Re: 2.6.30-git(16 and 17) system hangs after resume from suspend to disk, mce related?

[Andi Kleen <ak@linux.intel.com>]

Hidetoshi Seto wrote:
> Maciej Rutecki wrote:
>>> Also a "a few minutes" suggest something might be going wrong
>>> with the poll handler.  Does the problem still happen
>>> with you use CONFIG_X86_NEW_MCE again, but before
>>> resume do
>>>
>>> echo 0 > /sys/device/system/machinecheck/machinecheck0/check_interval
>>>
>>> On the other hand you should get a crash very fast with
>>>
>>> echo 1 > /sys/device/system/machinecheck/machinecheck0/check_interval
>> I didn't instructions from above, but I found something else. After
>> normal boot I try:
>>
>> echo 1 > /sys/devices/system/machinecheck/machinecheck0/check_interval
>>
>> I I found this in dmesg:
>>
>> [  141.704025] ------------[ cut here ]------------
>> [  141.704039] WARNING: at arch/x86/kernel/cpu/mcheck/mce.c:1102
>> mcheck_timer+0xf5/0x100()
> 
> I see.  At least this warning will be cleared by following patch.
>   WARN_ON(smp_processor_id() != data);
> 
> But I'm not sure whether this can cause system hangs or not.

It might actually. If two different handlers run on the same CPU
they could re-add a timer twice, which might cause loops in the timer
list etc.

Maciej, can you test Seto-san's patch please?

BTW this is probably related to

commit eea08f32adb3f97553d49a4f79a119833036000a
Author: Arun R Bharadwaj <arun@linux.vnet.ibm.com>
Date:   Thu Apr 16 12:16:41 2009 +0530

     timers: Logic to move non pinned timers

it might be also useful to test if reverting that patch makes
the problem go away. But with this patch we need the add_timer_on change.

-Andi