Messages in this thread | | | Date | Tue, 5 Sep 2006 22:33:09 +0200 | From | Mattia Dongili <> | Subject | JFS - real deadlock and lockdep warning (2.6.18-rc5-mm1) |
| |
Hello,
as the subject says it's some time[0] I'm experiencing deadlocks[1] (I'm only tracking -mm, and sporadically using the stable series). I have a couple of use cases that seem to reliably trigger the deadlock, namely using Eclipse and Firefox.
[0]: I'd say at least since 2.6.17-rc-something [1]: no sysrq, no logs, no nothing, netconsole doesn't log anything more, just screwed FS which grub can't read (Inconsistent filesystem - press any key to continue)
I decided to try out the LOCKDEP stuff to see if I can provide more useful info, and here we go, as soon as my /home (jfs) partition is written (at login): [ 72.420000] [ 72.420000] ============================================= [ 72.420000] [ INFO: possible recursive locking detected ] [ 72.420000] 2.6.18-rc5-mm1-2-lockdep #9 [ 72.420000] --------------------------------------------- [ 72.420000] zsh/3844 is trying to acquire lock: [ 72.420000] (&jfs_ip->commit_mutex){--..}, at: [<c0320948>] mutex_lock+0x8/0x10 [ 72.420000] [ 72.420000] but task is already holding lock: [ 72.420000] (&jfs_ip->commit_mutex){--..}, at: [<c0320948>] mutex_lock+0x8/0x10 [ 72.420000] [ 72.420000] other info that might help us debug this: [ 72.420000] 2 locks held by zsh/3844: [ 72.420000] #0: (&inode->i_mutex){--..}, at: [<c0320948>] mutex_lock+0x8/0x10 [ 72.420000] #1: (&jfs_ip->commit_mutex){--..}, at: [<c0320948>] mutex_lock+0x8/0x10 [ 72.420000] [ 72.420000] stack backtrace: [ 72.420000] [<c0103b8f>] dump_trace+0x1ef/0x230 [ 72.420000] [<c0103bf6>] show_trace_log_lvl+0x26/0x40 [ 72.420000] [<c01043cb>] show_trace+0x1b/0x20 [ 72.420000] [<c01044b4>] dump_stack+0x24/0x30 [ 72.420000] [<c0136050>] __lock_acquire+0x8e0/0xd80 [ 72.420000] [<c0136869>] lock_acquire+0x69/0x90 [ 72.420000] [<c0320705>] __mutex_lock_slowpath+0x75/0x2b0 [ 72.420000] [<c0320948>] mutex_lock+0x8/0x10 [ 72.420000] [<d137950b>] jfs_create+0xbb/0x420 [jfs] [ 72.420000] [<c016cb7b>] vfs_create+0xcb/0x120 [ 72.420000] [<c0170158>] open_namei+0x618/0x6f0 [ 72.420000] [<c0162368>] do_filp_open+0x38/0x60 [ 72.420000] [<c01623db>] do_sys_open+0x4b/0xf0 [ 72.420000] [<c01624d7>] sys_open+0x27/0x30 [ 72.420000] [<c01032af>] syscall_call+0x7/0xb [ 72.420000] [<b7ec8b8d>] 0xb7ec8b8d [ 72.420000] =======================
I suspect JFS is guilty, anyway my HD has these partitions:
/dev/hda1 on / type reiserfs (rw) /dev/hda3 on /usr type reiserfs (rw) /dev/hda5 on /home type jfs (rw)
bootlog: http://oioio.altervista.org/linux/dmesg-2.6.18-rc5-mm1-lockdep config: http://oioio.altervista.org/linux/config-2.6.18-rc5-mm1-lockdep
-- mattia :wq! - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |