Merge davem@nuts.ninka.net:/home/davem/src/BK/net-2.5

into kernel.bkbits.net:/home/davem/net-2.5

include/linux/if_bridge.h

@@ -101,7 +101,7 @@ struct __fdb_entry
 struct net_bridge;
 struct net_bridge_port;
 
-extern int (*br_ioctl_hook)(unsigned long arg);
+extern void brioctl_set(int (*ioctl_hook)(unsigned long));
 extern int (*br_handle_frame_hook)(struct sk_buff *skb);
 extern int (*br_should_route_hook)(struct sk_buff **pskb);
 

include/linux/igmp.h

@@ -32,13 +32,60 @@ struct igmphdr
 	__u32 group;
 };
 
+/* V3 group record types [grec_type] */
+#define IGMPV3_MODE_IS_INCLUDE		1
+#define IGMPV3_MODE_IS_EXCLUDE		2
+#define IGMPV3_CHANGE_TO_INCLUDE	3
+#define IGMPV3_CHANGE_TO_EXCLUDE	4
+#define IGMPV3_ALLOW_NEW_SOURCES	5
+#define IGMPV3_BLOCK_OLD_SOURCES	6
+
+struct igmpv3_grec {
+	__u8	grec_type;
+	__u8	grec_auxwords;
+	__u16	grec_nsrcs;
+	__u32	grec_mca;
+	__u32	grec_src[0];
+};
+
+struct igmpv3_report {
+	__u8 type;
+	__u8 resv1;
+	__u16 csum;
+	__u16 resv2;
+	__u16 ngrec;
+	struct igmpv3_grec grec[0];
+};
+
+struct igmpv3_query {
+	__u8 type;
+	__u8 code;
+	__u16 csum;
+	__u32 group;
+#if defined(__LITTLE_ENDIAN_BITFIELD)
+	__u8 qrv:3,
+	     suppress:1,
+	     resv:4;
+#elif defined(__BIG_ENDIAN_BITFIELD)
+	__u8 resv:4,
+	     suppress:1,
+	     qrv:3;
+#else
+#error "Please fix <asm/byteorder.h>"
+#endif
+	__u8 qqic;
+	__u16 nsrcs;
+	__u32 srcs[0];
+};
+
 #define IGMP_HOST_MEMBERSHIP_QUERY	0x11	/* From RFC1112 */
 #define IGMP_HOST_MEMBERSHIP_REPORT	0x12	/* Ditto */
 #define IGMP_DVMRP			0x13	/* DVMRP routing */
 #define IGMP_PIM			0x14	/* PIM routing */
 #define IGMP_TRACE			0x15
-#define IGMP_HOST_NEW_MEMBERSHIP_REPORT 0x16	/* New version of 0x11 */
+#define IGMPV2_HOST_MEMBERSHIP_REPORT	0x16	/* V2 version of 0x11 */
 #define IGMP_HOST_LEAVE_MESSAGE 	0x17
+#define IGMPV3_HOST_MEMBERSHIP_REPORT	0x22	/* V3 version of 0x11 */
 
 #define IGMP_MTRACE_RESP		0x1e
 #define IGMP_MTRACE			0x1f
@@ -68,6 +115,7 @@ struct igmphdr
 
 #define IGMP_ALL_HOSTS		htonl(0xE0000001L)
 #define IGMP_ALL_ROUTER 	htonl(0xE0000002L)
+#define IGMPV3_ALL_MCR	 	htonl(0xE0000016L)
 #define IGMP_LOCAL_GROUP	htonl(0xE0000000L)
 #define IGMP_LOCAL_GROUP_MASK	htonl(0xFFFFFF00L)
 
@@ -79,6 +127,18 @@ struct igmphdr
 #include <linux/skbuff.h>
 #include <linux/in.h>
 
+struct ip_sf_socklist
+{
+	unsigned int		sl_max;
+	unsigned int		sl_count;
+	__u32			sl_addr[0];
+};
+
+#define IP_SFLSIZE(count)	(sizeof(struct ip_sf_socklist) + \
+	(count) * sizeof(__u32))
+
+#define IP_SFBLOCK	10	/* allocate this many at once */
+
 /* ip_mc_socklist is real list now. Speed is not argument;
    this list never used in fast path code
  */
@@ -88,12 +148,28 @@ struct ip_mc_socklist
 	struct ip_mc_socklist	*next;
 	int			count;
 	struct ip_mreqn		multi;
+	unsigned int		sfmode;		/* MCAST_{INCLUDE,EXCLUDE} */
+	struct ip_sf_socklist	*sflist;
+};
+
+struct ip_sf_list
+{
+	struct ip_sf_list	*sf_next;
+	__u32			sf_inaddr;
+	unsigned long		sf_count[2];	/* include/exclude counts */
+	unsigned char		sf_gsresp;	/* include in g & s response? */
+	unsigned char		sf_oldin;	/* change state */
+	unsigned char		sf_crcount;	/* retrans. left to send */
 };
 
 struct ip_mc_list
 {
 	struct in_device	*interface;
 	unsigned long		multiaddr;
+	struct ip_sf_list	*sources;
+	struct ip_sf_list	*tomb;
+	unsigned int		sfmode;
+	unsigned long		sfcount[2];
 	struct ip_mc_list	*next;
 	struct timer_list	timer;
 	int			users;
@@ -103,13 +179,31 @@ struct ip_mc_list
 	char			reporter;
 	char			unsolicit_count;
 	char			loaded;
+	unsigned char		gsquery;	/* check source marks? */
+	unsigned char		crcount;
 };
 
-extern int ip_check_mc(struct in_device *dev, u32 mc_addr);
+/* V3 exponential field decoding */
+#define IGMPV3_MASK(value, nb) ((nb)>=32 ? (value) : ((1<<(nb))-1) & (value))
+#define IGMPV3_EXP(thresh, nbmant, nbexp, value) \
+	((value) < (thresh) ? (value) : \
+        ((IGMPV3_MASK(value, nbmant) | (1<<(nbmant+nbexp))) << \
+         (IGMPV3_MASK((value) >> (nbmant), nbexp) + (nbexp))))
+
+#define IGMPV3_QQIC(value) IGMPV3_EXP(0x80, 4, 3, value)
+#define IGMPV3_MRC(value) IGMPV3_EXP(0x8000, 12, 3, value)
+
+extern int ip_check_mc(struct in_device *dev, u32 mc_addr, u32 src_addr, u16 proto);
 extern int igmp_rcv(struct sk_buff *);
 extern int ip_mc_join_group(struct sock *sk, struct ip_mreqn *imr);
 extern int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr);
 extern void ip_mc_drop_socket(struct sock *sk);
+extern int ip_mc_source(int add, int omode, struct sock *sk,
+		struct ip_mreq_source *mreqs);
+extern int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf);
+extern int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
+		struct ip_msfilter *optval, int *optlen);
+extern int ip_mc_sf_allow(struct sock *sk, u32 local, u32 rmt, int dif);
 extern void ip_mr_init(void);
 extern void ip_mc_init_dev(struct in_device *);
 extern void ip_mc_destroy_dev(struct in_device *);

include/linux/in.h

@@ -85,6 +85,14 @@ struct in_addr {
 #define IP_MULTICAST_LOOP 		34
 #define IP_ADD_MEMBERSHIP		35
 #define IP_DROP_MEMBERSHIP		36
+#define IP_UNBLOCK_SOURCE		37
+#define IP_BLOCK_SOURCE			38
+#define IP_ADD_SOURCE_MEMBERSHIP	39
+#define IP_DROP_SOURCE_MEMBERSHIP	40
+#define IP_MSFILTER			41
+
+#define MCAST_EXCLUDE	0
+#define MCAST_INCLUDE	1
 
 /* These need to appear somewhere around here */
 #define IP_DEFAULT_MULTICAST_TTL        1
@@ -105,6 +113,24 @@ struct ip_mreqn
 	int		imr_ifindex;		/* Interface index */
 };
 
+struct ip_mreq_source {
+	__u32		imr_multiaddr;
+	__u32		imr_interface;
+	__u32		imr_sourceaddr;
+};
+
+struct ip_msfilter {
+	__u32		imsf_multiaddr;
+	__u32		imsf_interface;
+	__u32		imsf_fmode;
+	__u32		imsf_numsrc;
+	__u32		imsf_slist[1];
+};
+
+#define IP_MSFILTER_SIZE(numsrc) \
+	(sizeof(struct ip_msfilter) - sizeof(__u32) \
+	+ (numsrc) * sizeof(__u32))
+
 struct in_pktinfo
 {
 	int		ipi_ifindex;

include/linux/inetdevice.h

@@ -34,7 +34,17 @@ struct in_device
 	int			dead;
 	struct in_ifaddr	*ifa_list;	/* IP ifaddr chain		*/
 	struct ip_mc_list	*mc_list;	/* IP multicast filter chain    */
+	rwlock_t		mc_lock;	/* for mc_tomb */
+	struct ip_mc_list	*mc_tomb;
 	unsigned long		mr_v1_seen;
+	unsigned long		mr_v2_seen;
+	unsigned long		mr_maxdelay;
+	unsigned char		mr_qrv;
+	unsigned char		mr_gq_running;
+	unsigned char		mr_ifc_count;
+	struct timer_list	mr_gq_timer;	/* general query timer */
+	struct timer_list	mr_ifc_timer;	/* interface change timer */
+
 	struct neigh_parms	*arp_parms;
 	struct ipv4_devconf	cnf;
 };

include/net/ip.h

@@ -79,6 +79,7 @@ extern rwlock_t ip_ra_lock;
 extern void		ip_mc_dropsocket(struct sock *);
 extern void		ip_mc_dropdevice(struct net_device *dev);
 extern int		ip_mc_procinfo(char *, char **, off_t, int);
+extern int		ip_mcf_procinfo(char *, char **, off_t, int);
 
 /*
  *	Functions provided by ip.c

include/net/ip6_fib.h

@@ -72,6 +72,8 @@ struct rt6_info
 
 	struct rt6key			rt6i_dst;
 	struct rt6key			rt6i_src;
+
+	u8				rt6i_protocol;
 };
 
 struct fib6_walker_t
@@ -160,11 +162,14 @@ extern int			fib6_walk(struct fib6_walker_t *w);
 extern int			fib6_walk_continue(struct fib6_walker_t *w);
 
 extern int			fib6_add(struct fib6_node *root,
-					 struct rt6_info *rt);
+					 struct rt6_info *rt,
+					 struct nlmsghdr *nlh);
 
-extern int			fib6_del(struct rt6_info *rt);
+extern int			fib6_del(struct rt6_info *rt,
+					 struct nlmsghdr *nlh);
 
-extern void			inet6_rt_notify(int event, struct rt6_info *rt);
+extern void			inet6_rt_notify(int event, struct rt6_info *rt,
+						struct nlmsghdr *nlh);
 
 extern void			fib6_run_gc(unsigned long dummy);
 

include/net/ip6_route.h

@@ -37,8 +37,10 @@ extern void			ip6_route_cleanup(void);
 
 extern int			ipv6_route_ioctl(unsigned int cmd, void *arg);
 
-extern int			ip6_route_add(struct in6_rtmsg *rtmsg);
-extern int			ip6_del_rt(struct rt6_info *);
+extern int			ip6_route_add(struct in6_rtmsg *rtmsg,
+					      struct nlmsghdr *);
+extern int			ip6_del_rt(struct rt6_info *,
+					   struct nlmsghdr *);
 
 extern int			ip6_rt_addr_add(struct in6_addr *addr,
 						struct net_device *dev);

net/bridge/br.c

@@ -49,8 +49,9 @@ static int __init br_init(void)
 	if (br_netfilter_init())
 		return 1;
 #endif
+	brioctl_set(br_ioctl_deviceless_stub);
 	br_handle_frame_hook = br_handle_frame;
-	br_ioctl_hook = br_ioctl_deviceless_stub;
+
 #if defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE)
 	br_fdb_get_hook = br_fdb_get;
 	br_fdb_put_hook = br_fdb_put;
@@ -60,24 +61,18 @@ static int __init br_init(void)
 	return 0;
 }
 
-static void __br_clear_ioctl_hook(void)
-{
-	br_ioctl_hook = NULL;
-}
-
 static void __exit br_deinit(void)
 {
 #ifdef CONFIG_NETFILTER
 	br_netfilter_fini();
 #endif
 	unregister_netdevice_notifier(&br_device_notifier);
-	br_call_ioctl_atomic(__br_clear_ioctl_hook);
 
-	br_write_lock_bh(BR_NETPROTO_LOCK);
+	brioctl_set(NULL);
 	br_handle_frame_hook = NULL;
-	br_write_unlock_bh(BR_NETPROTO_LOCK);
 
 #if defined(CONFIG_ATM_LANE) || defined(CONFIG_ATM_LANE_MODULE)
+	/* FIX ME. move into hook structure with ref count */
 	br_fdb_get_hook = NULL;
 	br_fdb_put_hook = NULL;
 #endif

net/bridge/br_if.c

@@ -23,6 +23,7 @@
 #include "br_private.h"
 
 static struct net_bridge *bridge_list;
+static spinlock_t bridge_lock = SPIN_LOCK_UNLOCKED;
 
 static int br_initial_port_cost(struct net_device *dev)
 {
@@ -69,6 +70,7 @@ static int __br_del_if(struct net_bridge *br, struct net_device *dev)
 	return 0;
 }
 
+/* called with bridge_lock */
 static struct net_bridge **__find_br(char *name)
 {
 	struct net_bridge **b;
@@ -188,8 +190,10 @@ int br_add_bridge(char *name)
 		return -EEXIST;
 	}
 
+	spin_lock(&bridge_lock);
 	br->next = bridge_list;
 	bridge_list = br;
+	spin_unlock(&bridge_lock);
 
 	br_inc_use_count();
 	register_netdev(&br->dev);
@@ -202,17 +206,22 @@ int br_del_bridge(char *name)
 	struct net_bridge **b;
 	struct net_bridge *br;
 
-	if ((b = __find_br(name)) == NULL)
+	spin_lock(&bridge_lock);
+	if ((b = __find_br(name)) == NULL) {
+		spin_unlock(&bridge_lock);
 		return -ENXIO;
+	}
 
 	br = *b;
-
-	if (br->dev.flags & IFF_UP)
+	if (br->dev.flags & IFF_UP) {
+		spin_unlock(&bridge_lock);
 		return -EBUSY;
-
-	del_ifs(br);
+	}
 
 	*b = br->next;
+	spin_unlock(&bridge_lock);
+
+	del_ifs(br);
 
 	unregister_netdev(&br->dev);
 	kfree(br);
@@ -272,6 +281,7 @@ int br_get_bridge_ifindices(int *indices, int num)
 	struct net_bridge *br;
 	int i;
 
+	spin_lock(&bridge_lock);
 	br = bridge_list;
 	for (i=0;i<num;i++) {
 		if (br == NULL)
@@ -280,6 +290,7 @@ int br_get_bridge_ifindices(int *indices, int num)
 		indices[i] = br->dev.ifindex;
 		br = br->next;
 	}
+	spin_unlock(&bridge_lock);
 
 	return i;
 }
@@ -289,9 +300,11 @@ void br_get_port_ifindices(struct net_bridge *br, int *ifindices)
 {
 	struct net_bridge_port *p;
 
+	read_lock(&br->lock);
 	p = br->port_list;
 	while (p != NULL) {
 		ifindices[p->port_no] = p->dev->ifindex;
 		p = p->next;
 	}
+	read_unlock(&br->lock);
 }

net/bridge/br_ioctl.c

@@ -53,6 +53,7 @@ static int br_ioctl_device(struct net_bridge *br,
 	{
 		struct __bridge_info b;
 
+	        read_lock(&br->lock);
 		memset(&b, 0, sizeof(struct __bridge_info));
 		memcpy(&b.designated_root, &br->designated_root, 8);
 		memcpy(&b.bridge_id, &br->bridge_id, 8);
@@ -73,6 +74,7 @@ static int br_ioctl_device(struct net_bridge *br,
 		b.tcn_timer_value = br_timer_get_residue(&br->tcn_timer);
 		b.topology_change_timer_value = br_timer_get_residue(&br->topology_change_timer);
 		b.gc_timer_value = br_timer_get_residue(&br->gc_timer);
+	        read_unlock(&br->lock);
 
 		if (copy_to_user((void *)arg0, &b, sizeof(b)))
 			return -EFAULT;
@@ -96,21 +98,27 @@ static int br_ioctl_device(struct net_bridge *br,
 	}
 
 	case BRCTL_SET_BRIDGE_FORWARD_DELAY:
+		write_lock(&br->lock);
 		br->bridge_forward_delay = arg0;
 		if (br_is_root_bridge(br))
 			br->forward_delay = arg0;
+		write_unlock(&br->lock);
 		return 0;
 
 	case BRCTL_SET_BRIDGE_HELLO_TIME:
+		write_lock(&br->lock);
 		br->bridge_hello_time = arg0;
 		if (br_is_root_bridge(br))
 			br->hello_time = arg0;
+		write_unlock(&br->lock);
 		return 0;
 
 	case BRCTL_SET_BRIDGE_MAX_AGE:
+		write_lock(&br->lock);
 		br->bridge_max_age = arg0;
 		if (br_is_root_bridge(br))
 			br->max_age = arg0;
+		write_unlock(&br->lock);
 		return 0;
 
 	case BRCTL_SET_AGEING_TIME:
@@ -126,6 +134,7 @@ static int br_ioctl_device(struct net_bridge *br,
 		struct __port_info p;
 		struct net_bridge_port *pt;
 
+		read_lock(&br->lock);
 		if ((pt = br_get_port(br, arg1)) == NULL)
 			return -EINVAL;
 
@@ -143,6 +152,8 @@ static int br_ioctl_device(struct net_bridge *br,
 		p.forward_delay_timer_value = br_timer_get_residue(&pt->forward_delay_timer);
 		p.hold_timer_value = br_timer_get_residue(&pt->hold_timer);
 
+		read_unlock(&br->lock);
+
 		if (copy_to_user((void *)arg0, &p, sizeof(p)))
 			return -EFAULT;
 
@@ -154,16 +165,20 @@ static int br_ioctl_device(struct net_bridge *br,
 		return 0;
 
 	case BRCTL_SET_BRIDGE_PRIORITY:
+		write_lock(&br->lock);
 		br_stp_set_bridge_priority(br, arg0);
+		write_unlock(&br->lock);
 		return 0;
 
 	case BRCTL_SET_PORT_PRIORITY:
 	{
 		struct net_bridge_port *p;
 
+		write_lock(&br->lock);
 		if ((p = br_get_port(br, arg0)) == NULL)
 			return -EINVAL;
 		br_stp_set_port_priority(p, arg1);
+		write_unlock(&br->lock);
 		return 0;
 	}
 
@@ -171,9 +186,11 @@ static int br_ioctl_device(struct net_bridge *br,
 	{
 		struct net_bridge_port *p;
 
+		write_lock(&br->lock);
 		if ((p = br_get_port(br, arg0)) == NULL)
 			return -EINVAL;
 		br_stp_set_path_cost(p, arg1);
+		write_unlock(&br->lock);
 		return 0;
 	}
 
@@ -230,11 +247,9 @@ static int br_ioctl_deviceless(unsigned int cmd,
 	return -EOPNOTSUPP;
 }
 
-static DECLARE_MUTEX(ioctl_mutex);
 
 int br_ioctl_deviceless_stub(unsigned long arg)
 {
-	int err;
 	unsigned long i[3];
 
 	if (!capable(CAP_NET_ADMIN))
@@ -243,11 +258,7 @@ int br_ioctl_deviceless_stub(unsigned long arg)
 	if (copy_from_user(i, (void *)arg, 3*sizeof(unsigned long)))
 		return -EFAULT;
 
-	down(&ioctl_mutex);
-	err = br_ioctl_deviceless(i[0], i[1], i[2]);
-	up(&ioctl_mutex);
-
-	return err;
+	return br_ioctl_deviceless(i[0], i[1], i[2]);
 }
 
 int br_ioctl(struct net_bridge *br, unsigned int cmd, unsigned long arg0, unsigned long arg1, unsigned long arg2)
@@ -257,18 +268,9 @@ int br_ioctl(struct net_bridge *br, unsigned int cmd, unsigned long arg0, unsign
 	if (!capable(CAP_NET_ADMIN))
 		return -EPERM;
 
-	down(&ioctl_mutex);
 	err = br_ioctl_deviceless(cmd, arg0, arg1);
 	if (err == -EOPNOTSUPP)
 		err = br_ioctl_device(br, cmd, arg0, arg1, arg2);
-	up(&ioctl_mutex);
 
 	return err;
 }
-
-void br_call_ioctl_atomic(void (*fn)(void))
-{
-	down(&ioctl_mutex);
-	fn();
-	up(&ioctl_mutex);
-}

net/bridge/br_notify.c

@@ -21,15 +21,14 @@ static int br_device_event(struct notifier_block *unused, unsigned long event, v
 
 struct notifier_block br_device_notifier =
 {
-	br_device_event,
-	NULL,
-	0
+	.notifier_call = br_device_event
 };
 
 static int br_device_event(struct notifier_block *unused, unsigned long event, void *ptr)
 {
 	struct net_device *dev;
 	struct net_bridge_port *p;
+	struct net_bridge *br;
 
 	dev = ptr;
 	p = dev->br_port;
@@ -37,13 +36,15 @@ static int br_device_event(struct notifier_block *unused, unsigned long event, v
 	if (p == NULL)
 		return NOTIFY_DONE;
 
-	switch (event)
+	br = p->br;
+
+	switch (event) 
 	{
 	case NETDEV_CHANGEADDR:
-		read_lock(&p->br->lock);
+		write_lock_bh(&br->lock);
 		br_fdb_changeaddr(p, dev->dev_addr);
-		br_stp_recalculate_bridge_id(p->br);
-		read_unlock(&p->br->lock);
+		br_stp_recalculate_bridge_id(br);
+		write_unlock_bh(&br->lock);
 		break;
 
 	case NETDEV_GOING_DOWN:
@@ -51,23 +52,23 @@ static int br_device_event(struct notifier_block *unused, unsigned long event, v
 		break;
 
 	case NETDEV_DOWN:
-		if (p->br->dev.flags & IFF_UP) {
-			read_lock(&p->br->lock);
-			br_stp_disable_port(dev->br_port);
-			read_unlock(&p->br->lock);
+		if (br->dev.flags & IFF_UP) {
+			write_lock_bh(&br->lock);
+			br_stp_disable_port(p);
+			write_unlock_bh(&br->lock);
 		}
 		break;
 
 	case NETDEV_UP:
-		if (p->br->dev.flags & IFF_UP) {
-			read_lock(&p->br->lock);
-			br_stp_enable_port(dev->br_port);
-			read_unlock(&p->br->lock);
+		if (!(br->dev.flags & IFF_UP)) {
+			write_lock_bh(&br->lock);
+			br_stp_enable_port(p);
+			write_unlock_bh(&br->lock);
 		}
 		break;
 
 	case NETDEV_UNREGISTER:
-		br_del_if(dev->br_port->br, dev);
+		br_del_if(br, dev);
 		break;
 	}
 

net/core/dev.c

@@ -1433,9 +1433,8 @@ static void net_tx_action(struct softirq_action *h)
 	}
 }
 
-#if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
+#if defined(CONFIG_BRIDGE) || defined (CONFIG_BRIDGE_MODULE)
 int (*br_handle_frame_hook)(struct sk_buff *skb) = NULL;
-#endif
 
 static __inline__ int handle_bridge(struct sk_buff *skb,
 				     struct packet_type *pt_prev)
@@ -1454,6 +1453,7 @@ static __inline__ int handle_bridge(struct sk_buff *skb,
 	return ret;
 }
 
+#endif
 
 #ifdef CONFIG_NET_DIVERT
 static inline int handle_diverter(struct sk_buff *skb)
@@ -1510,12 +1510,13 @@ int netif_receive_skb(struct sk_buff *skb)
 #endif /* CONFIG_NET_DIVERT */
 
 #if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
-	if (skb->dev->br_port && br_handle_frame_hook) {
+	if (skb->dev->br_port) {
 		int ret;
 
 		ret = handle_bridge(skb, pt_prev);
-		if (br_handle_frame_hook(skb) == 0)
+		if (br_handle_frame_hook(skb) == 0) 
 			return ret;
+
 		pt_prev = NULL;
 	}
 #endif

net/ipv4/Kconfig

@@ -362,5 +362,13 @@ config INET_ESP
 
 	  If unsure, say Y.
 
+config INET_IPCOMP
+	tristate "IP: IPComp transformation"
+	---help---
+	  Support for IP Paylod Compression (RFC3173), typically needed
+	  for IPsec.
+	  
+	  If unsure, say Y.
+
 source "net/ipv4/netfilter/Kconfig"
 

net/ipv4/Makefile

@@ -18,6 +18,7 @@ obj-$(CONFIG_NET_IPGRE) += ip_gre.o
 obj-$(CONFIG_SYN_COOKIES) += syncookies.o
 obj-$(CONFIG_INET_AH) += ah.o
 obj-$(CONFIG_INET_ESP) += esp.o
+obj-$(CONFIG_INET_IPCOMP) += ipcomp.o
 obj-$(CONFIG_IP_PNP) += ipconfig.o
 obj-$(CONFIG_NETFILTER)	+= netfilter/
 

net/ipv4/igmp.c

@@ -68,6 +68,8 @@
  *		Alan Cox:		Forget to enable FDDI support earlier.
  *		Alexey Kuznetsov:	Fixed leaving groups on device down.
  *		Alexey Kuznetsov:	Accordance to igmp-v2-06 draft.
+ *		David L Stevens:	IGMPv3 support, with help from
+ *					Vinay Kulkarni
  */
 
 
@@ -101,12 +103,10 @@
 
 #define IP_MAX_MEMBERSHIPS 20
 
-#ifdef CONFIG_IP_MULTICAST
-
-
 /* Parameter names and values are taken from igmp-v2-06 draft */
 
 #define IGMP_V1_Router_Present_Timeout		(400*HZ)
+#define IGMP_V2_Router_Present_Timeout		(400*HZ)
 #define IGMP_Unsolicited_Report_Interval	(10*HZ)
 #define IGMP_Query_Response_Interval		(10*HZ)
 #define IGMP_Unsolicited_Report_Count		2
@@ -121,9 +121,21 @@
  * contradict to specs provided this delay is small enough.
  */
 
-#define IGMP_V1_SEEN(in_dev) ((in_dev)->mr_v1_seen && (long)(jiffies - (in_dev)->mr_v1_seen) < 0)
+#define IGMP_V1_SEEN(in_dev) ((in_dev)->mr_v1_seen && \
+		time_before(jiffies, (in_dev)->mr_v1_seen))
+#define IGMP_V2_SEEN(in_dev) ((in_dev)->mr_v2_seen && \
+		time_before(jiffies, (in_dev)->mr_v2_seen))
 
+#ifdef CONFIG_MULTICAST
+static void igmpv3_add_delrec(struct in_device *in_dev, struct ip_mc_list *im);
 #endif
+static void igmpv3_del_delrec(struct in_device *in_dev, __u32 multiaddr);
+static void igmpv3_clear_delrec(struct in_device *in_dev);
+static int sf_setstate(struct ip_mc_list *pmc);
+static void sf_markstate(struct ip_mc_list *pmc);
+static void ip_mc_clear_src(struct ip_mc_list *pmc);
+int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
+	int sfcount, __u32 *psfsrc, int delta);
 
 static void ip_ma_put(struct ip_mc_list *im)
 {
@@ -160,6 +172,23 @@ static void igmp_start_timer(struct ip_mc_list *im, int max_delay)
 		atomic_inc(&im->refcnt);
 }
 
+static void igmp_gq_start_timer(struct in_device *in_dev)
+{
+	int tv = net_random() % in_dev->mr_maxdelay;
+
+	in_dev->mr_gq_running = 1;
+	if (!mod_timer(&in_dev->mr_gq_timer, jiffies+tv+2))
+		atomic_inc(&in_dev->refcnt);
+}
+
+static void igmp_ifc_start_timer(struct in_device *in_dev, int delay)
+{
+	int tv = net_random() % delay;
+
+	if (!mod_timer(&in_dev->mr_ifc_timer, jiffies+tv+2))
+		atomic_inc(&in_dev->refcnt);
+}
+
 static void igmp_mod_timer(struct ip_mc_list *im, int max_delay)
 {
 	spin_lock_bh(&im->lock);
@@ -184,20 +213,396 @@ static void igmp_mod_timer(struct ip_mc_list *im, int max_delay)
 
 #define IGMP_SIZE (sizeof(struct igmphdr)+sizeof(struct iphdr)+4)
 
-static int igmp_send_report(struct net_device *dev, u32 group, int type)
+
+static int is_in(struct ip_mc_list *pmc, struct ip_sf_list *psf, int type,
+	int gdeleted, int sdeleted)
+{
+	switch (type) {
+	case IGMPV3_MODE_IS_INCLUDE:
+	case IGMPV3_MODE_IS_EXCLUDE:
+		if (gdeleted || sdeleted)
+			return 0;
+		return !(pmc->gsquery && !psf->sf_gsresp);
+	case IGMPV3_CHANGE_TO_INCLUDE:
+		if (gdeleted || sdeleted)
+			return 0;
+		return psf->sf_count[MCAST_INCLUDE] != 0;
+	case IGMPV3_CHANGE_TO_EXCLUDE:
+		if (gdeleted || sdeleted)
+			return 0;
+		if (pmc->sfcount[MCAST_EXCLUDE] == 0 ||
+		    psf->sf_count[MCAST_INCLUDE])
+			return 0;
+		return pmc->sfcount[MCAST_EXCLUDE] ==
+			psf->sf_count[MCAST_EXCLUDE];
+	case IGMPV3_ALLOW_NEW_SOURCES:
+		if (gdeleted || !psf->sf_crcount)
+			return 0;
+		return (pmc->sfmode == MCAST_INCLUDE) ^ sdeleted;
+	case IGMPV3_BLOCK_OLD_SOURCES:
+		if (pmc->sfmode == MCAST_INCLUDE)
+			return gdeleted || (psf->sf_crcount && sdeleted);
+		return psf->sf_crcount && !gdeleted && !sdeleted;
+	}
+	return 0;
+}
+
+static int
+igmp_scount(struct ip_mc_list *pmc, int type, int gdeleted, int sdeleted)
+{
+	struct ip_sf_list *psf;
+	int scount = 0;
+
+	for (psf=pmc->sources; psf; psf=psf->sf_next) {
+		if (!is_in(pmc, psf, type, gdeleted, sdeleted))
+			continue;
+		scount++;
+	}
+	return scount;
+}
+
+static struct sk_buff *igmpv3_newpack(struct net_device *dev, int size)
+{
+	struct sk_buff *skb;
+	struct rtable *rt;
+	struct iphdr *pip;
+	struct igmpv3_report *pig;
+
+	skb = alloc_skb(size + dev->hard_header_len + 15, GFP_ATOMIC);
+	if (skb == NULL)
+		return 0;
+
+	{
+		struct flowi fl = { .oif = dev->ifindex,
+				    .nl_u = { .ip4_u = {
+				    .daddr = IGMPV3_ALL_MCR } },
+				    .proto = IPPROTO_IGMP };
+		if (ip_route_output_key(&rt, &fl))
+			return 0;
+	}
+	if (rt->rt_src == 0) {
+		ip_rt_put(rt);
+		return 0;
+	}
+
+	skb->dst = &rt->u.dst;
+	skb->dev = dev;
+
+	skb_reserve(skb, (dev->hard_header_len+15)&~15);
+
+	skb->nh.iph = pip =(struct iphdr *)skb_put(skb, sizeof(struct iphdr)+4);
+
+	pip->version  = 4;
+	pip->ihl      = (sizeof(struct iphdr)+4)>>2;
+	pip->tos      = 0xc0;
+	pip->frag_off = htons(IP_DF);
+	pip->ttl      = 1;
+	pip->daddr    = rt->rt_dst;
+	pip->saddr    = rt->rt_src;
+	pip->protocol = IPPROTO_IGMP;
+	pip->tot_len  = 0;	/* filled in later */
+	ip_select_ident(pip, &rt->u.dst, NULL);
+	((u8*)&pip[1])[0] = IPOPT_RA;
+	((u8*)&pip[1])[1] = 4;
+	((u8*)&pip[1])[2] = 0;
+	((u8*)&pip[1])[3] = 0;
+
+	pig =(struct igmpv3_report *)skb_put(skb, sizeof(*pig));
+	skb->h.igmph = (struct igmphdr *)pig;
+	pig->type = IGMPV3_HOST_MEMBERSHIP_REPORT;
+	pig->resv1 = 0;
+	pig->csum = 0;
+	pig->resv2 = 0;
+	pig->ngrec = 0;
+	return skb;
+}
+
+static int igmpv3_sendpack(struct sk_buff *skb)
+{
+	struct iphdr *pip = skb->nh.iph;
+	struct igmphdr *pig = skb->h.igmph;
+	int iplen, igmplen;
+
+	iplen = skb->tail - (unsigned char *)skb->nh.iph;
+	pip->tot_len = htons(iplen);
+	ip_send_check(pip);
+
+	igmplen = skb->tail - (unsigned char *)skb->h.igmph;
+	pig->csum = ip_compute_csum((void *)skb->h.igmph, igmplen);
+
+	return NF_HOOK(PF_INET, NF_IP_LOCAL_OUT, skb, NULL, skb->dev,
+		       dst_output);
+}
+
+static int grec_size(struct ip_mc_list *pmc, int type, int gdel, int sdel)
+{
+	return sizeof(struct igmpv3_grec) + 4*igmp_scount(pmc,type,gdel,sdel);
+}
+
+static struct sk_buff *add_grhead(struct sk_buff *skb, struct ip_mc_list *pmc,
+	int type, struct igmpv3_grec **ppgr)
+{
+	struct net_device *dev = pmc->interface->dev;
+	struct igmpv3_report *pih;
+	struct igmpv3_grec *pgr;
+
+	if (!skb)
+		skb = igmpv3_newpack(dev, dev->mtu);
+	if (!skb)
+		return 0;
+	pgr = (struct igmpv3_grec *)skb_put(skb, sizeof(struct igmpv3_grec));
+	pgr->grec_type = type;
+	pgr->grec_auxwords = 0;
+	pgr->grec_nsrcs = 0;
+	pgr->grec_mca = pmc->multiaddr;
+	pih = (struct igmpv3_report *)skb->h.igmph;
+	pih->ngrec = htons(ntohs(pih->ngrec)+1);
+	*ppgr = pgr;
+	return skb;
+}
+
+#define AVAILABLE(skb) ((skb) ? ((skb)->dev ? (skb)->dev->mtu - (skb)->len : \
+	skb_tailroom(skb)) : 0)
+
+static struct sk_buff *add_grec(struct sk_buff *skb, struct ip_mc_list *pmc,
+	int type, int gdeleted, int sdeleted)
+{
+	struct net_device *dev = pmc->interface->dev;
+	struct igmpv3_report *pih;
+	struct igmpv3_grec *pgr = 0;
+	struct ip_sf_list *psf, *psf_next, *psf_prev, *psf_list;
+	int scount, first, isquery, truncate;
+
+	if (pmc->multiaddr == IGMP_ALL_HOSTS)
+		return skb;
+
+	isquery = type == IGMPV3_MODE_IS_INCLUDE ||
+		  type == IGMPV3_MODE_IS_EXCLUDE;
+	truncate = type == IGMPV3_MODE_IS_EXCLUDE ||
+		    type == IGMPV3_CHANGE_TO_EXCLUDE;
+
+	psf_list = sdeleted ? pmc->tomb : pmc->sources;
+
+	if (!psf_list) {
+		if (type == IGMPV3_ALLOW_NEW_SOURCES ||
+		    type == IGMPV3_BLOCK_OLD_SOURCES)
+			return skb;
+		if (pmc->crcount || isquery)
+			skb = add_grhead(skb, pmc, type, &pgr);
+		return skb;
+	}
+	pih = skb ? (struct igmpv3_report *)skb->h.igmph : 0;
+
+	/* EX and TO_EX get a fresh packet, if needed */
+	if (truncate) {
+		if (pih && pih->ngrec &&
+		    AVAILABLE(skb) < grec_size(pmc, type, gdeleted, sdeleted)) {
+			if (skb)
+				igmpv3_sendpack(skb);
+			skb = igmpv3_newpack(dev, dev->mtu);
+		}
+	}
+	first = 1;
+	scount = 0;
+	psf_prev = 0;
+	for (psf=psf_list; psf; psf=psf_next) {
+		u32 *psrc;
+
+		psf_next = psf->sf_next;
+
+		if (!is_in(pmc, psf, type, gdeleted, sdeleted)) {
+			psf_prev = psf;
+			continue;
+		}
+
+		/* clear marks on query responses */
+		if (isquery)
+			psf->sf_gsresp = 0;
+
+		if (AVAILABLE(skb) < sizeof(u32) +
+		    first*sizeof(struct igmpv3_grec)) {
+			if (truncate && !first)
+				break;	 /* truncate these */
+			if (pgr)
+				pgr->grec_nsrcs = htons(scount);
+			if (skb)
+				igmpv3_sendpack(skb);
+			skb = igmpv3_newpack(dev, dev->mtu);
+			first = 1;
+			scount = 0;
+		}
+		if (first) {
+			skb = add_grhead(skb, pmc, type, &pgr);
+			first = 0;
+		}
+		psrc = (u32 *)skb_put(skb, sizeof(u32));
+		*psrc = psf->sf_inaddr;
+		scount++;
+		if ((type == IGMPV3_ALLOW_NEW_SOURCES ||
+		     type == IGMPV3_BLOCK_OLD_SOURCES) && psf->sf_crcount) {
+			psf->sf_crcount--;
+			if ((sdeleted || gdeleted) && psf->sf_crcount == 0) {
+				if (psf_prev)
+					psf_prev->sf_next = psf->sf_next;
+				else
+					pmc->tomb = psf->sf_next;
+				kfree(psf);
+				continue;
+			}
+		}
+		psf_prev = psf;
+	}
+	if (pgr)
+		pgr->grec_nsrcs = htons(scount);
+
+	if (isquery)
+		pmc->gsquery = 0;	/* clear query state on report */
+	return skb;
+}
+
+static int igmpv3_send_report(struct in_device *in_dev, struct ip_mc_list *pmc)
+{
+	struct sk_buff *skb = 0;
+	int type;
+
+	if (!pmc) {
+		read_lock(&in_dev->lock);
+		for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
+			if (pmc->multiaddr == IGMP_ALL_HOSTS)
+				continue;
+			spin_lock_bh(&pmc->lock);
+			if (pmc->sfcount[MCAST_EXCLUDE])
+				type = IGMPV3_MODE_IS_EXCLUDE;
+			else
+				type = IGMPV3_MODE_IS_INCLUDE;
+			skb = add_grec(skb, pmc, type, 0, 0);
+			spin_unlock_bh(&pmc->lock);
+		}
+		read_unlock(&in_dev->lock);
+	} else {
+		spin_lock_bh(&pmc->lock);
+		if (pmc->sfcount[MCAST_EXCLUDE])
+			type = IGMPV3_MODE_IS_EXCLUDE;
+		else
+			type = IGMPV3_MODE_IS_INCLUDE;
+		skb = add_grec(skb, pmc, type, 0, 0);
+		spin_unlock_bh(&pmc->lock);
+	}
+	if (!skb)
+		return 0;
+	return igmpv3_sendpack(skb);
+}
+
+/*
+ * remove zero-count source records from a source filter list
+ */
+static void igmpv3_clear_zeros(struct ip_sf_list **ppsf)
+{
+	struct ip_sf_list *psf_prev, *psf_next, *psf;
+
+	psf_prev = 0;
+	for (psf=*ppsf; psf; psf = psf_next) {
+		psf_next = psf->sf_next;
+		if (psf->sf_crcount == 0) {
+			if (psf_prev)
+				psf_prev->sf_next = psf->sf_next;
+			else
+				*ppsf = psf->sf_next;
+			kfree(psf);
+		} else
+			psf_prev = psf;
+	}
+}
+
+static void igmpv3_send_cr(struct in_device *in_dev)
+{
+	struct ip_mc_list *pmc, *pmc_prev, *pmc_next;
+	struct sk_buff *skb = 0;
+	int type, dtype;
+
+	read_lock(&in_dev->lock);
+	write_lock_bh(&in_dev->mc_lock);
+
+	/* deleted MCA's */
+	pmc_prev = 0;
+	for (pmc=in_dev->mc_tomb; pmc; pmc=pmc_next) {
+		pmc_next = pmc->next;
+		if (pmc->sfmode == MCAST_INCLUDE) {
+			type = IGMPV3_BLOCK_OLD_SOURCES;
+			dtype = IGMPV3_BLOCK_OLD_SOURCES;
+			skb = add_grec(skb, pmc, type, 1, 0);
+			skb = add_grec(skb, pmc, dtype, 1, 1);
+		}
+		if (pmc->crcount) {
+			pmc->crcount--;
+			if (pmc->sfmode == MCAST_EXCLUDE) {
+				type = IGMPV3_CHANGE_TO_INCLUDE;
+				skb = add_grec(skb, pmc, type, 1, 0);
+			}
+			if (pmc->crcount == 0) {
+				igmpv3_clear_zeros(&pmc->tomb);
+				igmpv3_clear_zeros(&pmc->sources);
+			}
+		}
+		if (pmc->crcount == 0 && !pmc->tomb && !pmc->sources) {
+			if (pmc_prev)
+				pmc_prev->next = pmc_next;
+			else
+				in_dev->mc_tomb = pmc_next;
+			in_dev_put(pmc->interface);
+			kfree(pmc);
+		} else
+			pmc_prev = pmc;
+	}
+	write_unlock_bh(&in_dev->mc_lock);
+
+	/* change recs */
+	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
+		spin_lock_bh(&pmc->lock);
+		if (pmc->sfcount[MCAST_EXCLUDE]) {
+			type = IGMPV3_BLOCK_OLD_SOURCES;
+			dtype = IGMPV3_ALLOW_NEW_SOURCES;
+		} else {
+			type = IGMPV3_ALLOW_NEW_SOURCES;
+			dtype = IGMPV3_BLOCK_OLD_SOURCES;
+		}
+		skb = add_grec(skb, pmc, type, 0, 0);
+		skb = add_grec(skb, pmc, dtype, 0, 1);	/* deleted sources */
+
+		/* filter mode changes */
+		if (pmc->crcount) {
+			pmc->crcount--;
+			if (pmc->sfmode == MCAST_EXCLUDE)
+				type = IGMPV3_CHANGE_TO_EXCLUDE;
+			else
+				type = IGMPV3_CHANGE_TO_INCLUDE;
+			skb = add_grec(skb, pmc, type, 0, 0);
+		}
+		spin_unlock_bh(&pmc->lock);
+	}
+	read_unlock(&in_dev->lock);
+	if (!skb)
+		return;
+	(void) igmpv3_sendpack(skb);
+}
+
+static int igmp_send_report(struct in_device *in_dev, struct ip_mc_list *pmc,
+	int type)
 {
 	struct sk_buff *skb;
 	struct iphdr *iph;
 	struct igmphdr *ih;
 	struct rtable *rt;
+	struct net_device *dev = in_dev->dev;
+	u32	group = pmc ? pmc->multiaddr : 0;
 	u32	dst;
 
-	/* According to IGMPv2 specs, LEAVE messages are
-	 * sent to all-routers group.
-	 */
-	dst = group;
-	if (type == IGMP_HOST_LEAVE_MESSAGE)
+	if (type == IGMPV3_HOST_MEMBERSHIP_REPORT)
+		return igmpv3_send_report(in_dev, pmc);
+	else if (type == IGMP_HOST_LEAVE_MESSAGE)
 		dst = IGMP_ALL_ROUTER;
+	else
+		dst = group;
 
 	{
 		struct flowi fl = { .oif = dev->ifindex,
@@ -225,7 +630,7 @@ static int igmp_send_report(struct net_device *dev, u32 group, int type)
 
 	iph->version  = 4;
 	iph->ihl      = (sizeof(struct iphdr)+4)>>2;
-	iph->tos      = 0;
+	iph->tos      = 0xc0;
 	iph->frag_off = htons(IP_DF);
 	iph->ttl      = 1;
 	iph->daddr    = dst;
@@ -250,6 +655,34 @@ static int igmp_send_report(struct net_device *dev, u32 group, int type)
 		       dst_output);
 }
 
+static void igmp_gq_timer_expire(unsigned long data)
+{
+	struct in_device *in_dev = (struct in_device *)data;
+
+	in_dev->mr_gq_running = 0;
+	igmpv3_send_report(in_dev, 0);
+}
+
+static void igmp_ifc_timer_expire(unsigned long data)
+{
+	struct in_device *in_dev = (struct in_device *)data;
+
+	igmpv3_send_cr(in_dev);
+	if (in_dev->mr_ifc_count) {
+		in_dev->mr_ifc_count--;
+		igmp_ifc_start_timer(in_dev, IGMP_Unsolicited_Report_Interval);
+	}
+}
+
+static void igmp_ifc_event(struct in_device *in_dev)
+{
+	if (IGMP_V1_SEEN(in_dev) || IGMP_V2_SEEN(in_dev))
+		return;
+	in_dev->mr_ifc_count = in_dev->mr_qrv ? in_dev->mr_qrv : 
+		IGMP_Unsolicited_Report_Count;
+	igmp_ifc_start_timer(in_dev, 1);
+}
+
 
 static void igmp_timer_expire(unsigned long data)
 {
@@ -267,13 +700,33 @@ static void igmp_timer_expire(unsigned long data)
 	spin_unlock(&im->lock);
 
 	if (IGMP_V1_SEEN(in_dev))
-		igmp_send_report(in_dev->dev, im->multiaddr, IGMP_HOST_MEMBERSHIP_REPORT);
+		igmp_send_report(in_dev, im, IGMP_HOST_MEMBERSHIP_REPORT);
+	else if (IGMP_V2_SEEN(in_dev))
+		igmp_send_report(in_dev, im, IGMPV2_HOST_MEMBERSHIP_REPORT);
 	else
-		igmp_send_report(in_dev->dev, im->multiaddr, IGMP_HOST_NEW_MEMBERSHIP_REPORT);
+		igmp_send_report(in_dev, im, IGMPV3_HOST_MEMBERSHIP_REPORT);
 
 	ip_ma_put(im);
 }
 
+static void igmp_marksources(struct ip_mc_list *pmc, int nsrcs, __u32 *srcs)
+{
+	struct ip_sf_list *psf;
+	int i, scount;
+
+	scount = 0;
+	for (psf=pmc->sources; psf; psf=psf->sf_next) {
+		if (scount == nsrcs)
+			break;
+		for (i=0; i<nsrcs; i++)
+			if (srcs[i] == psf->sf_inaddr) {
+				psf->sf_gsresp = 1;
+				scount++;
+				break;
+			}
+	}
+}
+
 static void igmp_heard_report(struct in_device *in_dev, u32 group)
 {
 	struct ip_mc_list *im;
@@ -293,20 +746,46 @@ static void igmp_heard_report(struct in_device *in_dev, u32 group)
 	read_unlock(&in_dev->lock);
 }
 
-static void igmp_heard_query(struct in_device *in_dev, unsigned char max_resp_time,
-			     u32 group)
+static void igmp_heard_query(struct in_device *in_dev, struct igmphdr *ih,
+	int len)
 {
+	struct igmpv3_query *ih3 = (struct igmpv3_query *)ih;
 	struct ip_mc_list	*im;
+	u32			group = ih->group;
 	int			max_delay;
+	int			mark = 0;
 
-	max_delay = max_resp_time*(HZ/IGMP_TIMER_SCALE);
-
-	if (max_resp_time == 0) {
-		/* Alas, old v1 router presents here. */
 
-		max_delay = IGMP_Query_Response_Interval;
-		in_dev->mr_v1_seen = jiffies + IGMP_V1_Router_Present_Timeout;
-		group = 0;
+	if (len == 8) {
+		if (ih->code == 0) {
+			/* Alas, old v1 router presents here. */
+	
+			max_delay = IGMP_Query_Response_Interval;
+			in_dev->mr_v1_seen = jiffies +
+				IGMP_V1_Router_Present_Timeout;
+			group = 0;
+		} else {
+			/* v2 router present */
+			max_delay = ih->code*(HZ/IGMP_TIMER_SCALE);
+			in_dev->mr_v2_seen = jiffies +
+				IGMP_V2_Router_Present_Timeout;
+		}
+		igmpv3_clear_delrec(in_dev);
+	} else if (len < 12) {
+		return;	/* ignore bogus packet; freed by caller */
+	} else { /* v3 */
+		max_delay = IGMPV3_MRC(ih3->code)*(HZ/IGMP_TIMER_SCALE);
+		in_dev->mr_maxdelay = max_delay;
+		if (ih3->qrv)
+			in_dev->mr_qrv = ih3->qrv;
+		if (!group) { /* general query */
+			if (ih3->nsrcs)
+				return;	/* no sources allowed */
+			igmp_gq_start_timer(in_dev);
+			return;
+		}
+		/* mark sources to include, if group & source-specific */
+		mark = ih3->nsrcs != 0;
 	}
 
 	/*
@@ -325,6 +804,14 @@ static void igmp_heard_query(struct in_device *in_dev, unsigned char max_resp_ti
 			continue;
 		if (im->multiaddr == IGMP_ALL_HOSTS)
 			continue;
+		spin_lock_bh(&im->lock);
+		if (im->tm_running)
+			im->gsquery = im->gsquery && mark;
+		else
+			im->gsquery = mark;
+		if (im->gsquery)
+			igmp_marksources(im, ntohs(ih3->nsrcs), ih3->srcs);
+		spin_unlock_bh(&im->lock);
 		igmp_mod_timer(im, max_delay);
 	}
 	read_unlock(&in_dev->lock);
@@ -358,10 +845,11 @@ int igmp_rcv(struct sk_buff *skb)
 
 	switch (ih->type) {
 	case IGMP_HOST_MEMBERSHIP_QUERY:
-		igmp_heard_query(in_dev, ih->code, ih->group);
+		igmp_heard_query(in_dev, ih, len);
 		break;
 	case IGMP_HOST_MEMBERSHIP_REPORT:
-	case IGMP_HOST_NEW_MEMBERSHIP_REPORT:
+	case IGMPV2_HOST_MEMBERSHIP_REPORT:
+	case IGMPV3_HOST_MEMBERSHIP_REPORT:
 		/* Is it our report looped back? */
 		if (((struct rtable*)skb->dst)->fl.iif == 0)
 			break;
@@ -422,15 +910,105 @@ static void ip_mc_filter_del(struct in_device *in_dev, u32 addr)
 		dev_mc_delete(dev,buf,dev->addr_len,0);
 }
 
+#ifdef CONFIG_IP_MULTICAST
+/*
+ * deleted ip_mc_list manipulation
+ */
+static void igmpv3_add_delrec(struct in_device *in_dev, struct ip_mc_list *im)
+{
+	struct ip_mc_list *pmc;
+
+	/* this is an "ip_mc_list" for convenience; only the fields below
+	 * are actually used. In particular, the refcnt and users are not
+	 * used for management of the delete list. Using the same structure
+	 * for deleted items allows change reports to use common code with
+	 * non-deleted or query-response MCA's.
+	 */
+	pmc = (struct ip_mc_list *)kmalloc(sizeof(*pmc), GFP_KERNEL);
+	if (!pmc)
+		return;
+	memset(pmc, 0, sizeof(*pmc));
+	spin_lock_bh(&im->lock);
+	pmc->interface = im->interface;
+	in_dev_hold(in_dev);
+	pmc->multiaddr = im->multiaddr;
+	pmc->crcount = in_dev->mr_qrv ? in_dev->mr_qrv :
+		IGMP_Unsolicited_Report_Count;
+	pmc->sfmode = im->sfmode;
+	if (pmc->sfmode == MCAST_INCLUDE) {
+		struct ip_sf_list *psf;
+
+		pmc->tomb = im->tomb;
+		pmc->sources = im->sources;
+		im->tomb = im->sources = 0;
+		for (psf=pmc->sources; psf; psf=psf->sf_next)
+			psf->sf_crcount = pmc->crcount;
+	}
+	spin_unlock_bh(&im->lock);
+
+	write_lock_bh(&in_dev->mc_lock);
+	pmc->next = in_dev->mc_tomb;
+	in_dev->mc_tomb = pmc;
+	write_unlock_bh(&in_dev->mc_lock);
+}
+#endif
+
+static void igmpv3_del_delrec(struct in_device *in_dev, __u32 multiaddr)
+{
+	struct ip_mc_list *pmc, *pmc_prev;
+	struct ip_sf_list *psf, *psf_next;
+
+	write_lock_bh(&in_dev->mc_lock);
+	pmc_prev = 0;
+	for (pmc=in_dev->mc_tomb; pmc; pmc=pmc->next) {
+		if (pmc->multiaddr == multiaddr)
+			break;
+		pmc_prev = pmc;
+	}
+	if (pmc) {
+		if (pmc_prev)
+			pmc_prev->next = pmc->next;
+		else
+			in_dev->mc_tomb = pmc->next;
+	}
+	write_unlock_bh(&in_dev->mc_lock);
+	if (pmc) {
+		for (psf=pmc->tomb; psf; psf=psf_next) {
+			psf_next = psf->sf_next;
+			kfree(psf);
+		}
+		in_dev_put(pmc->interface);
+		kfree(pmc);
+	}
+}
+
+static void igmpv3_clear_delrec(struct in_device *in_dev)
+{
+	struct ip_mc_list *pmc, *nextpmc;
+
+	write_lock_bh(&in_dev->mc_lock);
+	pmc = in_dev->mc_tomb;
+	in_dev->mc_tomb = 0;
+	write_unlock_bh(&in_dev->mc_lock);
+
+	for (; pmc; pmc = nextpmc) {
+		nextpmc = pmc->next;
+		ip_mc_clear_src(pmc);
+		in_dev_put(pmc->interface);
+		kfree(pmc);
+	}
+}
+
 static void igmp_group_dropped(struct ip_mc_list *im)
 {
+	struct in_device *in_dev = im->interface;
 #ifdef CONFIG_IP_MULTICAST
 	int reporter;
 #endif
 
 	if (im->loaded) {
 		im->loaded = 0;
-		ip_mc_filter_del(im->interface, im->multiaddr);
+		ip_mc_filter_del(in_dev, im->multiaddr);
 	}
 
 #ifdef CONFIG_IP_MULTICAST
@@ -440,25 +1018,46 @@ static void igmp_group_dropped(struct ip_mc_list *im)
 	reporter = im->reporter;
 	igmp_stop_timer(im);
 
-	if (reporter && !IGMP_V1_SEEN(im->interface))
-		igmp_send_report(im->interface->dev, im->multiaddr, IGMP_HOST_LEAVE_MESSAGE);
+	if (IGMP_V1_SEEN(in_dev))
+		goto done;
+	if (IGMP_V2_SEEN(in_dev)) {
+		if (reporter)
+			igmp_send_report(in_dev, im, IGMP_HOST_LEAVE_MESSAGE);
+		goto done;
+	}
+	/* IGMPv3 */
+	igmpv3_add_delrec(in_dev, im);
+
+	igmp_ifc_event(in_dev);
+done:
+	ip_mc_clear_src(im);
 #endif
 }
 
 static void igmp_group_added(struct ip_mc_list *im)
 {
+	struct in_device *in_dev = im->interface;
+
 	if (im->loaded == 0) {
 		im->loaded = 1;
-		ip_mc_filter_add(im->interface, im->multiaddr);
+		ip_mc_filter_add(in_dev, im->multiaddr);
 	}
 
 #ifdef CONFIG_IP_MULTICAST
 	if (im->multiaddr == IGMP_ALL_HOSTS)
 		return;
 
-	spin_lock_bh(&im->lock);
-	igmp_start_timer(im, IGMP_Initial_Report_Delay);
-	spin_unlock_bh(&im->lock);
+	if (IGMP_V1_SEEN(in_dev) || IGMP_V2_SEEN(in_dev)) {
+		spin_lock_bh(&im->lock);
+		igmp_start_timer(im, IGMP_Initial_Report_Delay);
+		spin_unlock_bh(&im->lock);
+		return;
+	}
+	/* else, v3 */
+
+	im->crcount = in_dev->mr_qrv ? in_dev->mr_qrv :
+		IGMP_Unsolicited_Report_Count;
+	igmp_ifc_event(in_dev);
 #endif
 }
 
@@ -481,6 +1080,7 @@ void ip_mc_inc_group(struct in_device *in_dev, u32 addr)
 	for (im=in_dev->mc_list; im; im=im->next) {
 		if (im->multiaddr == addr) {
 			im->users++;
+			ip_mc_add_src(in_dev, &addr, MCAST_EXCLUDE, 0, 0, 0);
 			goto out;
 		}
 	}
@@ -493,6 +1093,13 @@ void ip_mc_inc_group(struct in_device *in_dev, u32 addr)
 	im->interface=in_dev;
 	in_dev_hold(in_dev);
 	im->multiaddr=addr;
+	/* initial mode is (EX, empty) */
+	im->sfmode = MCAST_EXCLUDE;
+	im->sfcount[MCAST_INCLUDE] = 0;
+	im->sfcount[MCAST_EXCLUDE] = 1;
+	im->sources = 0;
+	im->tomb = 0;
+	im->crcount = 0;
 	atomic_set(&im->refcnt, 1);
 	spin_lock_init(&im->lock);
 #ifdef  CONFIG_IP_MULTICAST
@@ -502,12 +1109,14 @@ void ip_mc_inc_group(struct in_device *in_dev, u32 addr)
 	im->timer.function=&igmp_timer_expire;
 	im->unsolicit_count = IGMP_Unsolicited_Report_Count;
 	im->reporter = 0;
+	im->gsquery = 0;
 #endif
 	im->loaded = 0;
 	write_lock_bh(&in_dev->lock);
 	im->next=in_dev->mc_list;
 	in_dev->mc_list=im;
 	write_unlock_bh(&in_dev->lock);
+	igmpv3_del_delrec(in_dev, im->multiaddr);
 	igmp_group_added(im);
 	if (in_dev->dev->flags & IFF_UP)
 		ip_rt_multicast_event(in_dev);
@@ -552,9 +1161,18 @@ void ip_mc_down(struct in_device *in_dev)
 
 	ASSERT_RTNL();
 
+	in_dev->mr_ifc_count = 0;
+	if (del_timer(&in_dev->mr_ifc_timer))
+		atomic_dec(&in_dev->refcnt);
+	in_dev->mr_gq_running = 0;
+	if (del_timer(&in_dev->mr_gq_timer))
+		atomic_dec(&in_dev->refcnt);
+
 	for (i=in_dev->mc_list; i; i=i->next)
 		igmp_group_dropped(i);
 
+	igmpv3_clear_delrec(in_dev);
+
 	ip_mc_dec_group(in_dev, IGMP_ALL_HOSTS);
 }
 
@@ -566,6 +1184,20 @@ void ip_mc_up(struct in_device *in_dev)
 
 	ASSERT_RTNL();
 
+#ifdef CONFIG_IP_MULTICAST
+	in_dev->mc_lock = RW_LOCK_UNLOCKED;
+	in_dev->mr_gq_running = 0;
+	init_timer(&in_dev->mr_gq_timer);
+	in_dev->mr_gq_timer.data=(unsigned long) in_dev;
+	in_dev->mr_gq_timer.function=&igmp_gq_timer_expire;
+	in_dev->mc_tomb = 0;
+	in_dev->mr_ifc_count = 0;
+	init_timer(&in_dev->mr_ifc_timer);
+	in_dev->mr_ifc_timer.data=(unsigned long) in_dev;
+	in_dev->mr_ifc_timer.function=&igmp_ifc_timer_expire;
+	in_dev->mr_qrv = IGMP_Unsolicited_Report_Count;
+#endif
+
 	ip_mc_inc_group(in_dev, IGMP_ALL_HOSTS);
 
 	for (i=in_dev->mc_list; i; i=i->next)
@@ -626,6 +1258,262 @@ static struct in_device * ip_mc_find_dev(struct ip_mreqn *imr)
  */
 int sysctl_igmp_max_memberships = IP_MAX_MEMBERSHIPS;
 
+
+static int ip_mc_del1_src(struct ip_mc_list *pmc, int sfmode,
+	__u32 *psfsrc)
+{
+	struct ip_sf_list *psf, *psf_prev;
+	int rv = 0;
+
+	psf_prev = 0;
+	for (psf=pmc->sources; psf; psf=psf->sf_next) {
+		if (psf->sf_inaddr == *psfsrc)
+			break;
+		psf_prev = psf;
+	}
+	if (!psf || psf->sf_count[sfmode] == 0) {
+		/* source filter not found, or count wrong =>  bug */
+		return -ESRCH;
+	}
+	psf->sf_count[sfmode]--;
+	if (psf->sf_count[sfmode] == 0) {
+		ip_rt_multicast_event(pmc->interface);
+	}
+	if (!psf->sf_count[MCAST_INCLUDE] && !psf->sf_count[MCAST_EXCLUDE]) {
+		struct in_device *in_dev = pmc->interface;
+
+		/* no more filters for this source */
+		if (psf_prev)
+			psf_prev->sf_next = psf->sf_next;
+		else
+			pmc->sources = psf->sf_next;
+		if (psf->sf_oldin &&
+		    !IGMP_V1_SEEN(in_dev) && !IGMP_V2_SEEN(in_dev)) {
+			psf->sf_crcount = in_dev->mr_qrv ? in_dev->mr_qrv : 
+				IGMP_Unsolicited_Report_Count;
+			psf->sf_next = pmc->tomb;
+			pmc->tomb = psf;
+			rv = 1;
+		} else
+			kfree(psf);
+	}
+	return rv;
+}
+
+#ifndef CONFIG_IP_MULTICAST
+#define igmp_ifc_event(x)	do { } while (0)
+#endif
+
+int ip_mc_del_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
+	int sfcount, __u32 *psfsrc, int delta)
+{
+	struct ip_mc_list *pmc;
+	int	changerec = 0;
+	int	i, err;
+
+	if (!in_dev)
+		return -ENODEV;
+	read_lock(&in_dev->lock);
+	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
+		if (*pmca == pmc->multiaddr)
+			break;
+	}
+	if (!pmc) {
+		/* MCA not found?? bug */
+		read_unlock(&in_dev->lock);
+		return -ESRCH;
+	}
+	spin_lock_bh(&pmc->lock);
+	read_unlock(&in_dev->lock);
+	sf_markstate(pmc);
+	if (!delta) {
+		if (!pmc->sfcount[sfmode])
+			return -EINVAL;
+		pmc->sfcount[sfmode]--;
+	}
+	err = 0;
+	for (i=0; i<sfcount; i++) {
+		int rv = ip_mc_del1_src(pmc, sfmode, &psfsrc[i]);
+
+		changerec |= rv > 0;
+		if (!err && rv < 0)
+			err = rv;
+	}
+	if (pmc->sfmode == MCAST_EXCLUDE &&
+	    pmc->sfcount[MCAST_EXCLUDE] == 0 &&
+	    pmc->sfcount[MCAST_INCLUDE]) {
+		struct ip_sf_list *psf;
+
+		/* filter mode change */
+		pmc->sfmode = MCAST_INCLUDE;
+		pmc->crcount = in_dev->mr_qrv ? in_dev->mr_qrv : 
+			IGMP_Unsolicited_Report_Count;
+		in_dev->mr_ifc_count = pmc->crcount;
+		for (psf=pmc->sources; psf; psf = psf->sf_next)
+			psf->sf_crcount = 0;
+		igmp_ifc_event(pmc->interface);
+	} else if (sf_setstate(pmc) || changerec) {
+		igmp_ifc_event(pmc->interface);
+	}
+	spin_unlock_bh(&pmc->lock);
+	return err;
+}
+
+/*
+ * Add multicast single-source filter to the interface list
+ */
+static int ip_mc_add1_src(struct ip_mc_list *pmc, int sfmode,
+	__u32 *psfsrc, int delta)
+{
+	struct ip_sf_list *psf, *psf_prev;
+
+	psf_prev = 0;
+	for (psf=pmc->sources; psf; psf=psf->sf_next) {
+		if (psf->sf_inaddr == *psfsrc)
+			break;
+		psf_prev = psf;
+	}
+	if (!psf) {
+		psf = (struct ip_sf_list *)kmalloc(sizeof(*psf), GFP_ATOMIC);
+		if (!psf)
+			return -ENOBUFS;
+		memset(psf, 0, sizeof(*psf));
+		psf->sf_inaddr = *psfsrc;
+		if (psf_prev) {
+			psf_prev->sf_next = psf;
+		} else
+			pmc->sources = psf;
+	}
+	psf->sf_count[sfmode]++;
+	if (psf->sf_count[sfmode] == 1) {
+		ip_rt_multicast_event(pmc->interface);
+	}
+	return 0;
+}
+
+static void sf_markstate(struct ip_mc_list *pmc)
+{
+	struct ip_sf_list *psf;
+	int mca_xcount = pmc->sfcount[MCAST_EXCLUDE];
+
+	for (psf=pmc->sources; psf; psf=psf->sf_next)
+		if (pmc->sfcount[MCAST_EXCLUDE]) {
+			psf->sf_oldin = mca_xcount ==
+				psf->sf_count[MCAST_EXCLUDE] &&
+				!psf->sf_count[MCAST_INCLUDE];
+		} else
+			psf->sf_oldin = psf->sf_count[MCAST_INCLUDE] != 0;
+}
+
+static int sf_setstate(struct ip_mc_list *pmc)
+{
+	struct ip_sf_list *psf;
+	int mca_xcount = pmc->sfcount[MCAST_EXCLUDE];
+	int qrv = pmc->interface->mr_qrv;
+	int new_in, rv;
+
+	rv = 0;
+	for (psf=pmc->sources; psf; psf=psf->sf_next) {
+		if (pmc->sfcount[MCAST_EXCLUDE]) {
+			new_in = mca_xcount == psf->sf_count[MCAST_EXCLUDE] &&
+				!psf->sf_count[MCAST_INCLUDE];
+		} else
+			new_in = psf->sf_count[MCAST_INCLUDE] != 0;
+		if (new_in != psf->sf_oldin) {
+			psf->sf_crcount = qrv;
+			rv++;
+		}
+	}
+	return rv;
+}
+
+/*
+ * Add multicast source filter list to the interface list
+ */
+int ip_mc_add_src(struct in_device *in_dev, __u32 *pmca, int sfmode,
+	int sfcount, __u32 *psfsrc, int delta)
+{
+	struct ip_mc_list *pmc;
+	int	isexclude;
+	int	i, err;
+
+	if (!in_dev)
+		return -ENODEV;
+	read_lock(&in_dev->lock);
+	for (pmc=in_dev->mc_list; pmc; pmc=pmc->next) {
+		if (*pmca == pmc->multiaddr)
+			break;
+	}
+	if (!pmc) {
+		/* MCA not found?? bug */
+		read_unlock(&in_dev->lock);
+		return -ESRCH;
+	}
+	spin_lock_bh(&pmc->lock);
+	read_unlock(&in_dev->lock);
+
+	sf_markstate(pmc);
+	isexclude = pmc->sfmode == MCAST_EXCLUDE;
+	if (!delta)
+		pmc->sfcount[sfmode]++;
+	err = 0;
+	for (i=0; i<sfcount; i++) {
+		err = ip_mc_add1_src(pmc, sfmode, &psfsrc[i], delta);
+		if (err)
+			break;
+	}
+	if (err) {
+		int j;
+
+		pmc->sfcount[sfmode]--;
+		for (j=0; j<i; j++)
+			(void) ip_mc_del1_src(pmc, sfmode, &psfsrc[i]);
+	} else if (isexclude != (pmc->sfcount[MCAST_EXCLUDE] != 0)) {
+		struct in_device *in_dev = pmc->interface;
+		struct ip_sf_list *psf;
+
+		/* filter mode change */
+		if (pmc->sfcount[MCAST_EXCLUDE])
+			pmc->sfmode = MCAST_EXCLUDE;
+		else if (pmc->sfcount[MCAST_INCLUDE])
+			pmc->sfmode = MCAST_INCLUDE;
+		/* else no filters; keep old mode for reports */
+
+		pmc->crcount = in_dev->mr_qrv ? in_dev->mr_qrv : 
+			IGMP_Unsolicited_Report_Count;
+		in_dev->mr_ifc_count = pmc->crcount;
+		for (psf=pmc->sources; psf; psf = psf->sf_next)
+			psf->sf_crcount = 0;
+		igmp_ifc_event(in_dev);
+	} else if (sf_setstate(pmc))
+		igmp_ifc_event(in_dev);
+	spin_unlock_bh(&pmc->lock);
+	return err;
+}
+
+static void ip_mc_clear_src(struct ip_mc_list *pmc)
+{
+	struct ip_sf_list *psf, *nextpsf;
+
+	for (psf=pmc->tomb; psf; psf=nextpsf) {
+		nextpsf = psf->sf_next;
+		kfree(psf);
+	}
+	pmc->tomb = 0;
+	for (psf=pmc->sources; psf; psf=nextpsf) {
+		nextpsf = psf->sf_next;
+		kfree(psf);
+	}
+	pmc->sources = 0;
+	pmc->sfmode = MCAST_EXCLUDE;
+	pmc->sfcount[MCAST_EXCLUDE] = 0;
+	pmc->sfcount[MCAST_EXCLUDE] = 1;
+}
+
+
+/*
+ * Join a multicast group
+ */
 int ip_mc_join_group(struct sock *sk , struct ip_mreqn *imr)
 {
 	int err;
@@ -674,6 +1562,8 @@ int ip_mc_join_group(struct sock *sk , struct ip_mreqn *imr)
 	memcpy(&iml->multi, imr, sizeof(*imr));
 	iml->next = inet->mc_list;
 	iml->count = 1;
+	iml->sflist = NULL;
+	iml->sfmode = MCAST_EXCLUDE;
 	inet->mc_list = iml;
 	ip_mc_inc_group(in_dev, addr);
 	iml = NULL;
@@ -686,6 +1576,24 @@ int ip_mc_join_group(struct sock *sk , struct ip_mreqn *imr)
 	return err;
 }
 
+int ip_mc_leave_src(struct sock *sk, struct ip_mc_socklist *iml,
+	struct in_device *in_dev)
+{
+	int err;
+
+	if (iml->sflist == 0) {
+		/* any-source empty exclude case */
+		return ip_mc_del_src(in_dev, &iml->multi.imr_multiaddr.s_addr,
+			iml->sfmode, 0, 0, 0);
+	}
+	err = ip_mc_del_src(in_dev, &iml->multi.imr_multiaddr.s_addr,
+			iml->sfmode, iml->sflist->sl_count,
+			iml->sflist->sl_addr, 0);
+	sock_kfree_s(sk, iml->sflist, IP_SFLSIZE(iml->sflist->sl_max));
+	iml->sflist = 0;
+	return err;
+}
+
 /*
  *	Ask a socket to leave a group.
  */
@@ -701,14 +1609,19 @@ int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr)
 		    iml->multi.imr_address.s_addr==imr->imr_address.s_addr &&
 		    (!imr->imr_ifindex || iml->multi.imr_ifindex==imr->imr_ifindex)) {
 			struct in_device *in_dev;
+
+			in_dev = inetdev_by_index(iml->multi.imr_ifindex);
+			if (in_dev)
+				(void) ip_mc_leave_src(sk, iml, in_dev);
 			if (--iml->count) {
 				rtnl_unlock();
+				if (in_dev)
+					in_dev_put(in_dev);
 				return 0;
 			}
 
 			*imlp = iml->next;
 
-			in_dev = inetdev_by_index(iml->multi.imr_ifindex);
 			if (in_dev) {
 				ip_mc_dec_group(in_dev, imr->imr_multiaddr.s_addr);
 				in_dev_put(in_dev);
@@ -722,6 +1635,283 @@ int ip_mc_leave_group(struct sock *sk, struct ip_mreqn *imr)
 	return -EADDRNOTAVAIL;
 }
 
+int ip_mc_source(int add, int omode, struct sock *sk, struct
+	ip_mreq_source *mreqs)
+{
+	int err;
+	struct ip_mreqn	imr;
+	u32 addr = mreqs->imr_multiaddr;
+	struct ip_mc_socklist *pmc;
+	struct in_device *in_dev;
+	struct inet_opt *inet = inet_sk(sk);
+	struct ip_sf_socklist *psl;
+	int i, j, rv;
+
+	if (!MULTICAST(addr))
+		return -EINVAL;
+
+	rtnl_shlock();
+
+	imr.imr_multiaddr.s_addr = mreqs->imr_multiaddr;
+	imr.imr_address.s_addr = mreqs->imr_interface;
+	imr.imr_ifindex = 0;
+	in_dev = ip_mc_find_dev(&imr);
+
+	if (!in_dev) {
+		err = -ENODEV;
+		goto done;
+	}
+	err = -EADDRNOTAVAIL;
+
+	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
+		if (memcmp(&pmc->multi, mreqs, 2*sizeof(__u32)) == 0)
+			break;
+	}
+	if (!pmc)		/* must have a prior join */
+		goto done;
+	/* if a source filter was set, must be the same mode as before */
+	if (pmc->sflist) {
+		if (pmc->sfmode != omode)
+			goto done;
+	} else if (pmc->sfmode != omode) {
+		/* allow mode switches for empty-set filters */
+		ip_mc_del_src(in_dev, &mreqs->imr_multiaddr, pmc->sfmode, 0, 
+			0, 0);
+		pmc->sfmode = omode;
+		ip_mc_add_src(in_dev, &mreqs->imr_multiaddr, pmc->sfmode, 0, 
+			0, 0);
+	}
+
+	psl = pmc->sflist;
+	if (!add) {
+		if (!psl)
+			goto done;
+		rv = !0;
+		for (i=0; i<psl->sl_count; i++) {
+			rv = memcmp(&psl->sl_addr, &mreqs->imr_multiaddr,
+				sizeof(__u32));
+			if (rv >= 0)
+				break;
+		}
+		if (!rv)	/* source not found */
+			goto done;
+
+		/* update the interface filter */
+		ip_mc_del_src(in_dev, &mreqs->imr_multiaddr, omode, 1, 
+			&mreqs->imr_sourceaddr, 1);
+
+		for (j=i+1; j<psl->sl_count; j++)
+			psl->sl_addr[j-1] = psl->sl_addr[j];
+		psl->sl_count--;
+		err = 0;
+		goto done;
+	}
+	/* else, add a new source to the filter */
+
+	if (!psl || psl->sl_count == psl->sl_max) {
+		struct ip_sf_socklist *newpsl;
+		int count = IP_SFBLOCK;
+
+		if (psl)
+			count += psl->sl_max;
+		newpsl = (struct ip_sf_socklist *)sock_kmalloc(sk,
+			IP_SFLSIZE(count), GFP_KERNEL);
+		if (!newpsl) {
+			err = -ENOBUFS;
+			goto done;
+		}
+		newpsl->sl_max = count;
+		newpsl->sl_count = count - IP_SFBLOCK;
+		if (psl) {
+			for (i=0; i<psl->sl_count; i++)
+				newpsl->sl_addr[i] = psl->sl_addr[i];
+			sock_kfree_s(sk, psl, IP_SFLSIZE(psl->sl_max));
+		}
+		pmc->sflist = psl = newpsl;
+	}
+	rv = 1;	/* > 0 for insert logic below if sl_count is 0 */
+	for (i=0; i<psl->sl_count; i++) {
+		rv = memcmp(&psl->sl_addr, &mreqs->imr_multiaddr,
+			sizeof(__u32));
+		if (rv >= 0)
+			break;
+	}
+	if (rv == 0)		/* address already there is an error */
+		goto done;
+	for (j=psl->sl_count-1; j>=i; j--)
+		psl->sl_addr[j+1] = psl->sl_addr[j];
+	psl->sl_addr[i] = mreqs->imr_sourceaddr;
+	psl->sl_count++;
+	err = 0;
+	/* update the interface list */
+	ip_mc_add_src(in_dev, &mreqs->imr_multiaddr, omode, 1, 
+		&mreqs->imr_sourceaddr, 1);
+done:
+	rtnl_shunlock();
+	return err;
+}
+
+int ip_mc_msfilter(struct sock *sk, struct ip_msfilter *msf)
+{
+	int err;
+	struct ip_mreqn	imr;
+	u32 addr = msf->imsf_multiaddr;
+	struct ip_mc_socklist *pmc;
+	struct in_device *in_dev;
+	struct inet_opt *inet = inet_sk(sk);
+	struct ip_sf_socklist *newpsl, *psl;
+
+	if (!MULTICAST(addr))
+		return -EINVAL;
+	if (msf->imsf_fmode != MCAST_INCLUDE &&
+	    msf->imsf_fmode != MCAST_EXCLUDE)
+		return -EINVAL;
+
+	rtnl_shlock();
+
+	imr.imr_multiaddr.s_addr = msf->imsf_multiaddr;
+	imr.imr_address.s_addr = msf->imsf_interface;
+	imr.imr_ifindex = 0;
+	in_dev = ip_mc_find_dev(&imr);
+
+	if (!in_dev) {
+		err = -ENODEV;
+		goto done;
+	}
+	err = -EADDRNOTAVAIL;
+
+	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
+		if (memcmp(&pmc->multi, &imr, sizeof(imr)) == 0)
+			break;
+	}
+	if (!pmc)		/* must have a prior join */
+		goto done;
+	if (msf->imsf_numsrc) {
+		newpsl = (struct ip_sf_socklist *)sock_kmalloc(sk,
+				IP_SFLSIZE(msf->imsf_numsrc), GFP_KERNEL);
+		if (!newpsl) {
+			err = -ENOBUFS;
+			goto done;
+		}
+		newpsl->sl_max = newpsl->sl_count = msf->imsf_numsrc;
+		memcpy(newpsl->sl_addr, msf->imsf_slist,
+			msf->imsf_numsrc * sizeof(msf->imsf_slist[0]));
+		err = ip_mc_add_src(in_dev, &msf->imsf_multiaddr,
+			msf->imsf_fmode, newpsl->sl_count, newpsl->sl_addr, 0);
+		if (err) {
+			sock_kfree_s(sk, newpsl, IP_SFLSIZE(newpsl->sl_max));
+			goto done;
+		}
+	} else
+		newpsl = 0;
+	psl = pmc->sflist;
+	if (psl) {
+		(void) ip_mc_del_src(in_dev, &msf->imsf_multiaddr, pmc->sfmode,
+			psl->sl_count, psl->sl_addr, 0);
+		sock_kfree_s(sk, psl, IP_SFLSIZE(psl->sl_max));
+	} else
+		(void) ip_mc_del_src(in_dev, &msf->imsf_multiaddr, pmc->sfmode,
+			0, 0, 0);
+	pmc->sflist = newpsl;
+	pmc->sfmode = msf->imsf_fmode;
+done:
+	rtnl_shunlock();
+	return err;
+}
+
+int ip_mc_msfget(struct sock *sk, struct ip_msfilter *msf,
+	struct ip_msfilter *optval, int *optlen)
+{
+	int err, len, count, copycount;
+	struct ip_mreqn	imr;
+	u32 addr = msf->imsf_multiaddr;
+	struct ip_mc_socklist *pmc;
+	struct in_device *in_dev;
+	struct inet_opt *inet = inet_sk(sk);
+	struct ip_sf_socklist *psl;
+
+	if (!MULTICAST(addr))
+		return -EINVAL;
+	if (msf->imsf_fmode != MCAST_INCLUDE &&
+	    msf->imsf_fmode != MCAST_EXCLUDE)
+		return -EINVAL;
+
+	rtnl_shlock();
+
+	imr.imr_multiaddr.s_addr = msf->imsf_multiaddr;
+	imr.imr_address.s_addr = msf->imsf_interface;
+	imr.imr_ifindex = 0;
+	in_dev = ip_mc_find_dev(&imr);
+
+	if (!in_dev) {
+		err = -ENODEV;
+		goto done;
+	}
+	err = -EADDRNOTAVAIL;
+
+	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
+		if (memcmp(&pmc->multi, &imr, sizeof(imr)) == 0)
+			break;
+	}
+	if (!pmc)		/* must have a prior join */
+		goto done;
+	msf->imsf_fmode = pmc->sfmode;
+	psl = pmc->sflist;
+	rtnl_shunlock();
+	if (!psl) {
+		len = 0;
+		count = 0;
+	} else {
+		count = psl->sl_count;
+	}
+	copycount = count < msf->imsf_numsrc ? count : msf->imsf_numsrc;
+	len = copycount * sizeof(psl->sl_addr[0]);
+	msf->imsf_numsrc = count;
+	if (put_user(IP_MSFILTER_SIZE(copycount), optlen) ||
+	    copy_to_user((void *)optval, msf, IP_MSFILTER_SIZE(0))) {
+		return -EFAULT;
+	}
+	if (len &&
+	    copy_to_user((void *)&optval->imsf_slist[0], psl->sl_addr, len))
+		return -EFAULT;
+	return 0;
+done:
+	rtnl_shunlock();
+	return err;
+}
+
+/*
+ * check if a multicast source filter allows delivery for a given <src,dst,intf>
+ */
+int ip_mc_sf_allow(struct sock *sk, u32 loc_addr, u32 rmt_addr, int dif)
+{
+	struct inet_opt *inet = inet_sk(sk);
+	struct ip_mc_socklist *pmc;
+	struct ip_sf_socklist *psl;
+	int i;
+
+	for (pmc=inet->mc_list; pmc; pmc=pmc->next) {
+		if (pmc->multi.imr_multiaddr.s_addr == loc_addr &&
+		    pmc->multi.imr_ifindex == dif)
+			break;
+	}
+	if (!pmc)
+		return 0;
+	psl = pmc->sflist;
+	if (!psl)
+		return pmc->sfmode == MCAST_EXCLUDE;
+
+	for (i=0; i<psl->sl_count; i++) {
+		if (psl->sl_addr[i] == rmt_addr)
+			break;
+	}
+	if (pmc->sfmode == MCAST_INCLUDE && i < psl->sl_count)
+		return 1;
+	if (pmc->sfmode == MCAST_EXCLUDE && i >= psl->sl_count)
+		return 1;
+	return 0;
+}
+
 /*
  *	A socket is closing.
  */
@@ -740,6 +1930,7 @@ void ip_mc_drop_socket(struct sock *sk)
 		inet->mc_list = iml->next;
 
 		if ((in_dev = inetdev_by_index(iml->multi.imr_ifindex)) != NULL) {
+			(void) ip_mc_leave_src(sk, iml, in_dev);
 			ip_mc_dec_group(in_dev, iml->multi.imr_multiaddr.s_addr);
 			in_dev_put(in_dev);
 		}
@@ -749,19 +1940,33 @@ void ip_mc_drop_socket(struct sock *sk)
 	rtnl_unlock();
 }
 
-int ip_check_mc(struct in_device *in_dev, u32 mc_addr)
+int ip_check_mc(struct in_device *in_dev, u32 mc_addr, u32 src_addr, u16 proto)
 {
 	struct ip_mc_list *im;
+	struct ip_sf_list *psf;
+	int rv = 0;
 
 	read_lock(&in_dev->lock);
 	for (im=in_dev->mc_list; im; im=im->next) {
-		if (im->multiaddr == mc_addr) {
-			read_unlock(&in_dev->lock);
-			return 1;
+		if (im->multiaddr == mc_addr)
+			break;
+	}
+	if (im && proto == IPPROTO_IGMP) {
+		rv = 1;
+	} else if (im) {
+		for (psf=im->sources; psf; psf=psf->sf_next) {
+			if (psf->sf_inaddr == src_addr)
+				break;
 		}
+		if (psf)
+			rv = psf->sf_count[MCAST_INCLUDE] ||
+				psf->sf_count[MCAST_EXCLUDE] !=
+				im->sfcount[MCAST_EXCLUDE];
+		else
+			rv = im->sfcount[MCAST_EXCLUDE] != 0;
 	}
 	read_unlock(&in_dev->lock);
-	return 0;
+	return rv;
 }
 
 
@@ -822,5 +2027,101 @@ int ip_mc_procinfo(char *buffer, char **start, off_t offset, int length)
 		len=0;
 	return len;
 }
+
+int ip_mcf_procinfo(char *buffer, char **start, off_t offset, int length)
+{
+	off_t pos=0, begin=0;
+	int len=0;
+	int first = 1;
+	struct net_device *dev;
+
+	read_lock(&dev_base_lock);
+	for(dev=dev_base; dev; dev=dev->next) {
+		struct in_device *in_dev = in_dev_get(dev);
+		struct ip_mc_list *imc;
+
+		if (in_dev == NULL)
+			continue;
+
+		read_lock(&in_dev->lock);
+
+		for (imc=in_dev->mc_list; imc; imc=imc->next) {
+			struct ip_sf_list *psf;
+			unsigned long icount, xcount;
+
+			spin_lock_bh(&imc->lock);
+			icount = imc->sfcount[MCAST_INCLUDE];
+			xcount = imc->sfcount[MCAST_EXCLUDE];
+			for (psf=imc->sources; psf; psf=psf->sf_next) {
+				if (first) {
+					len += sprintf(buffer+len, "%3s %6s "
+						"%10s %10s %6s %6s\n", "Idx",
+						"Device", "MCA", "SRC", "INC",
+						"EXC");
+					first = 0;
+				}
+				len += sprintf(buffer+len, "%3d %6.6s 0x%08x "
+					"0x%08x %6lu %6lu\n", dev->ifindex,
+					dev->name, ntohl(imc->multiaddr),
+					ntohl(psf->sf_inaddr),
+					psf->sf_count[MCAST_INCLUDE],
+					psf->sf_count[MCAST_EXCLUDE]);
+				pos=begin+len;
+				if(pos<offset)
+				{
+					len=0;
+					begin=pos;
+				}
+				if(pos>offset+length) {
+					spin_unlock_bh(&imc->lock);
+					read_unlock(&in_dev->lock);
+					in_dev_put(in_dev);
+					goto done;
+				}
+				icount -= psf->sf_count[MCAST_INCLUDE];
+				xcount -= psf->sf_count[MCAST_EXCLUDE];
+			}
+			if (icount > 0 || xcount > 0) {
+				if (first) {
+					len += sprintf(buffer+len, "%3s %6s "
+						"%10s %10s %6s %6s\n", "Idx",
+						"Device", "MCA", "SRC", "INC",
+						"EXC");
+					first = 0;
+				}
+				len += sprintf(buffer+len, "%3d %6.6s 0x%08x "
+					"%10s %6lu %6lu\n", dev->ifindex,
+					dev->name, ntohl(imc->multiaddr),
+					"NONE", icount, xcount);
+				pos=begin+len;
+				if(pos<offset)
+				{
+					len=0;
+					begin=pos;
+				}
+				if(pos>offset+length) {
+					spin_unlock_bh(&imc->lock);
+					read_unlock(&in_dev->lock);
+					in_dev_put(in_dev);
+					goto done;
+				}
+			}
+			spin_unlock_bh(&imc->lock);
+		}
+		read_unlock(&in_dev->lock);
+		in_dev_put(in_dev);
+	}
+done:
+	read_unlock(&dev_base_lock);
+
+	*start=buffer+(offset-begin);
+	len-=(offset-begin);
+	if(len>length)
+		len=length;
+	if(len<0)
+		len=0;
+	return len;
+}
+
 #endif
 

net/ipv4/ip_output.c

@@ -1312,5 +1312,6 @@ void __init ip_init(void)
 
 #ifdef CONFIG_IP_MULTICAST
 	proc_net_create("igmp", 0, ip_mc_procinfo);
+	proc_net_create("mcfilter", 0, ip_mcf_procinfo);
 #endif
 }

net/ipv4/ip_sockglue.c

@@ -610,9 +610,67 @@ int ip_setsockopt(struct sock *sk, int level, int optname, char *optval, int opt
 			}
 
 			if (optname == IP_ADD_MEMBERSHIP)
-				err = ip_mc_join_group(sk,&mreq);
+				err = ip_mc_join_group(sk, &mreq);
 			else
-				err = ip_mc_leave_group(sk,&mreq);
+				err = ip_mc_leave_group(sk, &mreq);
+			break;
+		}
+		case IP_MSFILTER:
+		{
+			struct ip_msfilter *msf;
+
+			if (optlen < IP_MSFILTER_SIZE(0))
+				goto e_inval;
+			msf = (struct ip_msfilter *)kmalloc(optlen, GFP_KERNEL);
+			if (msf == 0) {
+				err = -ENOBUFS;
+				break;
+			}
+			err = -EFAULT;
+			if (copy_from_user(msf, optval, optlen)) {
+				kfree(msf);
+				break;
+			}
+			err = ip_mc_msfilter(sk, msf);
+			kfree(msf);
+			break;
+		}
+		case IP_BLOCK_SOURCE:
+		case IP_UNBLOCK_SOURCE:
+		case IP_ADD_SOURCE_MEMBERSHIP:
+		case IP_DROP_SOURCE_MEMBERSHIP:
+		{
+			struct ip_mreq_source mreqs;
+			int omode, add;
+
+			if (optlen != sizeof(struct ip_mreq_source))
+				goto e_inval;
+			if (copy_from_user(&mreqs, optval, sizeof(mreqs))) {
+				err = -EFAULT;
+				break;
+			}
+			if (optname == IP_BLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 1;
+			} else if (optname == IP_UNBLOCK_SOURCE) {
+				omode = MCAST_EXCLUDE;
+				add = 0;
+			} else if (optname == IP_ADD_SOURCE_MEMBERSHIP) {
+				struct ip_mreqn mreq;
+
+				mreq.imr_multiaddr.s_addr = mreqs.imr_multiaddr;
+				mreq.imr_address.s_addr = mreqs.imr_interface;
+				mreq.imr_ifindex = 0;
+				err = ip_mc_join_group(sk, &mreq);
+				if (err)
+					break;
+				omode = MCAST_INCLUDE;
+				add = 1;
+			} else /*IP_DROP_SOURCE_MEMBERSHIP */ {
+				omode = MCAST_INCLUDE;
+				add = 0;
+			}
+			err = ip_mc_source(add, omode, sk, &mreqs);
 			break;
 		}
 		case IP_ROUTER_ALERT:	
@@ -763,6 +821,20 @@ int ip_getsockopt(struct sock *sk, int level, int optname, char *optval, int *op
 				return -EFAULT;
 			return 0;
 		}
+		case IP_MSFILTER:
+		{
+			struct ip_msfilter msf;
+			int err;
+
+			if (len < IP_MSFILTER_SIZE(0))
+				return -EINVAL;
+			if (copy_from_user(&msf, optval, IP_MSFILTER_SIZE(0)))
+				return -EFAULT;
+			err = ip_mc_msfget(sk, &msf,
+				(struct ip_msfilter *)optval, optlen);
+			release_sock(sk);
+			return err;
+		}
 		case IP_PKTOPTIONS:		
 		{
 			struct msghdr msg;

net/ipv4/ipcomp.c

+/*
+ * IP Payload Compression Protocol (IPComp) - RFC3713.
+ *
+ * Copyright (c) 2003 James Morris <jmorris@intercode.com.au>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the Free
+ * Software Foundation; either version 2 of the License, or (at your option) 
+ * any later version.
+ *
+ * Todo:
+ *   - Tunable compression parameters.
+ *   - Compression stats.
+ *   - Adaptive compression.
+ */
+#include <linux/config.h>
+#include <linux/module.h>
+#include <asm/scatterlist.h>
+#include <linux/crypto.h>
+#include <linux/pfkeyv2.h>
+#include <net/ip.h>
+#include <net/xfrm.h>
+#include <net/icmp.h>
+#include <net/esp.h>
+
+#define IPCOMP_SCRATCH_SIZE	65400
+
+struct ipcomp_hdr {
+	u8 nexthdr;
+	u8 flags;
+	u16 cpi;
+};
+
+struct ipcomp_data {
+	u16 threshold;
+	u8 *scratch;
+	struct crypto_tfm *tfm;
+};
+
+static int ipcomp_decompress(struct xfrm_state *x, struct sk_buff *skb)
+{
+	int err, plen, dlen;
+	struct iphdr *iph;
+	struct ipcomp_data *ipcd = x->data;
+	u8 *start, *scratch = ipcd->scratch;
+	
+	plen = skb->len;
+	dlen = IPCOMP_SCRATCH_SIZE;
+	start = skb->data;
+
+	err = crypto_comp_decompress(ipcd->tfm, start, plen, scratch, &dlen);
+	if (err)
+		goto out;
+
+	if (dlen < (plen + sizeof(struct ipcomp_hdr))) {
+		err = -EINVAL;
+		goto out;
+	}
+
+	err = pskb_expand_head(skb, 0, dlen - plen, GFP_ATOMIC);
+	if (err)
+		goto out;
+		
+	skb_put(skb, dlen - plen);
+	memcpy(skb->data, scratch, dlen);
+	iph = skb->nh.iph;
+	iph->tot_len = htons(dlen + iph->ihl * 4);
+out:	
+	return err;
+}
+
+static int ipcomp_input(struct xfrm_state *x,
+                        struct xfrm_decap_state *decap, struct sk_buff *skb)
+{
+	u8 nexthdr;
+	int err = 0;
+	struct iphdr *iph;
+	union {
+		struct iphdr	iph;
+		char 		buf[60];
+	} tmp_iph;
+
+
+	if ((skb_is_nonlinear(skb) || skb_cloned(skb)) &&
+	    skb_linearize(skb, GFP_ATOMIC) != 0) {
+	    	err = -ENOMEM;
+	    	goto out;
+	}
+
+	skb->ip_summed = CHECKSUM_NONE;
+
+	/* Remove ipcomp header and decompress original payload */	
+	iph = skb->nh.iph;
+	memcpy(&tmp_iph, iph, iph->ihl * 4);
+	nexthdr = *(u8 *)skb->data;
+	skb_pull(skb, sizeof(struct ipcomp_hdr));
+	memcpy(skb->nh.raw, &tmp_iph, tmp_iph.iph.ihl * 4);
+	iph->tot_len = htons(ntohs(iph->tot_len) - sizeof(struct ipcomp_hdr));
+	iph->protocol = nexthdr;
+	skb->h.raw = skb->data;
+	err = ipcomp_decompress(x, skb);
+
+out:	
+	return err;
+}
+
+static int ipcomp_compress(struct xfrm_state *x, struct sk_buff *skb)
+{
+	int err, plen, dlen, ihlen;
+	struct iphdr *iph = skb->nh.iph;
+	struct ipcomp_data *ipcd = x->data;
+	u8 *start, *scratch = ipcd->scratch;
+	
+	ihlen = iph->ihl * 4;
+	plen = skb->len - ihlen;
+	dlen = IPCOMP_SCRATCH_SIZE;
+	start = skb->data + ihlen;
+
+	err = crypto_comp_compress(ipcd->tfm, start, plen, scratch, &dlen);
+	if (err)
+		goto out;
+
+	if ((dlen + sizeof(struct ipcomp_hdr)) >= plen) {
+		err = -EMSGSIZE;
+		goto out;
+	}
+	
+	memcpy(start, scratch, dlen);
+	pskb_trim(skb, ihlen + dlen);
+	
+out:	
+	return err;
+}
+
+static void ipcomp_tunnel_encap(struct xfrm_state *x,
+                                struct sk_buff *skb, int compress)
+{
+	struct dst_entry *dst = skb->dst;
+	struct iphdr *iph, *top_iph;
+
+	iph = skb->nh.iph;
+	top_iph = (struct iphdr *)skb_push(skb, sizeof(struct iphdr));
+	top_iph->ihl = 5;
+	top_iph->version = 4;
+	top_iph->tos = iph->tos;
+	top_iph->tot_len = htons(skb->len);
+	if (!(iph->frag_off&htons(IP_DF)))
+		__ip_select_ident(top_iph, dst, 0);
+	top_iph->ttl = iph->ttl;
+	top_iph->protocol = compress ? IPPROTO_COMP : IPPROTO_IPIP;
+	top_iph->check = 0;
+	top_iph->saddr = x->props.saddr.a4;
+	top_iph->daddr = x->id.daddr.a4;
+	top_iph->frag_off = iph->frag_off&~htons(IP_MF|IP_OFFSET);
+	memset(&(IPCB(skb)->opt), 0, sizeof(struct ip_options));
+	skb->nh.raw = skb->data;
+}
+
+static int ipcomp_output(struct sk_buff *skb)
+{
+	int err;
+	struct dst_entry *dst = skb->dst;
+	struct xfrm_state *x = dst->xfrm;
+	struct iphdr *iph, *top_iph;
+	struct ipcomp_hdr *ipch;
+	struct ipcomp_data *ipcd = x->data;
+	union {
+		struct iphdr	iph;
+		char 		buf[60];
+	} tmp_iph;
+
+	if (skb->ip_summed == CHECKSUM_HW && skb_checksum_help(skb) == NULL) {
+		err = -EINVAL;
+		goto error_nolock;
+	}
+
+	spin_lock_bh(&x->lock);
+	
+	if ((err = xfrm_state_check_expire(x)) != 0)
+		goto error;
+	if ((err = xfrm_state_check_space(x, skb)) != 0)
+		goto error;
+
+	/* Don't bother compressing */
+	if (skb->len < ipcd->threshold) {
+		if (x->props.mode) {
+			ipcomp_tunnel_encap(x, skb, 0);
+			iph = skb->nh.iph;
+			ip_send_check(iph);
+		}
+		goto out_ok;
+	}
+
+	if (x->props.mode) 
+		ipcomp_tunnel_encap(x, skb, 1);
+
+	if ((skb_is_nonlinear(skb) || skb_cloned(skb)) &&
+	    skb_linearize(skb, GFP_ATOMIC) != 0) {
+	    	err = -ENOMEM;
+	    	goto error;
+	}
+	
+	err = ipcomp_compress(x, skb);
+	if (err) {
+		if (err == -EMSGSIZE)
+			goto out_ok;
+		goto error;
+	}
+
+	/* Install ipcomp header, convert into ipcomp datagram. */
+	iph = skb->nh.iph;
+	memcpy(&tmp_iph, iph, iph->ihl * 4);
+	top_iph = (struct iphdr *)skb_push(skb, sizeof(struct ipcomp_hdr));
+	memcpy(top_iph, &tmp_iph, iph->ihl * 4);
+	iph = top_iph;
+	iph->tot_len = htons(skb->len);
+	iph->protocol = IPPROTO_COMP;
+	iph->check = 0;
+	ipch = (struct ipcomp_hdr *)((char *)iph + iph->ihl * 4);
+	ipch->nexthdr = x->props.mode ? IPPROTO_IPIP : tmp_iph.iph.protocol;
+	ipch->flags = 0;
+	ipch->cpi = htons((u16 )ntohl(x->id.spi));
+	ip_send_check(iph);
+	skb->nh.raw = skb->data;
+
+out_ok:
+	x->curlft.bytes += skb->len;
+	x->curlft.packets++;
+	spin_unlock_bh(&x->lock);
+	
+	if ((skb->dst = dst_pop(dst)) == NULL) {
+		err = -EHOSTUNREACH;
+		goto error_nolock;
+	}
+	err = NET_XMIT_BYPASS;
+
+out_exit:
+	return err;
+error:
+	spin_unlock_bh(&x->lock);
+error_nolock:
+	kfree_skb(skb);
+	goto out_exit;
+}
+
+static void ipcomp4_err(struct sk_buff *skb, u32 info)
+{
+	u32 spi;
+	struct iphdr *iph = (struct iphdr *)skb->data;
+	struct ipcomp_hdr *ipch = (struct ipcomp_hdr *)(skb->data+(iph->ihl<<2));
+	struct xfrm_state *x;
+
+	if (skb->h.icmph->type != ICMP_DEST_UNREACH ||
+	    skb->h.icmph->code != ICMP_FRAG_NEEDED)
+		return;
+
+	spi = ntohl(ntohs(ipch->cpi));
+	x = xfrm_state_lookup((xfrm_address_t *)&iph->daddr,
+	                      spi, IPPROTO_COMP, AF_INET);
+	if (!x)
+		return;
+	printk(KERN_DEBUG "pmtu discvovery on SA IPCOMP/%08x/%u.%u.%u.%u\n",
+	       spi, NIPQUAD(iph->daddr));
+	xfrm_state_put(x);
+}
+
+static void ipcomp_free_data(struct ipcomp_data *ipcd)
+{
+	if (ipcd->tfm)
+		crypto_free_tfm(ipcd->tfm);
+	if (ipcd->scratch)
+		kfree(ipcd->scratch);	
+}
+
+static void ipcomp_destroy(struct xfrm_state *x)
+{
+	struct ipcomp_data *ipcd = x->data;
+	ipcomp_free_data(ipcd);
+	kfree(ipcd);
+}
+
+static int ipcomp_init_state(struct xfrm_state *x, void *args)
+{
+	int err = -ENOMEM;
+	struct ipcomp_data *ipcd;
+	struct xfrm_algo_desc *calg_desc;
+
+	ipcd = kmalloc(sizeof(*ipcd), GFP_KERNEL);
+	if (!ipcd)
+		goto error;
+
+	memset(ipcd, 0, sizeof(*ipcd));
+	x->props.header_len = sizeof(struct ipcomp_hdr);
+	if (x->props.mode)
+		x->props.header_len += sizeof(struct iphdr);
+	x->data = ipcd;
+
+	ipcd->scratch = kmalloc(IPCOMP_SCRATCH_SIZE, GFP_KERNEL);
+	if (!ipcd->scratch)
+		goto error;
+	
+	ipcd->tfm = crypto_alloc_tfm(x->calg->alg_name, 0);
+	if (!ipcd->tfm)
+		goto error;
+
+	calg_desc = xfrm_calg_get_byname(x->calg->alg_name);
+	BUG_ON(!calg_desc);
+	ipcd->threshold = calg_desc->uinfo.comp.threshold;
+	err = 0;
+out:
+	return err;
+
+error:
+	if (ipcd) {
+		ipcomp_free_data(ipcd);
+		kfree(ipcd);
+	}
+	goto out;
+}
+
+static struct xfrm_type ipcomp_type =
+{
+	.description	= "IPCOMP4",
+	.proto	     	= IPPROTO_COMP,
+	.init_state	= ipcomp_init_state,
+	.destructor	= ipcomp_destroy,
+	.input		= ipcomp_input,
+	.output		= ipcomp_output
+};
+
+static struct inet_protocol ipcomp4_protocol = {
+	.handler	=	xfrm4_rcv,
+	.err_handler	=	ipcomp4_err,
+	.no_policy	=	1,
+};
+
+static int __init ipcomp4_init(void)
+{
+	SET_MODULE_OWNER(&ipcomp_type);
+	if (xfrm_register_type(&ipcomp_type, AF_INET) < 0) {
+		printk(KERN_INFO "ipcomp init: can't add xfrm type\n");
+		return -EAGAIN;
+	}
+	if (inet_add_protocol(&ipcomp4_protocol, IPPROTO_COMP) < 0) {
+		printk(KERN_INFO "ipcomp init: can't add protocol\n");
+		xfrm_unregister_type(&ipcomp_type, AF_INET);
+		return -EAGAIN;
+	}
+	return 0;
+}
+
+static void __exit ipcomp4_fini(void)
+{
+	if (inet_del_protocol(&ipcomp4_protocol, IPPROTO_COMP) < 0)
+		printk(KERN_INFO "ip ipcomp close: can't remove protocol\n");
+	if (xfrm_unregister_type(&ipcomp_type, AF_INET) < 0)
+		printk(KERN_INFO "ip ipcomp close: can't remove xfrm type\n");
+}
+
+module_init(ipcomp4_init);
+module_exit(ipcomp4_fini);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("IP Payload Compression Protocol (IPComp) - RFC3713");
+MODULE_AUTHOR("James Morris <jmorris@intercode.com.au>");
+

net/ipv4/netfilter/ip_conntrack_core.c

@@ -273,6 +273,8 @@ static void remove_expectations(struct ip_conntrack *ct)
 		 * the un-established ones only */
 		if (exp->sibling) {
 			DEBUGP("remove_expectations: skipping established %p of %p\n", exp->sibling, ct);
+			/* Indicate that this expectations parent is dead */
+			exp->expectant = NULL;
 			continue;
 		}
 
@@ -324,6 +326,9 @@ destroy_conntrack(struct nf_conntrack *nfct)
 		ip_conntrack_destroyed(ct);
 
 	WRITE_LOCK(&ip_conntrack_lock);
+	/* Delete us from our own list to prevent corruption later */
+	list_del(&ct->sibling_list);
+
 	/* Delete our master expectation */
 	if (ct->master) {
 		/* can't call __unexpect_related here,

net/ipv4/route.c

@@ -1790,7 +1790,8 @@ int ip_route_input(struct sk_buff *skb, u32 daddr, u32 saddr,
 
 		read_lock(&inetdev_lock);
 		if ((in_dev = __in_dev_get(dev)) != NULL) {
-			int our = ip_check_mc(in_dev, daddr);
+			int our = ip_check_mc(in_dev, daddr, saddr,
+				skb->nh.iph->protocol);
 			if (our
 #ifdef CONFIG_IP_MROUTE
 			    || (!LOCAL_MCAST(daddr) && IN_DEV_MFORWARD(in_dev))
@@ -2020,7 +2021,7 @@ int ip_route_output_slow(struct rtable **rp, const struct flowi *oldflp)
 		}
 	} else if (res.type == RTN_MULTICAST) {
 		flags |= RTCF_MULTICAST|RTCF_LOCAL;
-		if (!ip_check_mc(in_dev, oldflp->fl4_dst))
+		if (!ip_check_mc(in_dev, oldflp->fl4_dst, oldflp->fl4_src, oldflp->proto))
 			flags &= ~RTCF_LOCAL;
 		/* If multicast route do not exist use
 		   default one, but do not gateway in this case.

net/ipv4/udp.c

@@ -298,6 +298,8 @@ static inline struct sock *udp_v4_mcast_next(struct sock *sk,
 		    ipv6_only_sock(s)					||
 		    (s->bound_dev_if && s->bound_dev_if != dif))
 			continue;
+		if (!ip_mc_sf_allow(sk, loc_addr, rmt_addr, dif))
+			continue;
 		break;
   	}
   	return s;

net/ipv4/xfrm4_input.c

@@ -48,6 +48,9 @@ int xfrm4_rcv_encap(struct sk_buff *skb, __u16 encap_type)
 		if (x->props.replay_window && xfrm_replay_check(x, seq))
 			goto drop_unlock;
 
+		if (xfrm_state_check_expire(x))
+			goto drop_unlock;
+
 		xfrm_vec[xfrm_nr].decap.decap_type = encap_type;
 		if (x->type->input(x, &(xfrm_vec[xfrm_nr].decap), skb))
 			goto drop_unlock;

net/ipv4/xfrm4_policy.c

@@ -201,6 +201,13 @@ _decode_session4(struct sk_buff *skb, struct flowi *fl)
 			}
 			break;
 
+		case IPPROTO_COMP:
+			if (pskb_may_pull(skb, xprth + 4 - skb->data)) {
+				u16 *ipcomp_hdr = (u16 *)xprth;
+
+				fl->uli_u.spi = ntohl(ntohs(ipcomp_hdr[1]));
+			}
+			break;
 		default:
 			fl->uli_u.spi = 0;
 			break;

net/ipv6/addrconf.c

@@ -1202,7 +1202,7 @@ addrconf_prefix_route(struct in6_addr *pfx, int plen, struct net_device *dev,
 	if (dev->type == ARPHRD_SIT && (dev->flags&IFF_POINTOPOINT))
 		rtmsg.rtmsg_flags |= RTF_NONEXTHOP;
 
-	ip6_route_add(&rtmsg);
+	ip6_route_add(&rtmsg, NULL);
 }
 
 /* Create "default" multicast route to the interface */
@@ -1219,7 +1219,7 @@ static void addrconf_add_mroute(struct net_device *dev)
 	rtmsg.rtmsg_ifindex = dev->ifindex;
 	rtmsg.rtmsg_flags = RTF_UP|RTF_ADDRCONF;
 	rtmsg.rtmsg_type = RTMSG_NEWROUTE;
-	ip6_route_add(&rtmsg);
+	ip6_route_add(&rtmsg, NULL);
 }
 
 static void sit_route_add(struct net_device *dev)
@@ -1236,7 +1236,7 @@ static void sit_route_add(struct net_device *dev)
 	rtmsg.rtmsg_flags	= RTF_UP|RTF_NONEXTHOP;
 	rtmsg.rtmsg_ifindex	= dev->ifindex;
 
-	ip6_route_add(&rtmsg);
+	ip6_route_add(&rtmsg, NULL);
 }
 
 static void addrconf_add_lroute(struct net_device *dev)
@@ -1328,7 +1328,7 @@ void addrconf_prefix_rcv(struct net_device *dev, u8 *opt, int len)
 	if (rt && ((rt->rt6i_flags & (RTF_GATEWAY | RTF_DEFAULT)) == 0)) {
 		if (rt->rt6i_flags&RTF_EXPIRES) {
 			if (pinfo->onlink == 0 || valid_lft == 0) {
-				ip6_del_rt(rt);
+				ip6_del_rt(rt, NULL);
 				rt = NULL;
 			} else {
 				rt->rt6i_expires = rt_expires;
@@ -1952,7 +1952,7 @@ static void addrconf_rs_timer(unsigned long data)
 
 		rtmsg.rtmsg_ifindex = ifp->idev->dev->ifindex;
 
-		ip6_route_add(&rtmsg);
+		ip6_route_add(&rtmsg, NULL);
 	}
 
 out:

net/ipv6/ip6_fib.c

@@ -423,7 +423,8 @@ static struct fib6_node * fib6_add_1(struct fib6_node *root, void *addr,
  *	Insert routing information in a node.
  */
 
-static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt)
+static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt,
+    struct nlmsghdr *nlh)
 {
 	struct rt6_info *iter = NULL;
 	struct rt6_info **ins;
@@ -480,7 +481,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct rt6_info *rt)
 	*ins = rt;
 	rt->rt6i_node = fn;
 	atomic_inc(&rt->rt6i_ref);
-	inet6_rt_notify(RTM_NEWROUTE, rt);
+	inet6_rt_notify(RTM_NEWROUTE, rt, nlh);
 	rt6_stats.fib_rt_entries++;
 
 	if ((fn->fn_flags & RTN_RTINFO) == 0) {
@@ -504,7 +505,7 @@ static __inline__ void fib6_start_gc(struct rt6_info *rt)
  *	with source addr info in sub-trees
  */
 
-int fib6_add(struct fib6_node *root, struct rt6_info *rt)
+int fib6_add(struct fib6_node *root, struct rt6_info *rt, struct nlmsghdr *nlh)
 {
 	struct fib6_node *fn;
 	int err = -ENOMEM;
@@ -577,7 +578,7 @@ int fib6_add(struct fib6_node *root, struct rt6_info *rt)
 	}
 #endif
 
-	err = fib6_add_rt2node(fn, rt);
+	err = fib6_add_rt2node(fn, rt, nlh);
 
 	if (err == 0) {
 		fib6_start_gc(rt);
@@ -885,7 +886,8 @@ static struct fib6_node * fib6_repair_tree(struct fib6_node *fn)
 	}
 }
 
-static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp)
+static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp,
+    struct nlmsghdr *nlh)
 {
 	struct fib6_walker_t *w;
 	struct rt6_info *rt = *rtp;
@@ -940,11 +942,11 @@ static void fib6_del_route(struct fib6_node *fn, struct rt6_info **rtp)
 		if (atomic_read(&rt->rt6i_ref) != 1) BUG();
 	}
 
-	inet6_rt_notify(RTM_DELROUTE, rt);
+	inet6_rt_notify(RTM_DELROUTE, rt, nlh);
 	rt6_release(rt);
 }
 
-int fib6_del(struct rt6_info *rt)
+int fib6_del(struct rt6_info *rt, struct nlmsghdr *nlh)
 {
 	struct fib6_node *fn = rt->rt6i_node;
 	struct rt6_info **rtp;
@@ -969,7 +971,7 @@ int fib6_del(struct rt6_info *rt)
 
 	for (rtp = &fn->leaf; *rtp; rtp = &(*rtp)->u.next) {
 		if (*rtp == rt) {
-			fib6_del_route(fn, rtp);
+			fib6_del_route(fn, rtp, nlh);
 			return 0;
 		}
 	}
@@ -1098,7 +1100,7 @@ static int fib6_clean_node(struct fib6_walker_t *w)
 		res = c->func(rt, c->arg);
 		if (res < 0) {
 			w->leaf = rt;
-			res = fib6_del(rt);
+			res = fib6_del(rt, NULL);
 			if (res) {
 #if RT6_DEBUG >= 2
 				printk(KERN_DEBUG "fib6_clean_node: del failed: rt=%p@%p err=%d\n", rt, rt->rt6i_node, res);

net/ipv6/ndisc.c

@@ -961,7 +961,7 @@ void ndisc_recv_na(struct sk_buff *skb)
 				struct rt6_info *rt;
 				rt = rt6_get_dflt_router(saddr, dev);
 				if (rt)
-					ip6_del_rt(rt);
+					ip6_del_rt(rt, NULL);
 			}
 		} else {
 			if (msg->icmph.icmp6_router)
@@ -1035,7 +1035,7 @@ static void ndisc_router_discovery(struct sk_buff *skb)
 	rt = rt6_get_dflt_router(&skb->nh.ipv6h->saddr, skb->dev);
 
 	if (rt && lifetime == 0) {
-		ip6_del_rt(rt);
+		ip6_del_rt(rt, NULL);
 		rt = NULL;
 	}
 

net/ipv6/route.c

@@ -318,12 +318,12 @@ struct rt6_info *rt6_lookup(struct in6_addr *daddr, struct in6_addr *saddr,
    be destroyed.
  */
 
-static int rt6_ins(struct rt6_info *rt)
+static int rt6_ins(struct rt6_info *rt, struct nlmsghdr *nlh)
 {
 	int err;
 
 	write_lock_bh(&rt6_lock);
-	err = fib6_add(&ip6_routing_table, rt);
+	err = fib6_add(&ip6_routing_table, rt, nlh);
 	write_unlock_bh(&rt6_lock);
 
 	return err;
@@ -366,7 +366,7 @@ static struct rt6_info *rt6_cow(struct rt6_info *ort, struct in6_addr *daddr,
 
 		dst_hold(&rt->u.dst);
 
-		err = rt6_ins(rt);
+		err = rt6_ins(rt, NULL);
 		if (err == 0)
 			return rt;
 
@@ -522,7 +522,7 @@ static struct dst_entry *ip6_negative_advice(struct dst_entry *dst)
 
 	if (rt) {
 		if (rt->rt6i_flags & RTF_CACHE)
-			ip6_del_rt(rt);
+			ip6_del_rt(rt, NULL);
 		else
 			dst_release(dst);
 	}
@@ -625,9 +625,10 @@ static int ipv6_get_hoplimit(struct net_device *dev)
  *
  */
 
-int ip6_route_add(struct in6_rtmsg *rtmsg)
+int ip6_route_add(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh)
 {
 	int err;
+	struct rtmsg *r;
 	struct rt6_info *rt;
 	struct net_device *dev = NULL;
 	int addr_type;
@@ -648,6 +649,11 @@ int ip6_route_add(struct in6_rtmsg *rtmsg)
 
 	rt->u.dst.obsolete = -1;
 	rt->rt6i_expires = rtmsg->rtmsg_info;
+	if (nlh && (r = NLMSG_DATA(nlh))) {
+		rt->rt6i_protocol = r->rtm_protocol;
+	} else {
+		rt->rt6i_protocol = RTPROT_BOOT;
+	}
 
 	addr_type = ipv6_addr_type(&rtmsg->rtmsg_dst);
 
@@ -772,7 +778,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg)
 	if (dst_metric(&rt->u.dst, RTAX_ADVMSS) > 65535-20)
 		rt->u.dst.metrics[RTAX_ADVMSS-1] = 65535;
 	rt->u.dst.dev = dev;
-	return rt6_ins(rt);
+	return rt6_ins(rt, nlh);
 
 out:
 	if (dev)
@@ -781,7 +787,7 @@ int ip6_route_add(struct in6_rtmsg *rtmsg)
 	return err;
 }
 
-int ip6_del_rt(struct rt6_info *rt)
+int ip6_del_rt(struct rt6_info *rt, struct nlmsghdr *nlh)
 {
 	int err;
 
@@ -793,13 +799,13 @@ int ip6_del_rt(struct rt6_info *rt)
 
 	dst_release(&rt->u.dst);
 
-	err = fib6_del(rt);
+	err = fib6_del(rt, nlh);
 	write_unlock_bh(&rt6_lock);
 
 	return err;
 }
 
-static int ip6_route_del(struct in6_rtmsg *rtmsg)
+static int ip6_route_del(struct in6_rtmsg *rtmsg, struct nlmsghdr *nlh)
 {
 	struct fib6_node *fn;
 	struct rt6_info *rt;
@@ -826,7 +832,7 @@ static int ip6_route_del(struct in6_rtmsg *rtmsg)
 			dst_hold(&rt->u.dst);
 			read_unlock_bh(&rt6_lock);
 
-			return ip6_del_rt(rt);
+			return ip6_del_rt(rt, nlh);
 		}
 	}
 	read_unlock_bh(&rt6_lock);
@@ -928,11 +934,11 @@ void rt6_redirect(struct in6_addr *dest, struct in6_addr *saddr,
 		nrt->u.dst.metrics[RTAX_ADVMSS-1] = 65535;
 	nrt->rt6i_hoplimit = ipv6_get_hoplimit(neigh->dev);
 
-	if (rt6_ins(nrt))
+	if (rt6_ins(nrt, NULL))
 		goto out;
 
 	if (rt->rt6i_flags&RTF_CACHE) {
-		ip6_del_rt(rt);
+		ip6_del_rt(rt, NULL);
 		return;
 	}
 
@@ -1018,7 +1024,7 @@ 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;
-		rt6_ins(nrt);
+		rt6_ins(nrt, NULL);
 	}
 
 out:
@@ -1091,7 +1097,7 @@ struct rt6_info *rt6_add_dflt_router(struct in6_addr *gwaddr,
 
 	rtmsg.rtmsg_ifindex = dev->ifindex;
 
-	ip6_route_add(&rtmsg);
+	ip6_route_add(&rtmsg, NULL);
 	return rt6_get_dflt_router(gwaddr, dev);
 }
 
@@ -1117,7 +1123,7 @@ void rt6_purge_dflt_routers(int last_resort)
 
 			read_unlock_bh(&rt6_lock);
 
-			ip6_del_rt(rt);
+			ip6_del_rt(rt, NULL);
 
 			goto restart;
 		}
@@ -1143,10 +1149,10 @@ int ipv6_route_ioctl(unsigned int cmd, void *arg)
 		rtnl_lock();
 		switch (cmd) {
 		case SIOCADDRT:
-			err = ip6_route_add(&rtmsg);
+			err = ip6_route_add(&rtmsg, NULL);
 			break;
 		case SIOCDELRT:
-			err = ip6_route_del(&rtmsg);
+			err = ip6_route_del(&rtmsg, NULL);
 			break;
 		default:
 			err = -EINVAL;
@@ -1203,7 +1209,7 @@ int ip6_rt_addr_add(struct in6_addr *addr, struct net_device *dev)
 
 	ipv6_addr_copy(&rt->rt6i_dst.addr, addr);
 	rt->rt6i_dst.plen = 128;
-	rt6_ins(rt);
+	rt6_ins(rt, NULL);
 
 	return 0;
 }
@@ -1220,7 +1226,7 @@ int ip6_rt_addr_del(struct in6_addr *addr, struct net_device *dev)
 	rt = rt6_lookup(addr, NULL, loopback_dev.ifindex, 1);
 	if (rt) {
 		if (rt->rt6i_dst.plen == 128)
-			err = ip6_del_rt(rt);
+			err = ip6_del_rt(rt, NULL);
 		else
 			dst_release(&rt->u.dst);
 	}
@@ -1350,7 +1356,7 @@ int inet6_rtm_delroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
 		return -EINVAL;
-	return ip6_route_del(&rtmsg);
+	return ip6_route_del(&rtmsg, nlh);
 }
 
 int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
@@ -1360,7 +1366,7 @@ int inet6_rtm_newroute(struct sk_buff *skb, struct nlmsghdr* nlh, void *arg)
 
 	if (inet6_rtm_to_rtmsg(r, arg, &rtmsg))
 		return -EINVAL;
-	return ip6_route_add(&rtmsg);
+	return ip6_route_add(&rtmsg, nlh);
 }
 
 struct rt6_rtnl_dump_arg
@@ -1373,13 +1379,18 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 			 struct in6_addr *dst,
 			 struct in6_addr *src,
 			 int iif,
-			 int type, u32 pid, u32 seq)
+			 int type, u32 pid, u32 seq,
+			 struct nlmsghdr *in_nlh)
 {
 	struct rtmsg *rtm;
 	struct nlmsghdr  *nlh;
 	unsigned char	 *b = skb->tail;
 	struct rta_cacheinfo ci;
 
+	if (!pid && in_nlh) {
+		pid = in_nlh->nlmsg_pid;
+	}
+
 	nlh = NLMSG_PUT(skb, pid, seq, type, sizeof(*rtm));
 	rtm = NLMSG_DATA(nlh);
 	rtm->rtm_family = AF_INET6;
@@ -1395,7 +1406,7 @@ static int rt6_fill_node(struct sk_buff *skb, struct rt6_info *rt,
 		rtm->rtm_type = RTN_UNICAST;
 	rtm->rtm_flags = 0;
 	rtm->rtm_scope = RT_SCOPE_UNIVERSE;
-	rtm->rtm_protocol = RTPROT_BOOT;
+	rtm->rtm_protocol = rt->rt6i_protocol;
 	if (rt->rt6i_flags&RTF_DYNAMIC)
 		rtm->rtm_protocol = RTPROT_REDIRECT;
 	else if (rt->rt6i_flags&(RTF_ADDRCONF|RTF_ALLONLINK))
@@ -1458,7 +1469,8 @@ static int rt6_dump_route(struct rt6_info *rt, void *p_arg)
 	struct rt6_rtnl_dump_arg *arg = (struct rt6_rtnl_dump_arg *) p_arg;
 
 	return rt6_fill_node(arg->skb, rt, NULL, NULL, 0, RTM_NEWROUTE,
-			     NETLINK_CB(arg->cb->skb).pid, arg->cb->nlh->nlmsg_seq);
+		     NETLINK_CB(arg->cb->skb).pid, arg->cb->nlh->nlmsg_seq,
+		     NULL);
 }
 
 static int fib6_dump_node(struct fib6_walker_t *w)
@@ -1608,7 +1620,8 @@ int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 			    fl.nl_u.ip6_u.daddr,
 			    fl.nl_u.ip6_u.saddr,
 			    iif,
-			    RTM_NEWROUTE, NETLINK_CB(in_skb).pid, nlh->nlmsg_seq);
+			    RTM_NEWROUTE, NETLINK_CB(in_skb).pid,
+			    nlh->nlmsg_seq, nlh);
 	if (err < 0) {
 		err = -EMSGSIZE;
 		goto out_free;
@@ -1624,7 +1637,7 @@ int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr* nlh, void *arg)
 	goto out;	
 }
 
-void inet6_rt_notify(int event, struct rt6_info *rt)
+void inet6_rt_notify(int event, struct rt6_info *rt, struct nlmsghdr *nlh)
 {
 	struct sk_buff *skb;
 	int size = NLMSG_SPACE(sizeof(struct rtmsg)+256);
@@ -1634,7 +1647,7 @@ void inet6_rt_notify(int event, struct rt6_info *rt)
 		netlink_set_err(rtnl, 0, RTMGRP_IPV6_ROUTE, ENOBUFS);
 		return;
 	}
-	if (rt6_fill_node(skb, rt, NULL, NULL, 0, event, 0, 0) < 0) {
+	if (rt6_fill_node(skb, rt, NULL, NULL, 0, event, 0, 0, nlh) < 0) {
 		kfree_skb(skb);
 		netlink_set_err(rtnl, 0, RTMGRP_IPV6_ROUTE, EINVAL);
 		return;

net/ipv6/xfrm6_input.c

@@ -172,6 +172,9 @@ int xfrm6_rcv(struct sk_buff **pskb)
 		if (x->props.replay_window && xfrm_replay_check(x, seq))
 			goto drop_unlock;
 
+		if (xfrm_state_check_expire(x))
+			goto drop_unlock;
+
 		nexthdr = x->type->input(x, &(xfrm_vec[xfrm_nr].decap), skb);
 		if (nexthdr <= 0)
 			goto drop_unlock;

net/ipv6/xfrm6_policy.c

@@ -203,6 +203,7 @@ _decode_session6(struct sk_buff *skb, struct flowi *fl)
 		/* XXX Why are there these headers? */
 		case IPPROTO_AH:
 		case IPPROTO_ESP:
+		case IPPROTO_COMP:
 		default:
 			fl->uli_u.spi = 0;
 			return;

net/netsyms.c

@@ -234,8 +234,8 @@ EXPORT_SYMBOL(scm_detach_fds);
 
 #if defined(CONFIG_BRIDGE) || defined(CONFIG_BRIDGE_MODULE)
 EXPORT_SYMBOL(br_handle_frame_hook);
+EXPORT_SYMBOL(brioctl_set);
 #endif
-EXPORT_SYMBOL(br_ioctl_hook);
 
 #ifdef CONFIG_NET_DIVERT
 EXPORT_SYMBOL(alloc_divert_blk);

net/sched/sch_csz.c

@@ -749,6 +749,14 @@ csz_reset(struct Qdisc* sch)
 static void
 csz_destroy(struct Qdisc* sch)
 {
+	struct csz_sched_data *q = (struct csz_sched_data *)sch->data;
+	struct tcf_proto *tp;
+
+	while ((tp = q->filter_list) != NULL) {
+		q->filter_list = tp->next;
+		tp->ops->destroy(tp);
+	}
+
 	MOD_DEC_USE_COUNT;
 }
 

net/sched/sch_htb.c

@@ -102,7 +102,9 @@
 #define HTB_PASSQ q,
 #define HTB_ARGQ struct htb_sched *q,
 #define static
+#undef __inline__
 #define __inline__
+#undef inline
 #define inline
 #define HTB_CMAGIC 0xFEFAFEF1
 #define htb_safe_rb_erase(N,R) do { BUG_TRAP((N)->rb_color != -1); \

net/sched/sch_prio.c

@@ -158,6 +158,12 @@ prio_destroy(struct Qdisc* sch)
 {
 	int prio;
 	struct prio_sched_data *q = (struct prio_sched_data *)sch->data;
+	struct tcf_proto *tp;
+
+	while ((tp = q->filter_list) != NULL) {
+		q->filter_list = tp->next;
+		tp->ops->destroy(tp);
+	}
 
 	for (prio=0; prio<q->bands; prio++) {
 		qdisc_destroy(q->queues[prio]);

net/socket.c

@@ -71,6 +71,7 @@
 #include <linux/wanrouter.h>
 #include <linux/netlink.h>
 #include <linux/rtnetlink.h>
+#include <linux/if_bridge.h>
 #include <linux/init.h>
 #include <linux/poll.h>
 #include <linux/cache.h>
@@ -712,7 +713,18 @@ static ssize_t sock_writev(struct file *file, const struct iovec *vector,
 				 file, vector, count, tot_len);
 }
 
-int (*br_ioctl_hook)(unsigned long arg);
+
+static DECLARE_MUTEX(br_ioctl_mutex);
+static int (*br_ioctl_hook)(unsigned long arg) = NULL;
+
+void brioctl_set(int (*hook)(unsigned long))
+{
+	down(&br_ioctl_mutex);
+	br_ioctl_hook = hook;
+	up(&br_ioctl_mutex);
+}
+
+
 int (*vlan_ioctl_hook)(unsigned long arg);
 
 #ifdef CONFIG_DLCI
@@ -759,12 +771,16 @@ static int sock_ioctl(struct inode *inode, struct file *file, unsigned int cmd,
 		case SIOCGIFBR:
 		case SIOCSIFBR:
 			err = -ENOPKG;
+			
 #ifdef CONFIG_KMOD
 			if (!br_ioctl_hook)
 				request_module("bridge");
 #endif
-			if (br_ioctl_hook)
+
+			down(&br_ioctl_mutex);
+			if (br_ioctl_hook) 
 				err = br_ioctl_hook(arg);
+			up(&br_ioctl_mutex);
 			break;
 		case SIOCGIFVLAN:
 		case SIOCSIFVLAN: