http://oss.sgi.com/bugzilla/show_bug.cgi?id=860
Dave Chinner <david@xxxxxxxxxxxxx> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |david@xxxxxxxxxxxxx
--- Comment #2 from Dave Chinner <david@xxxxxxxxxxxxx> 2010-01-04 22:04:55 CST
---
This looks like a DM lockup. From the output of dmesg you posted,
we have a DM barrier flush waiting for completion:
kdmflush D fa0f2705 0 21539 2 0x00000000
c485dec8 00000046 f09d2000 fa0f2705 c09f0080 d760e57c c09f0080 c09f0080
3c9701db 0000d1b8 c09f0080 c09f0080 d760e57c 00000000 0000d1b8 c09f0080
ef027e40 d760e2f0 6f3bdccc c113c080 ea1daa08 00000002 c485ded8 c069388b
Call Trace:
[<c069388b>] io_schedule+0x5b/0xb0
[<fa0f0818>] dm_wait_for_completion+0x78/0x130 [dm_mod]
[<fa0f1ddd>] dm_flush+0x5d/0x80 [dm_mod]
[<fa0f1e2f>] process_barrier+0x2f/0xd0 [dm_mod]
[<fa0f1f09>] dm_wq_work+0x39/0xe0 [dm_mod]
[<c0268b8b>] run_workqueue+0x7b/0x190
[<c026a6a0>] worker_thread+0x90/0x100
[<c026e5c4>] kthread+0x84/0x90
[<c0204d8b>] kernel_thread_helper+0x7/0x1c
Which was issued by pdflush via an XFS log write:
pdflush D 00000000 0 13231 2 0x00000000
d29d17d8 00000046 f006c000 00000000 c09f0080 d59ae8fc c09f0080 c09f0080
2056bae5 0000d1b8 c09f0080 c09f0080 d59ae8fc 00000000 0000d1b8 c09f0080
f156fc80 d59ae670 4a65614c c9eeb8cc 00001705 00000001 d29d181c fa22536b
Call Trace:
[<fa22536b>] xlog_state_sync+0x12b/0x200 [xfs]
[<fa2254ae>] _xfs_log_force+0x6e/0xa0 [xfs]
[<fa225509>] xfs_log_force+0x29/0x70 [xfs]
[<fa1e6204>] xfs_alloc_search_busy+0xd4/0xe0 [xfs]
[<fa1e812d>] xfs_alloc_ag_vextent+0x13d/0x150 [xfs]
[<fa1e88a9>] xfs_alloc_vextent+0x1e9/0x4c0 [xfs]
[<fa1f9d8c>] xfs_bmap_btalloc+0x1ac/0xb60 [xfs]
[<fa1fa769>] xfs_bmap_alloc+0x29/0x60 [xfs]
[<fa1faf4f>] xfs_bmapi+0x7af/0x1810 [xfs]
[<fa21ef8a>] xfs_iomap_write_allocate+0x23a/0x460 [xfs]
[<fa22024b>] xfs_iomap+0x35b/0x3a0 [xfs]
[<fa23c7ff>] xfs_map_blocks+0x3f/0x60 [xfs]
[<fa23dea2>] xfs_page_state_convert+0x432/0x7f0 [xfs]
[<fa23e3c3>] xfs_vm_writepage+0x73/0x140 [xfs]
[<c02db48a>] __writepage+0x1a/0x50
[<c02dc5e8>] write_cache_pages+0x178/0x3c0
[<c02dc859>] generic_writepages+0x29/0x50
[<fa23d31a>] xfs_vm_writepages+0x4a/0x70 [xfs]
[<c02dc8ae>] do_writepages+0x2e/0x60
[<c0337041>] writeback_single_inode+0x171/0x400
[<c0337685>] generic_sync_sb_inodes+0x155/0x4a0
[<c0337a15>] sync_sb_inodes+0x45/0x60
[<c0337b24>] writeback_inodes+0x54/0xf0
[<c02db812>] wb_kupdate+0xa2/0x120
[<c02dd719>] __pdflush+0xf9/0x200
[<c02dd863>] pdflush+0x43/0x60
[<c026e5c4>] kthread+0x84/0x90
And pdflush also is holding an inode lock, which means that the xfsdatad/0 is
stuck in completion waiting for that lock:
xfsdatad/0 D 00000000 0 813 2 0x00000000
f006dea8 00000046 f09d2000 00000000 c09f0080 f155f58c c09f0080 c09f0080
34a1dcfa 0000d1b8 c09f0080 c09f0080 f155f58c 00000000 0000d1b8 c09f0080
ef027e40 f155f300 ff1bb4c0 ef5e3be4 f006dee0 f155f300 f006ded8 c0695525
Call Trace:
[<c0695525>] rwsem_down_failed_common+0x75/0x1c0
[<c069569a>] rwsem_down_write_failed+0x2a/0x50
[<c069574a>] call_rwsem_down_write_failed+0x6/0x8
[<c0694b54>] down_write+0x24/0x40
[<fa2177fa>] xfs_ilock+0x7a/0xa0 [xfs]
[<fa23c223>] xfs_setfilesize+0x43/0x110 [xfs]
[<fa23c57d>] xfs_end_bio_written+0x1d/0x40 [xfs]
[<c0268b8b>] run_workqueue+0x7b/0x190
[<c026a6a0>] worker_thread+0x90/0x100
[<c026e5c4>] kthread+0x84/0x90
[<c0204d8b>] kernel_thread_helper+0x7/0x1c
and all the other threads in XFS are waiting on IO completion that is blocked
by this. Hence I think this is either a DM bug or you've got flakey hardware
that has dropped an IO. Either way, it does not look like an XFS problem....
--
Configure bugmail: http://oss.sgi.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
|