Re: [PATCH] cpufreq: Set cpufreq_cpu_data to NULL before putting kobject

[ethan zhao <ethan.zhao@oracle.com>]

Viresh,

On 2015/1/30 10:05, Viresh Kumar wrote:
> On 30 January 2015 at 07:00, ethan zhao <ethan.zhao@oracle.com> wrote:
>
>>   It is another bug.
> Hmm, lets see..
>
>>   Yes, here is one bug should be fix. but seems not enough to avoid the issue
>> completely,
> Did you test it ?
For a PPC notification and xen-bus thread race, could you tell me a way how
to reproduce it by trigger the PPC notification and xen-bus events 
manually ?
You really want me write some code into a test kernel to flood the PPC 
and xen-bus at the same time ? if we could analysis code and get the 
issue clearly,
we wouldn't wait the users to yell out.

Thanks,
Ethan

>
>>   how about the Thread B running here
>>
>>   Thread B: xenbus_thread()
>>
>>   xenbus_thread()
>>     msg->u.watch.handle->callback()
>>       handle_vcpu_hotplug_event()
>>         vcpu_hotplug()
>>           cpu_down()
>>             __cpu_notify(CPU_POST_DEAD..)
>>               cpufreq_cpu_callback()
>>                 __cpufreq_remove_dev_prepare
>>                   update_policy_cpu(){
>>                     ...
>>                     down_write(&policy->rwsem);
>>                     policy->last_cpu = policy->cpu;
>>                     policy->cpu = cpu;
>>                     up_write(&policy->rwsem);
>>                     ---->
>>                  }
>>
>> And thread A run to the same position as above, don't ignore my work
>> blindly, that piece of bandage
>> could save your time
> Oh, please don't misunderstand me. I didn't had any intention of showing
> any kind of disrespect.
>
> Okay, why do you say that the above thread you shown has a bug in there?
> Its juse updating policy->cpu and that shouldn't make anything unstable.
>
> Please explain again one more time, in details why do you think you hit a
> different bug. Also look at kref.h to see which piece of code has hit that
> WARN() and it will happen only in the case I have shown. Lets see if I
> missed something :)
>
> --
> viresh