dann frazier wrote:
On Thu, Aug 06, 2009 at 08:24:40AM +0100, Berni Elbourn wrote:
See also Bug #517449
Note that soft lockups are a class of bug, they can have many
causes. If you think this is related to the SCHED_IDLE issue, please
let us know why - I don't see a relation myself.
I am sorry if this is noise.
I arrived here due to a reference in #517449. #517449 seems to relate to
virtual hosts. To my untrained eye the logs here look similar...
Here is an detail example of the log round the pdflush task hangup:
Aug 6 06:26:46 sv28 kernel: [1300761.990362] INFO: task pdflush:18830
blocked for more than 120 seconds.
Aug 6 06:26:46 sv28 kernel: [1300761.990362] "echo 0 >
/proc/sys/kernel/hung_task_timeout_secs" disables this message.
Aug 6 06:26:46 sv28 kernel: [1300761.990362] pdflush D
0000000000000000 0 18830 2
Aug 6 06:26:46 sv28 kernel: [1300761.990362] ffff81008bdf7a50
0000000000000046 0000000000000000 ffffffff802284eb
Aug 6 06:26:46 sv28 kernel: [1300761.990362] ffff8102246e15f0
ffff8102270b1470 ffff8102246e1878 0000000100000001
Aug 6 06:26:46 sv28 kernel: [1300761.990362] 0000000000000282
0000000000000000 00000000ffffffff 0000000000000000
Aug 6 06:26:46 sv28 kernel: [1300761.990362] Call Trace:
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802284eb>]
__wake_up_common+0x41/0x74
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffffa013c360>]
:raid1:wait_barrier+0x87/0xc8
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8022c257>]
default_wake_function+0x0/0xe
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffffa013dcfd>]
:raid1:make_request+0x73/0x58e
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8031da6b>]
__up_read+0x13/0x8a
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8030c46e>]
generic_make_request+0x2fe/0x339
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff80273374>]
mempool_alloc+0x24/0xda
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8030d82c>]
submit_bio+0xdb/0xe2
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802b9fb6>]
submit_bh+0xd9/0xf9
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802bb9f0>]
__block_write_full_page+0x1bd/0x2ac
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802bf6c1>]
blkdev_get_block+0x0/0x46
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802770c7>]
__writepage+0xa/0x23
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8027758c>]
write_cache_pages+0x182/0x2b1
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802770bd>]
__writepage+0x0/0x23
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802776f7>]
do_writepages+0x20/0x2d
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802b6450>]
__writeback_single_inode+0x144/0x29d
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffffa01459c1>]
:dm_mod:dm_table_any_congested+0x46/0x69
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802b6914>]
sync_sb_inodes+0x1b1/0x293
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802b6d42>]
writeback_inodes+0x62/0xb3
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff80278053>]
wb_kupdate+0x9e/0x114
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff80278422>]
pdflush+0x0/0x211
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff80278586>]
pdflush+0x164/0x211
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff80277fb5>]
wb_kupdate+0x0/0x114
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8024608b>]
kthread+0x47/0x74
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff802301ec>]
schedule_tail+0x27/0x5c
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8020cf28>]
child_rip+0xa/0x12
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8021a82a>]
lapic_next_event+0xf/0x13
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff80246044>]
kthread+0x0/0x74
Aug 6 06:26:46 sv28 kernel: [1300761.990362] [<ffffffff8020cf1e>]
child_rip+0x0/0x12
And cat /proc/version:
Linux version 2.6.26-2-amd64 (Debian 2.6.26-15) ([email protected]) (gcc
version 4.1.3 20080704 (prerelease) (Debian 4.1.2-25)) #1 SMP Fri Mar 27
04:02:59 UTC 2009
Cheers,
Berni
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
