Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755597AbcCBBcd (ORCPT ); Tue, 1 Mar 2016 20:32:33 -0500 Received: from mail177-1.suw61.mandrillapp.com ([198.2.177.1]:43852 "EHLO mail177-1.suw61.mandrillapp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754844AbcCAXyh (ORCPT ); Tue, 1 Mar 2016 18:54:37 -0500 From: Greg Kroah-Hartman Subject: [PATCH 4.4 041/342] route: check and remove route cache when we get route X-Mailer: git-send-email 2.7.2 To: Cc: Greg Kroah-Hartman , , Xin Long , Hannes Frederic Sowa , "David S. Miller" Message-Id: <20160301234529.299816329@linuxfoundation.org> In-Reply-To: <20160301234527.990448862@linuxfoundation.org> References: <20160301234527.990448862@linuxfoundation.org> X-Report-Abuse: Please forward a copy of this message, including all headers, to abuse@mandrill.com X-Report-Abuse: You can also report abuse here: http://mandrillapp.com/contact/abuse?id=30481620.7baac11b0d6b4403b4c1b7926076198b X-Mandrill-User: md_30481620 Date: Tue, 01 Mar 2016 23:53:56 +0000 MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4698 Lines: 160 4.4-stable review patch. If anyone has any objections, please let me know. ------------------ From: Xin Long [ Upstream commit deed49df7390d5239024199e249190328f1651e7 ] Since the gc of ipv4 route was removed, the route cached would has no chance to be removed, and even it has been timeout, it still could be used, cause no code to check it's expires. Fix this issue by checking and removing route cache when we get route. Signed-off-by: Xin Long Acked-by: Hannes Frederic Sowa Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/net/ip_fib.h | 1 net/ipv4/route.c | 77 +++++++++++++++++++++++++++++++++++++++++---------- 2 files changed, 64 insertions(+), 14 deletions(-) --- a/include/net/ip_fib.h +++ b/include/net/ip_fib.h @@ -61,6 +61,7 @@ struct fib_nh_exception { struct rtable __rcu *fnhe_rth_input; struct rtable __rcu *fnhe_rth_output; unsigned long fnhe_stamp; + struct rcu_head rcu; }; struct fnhe_hash_bucket { --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -129,6 +129,7 @@ static int ip_rt_mtu_expires __read_most static int ip_rt_min_pmtu __read_mostly = 512 + 20 + 20; static int ip_rt_min_advmss __read_mostly = 256; +static int ip_rt_gc_timeout __read_mostly = RT_GC_TIMEOUT; /* * Interface to generic destination cache. */ @@ -755,7 +756,7 @@ static void __ip_do_redirect(struct rtab struct fib_nh *nh = &FIB_RES_NH(res); update_or_create_fnhe(nh, fl4->daddr, new_gw, - 0, 0); + 0, jiffies + ip_rt_gc_timeout); } if (kill_route) rt->dst.obsolete = DST_OBSOLETE_KILL; @@ -1556,6 +1557,36 @@ static void ip_handle_martian_source(str #endif } +static void ip_del_fnhe(struct fib_nh *nh, __be32 daddr) +{ + struct fnhe_hash_bucket *hash; + struct fib_nh_exception *fnhe, __rcu **fnhe_p; + u32 hval = fnhe_hashfun(daddr); + + spin_lock_bh(&fnhe_lock); + + hash = rcu_dereference_protected(nh->nh_exceptions, + lockdep_is_held(&fnhe_lock)); + hash += hval; + + fnhe_p = &hash->chain; + fnhe = rcu_dereference_protected(*fnhe_p, lockdep_is_held(&fnhe_lock)); + while (fnhe) { + if (fnhe->fnhe_daddr == daddr) { + rcu_assign_pointer(*fnhe_p, rcu_dereference_protected( + fnhe->fnhe_next, lockdep_is_held(&fnhe_lock))); + fnhe_flush_routes(fnhe); + kfree_rcu(fnhe, rcu); + break; + } + fnhe_p = &fnhe->fnhe_next; + fnhe = rcu_dereference_protected(fnhe->fnhe_next, + lockdep_is_held(&fnhe_lock)); + } + + spin_unlock_bh(&fnhe_lock); +} + /* called in rcu_read_lock() section */ static int __mkroute_input(struct sk_buff *skb, const struct fib_result *res, @@ -1609,11 +1640,20 @@ static int __mkroute_input(struct sk_buf fnhe = find_exception(&FIB_RES_NH(*res), daddr); if (do_cache) { - if (fnhe) + if (fnhe) { rth = rcu_dereference(fnhe->fnhe_rth_input); - else - rth = rcu_dereference(FIB_RES_NH(*res).nh_rth_input); + if (rth && rth->dst.expires && + time_after(jiffies, rth->dst.expires)) { + ip_del_fnhe(&FIB_RES_NH(*res), daddr); + fnhe = NULL; + } else { + goto rt_cache; + } + } + + rth = rcu_dereference(FIB_RES_NH(*res).nh_rth_input); +rt_cache: if (rt_cache_valid(rth)) { skb_dst_set_noref(skb, &rth->dst); goto out; @@ -2014,19 +2054,29 @@ static struct rtable *__mkroute_output(c struct fib_nh *nh = &FIB_RES_NH(*res); fnhe = find_exception(nh, fl4->daddr); - if (fnhe) + if (fnhe) { prth = &fnhe->fnhe_rth_output; - else { - if (unlikely(fl4->flowi4_flags & - FLOWI_FLAG_KNOWN_NH && - !(nh->nh_gw && - nh->nh_scope == RT_SCOPE_LINK))) { - do_cache = false; - goto add; + rth = rcu_dereference(*prth); + if (rth && rth->dst.expires && + time_after(jiffies, rth->dst.expires)) { + ip_del_fnhe(nh, fl4->daddr); + fnhe = NULL; + } else { + goto rt_cache; } - prth = raw_cpu_ptr(nh->nh_pcpu_rth_output); } + + if (unlikely(fl4->flowi4_flags & + FLOWI_FLAG_KNOWN_NH && + !(nh->nh_gw && + nh->nh_scope == RT_SCOPE_LINK))) { + do_cache = false; + goto add; + } + prth = raw_cpu_ptr(nh->nh_pcpu_rth_output); rth = rcu_dereference(*prth); + +rt_cache: if (rt_cache_valid(rth)) { dst_hold(&rth->dst); return rth; @@ -2569,7 +2619,6 @@ void ip_rt_multicast_event(struct in_dev } #ifdef CONFIG_SYSCTL -static int ip_rt_gc_timeout __read_mostly = RT_GC_TIMEOUT; static int ip_rt_gc_interval __read_mostly = 60 * HZ; static int ip_rt_gc_min_interval __read_mostly = HZ / 2; static int ip_rt_gc_elasticity __read_mostly = 8;