Segmentation faults since last update to php 5.2.17

[CrK01]

Hi all,

last night I updated a shared environment hosting ( apache, php,etc ). Since this update I got lots of problems with wordpress pages.

If I set the apache module to process PHP, pages won't load ( blank page , and no errors on error_log ) only segmentation faults.

So, the only way to work is put the PHP execution on CGI ( not Fast CGI ). It loads pages, but I prefer Apache module for better performance.

Here I put my php modules and version:

php -v
PHP 5.2.17 (cli) (built: Jan 7 2011 10:39:29)
Copyright (c) 1997-2010 The PHP Group
Zend Engine v2.2.0, Copyright (c) 1998-2010 Zend Technologies
with the ionCube PHP Loader v4.0.5, Copyright (c) 2002-2011, by ionCube Ltd., and
with Zend Optimizer v3.3.9, Copyright (c) 1998-2009, by Zend Technologies
with Suhosin v0.9.32.1, Copyright (c) 2007-2010, by SektionEins GmbH

/usr/sbin/httpd -v
Server version: Apache/2.2.3
Server built: Aug 30 2010 12:28:40

/usr/sbin/httpd -l
Compiled in modules:
core.c
prefork.c
http_core.c
mod_so.c

uname -a
Linux nls.es 2.6.34.6-xxxx-grs-ipv6-64 #3 SMP Fri Sep 17 16:06:38 UTC 2010 x86_64 x86_64 x86_64 GNU/Linux

It's a CentOS 5.5 with atomic repo.

I have lots of segaults with CGI anyway:

php-cgi[13670]: segfault at 3bc8 ip 00000000005adbcd sp 00007fff23833990 error 4 in php-cgi[400000+2bb000]
grsec: From 67.195.112.99: Segmentation fault occurred at 0000000000003bc8 in /usr/bin/php-cgi[php-cgi:13670] uid/euid:10001/10001 gid/egid:2522/2522, parent /usr/sbin/httpd[httpd:13666] uid/euid:48/48 gid/egid:48/48
php-cgi[13690]: segfault at 29 ip 00000000005adbae sp 00007fff996ec900 error 4 in php-cgi[400000+2bb000]
grsec: From 109.230.221.62: Segmentation fault occurred at 0000000000000029 in /usr/bin/php-cgi[php-cgi:13690] uid/euid:10001/10001 gid/egid:2522/2522, parent /usr/sbin/httpd[httpd:13558] uid/euid:48/48 gid/egid:48/48
php-cgi[13729]: segfault at 29 ip 00000000005adbae sp 00007fff4ff74900 error 4 in php-cgi[400000+2bb000]
grsec: From 109.230.221.62: Segmentation fault occurred at 0000000000000029 in /usr/bin/php-cgi[php-cgi:13729] uid/euid:10001/10001 gid/egid:2522/2522, parent /usr/sbin/httpd[httpd:13570] uid/euid:48/48 gid/egid:48/48

Mar 15 16:46:02 nls kernel: php-cgi[13926]: segfault at 29 ip 00000000005adbae sp 00007fffedf30900 error 4 in php-cgi[400000+2bb000]
Mar 15 16:46:02 nls kernel: grsec: From 109.230.221.62: Segmentation fault occurred at 0000000000000029 in /usr/bin/php-cgi[php-cgi:13926] uid/euid:10001/10001 gid/egid:2522/2522, parent /usr/sbin/httpd[httpd:13380] uid/euid:48/48 gid/egid:48/48
Mar 15 16:46:03 nls suhosin[13894]: ALERT - script tried to increase memory_limit to 402653184 bytes which is above the allowed value (attacker 'REMOTE_ADDR not set', file 'unknown')


This problem wasn't happening before

Thanks,

[biggles]

Would it be possible to try it without ioncube or zend? They are definitely on top of my "we like to cause segfaults in apache"-list...

[CrK01]

It seems that the wordpress admin requirement of 256MB in memory limit has something to do with Suhosin module.

I raise the memory_limit value en suhosin to 128MB ( and also in /etc/php.ini ) and now I can put that wordpress web with Apache module, no more segfaults.

I hope this would help anyone with the same problem