rcu_kthread hang on 3.0.0-rc1-next-20110602

rcu_kthread hang on 3.0.0-rc1-next-20110602

[Markus Hästbacka]

Hello!

After updating my debian squeeze box to 3.0.0-rc1-next-20110602 I'm experiencing a weird problem. If I run dmesg  the computer hangs for 15 minutes, and I cannot even login at this time, if I leave top (the top window updates very slowly, like every 5 minutes during this period) running I can see that the process rcu_kthread takes 100% CPU, the computer load average is at around 10.

I also sometimes get these disk errors in syslog:
Jun  4 02:01:39 midi3g kernel: ata1.01: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6
Jun  4 02:01:39 midi3g kernel: ata1.01: BMDMA stat 0x44
Jun  4 02:01:39 midi3g kernel: ata1.01: failed command: WRITE DMA
Jun  4 02:01:39 midi3g kernel: ata1.01: cmd ca/00:00:1c:7b:c6/00:00:00:00:00/f0 tag 0 dma 131072 out
Jun  4 02:01:39 midi3g kernel:         res 51/84:00:fe:7b:c6/00:00:00:00:00/f0 Emask 0x10 (ATA bus error)
Jun  4 02:01:39 midi3g kernel: ata1.01: status: { DRDY ERR }
Jun  4 02:01:39 midi3g kernel: ata1.01: error: { ICRC ABRT }
Jun  4 02:01:39 midi3g kernel: ata1: soft resetting link
Jun  4 02:01:39 midi3g kernel: ata1.01: configured for UDMA/66
Jun  4 02:01:39 midi3g kernel: ata1: EH complete

but for example the last time I got the hang, I didn't get this error, nor any other error for that matter.

The OS is installed on a CF-card. The computer is a eBox computer with VIA Nehemiah CPU, 256MB RAM.

what more information do you need? Or what more can I test?

M

Re: rcu_kthread hang on 3.0.0-rc1-next-20110602

[Manuel Lauss]

Hello,

Markus Hästbacka <mhastbacka <at> gmail.com> writes:
> If I
> run dmesg  the computer hangs for 15 minutes, and I cannot even login at this
time, if I leave top (the top
> window updates very slowly, like every 5 minutes during this period) running I
can see that the process
> rcu_kthread takes 100% CPU, the computer load average is at around 10.

> but for example the last time I got the hang, I didn't get this error, nor any
other error for that matter.
> 
> The OS is installed on a CF-card. The computer is a eBox computer with VIA
Nehemiah CPU, 256MB RAM.
> 
> what more information do you need? Or what more can I test?

I too experience this weird 15 to 20-minute hang, but only on non-preempt
uniprocessor kernels; it still responds to pings, though.
With either preempt or SMP enabled it's not as visible or simply gone.

Manuel

Re: rcu_kthread hang on 3.0.0-rc1-next-20110602

[Markus Hästbacka]

Hi!

I compiled the kernel with preempt and the problem is indeed gone, the hd errors are still present from time to time, but thats probably a hardware problem.

M

On Jun 4, 2011, at 9:17 AM, Manuel Lauss wrote:

> Hello,
> 
> Markus Hästbacka <mhastbacka <at> gmail.com> writes:
>> If I
>> run dmesg  the computer hangs for 15 minutes, and I cannot even login at this
> time, if I leave top (the top
>> window updates very slowly, like every 5 minutes during this period) running I
> can see that the process
>> rcu_kthread takes 100% CPU, the computer load average is at around 10.
> 
>> but for example the last time I got the hang, I didn't get this error, nor any
> other error for that matter.
>> 
>> The OS is installed on a CF-card. The computer is a eBox computer with VIA
> Nehemiah CPU, 256MB RAM.
>> 
>> what more information do you need? Or what more can I test?
> 
> I too experience this weird 15 to 20-minute hang, but only on non-preempt
> uniprocessor kernels; it still responds to pings, though.
> With either preempt or SMP enabled it's not as visible or simply gone.
> 
> Manuel
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/