Re: INFO: task blocked for more than 120 seconds

[Aneesh Kumar K.V]

On Mon, Aug 11, 2008 at 11:27:12AM -0700, Randy Dunlap wrote:
> On 2.6.27-rc2-git4 and several previous kernels, I see several
> of these messages.  E.g.:
> 
> INFO: task kjournald:665 blocked for more than 120 seconds.
> INFO: task stress:17797 blocked for more than 120 seconds.
> INFO: task stress:17805 blocked for more than 120 seconds.
> 
> 
> Has anyone tracked this down?  Should I attempt to bisect it?
> (on x86_64, SMP, 8 GB RAM)
> 
> 
> 
> INFO: task kjournald:665 blocked for more than 120 seconds.
> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> kjournald     D ffff88027e04be30  4592   665      2
>  ffff88027e04bdd0 0000000000000046 ffff88027e04bd90 ffffffff8022b5f8
>  ffff88027e703090 ffff880178c91bc0 ffff88027e7033d0 0000000178c91c08
>  ffff88027e04bdb0 ffff88027e04be30 ffff88017eaf80f0 0000000000000246
> Call Trace:
>  [<ffffffff8022b5f8>] ? __wake_up_common+0x41/0x74
>  [<ffffffff802f6eef>] journal_commit_transaction+0xe9/0xd7e
>  [<ffffffff8023db06>] ? lock_timer_base+0x26/0x4a
>  [<ffffffff80247240>] ? autoremove_wake_function+0x0/0x38
>  [<ffffffff8023db80>] ? try_to_del_timer_sync+0x56/0x62
>  [<ffffffff802fa388>] kjournald+0xc3/0x1fb
>  [<ffffffff80247240>] ? autoremove_wake_function+0x0/0x38
>  [<ffffffff802fa2c5>] ? kjournald+0x0/0x1fb
>  [<ffffffff80247107>] kthread+0x49/0x76
>  [<ffffffff8020ce39>] child_rip+0xa/0x11
>  [<ffffffff802470be>] ? kthread+0x0/0x76
>  [<ffffffff8020ce2f>] ? child_rip+0x0/0x11
> 
> INFO: task stress:17797 blocked for more than 120 seconds.
> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> stress        D ffff88017eaf8024  5088 17797  17795
>  ffff8801f4055cd8 0000000000000082 0000000000000086 ffff88027e04bec0
>  ffff880178c93090 ffff88017faf75f0 ffff880178c933d0 0000000300000001
>  0000000000000292 ffff8801f4055ce8 ffff88017eaf80a8 0000000000000246
> Call Trace:
>  [<ffffffff802f9c04>] log_wait_commit+0xa4/0xf4
>  [<ffffffff80247240>] ? autoremove_wake_function+0x0/0x38
>  [<ffffffff802f5798>] journal_stop+0x17c/0x1a9
>  [<ffffffff802f5fe6>] journal_force_commit+0x23/0x25
>  [<ffffffff802eee53>] ext3_force_commit+0x26/0x28
>  [<ffffffff802e91d2>] ext3_write_inode+0x39/0x3f
>  [<ffffffff802b58cf>] __writeback_single_inode+0x180/0x284
>  [<ffffffff80247278>] ? wake_bit_function+0x0/0x2a
>  [<ffffffff802b5db1>] generic_sync_sb_inodes+0x1c3/0x29e
>  [<ffffffff802b5e95>] sync_sb_inodes+0x9/0xb
>  [<ffffffff802b5f2c>] sync_inodes_sb+0x95/0x9c
>  [<ffffffff802b5f95>] __sync_inodes+0x62/0xaf
>  [<ffffffff802b6010>] sync_inodes+0x2e/0x33
>  [<ffffffff802b8908>] do_sync+0x34/0x59
>  [<ffffffff802b893b>] sys_sync+0xe/0x13
>  [<ffffffff8020beeb>] system_call_fastpath+0x16/0x1b
> 

Committing a transaction would means writing rest of the meta-data in
the transaction. And that would imply forcing most of the buffer_heads
to disk in ordered mode. This can result a lot of seeks and make take
more thatn 120 seconds.


-aneesh
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/