Re: [RFC v2] net: use atomic allocation for order-3 page allocation

2015-06-11 Thread Eric Dumazet
On Thu, 2015-06-11 at 16:32 -0700, Shaohua Li wrote: > > Ok, looks similar, added. Didn't trigger this one though. Probably because you do not use af_unix with big enough messages. > diff --git a/net/core/skbuff.c b/net/core/skbuff.c > index 3cfff2a..9856c7a 100644 > --- a/net/core/skbuff.c > +

Re: [RFC v2] net: use atomic allocation for order-3 page allocation

2015-06-11 Thread Shaohua Li
On Thu, Jun 11, 2015 at 03:53:04PM -0700, Eric Dumazet wrote: > On Thu, 2015-06-11 at 15:27 -0700, Shaohua Li wrote: > > We saw excessive direct memory compaction triggered by skb_page_frag_refill. > > This causes performance issues and add latency. Commit 5640f7685831e0 > > introduces the order-3

Re: [RFC v2] net: use atomic allocation for order-3 page allocation

2015-06-11 Thread Eric Dumazet
On Thu, 2015-06-11 at 15:27 -0700, Shaohua Li wrote: > We saw excessive direct memory compaction triggered by skb_page_frag_refill. > This causes performance issues and add latency. Commit 5640f7685831e0 > introduces the order-3 allocation. According to the changelog, the order-3 > allocation isn't