Re: [PATCH RFC V2 2/2] net: Optimize snmp stat aggregation by walking all the percpu data at once

[Raghavendra K T <raghavendra.kt@linux.vnet.ibm.com>]

On 08/29/2015 08:56 AM, Eric Dumazet wrote:
> On Sat, 2015-08-29 at 08:27 +0530, Raghavendra K T wrote:
>>
>>   	/* Use put_unaligned() because stats may not be aligned for u64. */
>>   	put_unaligned(items, &stats[0]);
>
>
>>   	for (i = 1; i < items; i++)
>> -		put_unaligned(snmp_fold_field64(mib, i, syncpoff), &stats[i]);
>> +		put_unaligned(buff[i], &stats[i]);
>>
>
> I believe Joe suggested following code instead :
>
> buff[0] = items;
> memcpy(stats, buff, items * sizeof(u64));

Thanks. Sure, will use this.

(I missed that. I thought that it was applicable only when we have
aligned data,and for power, put_aunaligned was not a nop unlike intel).

>
> Also please move buff[] array into __snmp6_fill_stats64() to make it
> clear it is used in a 'leaf' function.

Correct.

>
> (even if calling memcpy()/memset() makes it not a leaf function)
>