2.6.19-rc6 -- dm: possible recursive locking detected

2.6.19-rc6 -- dm: possible recursive locking detected

[Heiko Carstens]

Sent this a few days ago, but got no response. Probably because I addressed
this to the wrong persons. So, next try.

With current git tree as of today I get this:

=============================================
[ INFO: possible recursive locking detected ]
2.6.19-rc6-g1b9bb3c1 #28
---------------------------------------------
kpartx/945 is trying to acquire lock:
 (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c

but task is already holding lock:
 (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c

other info that might help us debug this:
1 lock held by kpartx/945:
 #0:  (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c

stack backtrace:
0000000000000000 0000000000000000 0000000000000000 0000000000000000 
       0000000000000000 00000000003a4d68 00000000003a4d68 0000000000016272 
       040000000d983280 00000000006d93e8 0000000000000000 0000000000477620 
       0000000000000000 000000000000000d 000000000d983380 000000000d9833f8 
       0000000000372450 0000000000016272 000000000d983380 000000000d9833d0 
Call Trace:
([<00000000000161d8>] show_trace+0xc0/0xdc)
 [<0000000000016294>] show_stack+0xa0/0xd0
 [<00000000000162f2>] dump_stack+0x2e/0x3c
 [<0000000000063244>] __lock_acquire+0xa80/0xeac
 [<0000000000063c10>] lock_acquire+0x90/0xc0
 [<000000000005d540>] down_read+0x54/0x9c
 [<00000000001f8faa>] dm_request+0x42/0x24c
 [<000000000017e528>] generic_make_request+0x174/0x1fc
 [<00000000001f88ee>] __map_bio+0x7e/0xec
 [<00000000001f8f44>] __split_bio+0x510/0x534
 [<00000000001f915c>] dm_request+0x1f4/0x24c
 [<000000000017e528>] generic_make_request+0x174/0x1fc
 [<000000000017e672>] submit_bio+0xc2/0x18c
 [<00000000000d7a00>] submit_bh+0x13c/0x1d8
 [<00000000000da95c>] block_read_full_page+0x3f0/0x470
 [<00000000000dd2e0>] blkdev_readpage+0x30/0x40
 [<0000000000084638>] __do_page_cache_readahead+0x190/0x300
 [<0000000000084976>] blockable_page_cache_readahead+0x7a/0x104
 [<0000000000084db4>] page_cache_readahead+0x29c/0x328
 [<000000000007c14c>] do_generic_mapping_read+0x428/0x51c
 [<000000000007e8fc>] generic_file_aio_read+0x160/0x244
 [<00000000000a9cac>] do_sync_read+0xd8/0x130
 [<00000000000a9dac>] vfs_read+0xa8/0x188
 [<00000000000aa19a>] sys_read+0x56/0x88
 [<0000000000020d40>] sysc_noemu+0x10/0x16

Re: 2.6.19-rc6 -- dm: possible recursive locking detected

[Alasdair G Kergon]

On Fri, Nov 17, 2006 at 04:30:03PM +0100, Heiko Carstens wrote:
> =============================================
> [ INFO: possible recursive locking detected ]
> 2.6.19-rc6-g1b9bb3c1 #28
> ---------------------------------------------
> kpartx/945 is trying to acquire lock:
>  (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c
> 
> but task is already holding lock:
>  (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c
 
Known problem - nobody's written a patch to fix it yet to my knowledge,
but I don't recall any reports of real world lock-ups related to this - just
several reports of the debug warning messages:-)

Simply ignore the message for now - or submit a patch that fixes the
underlying problem:-)

Alasdair
-- 
agk@redhat.com

Re: 2.6.19-rc6 -- dm: possible recursive locking detected

[Heiko Carstens]

On Fri, Nov 17, 2006 at 07:43:03PM +0000, Alasdair G Kergon wrote:
> On Fri, Nov 17, 2006 at 04:30:03PM +0100, Heiko Carstens wrote:
> > =============================================
> > [ INFO: possible recursive locking detected ]
> > 2.6.19-rc6-g1b9bb3c1 #28
> > ---------------------------------------------
> > kpartx/945 is trying to acquire lock:
> >  (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c
> > 
> > but task is already holding lock:
> >  (&md->io_lock){----}, at: [<00000000001f8faa>] dm_request+0x42/0x24c
> 
> Known problem - nobody's written a patch to fix it yet to my knowledge,
> but I don't recall any reports of real world lock-ups related to this - just
> several reports of the debug warning messages:-)
> 
> Simply ignore the message for now - or submit a patch that fixes the
> underlying problem:-)

Ignoring would be good, but this problem renders lockdep useless since it
stops working after it found the first possible bug.

While we are at it:

this one is sitting since ages in -mm and probably waiting for your approval:
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.19-rc5/2.6.19-rc5-mm2/broken-out/md-dm-reduce-stack-usage-with-stacked-block-devices.patch