Merge

Documentation/filesystems/proc.txt

@@ -1754,18 +1754,25 @@ settings contain additional options to set garbage collection parameters.
 
 In the interface directories you'll find the following entries:
 
-base_reachable_time
--------------------
+base_reachable_time, base_reachable_time_ms
+-------------------------------------------
 
 A base  value  used for computing the random reachable time value as specified
 in RFC2461.
 
-retrans_time
-------------
+Expression of base_reachable_time, which is deprecated, is in seconds.
+Expression of base_reachable_time_ms is in milliseconds.
+
+retrans_time, retrans_time_ms
+-----------------------------
+
+The time between retransmitted Neighbor Solicitation messages.
+Used for address resolution and to determine if a neighbor is
+unreachable.
 
-The time,  expressed  in  jiffies  (1/100 sec), between retransmitted Neighbor
-Solicitation messages.  Used  for  address  resolution  and  to determine if a
-neighbor is unreachable.
+Expression of retrans_time, which is deprecated, is in 1/100 seconds (for
+IPv4) or in jiffies (for IPv6).
+Expression of retrans_time_ms is in milliseconds.
 
 unres_qlen
 ----------

include/linux/ip.h

@@ -152,6 +152,7 @@ struct inet_sock {
 };
 
 #define IPCORK_OPT	1	/* ip-options has been held in ipcork.opt */
+#define IPCORK_ALLFRAG	2	/* always fragment (for ipv6 for now) */
 
 static inline struct inet_sock *inet_sk(const struct sock *sk)
 {

include/linux/rtnetlink.h

@@ -346,6 +346,7 @@ enum
 #define RTAX_FEATURE_ECN	0x00000001
 #define RTAX_FEATURE_SACK	0x00000002
 #define RTAX_FEATURE_TIMESTAMP	0x00000004
+#define RTAX_FEATURE_ALLFRAG	0x00000008
 
 struct rta_session
 {

include/linux/sysctl.h

@@ -399,6 +399,7 @@ enum
 	NET_IPV4_CONF_FORCE_IGMP_VERSION=17,
 	NET_IPV4_CONF_ARP_ANNOUNCE=18,
 	NET_IPV4_CONF_ARP_IGNORE=19,
+	__NET_IPV4_CONF_MAX
 };
 
 /* /proc/sys/net/ipv4/netfilter */
@@ -456,7 +457,8 @@ enum {
 	NET_IPV6_ROUTE_GC_INTERVAL=6,
 	NET_IPV6_ROUTE_GC_ELASTICITY=7,
 	NET_IPV6_ROUTE_MTU_EXPIRES=8,
-	NET_IPV6_ROUTE_MIN_ADVMSS=9
+	NET_IPV6_ROUTE_MIN_ADVMSS=9,
+	NET_IPV6_ROUTE_GC_MIN_INTERVAL_MS=10
 };
 
 enum {
@@ -476,7 +478,8 @@ enum {
 	NET_IPV6_REGEN_MAX_RETRY=14,
 	NET_IPV6_MAX_DESYNC_FACTOR=15,
 	NET_IPV6_MAX_ADDRESSES=16,
-	NET_IPV6_FORCE_MLD_VERSION=17
+	NET_IPV6_FORCE_MLD_VERSION=17,
+	__NET_IPV6_MAX
 };
 
 /* /proc/sys/net/ipv6/icmp */
@@ -501,7 +504,10 @@ enum {
 	NET_NEIGH_GC_INTERVAL=13,
 	NET_NEIGH_GC_THRESH1=14,
 	NET_NEIGH_GC_THRESH2=15,
-	NET_NEIGH_GC_THRESH3=16
+	NET_NEIGH_GC_THRESH3=16,
+	NET_NEIGH_RETRANS_TIME_MS=17,
+	NET_NEIGH_REACHABLE_TIME_MS=18,
+	__NET_NEIGH_MAX
 };
 
 /* /proc/sys/net/ipx */

include/net/addrconf.h

@@ -102,6 +102,8 @@ extern int ipv6_is_mld(struct sk_buff *skb, int nexthdr);
 
 extern void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len);
 
+extern int ipv6_get_hoplimit(struct net_device *dev);
+
 /*
  *	anycast prototypes (anycast.c)
  */

include/net/dst.h

@@ -134,6 +134,15 @@ static inline u32 dst_mtu(const struct dst_entry *dst)
 	return mtu;
 }
 
+static inline u32
+dst_allfrag(const struct dst_entry *dst)
+{
+	int ret = dst_path_metric(dst, RTAX_FEATURES) & RTAX_FEATURE_ALLFRAG;
+	/* Yes, _exactly_. This is paranoia. */
+	barrier();
+	return ret;
+}
+
 static inline int
 dst_metric_locked(struct dst_entry *dst, int metric)
 {

include/net/ipv6.h

@@ -305,6 +305,32 @@ static inline int ipv6_addr_equal(const struct in6_addr *a1,
 		a1->s6_addr32[3] == a2->s6_addr32[3]);
 }
 
+static inline int __ipv6_prefix_equal(const u32 *a1, const u32 *a2,
+				      unsigned int prefixlen)
+{
+	unsigned pdw, pbi;
+
+	/* check complete u32 in prefix */
+	pdw = prefixlen >> 5;
+	if (pdw && memcmp(a1, a2, pdw << 2))
+		return 0;
+
+	/* check incomplete u32 in prefix */
+	pbi = prefixlen & 0x1f;
+	if (pbi && ((a1[pdw] ^ a2[pdw]) & htonl((0xffffffff) << (32 - pbi))))
+		return 0;
+
+	return 1;
+}
+
+static inline int ipv6_prefix_equal(const struct in6_addr *a1,
+				    const struct in6_addr *a2,
+				    unsigned int prefixlen)
+{
+	return __ipv6_prefix_equal(a1->s6_addr32, a2->s6_addr32,
+				   prefixlen);
+}
+
 static inline int ipv6_addr_any(const struct in6_addr *a)
 {
 	return ((a->s6_addr32[0] | a->s6_addr32[1] | 

include/net/ndisc.h

@@ -15,11 +15,15 @@
  *	ndisc options
  */
 
-#define ND_OPT_SOURCE_LL_ADDR		1
-#define ND_OPT_TARGET_LL_ADDR		2
-#define ND_OPT_PREFIX_INFO		3
-#define ND_OPT_REDIRECT_HDR		4
-#define ND_OPT_MTU			5
+enum {
+	__ND_OPT_PREFIX_INFO_END = 0,
+	ND_OPT_SOURCE_LL_ADDR = 1,	/* RFC2461 */
+	ND_OPT_TARGET_LL_ADDR = 2,	/* RFC2461 */
+	ND_OPT_PREFIX_INFO = 3,		/* RFC2461 */
+	ND_OPT_REDIRECT_HDR = 4,	/* RFC2461 */
+	ND_OPT_MTU = 5,			/* RFC2461 */
+	__ND_OPT_MAX
+};
 
 #define MAX_RTR_SOLICITATION_DELAY	HZ
 

include/net/neighbour.h

@@ -274,7 +274,8 @@ extern int			neigh_sysctl_register(struct net_device *dev,
 						      struct neigh_parms *p,
 						      int p_id, int pdev_id,
 						      char *p_name,
-						      proc_handler *proc_handler);
+						      proc_handler *proc_handler,
+						      ctl_handler *strategy);
 extern void			neigh_sysctl_unregister(struct neigh_parms *p);
 
 static inline void __neigh_parms_put(struct neigh_parms *parms)

net/Kconfig

@@ -107,28 +107,23 @@ source "net/ipv4/Kconfig"
 
 #   IPv6 as module will cause a CRASH if you try to unload it
 config IPV6
-	tristate "The IPv6 protocol (EXPERIMENTAL)"
-	depends on INET && EXPERIMENTAL
+	tristate "The IPv6 protocol"
+	depends on INET
 	select CRYPTO if IPV6_PRIVACY
 	select CRYPTO_MD5 if IPV6_PRIVACY
 	---help---
-	  This is experimental support for the IP version 6 (formerly called
-	  IPng "IP next generation").  You will still be able to do
-	  regular IPv4 networking as well.
-
-	  Features of this new protocol include: expanded address space,
-	  authentication and privacy, and seamless interoperability with the
-	  current version of IP (IP version 4). For general information about
-	  IPv6, see <http://playground.sun.com/pub/ipng/html/ipng-main.html>;
-	  for specific information about IPv6 under Linux read the HOWTO at
-	  <http://www.bieringer.de/linux/IPv6/> and the file net/ipv6/README
-	  in the kernel source.
+	  This is complemental support for the IP version 6.
+	  You will still be able to do traditional IPv4 networking as well.
+
+	  For general information about IPv6, see
+	  <http://playground.sun.com/pub/ipng/html/ipng-main.html>.
+	  For Linux IPv6 development information, see <http://www.linux-ipv6.org>.
+	  For specific information about IPv6 under Linux, read the HOWTO at
+	  <http://www.bieringer.de/linux/IPv6/>.
 
 	  To compile this protocol support as a module, choose M here: the 
 	  module will be called ipv6.
 
-	  It is safe to say N here for now.
-
 source "net/ipv6/Kconfig"
 
 menuconfig NETFILTER

net/core/neighbour.c

@@ -2052,7 +2052,7 @@ static void neigh_app_notify(struct neighbour *n)
 
 static struct neigh_sysctl_table {
 	struct ctl_table_header *sysctl_header;
-	ctl_table		neigh_vars[17];
+	ctl_table		neigh_vars[__NET_NEIGH_MAX];
 	ctl_table		neigh_dev[2];
 	ctl_table		neigh_neigh_dir[2];
 	ctl_table		neigh_proto_dir[2];
@@ -2175,6 +2175,22 @@ static struct neigh_sysctl_table {
 			.mode		= 0644,
 			.proc_handler	= &proc_dointvec,
 		},
+		{
+			.ctl_name	= NET_NEIGH_RETRANS_TIME_MS,
+			.procname	= "retrans_time_ms",
+			.maxlen		= sizeof(int),
+			.mode		= 0644,
+			.proc_handler	= &proc_dointvec_ms_jiffies,
+			.strategy	= &sysctl_ms_jiffies,
+		},
+		{
+			.ctl_name	= NET_NEIGH_REACHABLE_TIME_MS,
+			.procname	= "base_reachable_time_ms",
+			.maxlen		= sizeof(int),
+			.mode		= 0644,
+			.proc_handler	= &proc_dointvec_ms_jiffies,
+			.strategy	= &sysctl_ms_jiffies,
+		},
 	},
 	.neigh_dev = {
 		{
@@ -2205,7 +2221,7 @@ static struct neigh_sysctl_table {
 
 int neigh_sysctl_register(struct net_device *dev, struct neigh_parms *p,
 			  int p_id, int pdev_id, char *p_name, 
-			  proc_handler *handler)
+			  proc_handler *handler, ctl_handler *strategy)
 {
 	struct neigh_sysctl_table *t = kmalloc(sizeof(*t), GFP_KERNEL);
 	const char *dev_name_source = NULL;
@@ -2219,10 +2235,6 @@ int neigh_sysctl_register(struct net_device *dev, struct neigh_parms *p,
 	t->neigh_vars[1].data  = &p->ucast_probes;
 	t->neigh_vars[2].data  = &p->app_probes;
 	t->neigh_vars[3].data  = &p->retrans_time;
-	if (handler) {
-		t->neigh_vars[3].proc_handler = handler;
-		t->neigh_vars[3].extra1 = dev;
-	}
 	t->neigh_vars[4].data  = &p->base_reachable_time;
 	t->neigh_vars[5].data  = &p->delay_probe_time;
 	t->neigh_vars[6].data  = &p->gc_staletime;
@@ -2232,18 +2244,43 @@ int neigh_sysctl_register(struct net_device *dev, struct neigh_parms *p,
 	t->neigh_vars[10].data = &p->proxy_delay;
 	t->neigh_vars[11].data = &p->locktime;
 
- 	dev_name_source = t->neigh_dev[0].procname;
 	if (dev) {
 		dev_name_source = dev->name;
 		t->neigh_dev[0].ctl_name = dev->ifindex;
-		memset(&t->neigh_vars[12], 0, sizeof(ctl_table));
+		t->neigh_vars[12].procname = NULL;
+		t->neigh_vars[13].procname = NULL;
+		t->neigh_vars[14].procname = NULL;
+		t->neigh_vars[15].procname = NULL;
 	} else {
+ 		dev_name_source = t->neigh_dev[0].procname;
 		t->neigh_vars[12].data = (int *)(p + 1);
 		t->neigh_vars[13].data = (int *)(p + 1) + 1;
 		t->neigh_vars[14].data = (int *)(p + 1) + 2;
 		t->neigh_vars[15].data = (int *)(p + 1) + 3;
 	}
 
+	t->neigh_vars[16].data  = &p->retrans_time;
+	t->neigh_vars[17].data  = &p->base_reachable_time;
+
+	if (handler || strategy) {
+		/* RetransTime */
+		t->neigh_vars[3].proc_handler = handler;
+		t->neigh_vars[3].strategy = strategy;
+		t->neigh_vars[3].extra1 = dev;
+		/* ReachableTime */
+		t->neigh_vars[4].proc_handler = handler;
+		t->neigh_vars[4].strategy = strategy;
+		t->neigh_vars[4].extra1 = dev;
+		/* RetransTime (in milliseconds)*/
+		t->neigh_vars[16].proc_handler = handler;
+		t->neigh_vars[16].strategy = strategy;
+		t->neigh_vars[16].extra1 = dev;
+		/* ReachableTime (in milliseconds) */
+		t->neigh_vars[17].proc_handler = handler;
+		t->neigh_vars[17].strategy = strategy;
+		t->neigh_vars[17].extra1 = dev;
+	}
+
 	dev_name = net_sysctl_strdup(dev_name_source);
 	if (!dev_name) {
 		err = -ENOBUFS;

net/ipv4/arp.c

@@ -1243,7 +1243,7 @@ void __init arp_init(void)
 	arp_proc_init();
 #ifdef CONFIG_SYSCTL
 	neigh_sysctl_register(NULL, &arp_tbl.parms, NET_IPV4,
-			      NET_IPV4_NEIGH, "ipv4", NULL);
+			      NET_IPV4_NEIGH, "ipv4", NULL, NULL);
 #endif
 	register_netdevice_notifier(&arp_netdev_notifier);
 }

net/ipv4/devinet.c

@@ -153,7 +153,7 @@ struct in_device *inetdev_init(struct net_device *dev)
 	dev_hold(dev);
 #ifdef CONFIG_SYSCTL
 	neigh_sysctl_register(dev, in_dev->arp_parms, NET_IPV4,
-			      NET_IPV4_NEIGH, "ipv4", NULL);
+			      NET_IPV4_NEIGH, "ipv4", NULL, NULL);
 #endif
 
 	/* Account for reference dev->ip_ptr */
@@ -1001,7 +1001,7 @@ static int inetdev_event(struct notifier_block *this, unsigned long event,
 		devinet_sysctl_unregister(&in_dev->cnf);
 		neigh_sysctl_unregister(in_dev->arp_parms);
 		neigh_sysctl_register(dev, in_dev->arp_parms, NET_IPV4,
-				      NET_IPV4_NEIGH, "ipv4", NULL);
+				      NET_IPV4_NEIGH, "ipv4", NULL, NULL);
 		devinet_sysctl_register(in_dev, &in_dev->cnf);
 #endif
 		break;
@@ -1221,7 +1221,7 @@ int ipv4_doint_and_flush_strategy(ctl_table *table, int __user *name, int nlen,
 
 static struct devinet_sysctl_table {
 	struct ctl_table_header *sysctl_header;
-	ctl_table		devinet_vars[20];
+	ctl_table		devinet_vars[__NET_IPV4_CONF_MAX];
 	ctl_table		devinet_dev[2];
 	ctl_table		devinet_conf_dir[2];
 	ctl_table		devinet_proto_dir[2];

net/ipv6/README

-To join in the work on Linux IPv6 send mail to:
-
-        majordomo@oss.sgi.com
-
-and in the body of the message include:
-
-subscribe netdev
-

net/ipv6/addrconf.c

@@ -391,7 +391,9 @@ static struct inet6_dev * ipv6_add_dev(struct net_device *dev)
 		ndev->tstamp = jiffies;
 #ifdef CONFIG_SYSCTL
 		neigh_sysctl_register(dev, ndev->nd_parms, NET_IPV6, 
-			NET_IPV6_NEIGH, "ipv6", &ndisc_ifinfo_sysctl_change);
+				      NET_IPV6_NEIGH, "ipv6",
+				      &ndisc_ifinfo_sysctl_change,
+				      NULL);
 		addrconf_sysctl_register(ndev, &ndev->cnf);
 #endif
 	}
@@ -589,6 +591,8 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp)
 	struct inet6_ifaddr *ifa, **ifap;
 	struct inet6_dev *idev = ifp->idev;
 	int hash;
+	int deleted = 0, onlink = 0;
+	unsigned long expires = jiffies;
 
 	hash = ipv6_addr_hash(&ifp->addr);
 
@@ -631,7 +635,31 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp)
 			*ifap = ifa->if_next;
 			__in6_ifa_put(ifp);
 			ifa->if_next = NULL;
-			break;
+			if (!(ifp->flags & IFA_F_PERMANENT) || onlink > 0)
+				break;
+			deleted = 1;
+		} else if (ifp->flags & IFA_F_PERMANENT) {
+			if (ipv6_prefix_equal(&ifa->addr, &ifp->addr,
+					      ifp->prefix_len)) {
+				if (ifa->flags & IFA_F_PERMANENT) {
+					onlink = 1;
+					if (deleted)
+						break;
+				} else {
+					unsigned long lifetime;
+
+					if (!onlink)
+						onlink = -1;
+
+					spin_lock(&ifa->lock);
+					lifetime = min_t(unsigned long,
+							 ifa->valid_lft, 0x7fffffffUL/HZ);
+					if (time_before(expires,
+							ifa->tstamp + lifetime * HZ))
+						expires = ifa->tstamp + lifetime * HZ;
+					spin_unlock(&ifa->lock);
+				}
+			}
 		}
 	}
 	write_unlock_bh(&idev->lock);
@@ -642,6 +670,40 @@ static void ipv6_del_addr(struct inet6_ifaddr *ifp)
 
 	addrconf_del_timer(ifp);
 
+	/*
+	 * Purge or update corresponding prefix
+	 *
+	 * 1) we don't purge prefix here if address was not permanent.
+	 *    prefix is managed by its own lifetime.
+	 * 2) if there're no addresses, delete prefix.
+	 * 3) if there're still other permanent address(es),
+	 *    corresponding prefix is still permanent.
+	 * 4) otherwise, update prefix lifetime to the
+	 *    longest valid lifetime among the corresponding
+	 *    addresses on the device.
+	 *    Note: subsequent RA will update lifetime.
+	 *
+	 * --yoshfuji
+	 */
+	if ((ifp->flags & IFA_F_PERMANENT) && onlink < 1) {
+		struct in6_addr prefix;
+		struct rt6_info *rt;
+
+		ipv6_addr_prefix(&prefix, &ifp->addr, ifp->prefix_len);
+		rt = rt6_lookup(&prefix, NULL, ifp->idev->dev->ifindex, 1);
+
+		if (rt && ((rt->rt6i_flags & (RTF_GATEWAY | RTF_DEFAULT)) == 0)) {
+			if (onlink == 0) {
+				ip6_del_rt(rt, NULL, NULL);
+				rt = NULL;
+			} else if (!(rt->rt6i_flags & RTF_EXPIRES)) {
+				rt->rt6i_expires = expires;
+				rt->rt6i_flags |= RTF_EXPIRES;
+			}
+		}
+		dst_release(&rt->u.dst);
+	}
+
 	in6_ifa_put(ifp);
 }
 
@@ -1982,7 +2044,10 @@ static int addrconf_notify(struct notifier_block *this, unsigned long event,
 		if (idev) {
 			addrconf_sysctl_unregister(&idev->cnf);
 			neigh_sysctl_unregister(idev->nd_parms);
-			neigh_sysctl_register(dev, idev->nd_parms, NET_IPV6, NET_IPV6_NEIGH, "ipv6", &ndisc_ifinfo_sysctl_change);
+			neigh_sysctl_register(dev, idev->nd_parms,
+					      NET_IPV6, NET_IPV6_NEIGH, "ipv6",
+					      &ndisc_ifinfo_sysctl_change,
+					      NULL);
 			addrconf_sysctl_register(idev, &idev->cnf);
 		}
 #endif
@@ -3150,7 +3215,7 @@ static int addrconf_sysctl_forward_strategy(ctl_table *table,
 static struct addrconf_sysctl_table
 {
 	struct ctl_table_header *sysctl_header;
-	ctl_table addrconf_vars[18];
+	ctl_table addrconf_vars[__NET_IPV6_MAX];
 	ctl_table addrconf_dev[2];
 	ctl_table addrconf_conf_dir[2];
 	ctl_table addrconf_proto_dir[2];

net/ipv6/anycast.c

@@ -48,32 +48,6 @@ static int ipv6_dev_ac_dec(struct net_device *dev, struct in6_addr *addr);
 /* Big ac list lock for all the sockets */
 static DEFINE_RWLOCK(ipv6_sk_ac_lock);
 
-/* XXX ip6_addr_match() and ip6_onlink() really belong in net/core.c */
-
-static int
-ip6_addr_match(struct in6_addr *addr1, struct in6_addr *addr2, int prefix)
-{
-	__u32	mask;
-	int	i;
-
-	if (prefix > 128 || prefix < 0)
-		return 0;
-	if (prefix == 0)
-		return 1;
-	for (i=0; i<4; ++i) {
-		if (prefix >= 32)
-			mask = ~0;
-		else
-			mask = htonl(~0 << (32 - prefix));
-		if ((addr1->s6_addr32[i] ^ addr2->s6_addr32[i]) & mask)
-			return 0;
-		prefix -= 32;
-		if (prefix <= 0)
-			break;
-	}
-	return 1;
-}
-
 static int
 ip6_onlink(struct in6_addr *addr, struct net_device *dev)
 {
@@ -87,8 +61,8 @@ ip6_onlink(struct in6_addr *addr, struct net_device *dev)
 	if (idev) {
 		read_lock_bh(&idev->lock);
 		for (ifa=idev->addr_list; ifa; ifa=ifa->if_next) {
-			onlink = ip6_addr_match(addr, &ifa->addr,
-					ifa->prefix_len);
+			onlink = ipv6_prefix_equal(addr, &ifa->addr,
+						   ifa->prefix_len);
 			if (onlink)
 				break;
 		}

net/ipv6/icmp.c

@@ -381,6 +381,8 @@ void icmpv6_send(struct sk_buff *skb, int type, int code, __u32 info,
 		hlimit = np->hop_limit;
 	if (hlimit < 0)
 		hlimit = dst_metric(dst, RTAX_HOPLIMIT);
+	if (hlimit < 0)
+		hlimit = ipv6_get_hoplimit(dst->dev);
 
 	msg.skb = skb;
 	msg.offset = skb->nh.raw - skb->data;
@@ -467,6 +469,8 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
 		hlimit = np->hop_limit;
 	if (hlimit < 0)
 		hlimit = dst_metric(dst, RTAX_HOPLIMIT);
+	if (hlimit < 0)
+		hlimit = ipv6_get_hoplimit(dst->dev);
 
 	idev = in6_dev_get(skb->dev);
 

net/ipv6/ip6_fib.c

@@ -116,36 +116,6 @@ static __inline__ u32 fib6_new_sernum(void)
  *	64bit processors)
  */
 
-/*
- *	compare "prefix length" bits of an address
- */
-
-static __inline__ int addr_match(void *token1, void *token2, int prefixlen)
-{
-	__u32 *a1 = token1;
-	__u32 *a2 = token2;
-	int pdw;
-	int pbi;
-
-	pdw = prefixlen >> 5;	  /* num of whole __u32 in prefix */
-	pbi = prefixlen &  0x1f;  /* num of bits in incomplete u32 in prefix */
-
-	if (pdw)
-		if (memcmp(a1, a2, pdw << 2))
-			return 0;
-
-	if (pbi) {
-		__u32 mask;
-
-		mask = htonl((0xffffffff) << (32 - pbi));
-
-		if ((a1[pdw] ^ a2[pdw]) & mask)
-			return 0;
-	}
-
-	return 1;
-}
-
 /*
  *	test bit
  */
@@ -261,7 +231,7 @@ static struct fib6_node * fib6_add_1(struct fib6_node *root, void *addr,
 		 *	Prefix match
 		 */
 		if (plen < fn->fn_bit ||
-		    !addr_match(&key->addr, addr, fn->fn_bit))
+		    !ipv6_prefix_equal(&key->addr, addr, fn->fn_bit))
 			goto insert_above;
 		
 		/*
@@ -667,7 +637,7 @@ static struct fib6_node * fib6_lookup_1(struct fib6_node *root,
 			key = (struct rt6key *) ((u8 *) fn->leaf +
 						 args->offset);
 
-			if (addr_match(&key->addr, args->addr, key->plen))
+			if (ipv6_prefix_equal(&key->addr, args->addr, key->plen))
 				return fn;
 		}
 
@@ -718,7 +688,7 @@ static struct fib6_node * fib6_locate_1(struct fib6_node *root,
 		 *	Prefix match
 		 */
 		if (plen < fn->fn_bit ||
-		    !addr_match(&key->addr, addr, fn->fn_bit))
+		    !ipv6_prefix_equal(&key->addr, addr, fn->fn_bit))
 			return NULL;
 
 		if (plen == fn->fn_bit)
@@ -1211,7 +1181,7 @@ void fib6_run_gc(unsigned long dummy)
 {
 	if (dummy != ~0UL) {
 		spin_lock_bh(&fib6_gc_lock);
-		gc_args.timeout = (int)dummy;
+		gc_args.timeout = dummy ? (int)dummy : ip6_rt_gc_interval;
 	} else {
 		local_bh_disable();
 		if (!spin_trylock(&fib6_gc_lock)) {

net/ipv6/ip6_output.c

@@ -147,7 +147,7 @@ static int ip6_output2(struct sk_buff *skb)
 
 int ip6_output(struct sk_buff *skb)
 {
-	if (skb->len > dst_pmtu(skb->dst))
+	if (skb->len > dst_pmtu(skb->dst) || dst_allfrag(skb->dst))
 		return ip6_fragment(skb, ip6_output2);
 	else
 		return ip6_output2(skb);
@@ -253,6 +253,8 @@ int ip6_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl,
 		hlimit = np->hop_limit;
 	if (hlimit < 0)
 		hlimit = dst_metric(dst, RTAX_HOPLIMIT);
+	if (hlimit < 0)
+		hlimit = ipv6_get_hoplimit(dst->dev);
 
 	hdr->payload_len = htons(seg_len);
 	hdr->nexthdr = proto;
@@ -848,6 +850,8 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, int offse
 		inet->cork.fl = *fl;
 		np->cork.hop_limit = hlimit;
 		inet->cork.fragsize = mtu = dst_pmtu(&rt->u.dst);
+		if (dst_allfrag(&rt->u.dst))
+			inet->cork.flags |= IPCORK_ALLFRAG;
 		inet->cork.length = 0;
 		sk->sk_sndmsg_page = NULL;
 		sk->sk_sndmsg_off = 0;
@@ -899,7 +903,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, int offse
 
 	while (length > 0) {
 		/* Check if the remaining data fits into current packet. */
-		copy = mtu - skb->len;
+		copy = (inet->cork.length <= mtu && !(inet->cork.flags & IPCORK_ALLFRAG) ? mtu : maxfraglen) - skb->len;
 		if (copy < length)
 			copy = maxfraglen - skb->len;
 
@@ -924,7 +928,7 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, int offse
 			 * we know we need more fragment(s).
 			 */
 			datalen = length + fraggap;
-			if (datalen > mtu - fragheaderlen)
+			if (datalen > (inet->cork.length <= mtu && !(inet->cork.flags & IPCORK_ALLFRAG) ? mtu : maxfraglen) - fragheaderlen)
 				datalen = maxfraglen - fragheaderlen;
 
 			fraglen = datalen + fragheaderlen;
@@ -1158,6 +1162,7 @@ int ip6_push_pending_frames(struct sock *sk)
 	if (np->cork.rt) {
 		dst_release(&np->cork.rt->u.dst);
 		np->cork.rt = NULL;
+		inet->cork.flags &= ~IPCORK_ALLFRAG;
 	}
 	memset(&inet->cork.fl, 0, sizeof(inet->cork.fl));
 	return err;
@@ -1185,6 +1190,7 @@ void ip6_flush_pending_frames(struct sock *sk)
 	if (np->cork.rt) {
 		dst_release(&np->cork.rt->u.dst);
 		np->cork.rt = NULL;
+		inet->cork.flags &= ~IPCORK_ALLFRAG;
 	}
 	memset(&inet->cork.fl, 0, sizeof(inet->cork.fl));
 }

net/ipv6/netfilter/Kconfig

@@ -2,8 +2,8 @@
 # IP netfilter configuration
 #
 
-menu "IPv6: Netfilter Configuration"
-	depends on INET && IPV6 && NETFILTER
+menu "IPv6: Netfilter Configuration (EXPERIMENTAL)"
+	depends on INET && IPV6 && NETFILTER && EXPERIMENTAL
 
 #tristate 'Connection tracking (required for masq/NAT)' CONFIG_IP6_NF_CONNTRACK
 #if [ "$CONFIG_IP6_NF_CONNTRACK" != "n" ]; then

net/ipv6/raw.c

@@ -756,6 +756,8 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk,
 			hlimit = np->hop_limit;
 		if (hlimit < 0)
 			hlimit = dst_metric(dst, RTAX_HOPLIMIT);
+		if (hlimit < 0)
+			hlimit = ipv6_get_hoplimit(dst->dev);
 	}
 
 	if (msg->msg_flags&MSG_CONFIRM)

net/ipv6/route.c

@@ -628,8 +628,10 @@ static void ip6_rt_update_pmtu(struct dst_entry *dst, u32 mtu)
 
 	if (mtu < dst_pmtu(dst) && rt6->rt6i_dst.plen == 128) {
 		rt6->rt6i_flags |= RTF_MODIFIED;
-		if (mtu < IPV6_MIN_MTU)
+		if (mtu < IPV6_MIN_MTU) {
 			mtu = IPV6_MIN_MTU;
+			dst->metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
+		}
 		dst->metrics[RTAX_MTU-1] = mtu;
 	}
 }
@@ -769,7 +771,7 @@ static int ipv6_get_mtu(struct net_device *dev)
 	return mtu;
 }
 
-static int ipv6_get_hoplimit(struct net_device *dev)
+int ipv6_get_hoplimit(struct net_device *dev)
 {
 	int hoplimit = ipv6_devconf.hop_limit;
 	struct inet6_dev *idev;
@@ -965,15 +967,8 @@ int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh, void *_rtattr)
 		}
 	}
 
-	if (rt->u.dst.metrics[RTAX_HOPLIMIT-1] == 0) {
-		if (ipv6_addr_is_multicast(&rt->rt6i_dst.addr))
-			rt->u.dst.metrics[RTAX_HOPLIMIT-1] =
-				IPV6_DEFAULT_MCASTHOPS;
-		else
-			rt->u.dst.metrics[RTAX_HOPLIMIT-1] =
-				ipv6_get_hoplimit(dev);
-	}
-
+	if (rt->u.dst.metrics[RTAX_HOPLIMIT-1] == 0)
+		rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1;
 	if (!rt->u.dst.metrics[RTAX_MTU-1])
 		rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(dev);
 	if (!rt->u.dst.metrics[RTAX_ADVMSS-1])
@@ -1164,26 +1159,26 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 			struct net_device *dev, u32 pmtu)
 {
 	struct rt6_info *rt, *nrt;
-
-	if (pmtu < IPV6_MIN_MTU) {
-		if (net_ratelimit())
-			printk(KERN_DEBUG "rt6_pmtu_discovery: invalid MTU value %d\n",
-			       pmtu);
-		/* According to RFC1981, the PMTU is set to the IPv6 minimum
-		   link MTU if the node receives a Packet Too Big message
-		   reporting next-hop MTU that is less than the IPv6 minimum MTU.
-		   */
-		pmtu = IPV6_MIN_MTU;
-	}
+	int allfrag = 0;
 
 	rt = rt6_lookup(daddr, saddr, dev->ifindex, 0);
-
 	if (rt == NULL)
 		return;
 
 	if (pmtu >= dst_pmtu(&rt->u.dst))
 		goto out;
 
+	if (pmtu < IPV6_MIN_MTU) {
+		/*
+		 * According to RFC2460, PMTU is set to the IPv6 Minimum Link 
+		 * MTU (1280) and a fragment header should always be included
+		 * after a node receiving Too Big message reporting PMTU is
+		 * less than the IPv6 Minimum Link MTU.
+		 */
+		pmtu = IPV6_MIN_MTU;
+		allfrag = 1;
+	}
+
 	/* New mtu received -> path was valid.
 	   They are sent only in response to data packets,
 	   so that this nexthop apparently is reachable. --ANK
@@ -1197,6 +1192,8 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 	 */
 	if (rt->rt6i_flags & RTF_CACHE) {
 		rt->u.dst.metrics[RTAX_MTU-1] = pmtu;
+		if (allfrag)
+			rt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
 		dst_set_expires(&rt->u.dst, ip6_rt_mtu_expires);
 		rt->rt6i_flags |= RTF_MODIFIED|RTF_EXPIRES;
 		goto out;
@@ -1211,6 +1208,8 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 		nrt = rt6_cow(rt, daddr, saddr);
 		if (!nrt->u.dst.error) {
 			nrt->u.dst.metrics[RTAX_MTU-1] = pmtu;
+			if (allfrag)
+				nrt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
 			/* According to RFC 1981, detecting PMTU increase shouldn't be
 			   happened within 5 mins, the recommended timer is 10 mins.
 			   Here this route expiration time is set to ip6_rt_mtu_expires
@@ -1232,6 +1231,8 @@ void rt6_pmtu_discovery(struct in6_addr *daddr, struct in6_addr *saddr,
 		dst_set_expires(&nrt->u.dst, ip6_rt_mtu_expires);
 		nrt->rt6i_flags |= RTF_DYNAMIC|RTF_CACHE|RTF_EXPIRES;
 		nrt->u.dst.metrics[RTAX_MTU-1] = pmtu;
+		if (allfrag)
+			nrt->u.dst.metrics[RTAX_FEATURES-1] |= RTAX_FEATURE_ALLFRAG;
 		ip6_ins_rt(nrt, NULL, NULL);
 	}
 
@@ -1406,7 +1407,7 @@ struct rt6_info *addrconf_dst_alloc(struct inet6_dev *idev,
 	rt->rt6i_idev = idev;
 	rt->u.dst.metrics[RTAX_MTU-1] = ipv6_get_mtu(rt->rt6i_dev);
 	rt->u.dst.metrics[RTAX_ADVMSS-1] = ipv6_advmss(dst_pmtu(&rt->u.dst));
-	rt->u.dst.metrics[RTAX_HOPLIMIT-1] = ipv6_get_hoplimit(rt->rt6i_dev);
+	rt->u.dst.metrics[RTAX_HOPLIMIT-1] = -1;
 	rt->u.dst.obsolete = -1;
 
 	rt->rt6i_flags = RTF_UP | RTF_NONEXTHOP;
@@ -1993,9 +1994,7 @@ int ipv6_sysctl_rtcache_flush(ctl_table *ctl, int write, struct file * filp,
 {
 	if (write) {
 		proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
-		if (flush_delay < 0)
-			flush_delay = 0;
-		fib6_run_gc((unsigned long)flush_delay);
+		fib6_run_gc(flush_delay <= 0 ? ~0UL : (unsigned long)flush_delay);
 		return 0;
 	} else
 		return -EINVAL;
@@ -2080,6 +2079,15 @@ ctl_table ipv6_route_table[] = {
          	.proc_handler	=	&proc_dointvec_jiffies,
 		.strategy	=	&sysctl_jiffies,
 	},
+	{
+		.ctl_name	=	NET_IPV6_ROUTE_GC_MIN_INTERVAL_MS,
+		.procname	=	"gc_min_interval_ms",
+         	.data		=	&ip6_rt_gc_min_interval,
+		.maxlen		=	sizeof(int),
+		.mode		=	0644,
+         	.proc_handler	=	&proc_dointvec_ms_jiffies,
+		.strategy	=	&sysctl_ms_jiffies,
+	},
 	{ .ctl_name = 0 }
 };
 

net/ipv6/udp.c

@@ -811,6 +811,8 @@ static int udpv6_sendmsg(struct kiocb *iocb, struct sock *sk,
 			hlimit = np->hop_limit;
 		if (hlimit < 0)
 			hlimit = dst_metric(dst, RTAX_HOPLIMIT);
+		if (hlimit < 0)
+			hlimit = ipv6_get_hoplimit(dst->dev);
 	}
 
 	if (msg->msg_flags&MSG_CONFIRM)