On Wed, Jul 23, 2014 at 10:36:28PM +0200, Borislav Petkov wrote:
>   /*
>    * At CPU state changes, update the x2apic cluster sibling info.
>    */
> @@ -151,9 +166,14 @@ static int
>  update_clusterinfo(struct notifier_block *nfb, unsigned long action, void *hcpu)
>  {
>  	unsigned int this_cpu = (unsigned long)hcpu;
> -	unsigned int cpu;
>  	int err = 0;
>  
> +	if (action & CPU_TASKS_FROZEN) {
> +		if ((action & ~CPU_TASKS_FROZEN) == CPU_UP_CANCELED)
> +			__update_clusterinfo(this_cpu);
> +		return NOTIFY_OK;
> +	}
> +
Hi, Boris

Here you assume once the action contains FROZEN flag, it will return
NOTIFY_OK. It looks like not quite comply with original logic. Once
new FROZEN logic is added, we have to update this code again. How
about using following code snippet:

+       if ((action & CPU_TASKS_FROZEN) &&
+           ((action & ~CPU_TASKS_FROZEN) == CPU_UP_CANCELED)) {
+               __update_clusterinfo(this_cpu);
+               return NOTIFY_OK;
+       }
+