Bug 770471

Summary: systemd dumps core when ntp is started after fsck
Product: [Fedora] Fedora Reporter: Sebastian Freundt <devel>
Component: systemdAssignee: systemd-maint
Status: CLOSED WORKSFORME QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 16CC: johannbg, lpoetter, metherid, mschmidt, notting, plautrba, systemd-maint
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2012-03-13 02:00:44 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Attachments:
Description Flags
core dump generated by udevd none

Description Sebastian Freundt 2011-12-26 21:46:42 UTC 
Description of problem:


Version-Release number of selected component (if applicable):
37-317.1

How reproducible:
Always

Steps to Reproduce:
1. Have a system with a relatively slower clock
2. Boot, fsck will report the super block from the future
3. Go to the emergency shell, set up the network
4. Trigger the ntp service
  
Actual results:
Fsck is started, some udev voodoo is performed -> segfault, system becomes
unusable.

Expected results:
The clock should have been set according to an ntp reference time.

Additional info:
root@kegel:~ # /etc/init.d/ntp start
redirecting to systemctl
[  126.021306] systemd[1]: systemd-modules-load.service: main process exited,
code=exited, status=1
[  126.040162] udevd[715]: starting version 173
[  126.044096] systemd[1]: Unit systemd-modules-load.service entered failed
state.
systemd-fsck[725]: /dev/sda1: Superblock last mount time (Mon Dec 26 18:53:00
2011,[  126.244629] systemd-fsck[725]: /dev/sda1: Superblock last mount time
(Mon Dec 26 18:53:00 2011,

systemd-fsck[725]: now = Mon Nov  7 00:02:04 2011) is in the future.[ 
126.247586] systemd-fsck[725]: now = Mon Nov  7 00:02:04 2011) is in the
future.

systemd-fsck[725]: /dev/sda1: UNEXPECTED INCONSISTENCY; RUN fsck MANUALLY.[ 
126.252062] systemd-fsck[725]: /dev/sda1: UNEXPECTED INCONSISTENCY; RUN fsck
MANUALLY.

systemd-fsck[725]: (i.e., without -a or -p options)
[  126.263233] systemd-fsck[725]: (i.e., without -a or -p options)
[  126.355229] systemd-fsck[725]: fsck failed with error code 4.
[  126.357360] systemd-fsck[725]: Running request
emergency.target/start/isolate
[  126.391924] systemd[766]: Failed at step EXEC spawning /bin/plymouth: No
such file or directory
Welcome to emergency mode. Use "systemctl default" or ^D to activate default
mode.
udevd[715]: worker [756] did not accept message -1 (Connection refused), kill
it

udevd[715]: worker [757] did not accept message -1 (Connection refused), kill
it

udevd[715]: worker [759] did not accept message -1 (Connection refused), kill
it

udevd[715]: worker [763] did not accept message -1 (Connection refused), kill
it

udevd[715]: worker [764] did not accept message -1 (Connection refused), kill
it

Give root password for login: [  126.568785] udevd[715]: segfault at 24 ip
0000000000404574 sp 00007fffe92090f0 error 6 in udevd[400000+21000]
[  126.575152] systemd[1]: udev.service: main process exited, code=killed,
status=11
[  126.579137] systemd[1]: Unit udev.service entered failed state.

I don't seem to be able to produce core dumps even though systemd.dump_core is set to true.  Will keep trying to come up with one.
Comment 1 Sebastian Freundt 2011-12-26 21:57:05 UTC 
The core dump is actually empty, I used my own dumper that doesn't respect limits.
Comment 2 Sebastian Freundt 2011-12-26 22:52:14 UTC 
Created attachment 549628 [details]
core dump generated by udevd

Core was generated by `/sbin/udevd'.
Program terminated with signal 11, Segmentation fault.
#0  worker_returned (fd_worker=<optimized out>) at udev/udevd.c:627
627     udev/udevd.c: No such file or directory.
        in udev/udevd.c
(gdb) bt
#0  worker_returned (fd_worker=<optimized out>) at udev/udevd.c:627
#1  main (argc=<optimized out>, argv=<optimized out>) at udev/udevd.c:1651
Comment 3 Michal Schmidt 2012-01-02 12:33:12 UTC 
(In reply to comment #0)
> Version-Release number of selected component (if applicable):
> 37-317.1

Are you sure you're using Fedora? This looks like Frederic Crozat's build of the systemd package for openSuse.
Comment 4 Sebastian Freundt 2012-01-02 12:46:46 UTC 
No I'm not using Fedora, I use opensuse's systemd, yes.  They asked me to report this to upstream however.  What is the upstream bug tracker?
Comment 5 Michal Schmidt 2012-01-02 15:19:43 UTC 
See the section "Bug Reports" at http://www.freedesktop.org/wiki/Software/systemd/

It is likely the bug is present in Fedora too, so I'm not going to close this BZ immediately. I'm setting NEEDINFO on anyone, because we need someone who can reproduce it or provide a core/backtrace of systemd.
Comment 6 Lennart Poettering 2012-03-13 02:00:44 UTC 
Hmm, let's close this here. THis is not a bug in Fedora and in an old version of systemd. Please report bugs regarding systemd to upstream only in the most recent versions, and don't report them in rhbz.