View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000400 | AlmaLinux-8 | autofs | public | 2023-05-25 16:12 | 2023-06-26 21:48 |
| Reporter | lantzer | Assigned To | |||
| Priority | high | Severity | crash | Reproducibility | always |
| Status | new | Resolution | open | ||
| Platform | x86_64 | OS | Alma Linux | OS Version | 8.8 |
| Summary | 0000400: Crash when restarting autofs when a direct map-ed directory is busy and cannot be unmounted | ||||
| Description | After upgrading to Alma Linux 8.8, the autofs 1:5.1.4-93.el8 package now crashes when the autofs service is restarted while a direct mapped mount is in use. This problem did not happen with Alma Linux 8.7 and autofs 1:5.1.4-83.el8. This problem also does not happen for an indirect mount, as it appears to be directly related to the autofs virtual mount and the target mount using the same directory. | ||||
| Steps To Reproduce | Create a direct map in autofs to mount an exported directory on an NFS server, change to that directory, then restart the autofs service. At this point the service crashes. /etc/auto.master: /- /etc/auto.home /etc/auto.home: /home nfsserver.local:/export cd /home systemctl restart autofs # autofs crashes | ||||
| Additional Information | May 25 10:56:51 localhost systemd[1]: Stopping Automounts filesystems on demand... May 25 10:56:53 localhost systemd[1]: autofs.service: Succeeded. May 25 10:56:53 localhost systemd[1]: Stopped Automounts filesystems on demand. May 25 10:56:53 localhost systemd[1]: Starting Automounts filesystems on demand... May 25 10:56:53 localhost systemd[1]: tmp-autozPOEVP.mount: Succeeded. May 25 10:56:53 localhost systemd[1]: tmp-autoYYS6Qj.mount: Succeeded. May 25 10:56:53 localhost kernel: automount[2193290]: segfault at 28 ip 000055a2e7004b39 sp 00007ff7f2968970 error 4 in automount[55a2e6fdb000+4b000] May 25 10:56:53 localhost kernel: Code: 01 00 31 c0 31 ed 48 8d 35 5d 5f 01 00 e8 af 18 00 00 41 83 fe 01 0f 84 d5 01 00 00 41 83 a4 24 90 00 00 00 fd 49 8b 44 24 30 <4c> 8b 40 28 41 80 38 2f 0f 84 99 01 00 00 4c 8d ac 24 00 01 00 00 May 25 10:56:53 localhost systemd[1]: Started Process Core Dump (PID 2193293/UID 0). May 25 10:56:53 localhost systemd-coredump[2193294]: Process 2193284 (automount) of user 0 dumped core. Stack trace of thread 2193290: #0 0x000055a2e7004b39 try_remount (automount) #1 0x000055a2e6feeb0c do_mount_autofs_direct (automount) 0000002 0x000055a2e6feef72 mount_autofs_direct (automount) 0000003 0x000055a2e6feb738 handle_mounts (automount) 0000004 0x00007ff7f6d3d1cf start_thread (libpthread.so.0) 0000005 0x00007ff7f52dce73 __clone (libc.so.6) Stack trace of thread 2193286: #0 0x00007ff7f6d4345c pthread_cond_wait@@GLIBC_2.3.2 (libpthread.so.0) #1 0x000055a2e7007ce4 alarm_handler (automount) 0000002 0x00007ff7f6d3d1cf start_thread (libpthread.so.0) 0000003 0x00007ff7f52dce73 __clone (libc.so.6) Stack trace of thread 2193284: #0 0x00007ff7f6d4345c pthread_cond_wait@@GLIBC_2.3.2 (libpthread.so.0) #1 0x000055a2e6ff57cb master_mount_mounts (automount) 0000002 0x000055a2e6ff5fcd master_read_master (automount) 0000003 0x000055a2e6fe81d4 main (automount) 0000004 0x00007ff7f52ddd85 __libc_start_main (libc.so.6) 0000005 0x000055a2e6fe89ee _start (automount) Stack trace of thread 2193287: #0 0x00007ff7f6d4345c pthread_cond_wait@@GLIBC_2.3.2 (libpthread.so.0) #1 0x000055a2e6ffdd3f st_queue_handler (automount) 0000002 0x00007ff7f6d3d1cf start_thread (libpthread.so.0) 0000003 0x00007ff7f52dce73 __clone (libc.so.6) May 25 10:56:53 localhost systemd[1]: systemd-coredump@2-2193293-0.service: Succeeded. May 25 10:56:53 localhost systemd[1]: autofs.service: Main process exited, code=killed, status=11/SEGV May 25 10:56:53 localhost systemd[1]: autofs.service: Failed with result 'signal'. May 25 10:56:53 localhost systemd[1]: Failed to start Automounts filesystems on demand. | ||||
| Tags | AlmaLinux, almalinux8, segfault | ||||
| abrt_hash | |||||
| URL | |||||
|
|
After finding a list of CentOS Stream 8 packages (http://mirror.centos.org/centos/8-stream/BaseOS/x86_64/os/Packages/), I tested several autofs RPMs from CentOS Stream 8, and found that this problem appears to have been introduced in autofs-5.1.4-91.el8, but it appears to be resolved in autofs-5.1.4-103.el8. It seems that one possible workaround is to install this package from CentOS Stream 8: http://mirror.centos.org/centos/8-stream/BaseOS/x86_64/os/Packages/autofs-5.1.4-103.el8.x86_64.rpm |
