1 files changed, 79 insertions, 17 deletions

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 85f184e429c6..60398a9370e7 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -129,6 +129,7 @@ static int ip_rt_mtu_expires __read_mostly	= 10 * 60 * HZ;
 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 rtable *rt, struct sk_buff *skb, struct flow
                                 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;
@@ -1437,9 +1438,9 @@ static void rt_set_nexthop(struct rtable *rt, __be32 daddr,
 #endif
 }
 
-static struct rtable *rt_dst_alloc(struct net_device *dev,
-                                   unsigned int flags, u16 type,
-                                   bool nopolicy, bool noxfrm, bool will_cache)
+struct rtable *rt_dst_alloc(struct net_device *dev,
+                            unsigned int flags, u16 type,
+                            bool nopolicy, bool noxfrm, bool will_cache)
 {
         struct rtable *rt;
 
@@ -1467,6 +1468,7 @@ static struct rtable *rt_dst_alloc(struct net_device *dev,
 
         return rt;
 }
+EXPORT_SYMBOL(rt_dst_alloc);
 
 /* called in rcu_read_lock() section */
 static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr,
@@ -1556,6 +1558,36 @@ static void ip_handle_martian_source(struct net_device *dev,
 #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 +1641,20 @@ static int __mkroute_input(struct sk_buff *skb,
 
         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;
@@ -2005,6 +2046,18 @@ static struct rtable *__mkroute_output(const struct fib_result *res,
                  */
                 if (fi && res->prefixlen < 4)
                         fi = NULL;
+        } else if ((type == RTN_LOCAL) && (orig_oif != 0) &&
+                   (orig_oif != dev_out->ifindex)) {
+                /* For local routes that require a particular output interface
+                 * we do not want to cache the result.  Caching the result
+                 * causes incorrect behaviour when there are multiple source
+                 * addresses on the interface, the end result being that if the
+                 * intended recipient is waiting on that interface for the
+                 * packet he won't receive it because it will be delivered on
+                 * the loopback interface and the IP_PKTINFO ipi_ifindex will
+                 * be set to the loopback interface as well.
+                 */
+                fi = NULL;
         }
 
         fnhe = NULL;
@@ -2014,19 +2067,29 @@ static struct rtable *__mkroute_output(const struct fib_result *res,
                 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 +2632,6 @@ void ip_rt_multicast_event(struct in_device *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;