I've not seen this reported...

[ 9564.772749] 
[ 9564.772752] =======================================================
[ 9564.772757] [ INFO: possible circular locking dependency detected ]
[ 9564.772760] 2.6.21-rc6-gc2481cc4 #9
[ 9564.772762] -------------------------------------------------------
[ 9564.772765] xfssyncd/6157 is trying to acquire lock:
[ 9564.772767]  (&(&ip->i_lock)->mr_lock){----}, at: [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.772775] 
[ 9564.772776] but task is already holding lock:
[ 9564.772778]  (&mp->m_ilock){--..}, at: [<ffffffff803348f6>] xfs_finish_reclaim_all+0x1d/0xc1
[ 9564.772785] 
[ 9564.772786] which lock already depends on the new lock.
[ 9564.772787] 
[ 9564.772789] 
[ 9564.772790] the existing dependency chain (in reverse order) is:
[ 9564.772792] 
[ 9564.772793] -> #1 (&mp->m_ilock){--..}:
[ 9564.772796]        [<ffffffff802478f1>] __lock_acquire+0xa3d/0xbe4
[ 9564.772804]        [<ffffffff8031b217>] xfs_iget_core+0x60e/0x6c4
[ 9564.772811]        [<ffffffff80247b13>] lock_acquire+0x7b/0x9f
[ 9564.772817]        [<ffffffff8031b217>] xfs_iget_core+0x60e/0x6c4
[ 9564.772824]        [<ffffffff804f6f19>] __mutex_lock_slowpath+0xe1/0x263
[ 9564.772832]        [<ffffffff8031b217>] xfs_iget_core+0x60e/0x6c4
[ 9564.772839]        [<ffffffff8031b371>] xfs_iget+0xa4/0x14b
[ 9564.772845]        [<ffffffff8032c9b8>] xfs_mountfs+0x708/0x954
[ 9564.772852]        [<ffffffff8033b37f>] xfs_setsize_buftarg_flags+0x30/0x9e
[ 9564.772859]        [<ffffffff80332a67>] xfs_mount+0x317/0x39d
[ 9564.772866]        [<ffffffff803427ff>] xfs_fs_fill_super+0x0/0x1a1
[ 9564.772873]        [<ffffffff8034287d>] xfs_fs_fill_super+0x7e/0x1a1
[ 9564.772880]        [<ffffffff804f8586>] _spin_unlock+0x17/0x20
[ 9564.772887]        [<ffffffff80277b65>] sget+0x377/0x389
[ 9564.772894]        [<ffffffff802774e0>] set_bdev_super+0x0/0xf
[ 9564.772900]        [<ffffffff802774ef>] test_bdev_super+0x0/0xd
[ 9564.772907]        [<ffffffff802784b2>] get_sb_bdev+0x121/0x17b
[ 9564.772913]        [<ffffffff80277ff5>] vfs_kern_mount+0x4f/0x8a
[ 9564.772920]        [<ffffffff80278072>] do_kern_mount+0x36/0x4d
[ 9564.772926]        [<ffffffff8028b3d6>] do_mount+0x625/0x699
[ 9564.772934]        [<ffffffff80289e83>] mntput_no_expire+0x1c/0x85
[ 9564.772941]        [<ffffffff8027ef8e>] link_path_walk+0xce/0xe0
[ 9564.772948]        [<ffffffff80236f20>] sigprocmask+0x28/0xc1
[ 9564.772956]        [<ffffffff80272a18>] kmem_cache_free+0xcf/0xd9
[ 9564.772963]        [<ffffffff80246a13>] trace_hardirqs_on+0x124/0x14f
[ 9564.772970]        [<ffffffff8025c17d>] get_page_from_freelist+0x1ef/0x35d
[ 9564.772977]        [<ffffffff80246a13>] trace_hardirqs_on+0x124/0x14f
[ 9564.772984]        [<ffffffff8025c358>] __alloc_pages+0x6d/0x2b4
[ 9564.772990]        [<ffffffff8028b4d4>] sys_mount+0x8a/0xd7
[ 9564.772997]        [<ffffffff8020962e>] system_call+0x7e/0x83
[ 9564.773004]        [<ffffffffffffffff>] 0xffffffffffffffff
[ 9564.773014] 
[ 9564.773015] -> #0 (&(&ip->i_lock)->mr_lock){----}:
[ 9564.773018]        [<ffffffff8024611f>] print_circular_bug_header+0xcc/0xd3
[ 9564.773025]        [<ffffffff802477ed>] __lock_acquire+0x939/0xbe4
[ 9564.773032]        [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.773039]        [<ffffffff80247b13>] lock_acquire+0x7b/0x9f
[ 9564.773045]        [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.773052]        [<ffffffff80242d08>] down_write_trylock+0x2f/0x35
[ 9564.773059]        [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.773066]        [<ffffffff80334919>] xfs_finish_reclaim_all+0x40/0xc1
[ 9564.773073]        [<ffffffff80332567>] xfs_syncsub+0x4e/0x237
[ 9564.773080]        [<ffffffff8023ff12>] keventd_create_kthread+0x0/0x6d
[ 9564.773087]        [<ffffffff80341e68>] vfs_sync_worker+0x19/0x38
[ 9564.773094]        [<ffffffff80342157>] xfssyncd+0x10c/0x15d
[ 9564.773101]        [<ffffffff8034204b>] xfssyncd+0x0/0x15d
[ 9564.773107]        [<ffffffff80240182>] kthread+0xd1/0x103
[ 9564.773114]        [<ffffffff8020a4b8>] child_rip+0xa/0x12
[ 9564.773121]        [<ffffffff804f884c>] _spin_unlock_irq+0x24/0x27
[ 9564.773127]        [<ffffffff80209bcc>] restore_args+0x0/0x30
[ 9564.773134]        [<ffffffff802400b1>] kthread+0x0/0x103
[ 9564.773140]        [<ffffffff8020a4ae>] child_rip+0x0/0x12
[ 9564.773147]        [<ffffffffffffffff>] 0xffffffffffffffff
[ 9564.773153] 
[ 9564.773154] other info that might help us debug this:
[ 9564.773155] 
[ 9564.773158] 1 lock held by xfssyncd/6157:
[ 9564.773159]  #0:  (&mp->m_ilock){--..}, at: [<ffffffff803348f6>] xfs_finish_reclaim_all+0x1d/0xc1
[ 9564.773166] 
[ 9564.773167] stack backtrace:
[ 9564.773169] 
[ 9564.773169] Call Trace:
[ 9564.773173]  [<ffffffff80245d6a>] print_circular_bug_tail+0x69/0x72
[ 9564.773177]  [<ffffffff8024611f>] print_circular_bug_header+0xcc/0xd3
[ 9564.773180]  [<ffffffff802477ed>] __lock_acquire+0x939/0xbe4
[ 9564.773183]  [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.773187]  [<ffffffff80247b13>] lock_acquire+0x7b/0x9f
[ 9564.773190]  [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.773194]  [<ffffffff80242d08>] down_write_trylock+0x2f/0x35
[ 9564.773197]  [<ffffffff8031abae>] xfs_ilock_nowait+0x6d/0xc8
[ 9564.773200]  [<ffffffff80334919>] xfs_finish_reclaim_all+0x40/0xc1
[ 9564.773204]  [<ffffffff80332567>] xfs_syncsub+0x4e/0x237
[ 9564.773207]  [<ffffffff8023ff12>] keventd_create_kthread+0x0/0x6d
[ 9564.773211]  [<ffffffff80341e68>] vfs_sync_worker+0x19/0x38
[ 9564.773214]  [<ffffffff80342157>] xfssyncd+0x10c/0x15d
[ 9564.773217]  [<ffffffff8034204b>] xfssyncd+0x0/0x15d
[ 9564.773220]  [<ffffffff80240182>] kthread+0xd1/0x103
[ 9564.773224]  [<ffffffff8020a4b8>] child_rip+0xa/0x12
[ 9564.773227]  [<ffffffff804f884c>] _spin_unlock_irq+0x24/0x27
[ 9564.773230]  [<ffffffff80209bcc>] restore_args+0x0/0x30
[ 9564.773233]  [<ffffffff802400b1>] kthread+0x0/0x103
[ 9564.773236]  [<ffffffff8020a4ae>] child_rip+0x0/0x12
[ 9564.773238] 


-- 
	Paolo Ornati
	Linux 2.6.21-rc6-gc2481cc4 on x86_64