Bug 1069764 - [abrt] lvm2: daemon_log(): lvmetad killed by SIGSEGV
Summary: [abrt] lvm2: daemon_log(): lvmetad killed by SIGSEGV
Keywords:
Status: CLOSED NEXTRELEASE
Alias: None
Product: Fedora
Classification: Fedora
Component: lvm2
Version: 20
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Petr Rockai
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard: abrt_hash:3dc4c7703d2ef9a49a108af7797...
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2014-02-25 15:48 UTC by Michael Hampton
Modified: 2014-06-09 07:10 UTC (History)
12 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2014-06-09 00:36:45 UTC
Type: ---
Embargoed:

Attachments (Terms of Use)
File: backtrace (10.54 KB, text/plain)
2014-02-25 15:48 UTC, Michael Hampton 		no flags Details
File: cgroup (163 bytes, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: core_backtrace (6.31 KB, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: dso_list (1.07 KB, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: environ (117 bytes, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: exploitable (82 bytes, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: limits (1.29 KB, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: maps (6.07 KB, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: open_fds (259 bytes, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: proc_pid_status (907 bytes, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
File: var_log_messages (20.69 KB, text/plain)
2014-02-25 15:49 UTC, Michael Hampton 		no flags Details
View All

Description Michael Hampton 2014-02-25 15:48:52 UTC 
Version-Release number of selected component:
lvm2-2.02.103-5.fc20

Additional info:
reporter:       libreport-2.1.12
backtrace_rating: 4
cmdline:        /usr/sbin/lvmetad
crash_function: daemon_log
executable:     /usr/sbin/lvmetad
kernel:         3.13.3-201.fc20.x86_64
runlevel:       unknown
type:           CCpp
uid:            0

Truncated backtrace:
Thread no. 1 (5 frames)
 #0 daemon_log at daemon-log.c:54
 #1 daemon_logf at daemon-log.c:76
 #2 pv_clear_all at lvmetad-core.c:829
 #3 handler at lvmetad-core.c:1094
 #4 client_thread at daemon-server.c:401
Comment 1 Michael Hampton 2014-02-25 15:48:58 UTC 
Created attachment 867483 [details]
File: backtrace
Comment 2 Michael Hampton 2014-02-25 15:49:03 UTC 
Created attachment 867484 [details]
File: cgroup
Comment 3 Michael Hampton 2014-02-25 15:49:14 UTC 
Created attachment 867485 [details]
File: core_backtrace
Comment 4 Michael Hampton 2014-02-25 15:49:15 UTC 
Created attachment 867486 [details]
File: dso_list
Comment 5 Michael Hampton 2014-02-25 15:49:19 UTC 
Created attachment 867487 [details]
File: environ
Comment 6 Michael Hampton 2014-02-25 15:49:21 UTC 
Created attachment 867488 [details]
File: exploitable
Comment 7 Michael Hampton 2014-02-25 15:49:22 UTC 
Created attachment 867489 [details]
File: limits
Comment 8 Michael Hampton 2014-02-25 15:49:24 UTC 
Created attachment 867490 [details]
File: maps
Comment 9 Michael Hampton 2014-02-25 15:49:26 UTC 
Created attachment 867491 [details]
File: open_fds
Comment 10 Michael Hampton 2014-02-25 15:49:28 UTC 
Created attachment 867492 [details]
File: proc_pid_status
Comment 11 Michael Hampton 2014-02-25 15:49:32 UTC 
Created attachment 867493 [details]
File: var_log_messages
Comment 12 Petr Rockai 2014-02-28 11:06:27 UTC 
Seems like a client thread is still trying to log while lvmetad is shutting down. We probably need to clean up worker threads before going into shutdown. The log seems to be related to #1069808 so that part of the problem should be already sorted out. (Seems like the user tried to obey that message, restarted lvmetad and it got a SEGV during shutdown as there was still traffic going in.)
Comment 13 Petr Rockai 2014-06-09 00:36:45 UTC 
Should be fixed upstream in 488f3085279e24af21e870d5a74d0953094552f7. The shutdown sequence now waits for client threads to finish before destroying resources (which they may be using).
Comment 14 Peter Rajnoha 2014-06-09 07:10:02 UTC 
The patch is part of upcoming lvm2 v2.02.107 release.
Note You need to log in before you can comment on or make changes to this bug.