Messages in this thread | | | Date | Wed, 15 Aug 2012 01:30:13 -0700 | From | Joel Becker <> | Subject | Re: 3.5-rc6 configfs BUG_ON |
| |
On Sun, Jul 15, 2012 at 07:58:31PM -0400, Dave Jones wrote: > On Sun, Jul 15, 2012 at 01:58:46PM -0700, Joel Becker wrote: > > > > Call Trace: > > > [<ffffffff811ec74f>] d_kill+0xaf/0x120 > > > [<ffffffff811ec8a2>] dput+0xe2/0x1d0 > > > [<ffffffff811dfaea>] path_put+0x1a/0x30 > > > [<ffffffff811d9705>] vfs_fstatat+0x55/0x70 > > > [<ffffffff811d973e>] vfs_lstat+0x1e/0x20 > > > [<ffffffff811d99ca>] sys_newlstat+0x1a/0x40 > > > [<ffffffff81696aed>] system_call_fastpath+0x1a/0x1f > > > Code: f6 43 40 01 75 eb 48 8b 7b 50 e8 41 4d f6 ff 48 8b 3d c2 bf 7f 01 48 89 de e8 a2 52 f6 ff 4c 89 e7 e8 1a b5 f9 ff 5b 41 5c 5d c3 <0f> 0b be 9c 00 00 00 48 c7 c7 30 a2 9e 81 e8 42 2e e1 ff eb a7 > > > RIP [<ffffffff812562fb>] configfs_d_iput+0x8b/0xa0 > > > RSP <ffff88014325be08> > > > ---[ end trace 65e130035db36f7d ]--- > > > > > > 58 if (sd) { > > > 59 BUG_ON(sd->s_dentry != dentry); > > > > > > > Dave, > > What were you doing at the time? Is this ocfs2, dlm, target, or > > what? > > > > Joel > > > > See http://codemonkey.org.uk/projects/trinity/ > Running that with -c newlstat should trigger it. > (load configfs module first, no further config necessary)
Cool tool. I presume that -c newlstat just randomly picks paths to newlstat. I think I see what's happening, but I can't see what gets the dirent into that state. Now to get a VM I can break.
Joel
--
"The trouble with being punctual is that nobody's there to appreciate it." - Franklin P. Jones
http://www.jlbec.org/ jlbec@evilplan.org
| |