[IPV4]: Aggregate route entries with different TOS values

When we get an ICMP need-to-frag message, the original TOS value in the
ICMP payload cannot be used as a key to look up the routes to update.
This is because the TOS field may have been modified by routers on the
way.  Similarly, ip_rt_redirect should also ignore the TOS as the router
that gave us the message may have modified the TOS value.

The patch achieves this objective by aggregating entries with different
TOS values (but are otherwise identical) into the same bucket.  This
makes it easy to update them at the same time when an ICMP message is
received.

In future we should use a twin-hashing scheme where teh aggregation
occurs at the entry level.  That is, the TOS goes back into the hash
for normal lookups while ICMP lookups will end up with a node that
gives us a list that contains all other route entries that differ
only by TOS.

Signed-off-by: Ilia Sotnikov <hostcc@gmail.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/include/net/route.h b/include/net/route.h
index 9c04f15090d2b6b6e75060ab56dfb69047ea5db7..98c915abdec8791c409ab811df55cdf50421a9ea 100644 (file)
--- a/include/net/route.h
+++ b/include/net/route.h
@@ -110,7 +110,7 @@ extern struct ip_rt_acct *ip_rt_acct;
 struct in_device;
 extern int             ip_rt_init(void);
 extern void            ip_rt_redirect(u32 old_gw, u32 dst, u32 new_gw,
-                                      u32 src, u8 tos, struct net_device *dev);
+                                      u32 src, struct net_device *dev);
 extern void            ip_rt_advice(struct rtable **rp, int advice);
 extern void            rt_cache_flush(int how);
 extern int             __ip_route_output_key(struct rtable **, const struct flowi *flp);

diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index e7bbff4340bb6c2d73c956b15c40a4b596296ecb..9831fd2c73a0a222290f504e44d413627c5fe216 100644 (file)
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -753,7 +753,7 @@ static void icmp_redirect(struct sk_buff *skb)
        case ICMP_REDIR_HOST:
        case ICMP_REDIR_HOSTTOS:
                ip_rt_redirect(skb->nh.iph->saddr, ip, skb->h.icmph->un.gateway,
-                              iph->saddr, iph->tos, skb->dev);
+                              iph->saddr, skb->dev);
                break;
        }
 out:

diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index fca5fe0cf94a397fe957f26a39ebd82054f9582c..94fcbc5e5a1b62cfd696a6751dfcd17816a74f4a 100644 (file)
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -55,6 +55,8 @@
  *     Robert Olsson           :       Added rt_cache statistics
  *     Arnaldo C. Melo         :       Convert proc stuff to seq_file
  *     Eric Dumazet            :       hashed spinlocks and rt_check_expire() fixes.
+ *     Ilia Sotnikov           :       Ignore TOS on PMTUD and Redirect
+ *     Ilia Sotnikov           :       Removed TOS from hash calculations
  *
  *             This program is free software; you can redistribute it and/or
  *             modify it under the terms of the GNU General Public License
@@ -247,9 +249,9 @@ static DEFINE_PER_CPU(struct rt_cache_stat, rt_cache_stat);
 static int rt_intern_hash(unsigned hash, struct rtable *rth,
                                struct rtable **res);
 
-static unsigned int rt_hash_code(u32 daddr, u32 saddr, u8 tos)
+static unsigned int rt_hash_code(u32 daddr, u32 saddr)
 {
-       return (jhash_3words(daddr, saddr, (u32) tos, rt_hash_rnd)
+       return (jhash_2words(daddr, saddr, rt_hash_rnd)
                & rt_hash_mask);
 }
 
@@ -1111,7 +1113,7 @@ static void rt_del(unsigned hash, struct rtable *rt)
 }
 
 void ip_rt_redirect(u32 old_gw, u32 daddr, u32 new_gw,
-                   u32 saddr, u8 tos, struct net_device *dev)
+                   u32 saddr, struct net_device *dev)
 {
        int i, k;
        struct in_device *in_dev = in_dev_get(dev);
@@ -1119,8 +1121,6 @@ void ip_rt_redirect(u32 old_gw, u32 daddr, u32 new_gw,
        u32  skeys[2] = { saddr, 0 };
        int  ikeys[2] = { dev->ifindex, 0 };
 
-       tos &= IPTOS_RT_MASK;
-
        if (!in_dev)
                return;
 
@@ -1141,8 +1141,7 @@ void ip_rt_redirect(u32 old_gw, u32 daddr, u32 new_gw,
        for (i = 0; i < 2; i++) {
                for (k = 0; k < 2; k++) {
                        unsigned hash = rt_hash_code(daddr,
-                                                    skeys[i] ^ (ikeys[k] << 5),
-                                                    tos);
+                                                    skeys[i] ^ (ikeys[k] << 5));
 
                        rthp=&rt_hash_table[hash].chain;
 
@@ -1152,7 +1151,6 @@ void ip_rt_redirect(u32 old_gw, u32 daddr, u32 new_gw,
 
                                if (rth->fl.fl4_dst != daddr ||
                                    rth->fl.fl4_src != skeys[i] ||
-                                   rth->fl.fl4_tos != tos ||
                                    rth->fl.oif != ikeys[k] ||
                                    rth->fl.iif != 0) {
                                        rthp = &rth->u.rt_next;
@@ -1232,10 +1230,9 @@ reject_redirect:
        if (IN_DEV_LOG_MARTIANS(in_dev) && net_ratelimit())
                printk(KERN_INFO "Redirect from %u.%u.%u.%u on %s about "
                        "%u.%u.%u.%u ignored.\n"
-                       "  Advised path = %u.%u.%u.%u -> %u.%u.%u.%u, "
-                       "tos %02x\n",
+                       "  Advised path = %u.%u.%u.%u -> %u.%u.%u.%u\n",
                       NIPQUAD(old_gw), dev->name, NIPQUAD(new_gw),
-                      NIPQUAD(saddr), NIPQUAD(daddr), tos);
+                      NIPQUAD(saddr), NIPQUAD(daddr));
 #endif
        in_dev_put(in_dev);
 }
@@ -1253,8 +1250,7 @@ static struct dst_entry *ipv4_negative_advice(struct dst_entry *dst)
                           rt->u.dst.expires) {
                        unsigned hash = rt_hash_code(rt->fl.fl4_dst,
                                                     rt->fl.fl4_src ^
-                                                       (rt->fl.oif << 5),
-                                                    rt->fl.fl4_tos);
+                                                       (rt->fl.oif << 5));
 #if RT_CACHE_DEBUG >= 1
                        printk(KERN_DEBUG "ip_rt_advice: redirect to "
                                          "%u.%u.%u.%u/%02x dropped\n",
@@ -1391,14 +1387,13 @@ unsigned short ip_rt_frag_needed(struct iphdr *iph, unsigned short new_mtu)
        struct rtable *rth;
        u32  skeys[2] = { iph->saddr, 0, };
        u32  daddr = iph->daddr;
-       u8   tos = iph->tos & IPTOS_RT_MASK;
        unsigned short est_mtu = 0;
 
        if (ipv4_config.no_pmtu_disc)
                return 0;
 
        for (i = 0; i < 2; i++) {
-               unsigned hash = rt_hash_code(daddr, skeys[i], tos);
+               unsigned hash = rt_hash_code(daddr, skeys[i]);
 
                rcu_read_lock();
                for (rth = rcu_dereference(rt_hash_table[hash].chain); rth;
@@ -1407,7 +1402,6 @@ unsigned short ip_rt_frag_needed(struct iphdr *iph, unsigned short new_mtu)
                            rth->fl.fl4_src == skeys[i] &&
                            rth->rt_dst  == daddr &&
                            rth->rt_src  == iph->saddr &&
-                           rth->fl.fl4_tos == tos &&
                            rth->fl.iif == 0 &&
                            !(dst_metric_locked(&rth->u.dst, RTAX_MTU))) {
                                unsigned short mtu = new_mtu;
@@ -1658,7 +1652,7 @@ static int ip_route_input_mc(struct sk_buff *skb, u32 daddr, u32 saddr,
        RT_CACHE_STAT_INC(in_slow_mc);
 
        in_dev_put(in_dev);
-       hash = rt_hash_code(daddr, saddr ^ (dev->ifindex << 5), tos);
+       hash = rt_hash_code(daddr, saddr ^ (dev->ifindex << 5));
        return rt_intern_hash(hash, rth, (struct rtable**) &skb->dst);
 
 e_nobufs:
@@ -1823,7 +1817,7 @@ static inline int ip_mkroute_input_def(struct sk_buff *skb,
                return err;
 
        /* put it into the cache */
-       hash = rt_hash_code(daddr, saddr ^ (fl->iif << 5), tos);
+       hash = rt_hash_code(daddr, saddr ^ (fl->iif << 5));
        return rt_intern_hash(hash, rth, (struct rtable**)&skb->dst);   
 }
 
@@ -1864,7 +1858,7 @@ static inline int ip_mkroute_input(struct sk_buff *skb,
                        return err;
 
                /* put it into the cache */
-               hash = rt_hash_code(daddr, saddr ^ (fl->iif << 5), tos);
+               hash = rt_hash_code(daddr, saddr ^ (fl->iif << 5));
                err = rt_intern_hash(hash, rth, &rtres);
                if (err)
                        return err;
@@ -2041,7 +2035,7 @@ local_input:
                rth->rt_flags   &= ~RTCF_LOCAL;
        }
        rth->rt_type    = res.type;
-       hash = rt_hash_code(daddr, saddr ^ (fl.iif << 5), tos);
+       hash = rt_hash_code(daddr, saddr ^ (fl.iif << 5));
        err = rt_intern_hash(hash, rth, (struct rtable**)&skb->dst);
        goto done;
 
@@ -2088,7 +2082,7 @@ int ip_route_input(struct sk_buff *skb, u32 daddr, u32 saddr,
        int iif = dev->ifindex;
 
        tos &= IPTOS_RT_MASK;
-       hash = rt_hash_code(daddr, saddr ^ (iif << 5), tos);
+       hash = rt_hash_code(daddr, saddr ^ (iif << 5));
 
        rcu_read_lock();
        for (rth = rcu_dereference(rt_hash_table[hash].chain); rth;
@@ -2286,10 +2280,8 @@ static inline int ip_mkroute_output_def(struct rtable **rp,
        int err = __mkroute_output(&rth, res, fl, oldflp, dev_out, flags);
        unsigned hash;
        if (err == 0) {
-               u32 tos = RT_FL_TOS(oldflp);
-
                hash = rt_hash_code(oldflp->fl4_dst, 
-                                   oldflp->fl4_src ^ (oldflp->oif << 5), tos);
+                                   oldflp->fl4_src ^ (oldflp->oif << 5));
                err = rt_intern_hash(hash, rth, rp);
        }
        
@@ -2304,7 +2296,6 @@ static inline int ip_mkroute_output(struct rtable** rp,
                                    unsigned flags)
 {
 #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED
-       u32 tos = RT_FL_TOS(oldflp);
        unsigned char hop;
        unsigned hash;
        int err = -EINVAL;
@@ -2334,7 +2325,7 @@ static inline int ip_mkroute_output(struct rtable** rp,
 
                        hash = rt_hash_code(oldflp->fl4_dst, 
                                            oldflp->fl4_src ^
-                                           (oldflp->oif << 5), tos);
+                                           (oldflp->oif << 5));
                        err = rt_intern_hash(hash, rth, rp);
 
                        /* forward hop information to multipath impl. */
@@ -2563,7 +2554,7 @@ int __ip_route_output_key(struct rtable **rp, const struct flowi *flp)
        unsigned hash;
        struct rtable *rth;
 
-       hash = rt_hash_code(flp->fl4_dst, flp->fl4_src ^ (flp->oif << 5), flp->fl4_tos);
+       hash = rt_hash_code(flp->fl4_dst, flp->fl4_src ^ (flp->oif << 5));
 
        rcu_read_lock_bh();
        for (rth = rcu_dereference(rt_hash_table[hash].chain); rth;