Re: [PATCH 5/8] x86, mce: introduce mce_sysdev_init()

[Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>]

(2011/06/18 1:32), Borislav Petkov wrote:
>> @@ -1953,6 +1953,39 @@ static __cpuinit void mce_sysdev_remove(unsigned int cpu)
>>  	cpumask_clear_cpu(cpu, mce_sysdev_initialized);
>>  }
>>  
>> +static __init int mce_sysdev_init(void)
>> +{
>> +	int err, i;
>> +
>> +	zalloc_cpumask_var(&mce_sysdev_initialized, GFP_KERNEL);
>> +
>> +	/* init attributes for each mce_banks */
>> +	for (i = 0; i < banks; i++) {
>> +		struct mce_bank *b = &mce_banks[i];
>> +		struct sysdev_attribute *a = &b->attr;
>> +
>> +		sysfs_attr_init(&a->attr);
>> +		a->attr.name	= b->attrname;
>> +		snprintf(b->attrname, ATTR_LEN, "bank%d", i);
>> +
>> +		a->attr.mode	= 0644;
>> +		a->show		= show_bank;
>> +		a->store	= set_bank;
>> +	}
>> +
>> +	err = sysdev_class_register(&mce_sysdev_class);
>> +	if (err)
>> +		return err;
>> +
>> +	for_each_online_cpu(i) {
>> +		err = mce_sysdev_create(i);
>> +		if (err)
>> +			return err;
>> +	}
> 
> Yeah, let's handle the error case here and unwind to a sane state:
> 
> 	if (err)
> 		goto sysdev_unwind;
> ...
> 
> sysdev_unwind:
> 	while (--i > 0)
> 		mce_sysdev_remove(i);
> 

Nice!
I'll do it.


Thanks,
H.Seto