Re: __GFP_REPEAT usage in fq_alloc_node

[Eric Dumazet]

On Fri, Jan 6, 2017 at 8:48 AM, Eric Dumazet <edumazet@xxxxxxxxxx> wrote:
> On Fri, Jan 6, 2017 at 8:31 AM, Vlastimil Babka <vbabka@xxxxxxx> wrote:
>
>>
>> I wonder what's that cause of the penalty (when accessing the vmapped
>> area I suppose?) Is it higher risk of collisions cache misses within the
>> area, compared to consecutive physical adresses?
>
> I believe tests were done with 48 fq qdisc, each having 2^16 slots.
> So I had 48 blocs,of 524288 bytes.
>
> Trying a bit harder at setup time to get 128 consecutive pages got
> less TLB pressure.

Forgot to mention tests include DDOS, so hitting a random hash bucket
for every packet.