgcc_s.so.1[7f0a7f6db000+16000]
Oct  1 06:28:01 anubis kernel: php[3998]: segfault at 7f5e80d10ed0 ip 00007f5e80d10ed0 sp 0000000041382128 error 14 in libtasn1.so.3.0.15[7f5e8372e000+f000]
Oct  1 06:29:01 anubis /USR/SBIN/CRON[4027]: (root) CMD (/root/minuter)
Oct  1 06:29:01 anubis kernel: php[4051]: segfault at 7fcbbf5caed0 ip 00007fcbbf5caed0 sp 00000000422af128 error 14 in libtasn1.so.3.0.15[7fcbc1fe8000+f000]
Oct  1 06:30:00 anubis named[7902]: too many timeouts resolving 'wonkavator.glbdns.microsoft.com/A' (in 'glbdns.microsoft.com'?): reducing the advertised EDNS
Oct  1 06:30:01 anubis /USR/SBIN/CRON[4080]: (root) CMD (/root/ipquotaupd)
Oct  1 06:30:01 anubis /USR/SBIN/CRON[4084]: (root) CMD (/root/minuter)
Oct  1 06:31:01 anubis /USR/SBIN/CRON[4153]: (root) CMD (/root/minuter)
Oct  1 06:31:01 anubis kernel: php[4158]: segfault at 7f3c238d1ed0 ip 00007f3c238d1ed0 sp 0000000041192128 error 14 in librt-2.7.so[7f3c26c32000+8000]
Oct  1 06:32:01 anubis /USR/SBIN/CRON[4206]: (root) CMD (/root/minuter)
Oct  1 06:32:01 anubis kernel: php[4229]: segfault at 7ffe51968ed0 ip 00007ffe51968ed0 sp 00000000423e9128 error 14 in libtasn1.so.3.0.15[7ffe54386000+f000]
Oct  1 06:33:01 anubis /USR/SBIN/CRON[4695]: (root) CMD (/root/minuter)
Oct  1 06:33:01 anubis kernel: php[4700]: segfault at 7ffbe2ca8ed0 ip 00007ffbe2ca8ed0 sp 000000004213b128 error 14 in librt-2.7.so[7ffbe6009000+8000]
Oct  1 06:33:01 anubis kernel: php[4719]: segfault at 7f11b9572ed0 ip 00007f11b9572ed0 sp 00000000411a7128 error 14 in librt-2.7.so[7f11bc8d3000+8000]
Oct  1 06:34:01 anubis /USR/SBIN/CRON[4748]: (root) CMD (/root/minuter)
Oct  1 06:35:01 anubis /USR/SBIN/CRON[4797]: (root) CMD (/root/ipquotaupd)
Oct  1 06:35:01 anubis /USR/SBIN/CRON[4800]: (root) CMD (/root/minuter)
Oct  1 06:35:01 anubis /USR/SBIN/CRON[4803]: (root) CMD ([ -x /usr/lib/sysstat/sa1 ] && { [ -r "$DEFAULT" ] && . "$DEFAULT" ; [ "$ENABLED" = "true" ] && exec
Oct  1 06:35:01 anubis kernel: php[4834]: segfault at 7f69c1155ed0 ip 00007f69c1155ed0 sp 00000000409e5128 error 14 in librt-2.7.so[7f69c44b6000+8000]

The problem is as i wrote earlier; a simple php script parsing the
output of iptables dumping numbers to mysql database, nothing fancy,
did work ok for long time.

Packages (updated yesterday):

anubis:/var/log# dpkg -l | grep mysql | awk '{print $2 " " $3}'
libdbd-mysql-perl 4.007-1
libmysqlclient15off 5.0.51a-24+lenny2
mysql-client-5.0 5.0.51a-24+lenny2
mysql-common 5.0.51a-24+lenny2
mysql-server-5.0 5.0.51a-24+lenny2
php5-mysql 5.2.6.dfsg.1-1+lenny3
zabbix-server-mysql 1:1.4.6-1

This issue is getting serious to me, could you guys please check out
whats wrong? Thank you!

-- 
Best regards:
 Zsolt Gurmai                          mailto:gurr@mellesleg.hu

Message #143 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Gurr <gurr@mellesleg.hu>, 493045@bugs.debian.org

Subject: Re: [php-maint] Bug#493045: the segfaults are back, and are at large

Date: Thu, 1 Oct 2009 14:36:56 +0200

[Message part 1 (text/plain, inline)]

On Thu, Oct 01, 2009 at 12:35:09PM +0200, Gurr wrote:
> anubis:/var/log# dpkg -l | grep mysql | awk '{print $2 " " $3}'
> libdbd-mysql-perl 4.007-1
> libmysqlclient15off 5.0.51a-24+lenny2
> mysql-client-5.0 5.0.51a-24+lenny2
> mysql-common 5.0.51a-24+lenny2
> mysql-server-5.0 5.0.51a-24+lenny2
> php5-mysql 5.2.6.dfsg.1-1+lenny3
> zabbix-server-mysql 1:1.4.6-1
> 
> This issue is getting serious to me, could you guys please check out
> whats wrong? Thank you!

have you tried the stable-proposed-updates version of the
libmysqlclient15off?  it seems this bug wasn't merged with the other
ones so i guess this bug didn't get the follow up messages :(

see #524366 or #513204, one of which should have the sources.list entry
to use.


	sean

[signature.asc (application/pgp-signature, inline)]

Message #148 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Gurr <gurr@mellesleg.hu>

To: sean finney <seanius@debian.org>

Cc: 493045@bugs.debian.org

Subject: Re[2]: [php-maint] Bug#493045: the segfaults are back, and are at large

Date: Fri, 2 Oct 2009 10:57:56 +0200

Hi!

I think i found the link you mentioned, altough i dld the package
manually. Upon install i got this:

dpkg - warning: downgrading libmysqlclient15off from 5.0.51a-24+lenny2 to 5.0.51a-24lenny1.

This means i guess, that i had more recent version of this package,
rigth? Is it possible, that mine was newer (from main debian repo),
but it does not contain the fix? Anyway i let it install (only this
single package), will see if it solves my problem.

Will report back. Thank you for your help!


Thursday, October 1, 2009, 2:36:56 PM, you wrote:

> have you tried the stable-proposed-updates version of the
> libmysqlclient15off?  it seems this bug wasn't merged with the other
> ones so i guess this bug didn't get the follow up messages :(

> see #524366 or #513204, one of which should have the sources.list entry
> to use.

>         sean

-- 
Best regards:
 Zsolt Gurmai                            mailto:gurr@mellesleg.hu

Message #153 received at 493045@bugs.debian.org (full text, mbox, reply):

From: sean finney <seanius@debian.org>

To: Gurr <gurr@mellesleg.hu>

Cc: 493045@bugs.debian.org

Subject: Re: [php-maint] Bug#493045: the segfaults are back, and are at large

Date: Fri, 2 Oct 2009 12:56:40 +0200

[Message part 1 (text/plain, inline)]

hi gurr,

On Fri, Oct 02, 2009 at 10:57:56AM +0200, Gurr wrote:
> This means i guess, that i had more recent version of this package,
> rigth? Is it possible, that mine was newer (from main debian repo),
> but it does not contain the fix? Anyway i let it install (only this
> single package), will see if it solves my problem.
> 
> Will report back. Thank you for your help!

look bit deeper into the bug log, there should be an updated package (the
version string will contain "spu" and will be strictly greater than what's
on your system) referenced with a sample sources.list.  the version you
downloaded should also contain the fix, but will be missing some important
security updates.


	sean

[signature.asc (application/pgp-signature, inline)]

Message #158 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Ondřej Surý <ondrej@debian.org>

To: Gurr <gurr@mellesleg.hu>, 493045@bugs.debian.org

Cc: sean finney <seanius@debian.org>

Subject: Re: [php-maint] Bug#493045: Re[2]: Bug#493045: the segfaults are back, and are at large

Date: Fri, 2 Oct 2009 12:51:30 +0200

Yep, lenny had two security updates meanwhile (compare changelogs).

Ondrej

On Fri, Oct 2, 2009 at 10:57, Gurr <gurr@mellesleg.hu> wrote:
>
> Hi!
>
> I think i found the link you mentioned, altough i dld the package
> manually. Upon install i got this:
>
> dpkg - warning: downgrading libmysqlclient15off from 5.0.51a-24+lenny2 to 5.0.51a-24lenny1.
>
> This means i guess, that i had more recent version of this package,
> rigth? Is it possible, that mine was newer (from main debian repo),
> but it does not contain the fix? Anyway i let it install (only this
> single package), will see if it solves my problem.
>
> Will report back. Thank you for your help!
>
>
> Thursday, October 1, 2009, 2:36:56 PM, you wrote:
>
>> have you tried the stable-proposed-updates version of the
>> libmysqlclient15off?  it seems this bug wasn't merged with the other
>> ones so i guess this bug didn't get the follow up messages :(
>
>> see #524366 or #513204, one of which should have the sources.list entry
>> to use.
>
>>         sean
>
> --
> Best regards:
>  Zsolt Gurmai                            mailto:gurr@mellesleg.hu
>
>
>
>
> _______________________________________________
> pkg-php-maint mailing list
> pkg-php-maint@lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/pkg-php-maint
>



-- 
Ondřej Surý <ondrej@sury.org>
http://blog.rfc1925.org/

Message #163 received at 493045@bugs.debian.org (full text, mbox, reply):

From: Gurr <gurr@mellesleg.hu>

To: sean finney <seanius@debian.org>

Cc: 493045@bugs.debian.org

Subject: Re[2]: [php-maint] Bug#493045: the segfaults are back, and are at large

Date: Fri, 2 Oct 2009 13:53:28 +0200

Thank you for both replies! You were right, when i checked
the log i "bailed out" upon finding the first source list :). I found
now the "spu" one. It required mysql-common aside from
libmysqlclient15off to be updated (again i choose to
download/install it manually). The older package seemed to solve my
problem (only checked 2-3 hours of log), so i expect, that the "spu"
one will work just the same, so i will not report back unless there is
some problem.

Thank you very much again for your help!

Friday, October 2, 2009, 12:56:40 PM, you wrote:

> hi gurr,

> look bit deeper into the bug log, there should be an updated package (the
> version string will contain "spu" and will be strictly greater than what's
> on your system) referenced with a sample sources.list.  the version you
> downloaded should also contain the fix, but will be missing some important
> security updates.

>         sean

-- 
Best regards:
 Zsolt Gurmai                            mailto:gurr@mellesleg.hu

Message #168 received at 493045@bugs.debian.org (full text, mbox, reply):

From: db <db.pub.mail@gmail.com>

To: 493045@bugs.debian.org

Subject: This bug still occurs for me :)

Date: Sat, 24 Oct 2009 18:22:54 +1100

This bug still occurs for me :)
Doing  echo "test" > test.php
and then
"
for f in `seq 0 100`; do  php test.php; done
"

While i have hyper threading on (this is a pentium 4 box), i get get
Segmentation fault every now and again.

If i do  echo 0 > /sys/devices/system/cpu/cpu1/online ...the
Segmentation fault  do not seem to occur.


____________________________
the above was done when the system was underload (i was compiling a
kernel at the same time).
____________________________
The fix that was attached in this bug report seems to have been
applied already... (unless i there is another one i am missing).
"  * New patch 64_fix-dummy-thread-race-condition.dpatch to back out
an" I believe is the exact patch....


____________________________
php -v
PHP 5.2.6-1+lenny3 with Suhosin-Patch 0.9.6.2 (cli) (built: Apr 26
2009 20:09:03)
Copyright (c) 1997-2008 The PHP Group
Zend Engine v2.2.0, Copyright (c) 1998-2008 Zend Technologies
   with Suhosin v0.9.27, Copyright (c) 2007, by SektionEins GmbH


dpkg -l php5-cli
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Inst/Cfg-files/Unpacked/Failed-cfg/Half-inst/trig-aWait/Trig-pend
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
ii  php5-cli       5.2.6.dfsg.1-1 command-line interpreter for the php5 script



uname -a
Linux X 2.6.31.3 #1 SMP Thu Oct 8 23:53:00 EST 2009 x86_64 GNU/Linux

Message #173 received at 493045-done@bugs.debian.org (full text, mbox, reply):

From: Ondřej Surý <ondrej@sury.org>

To: 465081-done@bugs.debian.org, 537794-done@bugs.debian.org, 553048-done@bugs.debian.org, 574610-done@bugs.debian.org, 584885-done@bugs.debian.org, 584957-done@bugs.debian.org, 594613-done@bugs.debian.org, 493045-done@bugs.debian.org, 549492-done@bugs.debian.org, 450581-done@bugs.debian.org, 502174-done@bugs.debian.org, 543177-done@bugs.debian.org, 547134-done@bugs.debian.org, 552089-done@bugs.debian.org, 556523-done@bugs.debian.org, 559273-done@bugs.debian.org, 576147-done@bugs.debian.org, 578754-done@bugs.debian.org, 601602-done@bugs.debian.org, 609355-done@bugs.debian.org, 419649-done@bugs.debian.org, 442063-done@bugs.debian.org, 500567-done@bugs.debian.org, 513429-done@bugs.debian.org, 528600-done@bugs.debian.org, 597650-done@bugs.debian.org, 603641-done@bugs.debian.org, 405067-done@bugs.debian.org, 430397-done@bugs.debian.org, 440775-done@bugs.debian.org, 591759-done@bugs.debian.org, 565387-done@bugs.debian.org, 507762-done@bugs.debian.org, 529278-done@bugs.debian.org, 556459-done@bugs.debian.org

Subject: Closing segfaults (and some other minor bugs) for version older than squeeze (5.3.3)

Date: Wed, 27 Apr 2011 10:28:24 +0200

Version: 5.3.3-7

Hi,

since lenny is oldstable it will not get any updates now (except
security)[1], I am closing all segfault bugs filled against php5 in
lenny. (This is kind of saying that we don't care much about php5 in
lenny anymore).

If you believe the bug is still there, please provide evidence[2] and
a (preferably complete) test case with up-to-date squeeze (and/or
testing or unstable) version of php5 and reopen the bug.

O.
1. http://wiki.debian.org/PHP#Notes_on_PHP_and_security
2. Install php5-dbg and provide backtrace:
http://bugs.php.net/bugs-generating-backtrace.php
-- 
Ondřej Surý <ondrej@sury.org>

Send a report that this bug log contains spam.