* EXT4: lockdep complaints on 3.1-rc4+
@ 2011-09-20 17:55 Dmitry Torokhov
2011-09-20 17:59 ` Josh Boyer
0 siblings, 1 reply; 3+ messages in thread
From: Dmitry Torokhov @ 2011-09-20 17:55 UTC (permalink / raw)
To: linux-ext4; +Cc: LKML
Hi,
Not sure if it has been reported but I just got the following lockdep
warning on 3.1-rc4+ + my input queue (which does not touch FS code at
all):
[ 327.735223] =======================================================
[ 327.735228] [ INFO: possible circular locking dependency detected ]
[ 327.735230] 3.1.0-rc4+ #664
[ 327.735232] -------------------------------------------------------
[ 327.735234] kontact/2570 is trying to acquire lock:
[ 327.735236] (&sb->s_type->i_mutex_key#14){+.+.+.}, at: [<ffffffff811f79a9>] ext4_evict_inode+0x29/0x5c0
[ 327.735247]
[ 327.735247] but task is already holding lock:
[ 327.735249] (&mm->mmap_sem){++++++}, at: [<ffffffff81142b78>] sys_munmap+0x48/0x80
[ 327.735256]
[ 327.735256] which lock already depends on the new lock.
[ 327.735257]
[ 327.735259]
[ 327.735259] the existing dependency chain (in reverse order) is:
[ 327.735262]
[ 327.735262] -> #1 (&mm->mmap_sem){++++++}:
[ 327.735266] [<ffffffff8109914a>] check_prevs_add+0xda/0x140
[ 327.735271] [<ffffffff81099788>] validate_chain+0x5d8/0x700
[ 327.735275] [<ffffffff8109a191>] __lock_acquire+0x461/0x9c0
[ 327.735278] [<ffffffff8109ada0>] lock_acquire+0x90/0x1f0
[ 327.735282] [<ffffffff811393c0>] might_fault+0x80/0xb0
[ 327.735285] [<ffffffff8118b187>] filldir+0x77/0xe0
[ 327.735289] [<ffffffff811eb71a>] call_filldir+0x9a/0xd0
[ 327.735292] [<ffffffff811eb8d2>] ext4_dx_readdir+0x182/0x240
[ 327.735296] [<ffffffff811ebb6f>] ext4_readdir+0x6f/0x3e0
[ 327.735299] [<ffffffff8118b458>] vfs_readdir+0xb8/0xf0
[ 327.735302] [<ffffffff8118b589>] sys_getdents+0x89/0x100
[ 327.735305] [<ffffffff81568beb>] system_call_fastpath+0x16/0x1b
[ 327.735311]
[ 327.735311] -> #0 (&sb->s_type->i_mutex_key#14){+.+.+.}:
[ 327.735316] [<ffffffff81099068>] check_prev_add+0x6b8/0x6c0
[ 327.735319] [<ffffffff8109914a>] check_prevs_add+0xda/0x140
[ 327.735323] [<ffffffff81099788>] validate_chain+0x5d8/0x700
[ 327.735326] [<ffffffff8109a191>] __lock_acquire+0x461/0x9c0
[ 327.735329] [<ffffffff8109ada0>] lock_acquire+0x90/0x1f0
[ 327.735332] [<ffffffff8155e903>] mutex_lock_nested+0x63/0x3a0
[ 327.735337] [<ffffffff811f79a9>] ext4_evict_inode+0x29/0x5c0
[ 327.735340] [<ffffffff811932d3>] evict+0xa3/0x1b0
[ 327.735343] [<ffffffff811934c2>] iput_final+0xe2/0x1e0
[ 327.735346] [<ffffffff811935fe>] iput+0x3e/0x50
[ 327.735349] [<ffffffff8118f028>] dentry_iput+0xa8/0x100
[ 327.735353] [<ffffffff8118fbd3>] dput+0x123/0x220
[ 327.735356] [<ffffffff81179293>] __fput+0x173/0x250
[ 327.735360] [<ffffffff81179395>] fput+0x25/0x30
[ 327.735362] [<ffffffff811402f1>] remove_vma+0x51/0x90
[ 327.735366] [<ffffffff81141743>] do_munmap+0x1f3/0x2f0
[ 327.735369] [<ffffffff81142b86>] sys_munmap+0x56/0x80
[ 327.735373] [<ffffffff81568beb>] system_call_fastpath+0x16/0x1b
[ 327.735377]
[ 327.735377] other info that might help us debug this:
[ 327.735378]
[ 327.735380] Possible unsafe locking scenario:
[ 327.735380]
[ 327.735382] CPU0 CPU1
[ 327.735384] ---- ----
[ 327.735385] lock(&mm->mmap_sem);
[ 327.735388] lock(&sb->s_type->i_mutex_key);
[ 327.735391] lock(&mm->mmap_sem);
[ 327.735394] lock(&sb->s_type->i_mutex_key);
[ 327.735397]
[ 327.735398] *** DEADLOCK ***
[ 327.735399]
[ 327.735401] 1 lock held by kontact/2570:
[ 327.735402] #0: (&mm->mmap_sem){++++++}, at: [<ffffffff81142b78>] sys_munmap+0x48/0x80
[ 327.735408]
[ 327.735409] stack backtrace:
[ 327.735412] Pid: 2570, comm: kontact Not tainted 3.1.0-rc4+ #664
[ 327.735414] Call Trace:
[ 327.735418] [<ffffffff8154b5cf>] print_circular_bug+0xd1/0xe2
[ 327.735423] [<ffffffff81099068>] check_prev_add+0x6b8/0x6c0
[ 327.735427] [<ffffffff8109914a>] check_prevs_add+0xda/0x140
[ 327.735430] [<ffffffff81099788>] validate_chain+0x5d8/0x700
[ 327.735433] [<ffffffff8109a191>] __lock_acquire+0x461/0x9c0
[ 327.735437] [<ffffffff8109ada0>] lock_acquire+0x90/0x1f0
[ 327.735441] [<ffffffff811f79a9>] ? ext4_evict_inode+0x29/0x5c0
[ 327.735444] [<ffffffff811f79a9>] ? ext4_evict_inode+0x29/0x5c0
[ 327.735448] [<ffffffff8155e903>] mutex_lock_nested+0x63/0x3a0
[ 327.735451] [<ffffffff811f79a9>] ? ext4_evict_inode+0x29/0x5c0
[ 327.735454] [<ffffffff8119329f>] ? evict+0x6f/0x1b0
[ 327.735457] [<ffffffff8109ab4d>] ? __lock_release+0xad/0xd0
[ 327.735460] [<ffffffff811932c1>] ? evict+0x91/0x1b0
[ 327.735464] [<ffffffff8104f0b3>] ? get_parent_ip+0x33/0x50
[ 327.735468] [<ffffffff811f79a9>] ext4_evict_inode+0x29/0x5c0
[ 327.735471] [<ffffffff811932d3>] evict+0xa3/0x1b0
[ 327.735474] [<ffffffff811934c2>] iput_final+0xe2/0x1e0
[ 327.735477] [<ffffffff811935fe>] iput+0x3e/0x50
[ 327.735480] [<ffffffff8118f028>] dentry_iput+0xa8/0x100
[ 327.735483] [<ffffffff8118fbd3>] dput+0x123/0x220
[ 327.735486] [<ffffffff81179293>] __fput+0x173/0x250
[ 327.735489] [<ffffffff81179395>] fput+0x25/0x30
[ 327.735492] [<ffffffff811402f1>] remove_vma+0x51/0x90
[ 327.735496] [<ffffffff81141743>] do_munmap+0x1f3/0x2f0
[ 327.735499] [<ffffffff81142b86>] sys_munmap+0x56/0x80
[ 327.735503] [<ffffffff81568beb>] system_call_fastpath+0x16/0x1b
Thanks.
--
Dmitry
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: EXT4: lockdep complaints on 3.1-rc4+
2011-09-20 17:55 EXT4: lockdep complaints on 3.1-rc4+ Dmitry Torokhov
@ 2011-09-20 17:59 ` Josh Boyer
2011-09-20 18:21 ` Dmitry Torokhov
0 siblings, 1 reply; 3+ messages in thread
From: Josh Boyer @ 2011-09-20 17:59 UTC (permalink / raw)
To: Dmitry Torokhov; +Cc: linux-ext4, LKML
On Tue, Sep 20, 2011 at 1:55 PM, Dmitry Torokhov
<dmitry.torokhov@gmail.com> wrote:
> Hi,
>
> Not sure if it has been reported but I just got the following lockdep
> warning on 3.1-rc4+ + my input queue (which does not touch FS code at
> all):
IIRC, that should be fixed with
8c0bec2151a47906bf779c6715a10ce04453ab77, which is in -rc5
josh
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: EXT4: lockdep complaints on 3.1-rc4+
2011-09-20 17:59 ` Josh Boyer
@ 2011-09-20 18:21 ` Dmitry Torokhov
0 siblings, 0 replies; 3+ messages in thread
From: Dmitry Torokhov @ 2011-09-20 18:21 UTC (permalink / raw)
To: Josh Boyer; +Cc: linux-ext4, LKML
On Tue, Sep 20, 2011 at 01:59:05PM -0400, Josh Boyer wrote:
> On Tue, Sep 20, 2011 at 1:55 PM, Dmitry Torokhov
> <dmitry.torokhov@gmail.com> wrote:
> > Hi,
> >
> > Not sure if it has been reported but I just got the following lockdep
> > warning on 3.1-rc4+ + my input queue (which does not touch FS code at
> > all):
>
> IIRC, that should be fixed with
> 8c0bec2151a47906bf779c6715a10ce04453ab77, which is in -rc5
>
Ah, excellent. Sorry for the noise.
--
Dmitry
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2011-09-20 18:21 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-09-20 17:55 EXT4: lockdep complaints on 3.1-rc4+ Dmitry Torokhov
2011-09-20 17:59 ` Josh Boyer
2011-09-20 18:21 ` Dmitry Torokhov
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).