2008-07-08 09:20:57

by Jesper Krogh

Subject: 2.6.26-rc5 call trace

Hi

I'm getting this call trace on the server: 2.6.26-rc5

[4571608.304232] INFO: task nfsd:7747 blocked for more than 120 seconds.
[4571608.304232] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
disables this message.
[4571608.304232] nfsd D ffff810303a20300 0 7747 2
[4571608.304232] ffff8102faa49d18 0000000000000046 0000000000004000
0000000000000001
[4571608.304232] ffffffff8069c300 ffffffff8069c300 ffffffff80698480
ffffffff8069c300
[4571608.304232] ffff8102fb5d4ee0 0000000000000000 ffffffff802b89f0
fffffffffffffffe
[4571608.304232] Call Trace:
[4571608.304232] [<ffffffff802b89f0>] permission+0xb0/0x160
[4571608.304232] [<ffffffffa03d02cf>] :nfsd:nfsd_permission+0x9f/0x130
[4571608.304232] [<ffffffffa03cf335>] :nfsd:fh_verify+0x105/0x530
[4571608.304232] [<ffffffff80476051>] __mutex_lock_slowpath+0x81/0xd0
[4571608.304232] [<ffffffff80475e3a>] mutex_lock+0xa/0x10
[4571608.304232] [<ffffffffa03d1f12>] :nfsd:nfsd_lookup_dentry+0x182/0x4b0
[4571608.304232] [<ffffffffa03d2270>] :nfsd:nfsd_lookup+0x30/0xd0
[4571608.304232] [<ffffffffa03da711>] :nfsd:nfsd3_proc_lookup+0xa1/0x120
[4571608.304232] [<ffffffffa03cc271>] :nfsd:nfsd_dispatch+0xb1/0x240
[4571608.304232] [<ffffffffa03782ca>] :sunrpc:svc_process+0x48a/0x7f0
[4571608.304232] [<ffffffff8022bd90>] default_wake_function+0x0/0x10
[4571608.304232] [<ffffffff80476f02>] __down_read+0x12/0xad
[4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
[4571608.304232] [<ffffffffa03cc99b>] :nfsd:nfsd+0x18b/0x2e0
[4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
[4571608.304232] [<ffffffff8020d0e8>] child_rip+0xa/0x12
[4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
[4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
[4571608.304232] [<ffffffff8020d0de>] child_rip+0x0/0x12

$ dmesg | grep -c INFO
73

Jesper

--
Jesper Krogh

2008-07-08 14:41:57

by J. Bruce Fields

Subject: Re: 2.6.26-rc5 call trace

On Tue, Jul 08, 2008 at 11:20:39AM +0200, Jesper Krogh wrote:
> Hi
>
> I'm getting this call trace on the server: 2.6.26-rc5
>
> [4571608.304232] INFO: task nfsd:7747 blocked for more than 120 seconds.
> [4571608.304232] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs"
> disables this message.
> [4571608.304232] nfsd D ffff810303a20300 0 7747 2
> [4571608.304232] ffff8102faa49d18 0000000000000046 0000000000004000
> 0000000000000001
> [4571608.304232] ffffffff8069c300 ffffffff8069c300 ffffffff80698480
> ffffffff8069c300
> [4571608.304232] ffff8102fb5d4ee0 0000000000000000 ffffffff802b89f0
> fffffffffffffffe
> [4571608.304232] Call Trace:
> [4571608.304232] [<ffffffff802b89f0>] permission+0xb0/0x160
> [4571608.304232] [<ffffffffa03d02cf>] :nfsd:nfsd_permission+0x9f/0x130
> [4571608.304232] [<ffffffffa03cf335>] :nfsd:fh_verify+0x105/0x530
> [4571608.304232] [<ffffffff80476051>] __mutex_lock_slowpath+0x81/0xd0
> [4571608.304232] [<ffffffff80475e3a>] mutex_lock+0xa/0x10
> [4571608.304232] [<ffffffffa03d1f12>] :nfsd:nfsd_lookup_dentry+0x182/0x4b0
> [4571608.304232] [<ffffffffa03d2270>] :nfsd:nfsd_lookup+0x30/0xd0
> [4571608.304232] [<ffffffffa03da711>] :nfsd:nfsd3_proc_lookup+0xa1/0x120
> [4571608.304232] [<ffffffffa03cc271>] :nfsd:nfsd_dispatch+0xb1/0x240
> [4571608.304232] [<ffffffffa03782ca>] :sunrpc:svc_process+0x48a/0x7f0
> [4571608.304232] [<ffffffff8022bd90>] default_wake_function+0x0/0x10
> [4571608.304232] [<ffffffff80476f02>] __down_read+0x12/0xad
> [4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
> [4571608.304232] [<ffffffffa03cc99b>] :nfsd:nfsd+0x18b/0x2e0
> [4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
> [4571608.304232] [<ffffffff8020d0e8>] child_rip+0xa/0x12
> [4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
> [4571608.304232] [<ffffffffa03cc810>] :nfsd:nfsd+0x0/0x2e0
> [4571608.304232] [<ffffffff8020d0de>] child_rip+0x0/0x12
>
> $ dmesg | grep -c INFO
> 73

Thanks for the report!

What filesystem are you exporting, with what export options? And with
which kernel version did you start seeing this?

--b.

2008-07-08 15:38:16

by Jesper Krogh

Subject: Re: 2.6.26-rc5 call trace

> What filesystem are you exporting, with what export options? And with
> which kernel version did you start seeing this?

ext3 filesystems with these exports:
*(rw,async,no_subtree_check)

I havent seen it running -rc4 but the trace first arrived after 29 days of
uptime on -rc5 and i dont think I have run -rc4 that long.

Jesper

--
Jesper Krogh