Description of problem: Child segmentations faults... Version-Release number of selected component (if applicable): [root@brimir tmp]# rpm -qa | grep httpd && rpm -qa | grep mod_python httpd-2.2.3-11.el5_1.3 httpd-manual-2.2.3-11.el5_1.3 httpd-devel-2.2.3-11.el5_1.3 httpd-devel-2.2.3-11.el5_1.3 system-config-httpd-1.3.3.3-1.el5 mod_python-3.2.8-3.1 How reproducible: Always. Steps to Reproduce: 1. Start the daemon and childs starts dying.. 2. 3. Actual results: Segmentation fault Expected results: No Segmentation fault Additional info: Core was generated by `/usr/sbin/httpd'. Program terminated with signal 11, Segmentation fault. #0 0x00002aaab9e6de1b in PyImport_GetModuleDict () from /usr/lib64/libpython2.4.so.1.0 (gdb) where #0 0x00002aaab9e6de1b in PyImport_GetModuleDict () from /usr/lib64/libpython2.4.so.1.0 #1 0x00002aaab9e6de6b in PyImport_AddModule () from /usr/lib64/libpython2.4.so.1.0 #2 0x00002aaab9e74578 in Py_InitModule4 () from /usr/lib64/libpython2.4.so.1.0 #3 0x00002af5a695b743 in inittime () from /usr/lib64/python2.4/lib-dynload/timemodule.so #4 0x00002aaabb8b9fed in ?? () #5 0x0000000000000000 in ?? ()
Created attachment 311253 [details] Childs Core dump
Can you be more specific about the cause of the problem? 1) are children dying in response to particular requests? 2) what mod_python configuration is being used?
I removed the httpd-manual rpm and hashed out <Directory "/var/www/manual/mod/mod_python> entry in /etc/httpd/conf.d/python.conf ( the only one that was unhased ) but kept the module loaded ( we are not even using it ) and segmentation faults stopped ( well atleast for the last 24 hours so far so good ).
I'm unable to reproduce this here. Can you do: # rpm -V httpd mod_python python # debuginfo-install --enablerepo=rhel-debuginfo mod_python httpd python ... then gdb /usr/sbin/httpd /path/to/core and then enter "bt" to get a backtrace?
Created attachment 311945 [details] rpm-yum-gdb output. Here you go :) As I mentioned after I removed httpd-manual and hashed out the relevant section in /etc/httpd/conf.d/python.conf apache stopped segfaulting
Ah, you've transposed "debuginfo-install" with "yum -y install" there! debuginfo-install will install the set of debuginfo packages necessary to debug this further.
Created attachment 311948 [details] debug-httpd-gdb.txt Here you go take2... :) Since I did not have yum-utils installed and never used debuginfo-install I presumed that you where meaning yum.
I have the same problem when i enable php support to httpd Reported in: /var/log/httpd/error_log [Tue Jun 16 22:31:29 2009] [debug] proxy_util.c(1801): proxy: grabbed scoreboard slot 0 in child 573 for worker proxy:reverse [Tue Jun 16 22:31:29 2009] [debug] proxy_util.c(1820): proxy: worker proxy:reverse already initialized [Tue Jun 16 22:31:29 2009] [debug] proxy_util.c(1914): proxy: initialized single connection worker 0 in child 573 for (*) [Tue Jun 16 22:31:29 2009] [notice] child pid 541 exit signal Segmentation fault (11) [Tue Jun 16 22:31:29 2009] [notice] child pid 542 exit signal Segmentation fault (11) [Tue Jun 16 22:31:29 2009] [notice] child pid 543 exit signal Segmentation fault (11) [Tue Jun 16 22:31:29 2009] [notice] child pid 544 exit signal Segmentation fault (11) [Tue Jun 16 22:31:29 2009] [notice] child pid 545 exit signal Segmentation fault (11) ... This problem appeared after upgrade to fedora 11. My configuration is accurately the same of my fedora 10.
Please, disrespects the commentary above, the true causer is mod_nss Reported in: /var/log/httpd/error_log [Wed Jun 17 13:44:01 2009] [debug] proxy_util.c(1820): proxy: worker proxy:reverse already initialized [Wed Jun 17 13:44:01 2009] [debug] proxy_util.c(1914): proxy: initialized single connection worker 0 in child 14832 for (*) [Wed Jun 17 13:44:01 2009] [notice] child pid 14827 exit signal Segmentation fault (11) [Wed Jun 17 13:44:01 2009] [notice] child pid 14828 exit signal Segmentation fault (11) [Wed Jun 17 13:44:01 2009] [notice] child pid 14829 exit signal Segmentation fault (11) [Wed Jun 17 13:44:01 2009] [notice] child pid 14830 exit signal Segmentation fault (11) [Wed Jun 17 13:44:01 2009] [notice] child pid 14832 exit signal Segmentation fault (11) [Wed Jun 17 13:44:01 2009] [notice] child pid 14833 exit signal Segmentation fault (11) [Wed Jun 17 13:44:01 2009] [notice] child pid 14834 exit signal Segmentation fault (11) If i remove mod_nss httpd (without https support) work fine! This problem appeared after upgrade to fedora 11. My configuration is accurately the same of my fedora 10.
I install nss from testing repository and and now httpd work fine! Packages installed from testing repository: nss-3.12.3.99.3-2.11.2.fc11.i586 nss-tools-3.12.3.99.3-2.11.2.fc11.i586 nss_ldap-264-2.fc11.i586 nss-softokn-freebl-3.12.3.99.3-2.11.2.fc11.i586
I had the same problem with 2 machines after an upgrade F10=>F11 the nss upgrade solved the issue Thanks
This request was evaluated by Red Hat Product Management for inclusion in the current release of Red Hat Enterprise Linux. Because the affected component is not scheduled to be updated in the current release, Red Hat is unable to address this request at this time. Red Hat invites you to ask your support representative to propose this request, if appropriate, in the next release of Red Hat Enterprise Linux.
Can you still reproduce it? If yes, what's the output of: ldd /etc/httpd/modules/mod_python.so Also, if you have mod_wsgi installed, can you paste also output of: ldd /etc/httpd/modules/mod_wsgi.so I'm not able to reproduce it so far.
We've been unable to reproduce this internally; I'm closing this bug. If you have further issues please contact Red Hat Support for assistance.