Bug 5846 - autofs: automount segfaults on startup
Summary: autofs: automount segfaults on startup
Status: RESOLVED OLD
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: x86_64 Linux
Priority: Normal critical
Target Milestone: ---
Assignee: Mageia Bug Squad
QA Contact:
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2012-05-11 05:58 CEST by Herbert Poetzl
Modified: 2012-08-03 21:12 CEST (History)
5 users (show)

See Also:
Source RPM: autofs-5.0.6-6.mga2.src.rpm
CVE:
Status comment:


Attachments

Description Herbert Poetzl 2012-05-11 05:58:00 CEST
Description of problem:
when starting the autofs service, automount crashes with a segfault

Version-Release number of selected component (if applicable):
autofs-5.0.6-6.mga2.x86_64

How reproducible:
at least on one system always

Steps to Reproduce:
1. configure autofs to do some nfs mounts
2. reboot system or restart autofs service
3.


Starting program: /usr/sbin/automount -v -f
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".
Starting automounter version 5.0.6, master map auto.master
using kernel protocol version 5.02
[New Thread 0x7ffff7ff9700 (LWP 1927)]
[New Thread 0x7ffff7fdb700 (LWP 1928)]
Detaching after fork from child process 1929.
Detaching after fork from child process 1930.
lookup(file): failed to read included master map auto.master
[New Thread 0x7ffff67da700 (LWP 1931)]
Detaching after fork from child process 1932.
Detaching after fork from child process 1933.
mounted indirect on /mnt with timeout 60, freq 15 seconds

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffff67da700 (LWP 1931)]
0x00005555555726b5 in do_remount_indirect (ap=0x55555579f120, fd=15, path=0x55555579f200 "/mnt") at mounts.c:1212
1212					free(de2[i]);
(gdb) where
#0  0x00005555555726b5 in do_remount_indirect (ap=0x55555579f120, fd=15, path=0x55555579f200 "/mnt") at mounts.c:1212
#1  0x0000555555572d1d in remount_active_mount (ap=0x55555579f120, mc=0x0, path=0x55555579f200 "/mnt", devid=<optimized out>, type=1, 
    ioctlfd=0x7ffff67d9af4) at mounts.c:1327
#2  0x0000555555572e38 in try_remount (ap=0x55555579f120, me=0x0, type=1) at mounts.c:1357
#3  0x0000555555561a79 in do_mount_autofs_indirect (root=0x7ffff00008c0 "/mnt", ap=0x55555579f120) at indirect.c:103
#4  mount_autofs_indirect (ap=0x55555579f120, root=0x7ffff00008c0 "/mnt") at indirect.c:213
#5  0x0000555555560549 in mount_autofs (root=0x7ffff00008c0 "/mnt", ap=0x55555579f120) at automount.c:1005
#6  handle_mounts (arg=0x7fffffffdd20) at automount.c:1561
#7  0x00007ffff7bcab99 in start_thread (arg=0x7ffff67da700) at pthread_create.c:301
#8  0x00007ffff6cee0cd in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112
#9  0x0000000000000000 in ?? ()
Manuel Hiebel 2012-05-12 19:51:05 CEST

CC: (none) => ennael1, guillomovitch, supp

Comment 1 Guillaume Rousse 2012-05-15 09:54:42 CEST
Works for me with 'some nfs mounts'. Please give more details about your configuration.
Comment 2 Herbert Poetzl 2012-05-15 17:03:12 CEST
I can not consistently recreate the issue either, it seems to be related to the mount already existing (e.g. from a previous autfs mount) when autofs is restarted ... I'd say source code analysis might help, it seems that the de2[] array  generates illegal addresses.

my setup is nfs v3 on overlapping mount points i.e. /mnt/{0,1,2} and /mnt/A/{3,4,5} with ghosting set to 60 seconds.

CC: (none) => herbert

Comment 3 Marja Van Waes 2012-05-26 13:06:13 CEST
Hi,

This bug was filed against cauldron, but we do not have cauldron at the moment.

Please report whether this bug is still valid for Mageia 2.

Thanks :)

Cheers,
marja

Keywords: (none) => NEEDINFO

Comment 4 Marja Van Waes 2012-08-03 21:12:10 CEST
2 months later, no reply, so closing as old

Feel free to reopen if needed and tell in which fully updated Mageia version(s) you still have this issue

Status: NEW => RESOLVED
CC: (none) => marja11
Resolution: (none) => OLD


Note You need to log in before you can comment on or make changes to this bug.