possible circular lock dependency in reiserfs
Hi, I got the following report in syslog: Oct 17 10:56:35 oenone kernel: = Oct 17 10:56:35 oenone kernel: [ INFO: possible recursive locking detected ] Oct 17 10:56:35 oenone kernel: 2.6.23-default #1 Oct 17 10:56:35 oenone kernel: - Oct 17 10:56:35 oenone kernel: beagled/4092 is trying to acquire lock: Oct 17 10:56:35 oenone kernel: (&inode->i_mutex){--..}, at: [] reiserfs_xattr_set+0xf3/0x30e [reiserfs] Oct 17 10:56:35 oenone kernel: Oct 17 10:56:35 oenone kernel: but task is already holding lock: Oct 17 10:56:35 oenone kernel: (&inode->i_mutex){--..}, at: [] vfs_setxattr+0x4d/0x1e0 Oct 17 10:56:35 oenone kernel: Oct 17 10:56:35 oenone kernel: other info that might help us debug this: Oct 17 10:56:35 oenone kernel: 3 locks held by beagled/4092: Oct 17 10:56:35 oenone kernel: #0: (&inode->i_mutex){--..}, at: [] vfs_setxattr+0x4d/0x1e0 Oct 17 10:56:35 oenone kernel: #1: (&REISERFS_I(inode)->xattr_sem){}, at: [] reiserfs_setxattr+0x6b/0x11f [reiserfs] Oct 17 10:56:35 oenone kernel: #2: (&REISERFS_SB(s)->xattr_dir_sem){}, at: [] reiserfs_setxattr+0x9e/0x11f [reiserfs] Oct 17 10:56:35 oenone kernel: Oct 17 10:56:35 oenone kernel: stack backtrace: Oct 17 10:56:35 oenone kernel: Oct 17 10:56:35 oenone kernel: Call Trace: Oct 17 10:56:35 oenone kernel: [] __lock_acquire+0x1b7/0xcaf Oct 17 10:56:35 oenone kernel: [] lock_acquire+0x51/0x6c Oct 17 10:56:35 oenone kernel: [] :reiserfs:reiserfs_xattr_set+0xf3/0x30e Oct 17 10:56:35 oenone kernel: [] debug_mutex_lock_common+0x16/0x23 Oct 17 10:56:35 oenone kernel: [] __mutex_lock_slowpath+0xe1/0x277 Oct 17 10:56:35 oenone kernel: [] :reiserfs:reiserfs_xattr_set+0xf3/0x30e Oct 17 10:56:35 oenone kernel: [] trace_hardirqs_on+0x115/0x139 Oct 17 10:56:35 oenone kernel: [] _spin_unlock_irq+0x24/0x27 Oct 17 10:56:35 oenone kernel: [] __down_write_nested+0x34/0x9e Oct 17 10:56:35 oenone kernel: [] :reiserfs:reiserfs_setxattr+0xc7/0x11f Oct 17 10:56:35 oenone kernel: [] vfs_setxattr+0xae/0x1e0 Oct 17 10:56:35 oenone kernel: [] setxattr+0xc0/0xdd Oct 17 10:56:35 oenone kernel: [] kmem_cache_free+0x1c4/0x1d6 Oct 17 10:56:35 oenone kernel: [] trace_hardirqs_on+0x115/0x139 Oct 17 10:56:35 oenone kernel: [] do_path_lookup+0x1ae/0x22c Oct 17 10:56:35 oenone kernel: [] __user_walk_fd+0x41/0x4c Oct 17 10:56:35 oenone kernel: [] sys_lsetxattr+0x49/0x6a Oct 17 10:56:35 oenone kernel: [] tracesys+0x71/0xe1 Oct 17 10:56:35 oenone kernel: [] tracesys+0xdc/0xe1 Oct 17 10:56:35 oenone kernel: - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [Jfs-discussion] [2/4] 2.6.23-rc6: known regressions
Am Donnerstag 13 September 2007 schrieb Dave Kleikamp: > On Wed, 2007-09-12 at 18:58 +0200, Michal Piotrowski wrote: > > > Subject : umount triggers a warning in jfs and takes almost a minute > > References : http://lkml.org/lkml/2007/9/4/73 > > Last known good : ? > > Submitter : Oliver Neukum <[EMAIL PROTECTED]> > > Caused-By : ? > > Handled-By : ? > > Status : unknown > > I'm still waiting to hear from Oliver whether or not this is actually a > regression. OK, I've done test. On 2.6.22 I was unable to trigger the warning. On 2.6.23-rc6 I get the following warning in about 3/4 of all unmounts: Sep 19 13:08:04 oenone kernel: WARNING: at fs/jfs/jfs_logmgr.c:1643 jfs_flush_journal() Sep 19 13:08:04 oenone kernel: Sep 19 13:08:04 oenone kernel: Call Trace: Sep 19 13:08:04 oenone kernel: [] :jfs:jfs_flush_journal+0x26a/0x27d Sep 19 13:08:04 oenone kernel: [] dispose_list+0xde/0xf7 Sep 19 13:08:04 oenone kernel: [] :jfs:jfs_umount+0x30/0xe5 Sep 19 13:08:04 oenone kernel: [] :jfs:jfs_put_super+0xd/0x5e Sep 19 13:08:04 oenone kernel: [] generic_shutdown_super+0x60/0xf0 Sep 19 13:08:04 oenone kernel: [] kill_block_super+0xd/0x1e Sep 19 13:08:04 oenone kernel: [] deactivate_super+0x6a/0x82 Sep 19 13:08:04 oenone kernel: [] sys_umount+0x249/0x25a Sep 19 13:08:04 oenone kernel: [] audit_syscall_entry+0x141/0x174 Sep 19 13:08:04 oenone kernel: [] tracesys+0xdc/0xe1 Sep 19 13:08:04 oenone kernel: This is on a partition of a usb mass storage device with 2K sector size. IMHO it is a regression. I am recompiling with CONFIG_JFS_DEBUG. Regards Oliver - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [2/3] 2.6.23-rc6: known regressions v2
Am Dienstag 18 September 2007 schrieb Jan Kara: > > Subject : umount triggers a warning in jfs and takes almost a minute > > References : http://lkml.org/lkml/2007/9/4/73 > > Last known good : ? > > Submitter : Oliver Neukum <[EMAIL PROTECTED]> > > Caused-By : ? > > Handled-By : ? > > Status : unknown > I thought Shaggy asked Oliver about some details (and he did not > answer so far) so I'd assume Shaggy is handling this. I was without access to the hardware for a week. Testing will start tomorrow. Regards Oliver - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] [-mm] FS: file name must be unique in the same dir in procfs
Am Montag 20 August 2007 schrieb Zhang Rui: > Files name must be unique in the same directory. > > Bug is reported here: > http://bugzilla.kernel.org/show_bug.cgi?id=8798 Then I'd say fix the callers. This will paper over bugs. Regards Oliver - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html