On Mon 24-03-14 10:42:25, Sasha Levin wrote: > On 03/10/2014 10:13 AM, Sasha Levin wrote: > >On 03/10/2014 06:43 AM, Jan Kara wrote: > >> By garbage, do you mean that it is a poison, completely random data or does > >>inode->i_sb look like a valid pointer but just superblock isn't where it > >>points to? > > > >It's poison. > > > >>>>Any way I could get anything useful any other way? > >> Hum, can you dump the whole contents of 'dentry' at that place? Maybe it > >>will tell us something. > > > >I'll give it a go, will update when it happens again. > > Okay, I've added this: > > diff --git a/fs/libfs.c b/fs/libfs.c > index a184424..2492dc4 100644 > --- a/fs/libfs.c > +++ b/fs/libfs.c > @@ -377,6 +377,7 @@ int simple_setattr(struct dentry *dentry, struct iattr *iattr) > if (iattr->ia_valid & ATTR_SIZE) > truncate_setsize(inode, iattr->ia_size); > setattr_copy(inode, iattr); > + printk(KERN_ERR "** %u %p %s %p %s %p %p %p\n", dentry->d_flags, dentry->d_pare > mark_inode_dirty(inode); > return 0; > } > > And got the following: > > [ 339.948946] ** 4194304 ffff8805ac03ba38 [eventpoll] ffff8806ec051fe0 > [eventpoll] ffffffff84666040 ffff88056c73e7b0 (null) OK, great. So finally we have something useful. We know we have problems with [eventpoll] dentry. That is actually a special filesystem not mounted anywhere - likely you get to that dentry through /proc/<pid>/fd/. Now eventpoll is interesting because it uses single anon inode for all eventpoll instances. And that inode should stay in place as long as eventpoll filesystem exists. So it's not clear how come that inode is freed. The basic check of handling of inode use count didn't find anything suspicious. But I can check in more detail and if I fail, we now have a pretty narrow area where to look... Honza > [ 339.956028] general protection fault: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC > [ 339.958199] Dumping ftrace buffer: > [ 339.959158] (ftrace buffer empty) > [ 339.960075] Modules linked in: > [ 339.960712] CPU: 2 PID: 9702 Comm: trinity-c2 Tainted: G W 3.14.0-rc7-next-20140321-sasha-00018-g0516fe6-dirty #266 > [ 339.962657] task: ffff88012b8b0000 ti: ffff88012b88a000 task.ti: ffff88012b88a000 > [ 339.964089] RIP: 0010:[<ffffffff8133af2c>] [<ffffffff8133af2c>] __mark_inode_dirty+0x10c/0x4a0 > [ 339.965274] RSP: 0018:ffff88012b88bdb8 EFLAGS: 00010206 > [ 339.965274] RAX: 6b6b6b6b6b6b6b6b RBX: ffff8806ec051fe0 RCX: 0000000000000006 > [ 339.965274] RDX: 0000000000008a90 RSI: 0000000000000007 RDI: ffff8806ec051fe0 > [ 339.965274] RBP: ffff88012b88bdd8 R08: 0000000000000000 R09: 0000000000000000 > [ 339.965274] R10: 0000000000000001 R11: 3330636135303838 R12: 0000000000000007 > [ 339.965274] R13: ffff8806ec051fe0 R14: ffff8806ec6bb3d8 R15: ffff8806ec051fe0 > [ 339.965274] FS: 00007f1993d82700(0000) GS:ffff8800bec00000(0000) knlGS:0000000000000000 > [ 339.965274] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b > [ 339.974036] CR2: 0000000000000000 CR3: 0000000123b16000 CR4: 00000000000006a0 > [ 339.974036] DR0: 0000000000698000 DR1: 0000000000000000 DR2: 0000000000000000 > [ 339.974036] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000600 > [ 339.974036] Stack: > [ 339.974036] 0000000000000000 ffff8805ac03ba38 ffff8806ec051fe0 ffff88012b88bec8 > [ 339.974036] ffff88012b88be28 ffffffff81336f7b ffffffff84666040 ffff88056c73e7b0 > [ 339.974036] 0000000000000000 ffff8806ec0520c8 0000000000000041 ffff88012b88bec8 > [ 339.974036] Call Trace: > [ 339.974036] [<ffffffff81336f7b>] simple_setattr+0xab/0xd0 > [ 339.974036] [<ffffffff8132b148>] notify_change+0x258/0x390 > [ 339.974036] [<ffffffff81307db2>] ? chmod_common+0x72/0x150 > [ 339.974036] [<ffffffff81307df4>] chmod_common+0xb4/0x150 > [ 339.974036] [<ffffffff8132c274>] ? __fget_light+0xe4/0x130 > [ 339.974036] [<ffffffff81309382>] SyS_fchmod+0x62/0xa0 > [ 339.974036] [<ffffffff84506a58>] tracesys+0xe1/0xe6 > [ 339.974036] Code: 8b 45 00 0f 1f 40 00 49 8b 7d 08 44 89 e2 49 83 c5 10 48 89 de ff d0 49 8b 45 00 48 85 c0 75 e7 eb c5 0f 1f 44 00 00 49 8b 46 30 <48> 8b 40 10 48 85 c0 74 08 44 89 e6 48 89 df ff d0 66 66 66 66 > [ 339.974036] RIP [<ffffffff8133af2c>] __mark_inode_dirty+0x10c/0x4a0 > [ 339.974036] RSP <ffff88012b88bdb8> > > > Thanks, > Sasha -- Jan Kara <jack@xxxxxxx> SUSE Labs, CR -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html