Bug 713322

Summary: systemd hang, lots of zombies
Product: [openSUSE] openSUSE 12.1 Reporter: Christoph Obexer <cobexer>
Component: BasesystemAssignee: Frederic Crozat <fcrozat>
Status: RESOLVED FIXED QA Contact: E-mail List <qa-bugs>
Severity: Critical    
Priority: P5 - None    
Version: Milestone 3   
Target Milestone: ---   
Hardware: i686   
OS: Other   
Whiteboard:
Found By: --- Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Attachments: foto of kernel panic stack trace

Description Christoph Obexer 2011-08-20 20:31:32 UTC
Created attachment 446796 [details]
foto of kernel panic stack trace

User-Agent:       Mozilla/5.0 (Windows NT 6.1; WOW64; rv:8.0a2) Gecko/20110819 Firefox/8.0a2

today systemd suddenly stopped working, initially i found out that the network went down, however upon examination why that happened i discovered _lots_ of zomie processes(dbus-daemon, wpa_supplicant, udisks,...), all of them were children of the init process (systemd).

knowing that the init process should clean up such zombies i thought i attach with gdb and get a backtrace of the hang... well that turned out to be impossible because gdb pretended the process was already being debugged.

gdb asked to kill the process and stupid as i was i answered yes: -> kernel panic xD
I'm not sure the backtrace i have is of use, but I don't have more =/

a few hours before i did a zypper dup that also updated systemd from 33.x to 33.y IIRC

Kernel panic - not syncing: Attempted to kill init!
Pid: 1, comm systemd Not tainted 3.0.0-2-desktop #1
Call Trace:
try_stack_unwind
dump_trace
show_trace_log_lvl
show_trace
dump_stack
panic
forget_original_parent
exit_notify
do_exit
do_group_exit
get_signal_to_deliver
do_signal




Reproducible: Didn't try

Steps to Reproduce:
1.
2.
3.
Comment 1 Christoph Obexer 2011-08-29 17:41:50 UTC
I just found out that systemd wrote a core file that day, sure enough the systemd binary and the pthread library don't match any more...

any way to get useful data out of that core dump?
(gdb currently does not even finish to load the core dump)
Comment 2 Frederic Crozat 2011-09-08 12:39:19 UTC
Can you reproduce this with latest systemd from Factory ?
Comment 3 Christoph Obexer 2011-09-10 17:59:03 UTC
I tried to reproduce today but was unable to do that.
Comment 4 Frederic Crozat 2011-10-12 12:04:51 UTC
closing as fixed then.