Re: [PATCH 3/3 -mm] generic-ipi: fix the race between generic_smp_call_function_*() and hotplug_cfd()

[Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>]

Peter Zijlstra wrote:

> It introduces this run_callbacks thing to two functions, but nothing
> actually uses that... makes me suspicious there's something missing.
> 

Hi Peter,

Thanks for your point out.
I'm sorry for making a mistake when I make this patch, I'll send the new
patch soon

Hi Andrew,

I can't apply "kernel/smp.c: relocate some code" because this patch are
base on my other patch: "generic-ipi: make struct call_function_data lockless",
which can be found at:
	http://lkml.org/lkml/2009/7/24/63

I'll rewrite your patch base on it

Thanks,
Xiao

> 
>> case CPU_DEAD_FROZEN:
>> +               local_irq_save(flags);
>> +               __generic_smp_call_function_interrupt(cpu, 0);
>> +               __generic_smp_call_function_single_interrupt(cpu, 0);
>> +               local_irq_restore(flags);
> 
> Doing the callbacks from a different cpu than they were queued on seems
> like a fine way to mess things up.
> 
>