From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH 0/2] AF_PACKET fanout support Date: Tue, 05 Jul 2011 23:55:37 -0700 (PDT) Message-ID: <20110705.235537.2209662830319369211.davem@davemloft.net> References: <20110705.201950.199250194399828543.davem@davemloft.net> <1309925266.2545.48.camel@edumazet-laptop> <20110705.234424.2099388019260070145.davem@davemloft.net> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: therbert@google.com, victor@inliniac.net, netdev@vger.kernel.org, willemb@google.com To: eric.dumazet@gmail.com Return-path: Received: from shards.monkeyblade.net ([198.137.202.13]:45334 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752197Ab1GFGzq (ORCPT ); Wed, 6 Jul 2011 02:55:46 -0400 In-Reply-To: <20110705.234424.2099388019260070145.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: From: David Miller Date: Tue, 05 Jul 2011 23:44:24 -0700 (PDT) > From: Eric Dumazet > Date: Wed, 06 Jul 2011 06:07:46 +0200 > >> I suspect this can be solved adding a third policy : hash by CPU only > > Agreed, I'll implement this policy. packet: Add 'cpu' fanout policy. Unfortunately we have to use a real modulus here as the multiply trick won't work as effectively with cpu numbers as it does with rxhash values. Signed-off-by: David S. Miller diff --git a/include/linux/if_packet.h b/include/linux/if_packet.h index 84e684e..c148606 100644 --- a/include/linux/if_packet.h +++ b/include/linux/if_packet.h @@ -53,6 +53,7 @@ struct sockaddr_ll { #define PACKET_FANOUT_HASH 0 #define PACKET_FANOUT_LB 1 +#define PACKET_FANOUT_CPU 2 #define PACKET_FANOUT_FLAG_DEFRAG 0x8000 struct tpacket_stats { diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 7ba6871..41f0489 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -447,6 +447,13 @@ static struct sock *fanout_demux_lb(struct packet_fanout *f, struct sk_buff *skb return f->arr[cur]; } +static struct sock *fanout_demux_cpu(struct packet_fanout *f, struct sk_buff *skb, unsigned int num) +{ + unsigned int cpu = smp_processor_id(); + + return f->arr[cpu % num]; +} + static struct sk_buff *fanout_check_defrag(struct sk_buff *skb) { const struct iphdr *iph; @@ -482,8 +489,8 @@ static struct sk_buff *fanout_check_defrag(struct sk_buff *skb) return skb; } -static int packet_rcv_fanout_hash(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt, struct net_device *orig_dev) +static int packet_rcv_fanout(struct sk_buff *skb, struct net_device *dev, + struct packet_type *pt, struct net_device *orig_dev) { struct packet_fanout *f = pt->af_packet_priv; unsigned int num = f->num_members; @@ -496,35 +503,25 @@ static int packet_rcv_fanout_hash(struct sk_buff *skb, struct net_device *dev, return 0; } - if (f->defrag) { - skb = fanout_check_defrag(skb); - if (!skb) - return 0; - } - - skb_get_rxhash(skb); - - sk = fanout_demux_hash(f, skb, num); - po = pkt_sk(sk); - - return po->prot_hook.func(skb, dev, &po->prot_hook, orig_dev); -} - -static int packet_rcv_fanout_lb(struct sk_buff *skb, struct net_device *dev, - struct packet_type *pt, struct net_device *orig_dev) -{ - struct packet_fanout *f = pt->af_packet_priv; - unsigned int num = f->num_members; - struct packet_sock *po; - struct sock *sk; - - if (!net_eq(dev_net(dev), read_pnet(&f->net)) || - !num) { - kfree_skb(skb); - return 0; + switch (f->type) { + case PACKET_FANOUT_HASH: + default: + if (f->defrag) { + skb = fanout_check_defrag(skb); + if (!skb) + return 0; + } + skb_get_rxhash(skb); + sk = fanout_demux_hash(f, skb, num); + break; + case PACKET_FANOUT_LB: + sk = fanout_demux_lb(f, skb, num); + break; + case PACKET_FANOUT_CPU: + sk = fanout_demux_cpu(f, skb, num); + break; } - sk = fanout_demux_lb(f, skb, num); po = pkt_sk(sk); return po->prot_hook.func(skb, dev, &po->prot_hook, orig_dev); @@ -571,6 +568,7 @@ static int fanout_add(struct sock *sk, u16 id, u16 type_flags) switch (type) { case PACKET_FANOUT_HASH: case PACKET_FANOUT_LB: + case PACKET_FANOUT_CPU: break; default: return -EINVAL; @@ -606,14 +604,7 @@ static int fanout_add(struct sock *sk, u16 id, u16 type_flags) atomic_set(&match->sk_ref, 0); match->prot_hook.type = po->prot_hook.type; match->prot_hook.dev = po->prot_hook.dev; - switch (type) { - case PACKET_FANOUT_HASH: - match->prot_hook.func = packet_rcv_fanout_hash; - break; - case PACKET_FANOUT_LB: - match->prot_hook.func = packet_rcv_fanout_lb; - break; - } + match->prot_hook.func = packet_rcv_fanout; match->prot_hook.af_packet_priv = match; dev_add_pack(&match->prot_hook); list_add(&match->list, &fanout_list);