From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ido Schimmel Subject: Re: [PATCH RFC net-next 4/4] ipv6: route: dissect flow in input path if fib rules need it Date: Mon, 12 Feb 2018 09:44:27 +0200 Message-ID: <20180212074426.GA22807@splinter> References: <1518387989-33735-1-git-send-email-roopa@cumulusnetworks.com> <1518387989-33735-5-git-send-email-roopa@cumulusnetworks.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: netdev@vger.kernel.org, dsa@cumulusnetworks.com, nikolay@cumulusnetworks.com To: Roopa Prabhu Return-path: Received: from out2-smtp.messagingengine.com ([66.111.4.26]:45657 "EHLO out2-smtp.messagingengine.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932720AbeBLHo3 (ORCPT ); Mon, 12 Feb 2018 02:44:29 -0500 Content-Disposition: inline In-Reply-To: <1518387989-33735-5-git-send-email-roopa@cumulusnetworks.com> Sender: netdev-owner@vger.kernel.org List-ID: On Sun, Feb 11, 2018 at 02:26:29PM -0800, Roopa Prabhu wrote: > From: Roopa Prabhu > > Dissect flow in fwd path if fib rules require it. Controlled by > a flag to avoid penatly for the common case. Flag is set when fib > rules with sport, dport and proto match that require flow dissect > are installed. Also passes the dissected hash keys to the multipath > hash function when applicable to avoid dissecting the flow again. > icmp packets will continue to use inner header for hash > calculations. > > Signed-off-by: Roopa Prabhu > --- > include/net/ip6_route.h | 3 ++- > include/net/netns/ipv6.h | 1 + > net/ipv6/fib6_rules.c | 12 +++++++++--- > net/ipv6/icmp.c | 2 +- > net/ipv6/route.c | 45 ++++++++++++++++++++++++++++++++++++--------- > 5 files changed, 49 insertions(+), 14 deletions(-) > > diff --git a/include/net/ip6_route.h b/include/net/ip6_route.h > index 27d23a6..218f89c 100644 > --- a/include/net/ip6_route.h > +++ b/include/net/ip6_route.h > @@ -127,7 +127,8 @@ static inline int ip6_route_get_saddr(struct net *net, struct rt6_info *rt, > > struct rt6_info *rt6_lookup(struct net *net, const struct in6_addr *daddr, > const struct in6_addr *saddr, int oif, int flags); > -u32 rt6_multipath_hash(const struct flowi6 *fl6, const struct sk_buff *skb); > +u32 rt6_multipath_hash(const struct flowi6 *fl6, const struct sk_buff *skb, > + struct flow_keys *hkeys); > > struct dst_entry *icmp6_dst_alloc(struct net_device *dev, struct flowi6 *fl6); > > diff --git a/include/net/netns/ipv6.h b/include/net/netns/ipv6.h > index 987cc45..7aca00e 100644 > --- a/include/net/netns/ipv6.h > +++ b/include/net/netns/ipv6.h > @@ -72,6 +72,7 @@ struct netns_ipv6 { > unsigned long ip6_rt_last_gc; > #ifdef CONFIG_IPV6_MULTIPLE_TABLES > bool fib6_has_custom_rules; > + bool fib6_rules_require_fldissect; > struct rt6_info *ip6_prohibit_entry; > struct rt6_info *ip6_blk_hole_entry; > struct fib6_table *fib6_local_tbl; [...] > @@ -1847,12 +1858,27 @@ void ip6_route_input(struct sk_buff *skb) > .flowi6_mark = skb->mark, > .flowi6_proto = iph->nexthdr, > }; > + struct flow_keys *flkeys = NULL, _flkeys; > > tun_info = skb_tunnel_info(skb); > if (tun_info && !(tun_info->mode & IP_TUNNEL_INFO_TX)) > fl6.flowi6_tun_key.tun_id = tun_info->key.tun_id; > + > +#ifdef CONFIG_IP_MULTIPLE_TABLES s/CONFIG_IP_MULTIPLE_TABLES/CONFIG_IPV6_MULTIPLE_TABLES/ ? > + if (net->ipv6.fib6_rules_require_fldissect) { > + unsigned int flag = FLOW_DISSECTOR_F_STOP_AT_ENCAP; > + > + memset(&_flkeys, 0, sizeof(_flkeys)); > + skb_flow_dissect_flow_keys(skb, &_flkeys, flag); > + fl6.fl6_sport = _flkeys.ports.src; > + fl6.fl6_dport = _flkeys.ports.dst; > + fl6.flowi6_proto = _flkeys.basic.ip_proto; > + flkeys = &_flkeys; > + } > +#endif