Christian Kujau
2010-Apr-09 21:19 UTC
2.6.34-rc3: task btrfs-transacti:4083 blocked for more than 120 seconds.
Hi,
while running some filesystem benchmarks[0] I noticed the following
message in my logs during bonnie++:
device fsid 944150ad12159fd6-cc6b5d7368bfb90 devid 1 transid 7 /dev/md0
INFO: task btrfs-transacti:4083 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this
message.
btrfs-transac D 0000000000000001 0 4083 2 0x00000000
ffff8801feb0dce0 0000000000000046 ffff8801feb0dc50 ffffffff81060e1d
00000000001d2240 00000000001d2240 ffff8801feb0dfd8 ffff8801feb0dfd8
ffff8801feb8b930 00000000001d2240 ffff8801feb0dfd8 00000000001d2240
Call Trace:
[<ffffffff81060e1d>] ? trace_hardirqs_off+0xd/0x10
[<ffffffff814cd055>] schedule_timeout+0x1c5/0x220
[<ffffffff814cd489>] ? mutex_unlock+0x9/0x10
[<ffffffff812e9414>] ? btrfs_run_ordered_operations+0x224/0x240
[<ffffffff8106467d>] ? trace_hardirqs_on+0xd/0x10
[<ffffffff812cf2fd>] btrfs_commit_transaction+0x24d/0x6f0
[<ffffffff81051440>] ? autoremove_wake_function+0x0/0x40
[<ffffffff812cfac0>] ? start_transaction+0x120/0x170
[<ffffffff812c9edb>] transaction_kthread+0x1fb/0x230
[<ffffffff812c9ce0>] ? transaction_kthread+0x0/0x230
[<ffffffff81050f9e>] kthread+0xae/0xc0
[<ffffffff81064625>] ? trace_hardirqs_on_caller+0x155/0x1a0
[<ffffffff81003b54>] kernel_thread_helper+0x4/0x10
[<ffffffff814cfa3c>] ? restore_args+0x0/0x30
[<ffffffff81050ef0>] ? kthread+0x0/0xc0
[<ffffffff81003b50>] ? kernel_thread_helper+0x0/0x10
1 lock held by btrfs-transacti/4083:
#0: (&fs_info->transaction_kthread_mutex){+.+...}, at:
[<ffffffff812c9dc5>] transaction_kthread+0xe5/0x230
I''ve run bonnie++ with earlier kernel version (2.6.33, etc) on this
very
same machine w/o thoese messages. Although bonnie++ has been
upgraded, I''m seeing the same when running tiobench on the same
machine:
device fsid 9b4ce407187a1f20-31cd51397b5b52bf devid 1 transid 7 /dev/md0
INFO: task btrfs-transacti:10530 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this
message.
btrfs-transac D 0000000000000001 0 10530 2 0x00000000
ffff8800b147bce0 0000000000000046 ffff8800b147bc50 ffffffff81060dfd
00000000001d2240 00000000001d2240 ffff8800b147bfd8 ffff8800b147bfd8
ffff880009f09310 00000000001d2240 ffff8800b147bfd8 00000000001d2240
Call Trace:
[<ffffffff81060dfd>] ? trace_hardirqs_off+0xd/0x10
[<ffffffff814ccf55>] schedule_timeout+0x1c5/0x220
[<ffffffff814cd389>] ? mutex_unlock+0x9/0x10
[<ffffffff812e9394>] ? btrfs_run_ordered_operations+0x224/0x240
[<ffffffff8106465d>] ? trace_hardirqs_on+0xd/0x10
[<ffffffff812cf27d>] btrfs_commit_transaction+0x24d/0x6f0
[<ffffffff81051420>] ? autoremove_wake_function+0x0/0x40
[<ffffffff812cfa40>] ? start_transaction+0x120/0x170
[<ffffffff812c9e5b>] transaction_kthread+0x1fb/0x230
[<ffffffff812c9c60>] ? transaction_kthread+0x0/0x230
[<ffffffff81050f7e>] kthread+0xae/0xc0
[<ffffffff81064605>] ? trace_hardirqs_on_caller+0x155/0x1a0
[<ffffffff81003b54>] kernel_thread_helper+0x4/0x10
[<ffffffff814cf93c>] ? restore_args+0x0/0x30
[<ffffffff81050ed0>] ? kthread+0x0/0xc0
[<ffffffff81003b50>] ? kernel_thread_helper+0x0/0x10
1 lock held by btrfs-transacti/10530:
#0: (&fs_info->transaction_kthread_mutex){+.+...}, at:
[<ffffffff812c9d45>] transaction_kthread+0xe5/0x230
Is this something to worry about or should I just go ahead and set
hung_task_timeout_secs to 0? Running the same version of bonnie++ on other
filesystems (on the same machine) does not prodcude this messages (well,
the other fs have different issues).
I''ve put a few details about this online:
http://nerdbynature.de/bits/2.6.34-rc3/btrfs/
http://nerdbynature.de/bits/2.6.34-rc3/err/
Christian.
[0] http://nerdbynature.de/benchmarks/v40z/2010-04-06/
--
BOFH excuse #428:
Firmware update in the coffee machine
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs"
in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html