X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=net%2Fnetlink%2Faf_netlink.c;h=87d8074b7757a534bd133bc84282abeefbd0b717;hb=refs%2Fheads%2Fvserver;hp=4aa05030ebae63cd589fc4f8aa98bda8b24c56e7;hpb=c7b5ebbddf7bcd3651947760f423e3783bbe6573;p=linux-2.6.git

diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 4aa05030e..87d8074b7 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -13,15 +13,19 @@
  *                               added netlink_proto_exit
  * Tue Jan 22 18:32:44 BRST 2002 Arnaldo C. de Melo <acme@conectiva.com.br>
  * 				 use nlk_sk, as sk->protinfo is on a diet 8)
- *
+ * Fri Jul 22 19:51:12 MEST 2005 Harald Welte <laforge@gnumonks.org>
+ * 				 - inc module use count of module that owns
+ * 				   the kernel socket in case userspace opens
+ * 				   socket of same protocol
+ * 				 - remove all module support, since netlink is
+ * 				   mandatory if CONFIG_NET=y these days
  */
 
-#include <linux/config.h>
 #include <linux/module.h>
 
+#include <linux/capability.h>
 #include <linux/kernel.h>
 #include <linux/init.h>
-#include <linux/major.h>
 #include <linux/signal.h>
 #include <linux/sched.h>
 #include <linux/errno.h>
@@ -44,52 +48,94 @@
 #include <linux/smp_lock.h>
 #include <linux/notifier.h>
 #include <linux/security.h>
-#include <linux/vs_base.h>
+#include <linux/jhash.h>
+#include <linux/jiffies.h>
+#include <linux/random.h>
+#include <linux/bitops.h>
+#include <linux/mm.h>
+#include <linux/types.h>
+#include <linux/audit.h>
+#include <linux/selinux.h>
 #include <linux/vs_context.h>
 #include <linux/vs_network.h>
 #include <linux/vs_limit.h>
+
 #include <net/sock.h>
 #include <net/scm.h>
+#include <net/netlink.h>
 
-#define Nprintk(a...)
-
-#if defined(CONFIG_NETLINK_DEV) || defined(CONFIG_NETLINK_DEV_MODULE)
-#define NL_EMULATE_DEV
-#endif
+#define NLGRPSZ(x)	(ALIGN(x, sizeof(unsigned long) * 8) / 8)
 
-struct netlink_opt
-{
+struct netlink_sock {
+	/* struct sock has to be the first member of netlink_sock */
+	struct sock		sk;
 	u32			pid;
-	unsigned		groups;
 	u32			dst_pid;
-	unsigned		dst_groups;
+	u32			dst_group;
+	u32			flags;
+	u32			subscriptions;
+	u32			ngroups;
+	unsigned long		*groups;
 	unsigned long		state;
-	int			(*handler)(int unit, struct sk_buff *skb);
 	wait_queue_head_t	wait;
 	struct netlink_callback	*cb;
 	spinlock_t		cb_lock;
 	void			(*data_ready)(struct sock *sk, int bytes);
+	struct module		*module;
 };
 
-#define nlk_sk(__sk) ((struct netlink_opt *)(__sk)->sk_protinfo)
+#define NETLINK_KERNEL_SOCKET	0x1
+#define NETLINK_RECV_PKTINFO	0x2
 
-static struct hlist_head nl_table[MAX_LINKS];
-static DECLARE_WAIT_QUEUE_HEAD(nl_table_wait);
-static unsigned nl_nonroot[MAX_LINKS];
+static inline struct netlink_sock *nlk_sk(struct sock *sk)
+{
+	return (struct netlink_sock *)sk;
+}
 
-#ifdef NL_EMULATE_DEV
-static struct socket *netlink_kernel[MAX_LINKS];
-#endif
+struct nl_pid_hash {
+	struct hlist_head *table;
+	unsigned long rehash_time;
+
+	unsigned int mask;
+	unsigned int shift;
+
+	unsigned int entries;
+	unsigned int max_shift;
+
+	u32 rnd;
+};
+
+struct netlink_table {
+	struct nl_pid_hash hash;
+	struct hlist_head mc_list;
+	unsigned long *listeners;
+	unsigned int nl_nonroot;
+	unsigned int groups;
+	struct module *module;
+	int registered;
+};
+
+static struct netlink_table *nl_table;
+
+static DECLARE_WAIT_QUEUE_HEAD(nl_table_wait);
 
 static int netlink_dump(struct sock *sk);
 static void netlink_destroy_callback(struct netlink_callback *cb);
 
-atomic_t netlink_sock_nr;
-
-static rwlock_t nl_table_lock = RW_LOCK_UNLOCKED;
+static DEFINE_RWLOCK(nl_table_lock);
 static atomic_t nl_table_users = ATOMIC_INIT(0);
 
-static struct notifier_block *netlink_chain;
+static ATOMIC_NOTIFIER_HEAD(netlink_chain);
+
+static u32 netlink_group_mask(u32 group)
+{
+	return group ? 1 << (group - 1) : 0;
+}
+
+static struct hlist_head *nl_pid_hashfn(struct nl_pid_hash *hash, u32 pid)
+{
+	return &hash->table[jhash_1word(pid, hash->rnd) & hash->mask];
+}
 
 static void netlink_sock_destruct(struct sock *sk)
 {
@@ -102,13 +148,7 @@ static void netlink_sock_destruct(struct sock *sk)
 	BUG_TRAP(!atomic_read(&sk->sk_rmem_alloc));
 	BUG_TRAP(!atomic_read(&sk->sk_wmem_alloc));
 	BUG_TRAP(!nlk_sk(sk)->cb);
-
-	kfree(nlk_sk(sk));
-
-	atomic_dec(&netlink_sock_nr);
-#ifdef NETLINK_REFCNT_DEBUG
-	printk(KERN_DEBUG "NETLINK %p released, %d are still alive\n", sk, atomic_read(&netlink_sock_nr));
-#endif
+	BUG_TRAP(!nlk_sk(sk)->groups);
 }
 
 /* This lock without WQ_FLAG_EXCLUSIVE is good on UP and it is _very_ bad on SMP.
@@ -119,7 +159,7 @@ static void netlink_sock_destruct(struct sock *sk)
 
 static void netlink_table_grab(void)
 {
-	write_lock_bh(&nl_table_lock);
+	write_lock_irq(&nl_table_lock);
 
 	if (atomic_read(&nl_table_users)) {
 		DECLARE_WAITQUEUE(wait, current);
@@ -129,9 +169,9 @@ static void netlink_table_grab(void)
 			set_current_state(TASK_UNINTERRUPTIBLE);
 			if (atomic_read(&nl_table_users) == 0)
 				break;
-			write_unlock_bh(&nl_table_lock);
+			write_unlock_irq(&nl_table_lock);
 			schedule();
-			write_lock_bh(&nl_table_lock);
+			write_lock_irq(&nl_table_lock);
 		}
 
 		__set_current_state(TASK_RUNNING);
@@ -141,7 +181,7 @@ static void netlink_table_grab(void)
 
 static __inline__ void netlink_table_ungrab(void)
 {
-	write_unlock_bh(&nl_table_lock);
+	write_unlock_irq(&nl_table_lock);
 	wake_up(&nl_table_wait);
 }
 
@@ -164,11 +204,14 @@ netlink_unlock_table(void)
 
 static __inline__ struct sock *netlink_lookup(int protocol, u32 pid)
 {
+	struct nl_pid_hash *hash = &nl_table[protocol].hash;
+	struct hlist_head *head;
 	struct sock *sk;
 	struct hlist_node *node;
 
 	read_lock(&nl_table_lock);
-	sk_for_each(sk, node, &nl_table[protocol]) {
+	head = nl_pid_hashfn(hash, pid);
+	sk_for_each(sk, node, head) {
 		if (nlk_sk(sk)->pid == pid) {
 			sock_hold(sk);
 			goto found;
@@ -180,27 +223,136 @@ found:
 	return sk;
 }
 
-static struct proto_ops netlink_ops;
+static inline struct hlist_head *nl_pid_hash_alloc(size_t size)
+{
+	if (size <= PAGE_SIZE)
+		return kmalloc(size, GFP_ATOMIC);
+	else
+		return (struct hlist_head *)
+			__get_free_pages(GFP_ATOMIC, get_order(size));
+}
+
+static inline void nl_pid_hash_free(struct hlist_head *table, size_t size)
+{
+	if (size <= PAGE_SIZE)
+		kfree(table);
+	else
+		free_pages((unsigned long)table, get_order(size));
+}
+
+static int nl_pid_hash_rehash(struct nl_pid_hash *hash, int grow)
+{
+	unsigned int omask, mask, shift;
+	size_t osize, size;
+	struct hlist_head *otable, *table;
+	int i;
+
+	omask = mask = hash->mask;
+	osize = size = (mask + 1) * sizeof(*table);
+	shift = hash->shift;
+
+	if (grow) {
+		if (++shift > hash->max_shift)
+			return 0;
+		mask = mask * 2 + 1;
+		size *= 2;
+	}
+
+	table = nl_pid_hash_alloc(size);
+	if (!table)
+		return 0;
+
+	memset(table, 0, size);
+	otable = hash->table;
+	hash->table = table;
+	hash->mask = mask;
+	hash->shift = shift;
+	get_random_bytes(&hash->rnd, sizeof(hash->rnd));
+
+	for (i = 0; i <= omask; i++) {
+		struct sock *sk;
+		struct hlist_node *node, *tmp;
+
+		sk_for_each_safe(sk, node, tmp, &otable[i])
+			__sk_add_node(sk, nl_pid_hashfn(hash, nlk_sk(sk)->pid));
+	}
+
+	nl_pid_hash_free(otable, osize);
+	hash->rehash_time = jiffies + 10 * 60 * HZ;
+	return 1;
+}
+
+static inline int nl_pid_hash_dilute(struct nl_pid_hash *hash, int len)
+{
+	int avg = hash->entries >> hash->shift;
+
+	if (unlikely(avg > 1) && nl_pid_hash_rehash(hash, 1))
+		return 1;
+
+	if (unlikely(len > avg) && time_after(jiffies, hash->rehash_time)) {
+		nl_pid_hash_rehash(hash, 0);
+		return 1;
+	}
+
+	return 0;
+}
+
+static const struct proto_ops netlink_ops;
+
+static void
+netlink_update_listeners(struct sock *sk)
+{
+	struct netlink_table *tbl = &nl_table[sk->sk_protocol];
+	struct hlist_node *node;
+	unsigned long mask;
+	unsigned int i;
+
+	for (i = 0; i < NLGRPSZ(tbl->groups)/sizeof(unsigned long); i++) {
+		mask = 0;
+		sk_for_each_bound(sk, node, &tbl->mc_list)
+			mask |= nlk_sk(sk)->groups[i];
+		tbl->listeners[i] = mask;
+	}
+	/* this function is only called with the netlink table "grabbed", which
+	 * makes sure updates are visible before bind or setsockopt return. */
+}
 
 static int netlink_insert(struct sock *sk, u32 pid)
 {
+	struct nl_pid_hash *hash = &nl_table[sk->sk_protocol].hash;
+	struct hlist_head *head;
 	int err = -EADDRINUSE;
 	struct sock *osk;
 	struct hlist_node *node;
+	int len;
 
 	netlink_table_grab();
-	sk_for_each(osk, node, &nl_table[sk->sk_protocol]) {
+	head = nl_pid_hashfn(hash, pid);
+	len = 0;
+	sk_for_each(osk, node, head) {
 		if (nlk_sk(osk)->pid == pid)
 			break;
+		len++;
 	}
-	if (!node) {
-		err = -EBUSY;
-		if (nlk_sk(sk)->pid == 0) {
-			nlk_sk(sk)->pid = pid;
-			sk_add_node(sk, &nl_table[sk->sk_protocol]);
-			err = 0;
-		}
-	}
+	if (node)
+		goto err;
+
+	err = -EBUSY;
+	if (nlk_sk(sk)->pid)
+		goto err;
+
+	err = -ENOMEM;
+	if (BITS_PER_LONG > 32 && unlikely(hash->entries >= UINT_MAX))
+		goto err;
+
+	if (len && nl_pid_hash_dilute(hash, len))
+		head = nl_pid_hashfn(hash, pid);
+	hash->entries++;
+	nlk_sk(sk)->pid = pid;
+	sk_add_node(sk, head);
+	err = 0;
+
+err:
 	netlink_table_ungrab();
 	return err;
 }
@@ -208,14 +360,47 @@ static int netlink_insert(struct sock *sk, u32 pid)
 static void netlink_remove(struct sock *sk)
 {
 	netlink_table_grab();
-	sk_del_node_init(sk);
+	if (sk_del_node_init(sk))
+		nl_table[sk->sk_protocol].hash.entries--;
+	if (nlk_sk(sk)->subscriptions)
+		__sk_del_bind_node(sk);
 	netlink_table_ungrab();
 }
 
-static int netlink_create(struct socket *sock, int protocol)
+static struct proto netlink_proto = {
+	.name	  = "NETLINK",
+	.owner	  = THIS_MODULE,
+	.obj_size = sizeof(struct netlink_sock),
+};
+
+static int __netlink_create(struct socket *sock, int protocol)
 {
 	struct sock *sk;
-	struct netlink_opt *nlk;
+	struct netlink_sock *nlk;
+
+	sock->ops = &netlink_ops;
+
+	sk = sk_alloc(PF_NETLINK, GFP_KERNEL, &netlink_proto, 1);
+	if (!sk)
+		return -ENOMEM;
+
+	sock_init_data(sock, sk);
+
+	nlk = nlk_sk(sk);
+	spin_lock_init(&nlk->cb_lock);
+	init_waitqueue_head(&nlk->wait);
+
+	sk->sk_destruct = netlink_sock_destruct;
+	sk->sk_protocol = protocol;
+	return 0;
+}
+
+static int netlink_create(struct socket *sock, int protocol)
+{
+	struct module *module = NULL;
+	struct netlink_sock *nlk;
+	unsigned int groups;
+	int err = 0;
 
 	sock->state = SS_UNCONNECTED;
 
@@ -225,41 +410,37 @@ static int netlink_create(struct socket *sock, int protocol)
 	if (protocol<0 || protocol >= MAX_LINKS)
 		return -EPROTONOSUPPORT;
 
-	sock->ops = &netlink_ops;
-
-	sk = sk_alloc(PF_NETLINK, GFP_KERNEL, 1, NULL);
-	if (!sk)
-		return -ENOMEM;
-
-	sock_init_data(sock,sk);
-	sk_set_owner(sk, THIS_MODULE);
-
-	nlk = sk->sk_protinfo = kmalloc(sizeof(*nlk), GFP_KERNEL);
-	if (!nlk) {
-		sk_free(sk);
-		return -ENOMEM;
+	netlink_lock_table();
+#ifdef CONFIG_KMOD
+	if (!nl_table[protocol].registered) {
+		netlink_unlock_table();
+		request_module("net-pf-%d-proto-%d", PF_NETLINK, protocol);
+		netlink_lock_table();
 	}
-	memset(nlk, 0, sizeof(*nlk));
+#endif
+	if (nl_table[protocol].registered &&
+	    try_module_get(nl_table[protocol].module))
+		module = nl_table[protocol].module;
+	groups = nl_table[protocol].groups;
+	netlink_unlock_table();
 
-	spin_lock_init(&nlk->cb_lock);
-	init_waitqueue_head(&nlk->wait);
-	sk->sk_destruct = netlink_sock_destruct;
-	atomic_inc(&netlink_sock_nr);
+	if ((err = __netlink_create(sock, protocol)) < 0)
+		goto out_module;
 
-	set_vx_info(&sk->sk_vx_info, current->vx_info);
-	sk->sk_xid = vx_current_xid();
-	vx_sock_inc(sk);
-	set_nx_info(&sk->sk_nx_info, current->nx_info);
-	sk->sk_nid = nx_current_nid();
+	nlk = nlk_sk(sock->sk);
+	nlk->module = module;
+out:
+	return err;
 
-	sk->sk_protocol = protocol;
-	return 0;
+out_module:
+	module_put(module);
+	goto out;
 }
 
 static int netlink_release(struct socket *sock)
 {
 	struct sock *sk = sock->sk;
-	struct netlink_opt *nlk;
+	struct netlink_sock *nlk;
 
 	if (!sk)
 		return 0;
@@ -269,10 +450,10 @@ static int netlink_release(struct socket *sock)
 
 	spin_lock(&nlk->cb_lock);
 	if (nlk->cb) {
-		nlk->cb->done(nlk->cb);
+		if (nlk->cb->done)
+			nlk->cb->done(nlk->cb);
 		netlink_destroy_callback(nlk->cb);
 		nlk->cb = NULL;
-		__sock_put(sk);
 	}
 	spin_unlock(&nlk->cb_lock);
 
@@ -285,19 +466,28 @@ static int netlink_release(struct socket *sock)
 
 	skb_queue_purge(&sk->sk_write_queue);
 
-	if (nlk->pid && !nlk->groups) {
+	if (nlk->pid && !nlk->subscriptions) {
 		struct netlink_notify n = {
 						.protocol = sk->sk_protocol,
 						.pid = nlk->pid,
 					  };
-		notifier_call_chain(&netlink_chain, NETLINK_URELEASE, &n);
+		atomic_notifier_call_chain(&netlink_chain,
+				NETLINK_URELEASE, &n);
 	}	
-	
-	vx_sock_dec(sk);
-	clr_vx_info(&sk->sk_vx_info);
-	sk->sk_xid = -1;
-	clr_nx_info(&sk->sk_nx_info);
-	sk->sk_nid = -1;
+
+	module_put(nlk->module);
+
+	netlink_table_grab();
+	if (nlk->flags & NETLINK_KERNEL_SOCKET) {
+		kfree(nl_table[sk->sk_protocol].listeners);
+		nl_table[sk->sk_protocol].module = NULL;
+		nl_table[sk->sk_protocol].registered = 0;
+	} else if (nlk->subscriptions)
+		netlink_update_listeners(sk);
+	netlink_table_ungrab();
+
+	kfree(nlk->groups);
+	nlk->groups = NULL;
 
 	sock_put(sk);
 	return 0;
@@ -306,19 +496,24 @@ static int netlink_release(struct socket *sock)
 static int netlink_autobind(struct socket *sock)
 {
 	struct sock *sk = sock->sk;
+	struct nl_pid_hash *hash = &nl_table[sk->sk_protocol].hash;
+	struct hlist_head *head;
 	struct sock *osk;
 	struct hlist_node *node;
-	s32 pid = current->pid;
+	s32 pid = current->tgid;
 	int err;
+	static s32 rover = -4097;
 
 retry:
+	cond_resched();
 	netlink_table_grab();
-	sk_for_each(osk, node, &nl_table[sk->sk_protocol]) {
+	head = nl_pid_hashfn(hash, pid);
+	sk_for_each(osk, node, head) {
 		if (nlk_sk(osk)->pid == pid) {
 			/* Bind collision, search negative pid values. */
-			if (pid > 0)
-				pid = -4096;
-			pid--;
+			pid = rover--;
+			if (rover > -4097)
+				rover = -4097;
 			netlink_table_ungrab();
 			goto retry;
 		}
@@ -328,20 +523,58 @@ retry:
 	err = netlink_insert(sk, pid);
 	if (err == -EADDRINUSE)
 		goto retry;
-	nlk_sk(sk)->groups = 0;
-	return 0;
+
+	/* If 2 threads race to autobind, that is fine.  */
+	if (err == -EBUSY)
+		err = 0;
+
+	return err;
 }
 
-static inline int netlink_capable(struct socket *sock, unsigned flag) 
+static inline int netlink_capable(struct socket *sock, unsigned int flag) 
 { 
-	return (nl_nonroot[sock->sk->sk_protocol] & flag) ||
+	return (nl_table[sock->sk->sk_protocol].nl_nonroot & flag) ||
 	       capable(CAP_NET_ADMIN);
 } 
 
+static void
+netlink_update_subscriptions(struct sock *sk, unsigned int subscriptions)
+{
+	struct netlink_sock *nlk = nlk_sk(sk);
+
+	if (nlk->subscriptions && !subscriptions)
+		__sk_del_bind_node(sk);
+	else if (!nlk->subscriptions && subscriptions)
+		sk_add_bind_node(sk, &nl_table[sk->sk_protocol].mc_list);
+	nlk->subscriptions = subscriptions;
+}
+
+static int netlink_alloc_groups(struct sock *sk)
+{
+	struct netlink_sock *nlk = nlk_sk(sk);
+	unsigned int groups;
+	int err = 0;
+
+	netlink_lock_table();
+	groups = nl_table[sk->sk_protocol].groups;
+	if (!nl_table[sk->sk_protocol].registered)
+		err = -ENOENT;
+	netlink_unlock_table();
+
+	if (err)
+		return err;
+
+	nlk->groups = kzalloc(NLGRPSZ(groups), GFP_KERNEL);
+	if (nlk->groups == NULL)
+		return -ENOMEM;
+	nlk->ngroups = groups;
+	return 0;
+}
+
 static int netlink_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
 {
 	struct sock *sk = sock->sk;
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 	struct sockaddr_nl *nladdr = (struct sockaddr_nl *)addr;
 	int err;
 	
@@ -349,27 +582,39 @@ static int netlink_bind(struct socket *sock, struct sockaddr *addr, int addr_len
 		return -EINVAL;
 
 	/* Only superuser is allowed to listen multicasts */
-	if (nladdr->nl_groups && !netlink_capable(sock, NL_NONROOT_RECV))
-		return -EPERM;
+	if (nladdr->nl_groups) {
+		if (!netlink_capable(sock, NL_NONROOT_RECV))
+			return -EPERM;
+		if (nlk->groups == NULL) {
+			err = netlink_alloc_groups(sk);
+			if (err)
+				return err;
+		}
+	}
 
 	if (nlk->pid) {
 		if (nladdr->nl_pid != nlk->pid)
 			return -EINVAL;
-		nlk->groups = nladdr->nl_groups;
-		return 0;
+	} else {
+		err = nladdr->nl_pid ?
+			netlink_insert(sk, nladdr->nl_pid) :
+			netlink_autobind(sock);
+		if (err)
+			return err;
 	}
 
-	if (nladdr->nl_pid == 0) {
-		err = netlink_autobind(sock);
-		if (err == 0)
-			nlk->groups = nladdr->nl_groups;
-		return err;
-	}
+	if (!nladdr->nl_groups && (nlk->groups == NULL || !(u32)nlk->groups[0]))
+		return 0;
 
-	err = netlink_insert(sk, nladdr->nl_pid);
-	if (err == 0)
-		nlk->groups = nladdr->nl_groups;
-	return err;
+	netlink_table_grab();
+	netlink_update_subscriptions(sk, nlk->subscriptions +
+	                                 hweight32(nladdr->nl_groups) -
+	                                 hweight32(nlk->groups[0]));
+	nlk->groups[0] = (nlk->groups[0] & ~0xffffffffUL) | nladdr->nl_groups; 
+	netlink_update_listeners(sk);
+	netlink_table_ungrab();
+
+	return 0;
 }
 
 static int netlink_connect(struct socket *sock, struct sockaddr *addr,
@@ -377,13 +622,13 @@ static int netlink_connect(struct socket *sock, struct sockaddr *addr,
 {
 	int err = 0;
 	struct sock *sk = sock->sk;
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 	struct sockaddr_nl *nladdr=(struct sockaddr_nl*)addr;
 
 	if (addr->sa_family == AF_UNSPEC) {
 		sk->sk_state	= NETLINK_UNCONNECTED;
 		nlk->dst_pid	= 0;
-		nlk->dst_groups = 0;
+		nlk->dst_group  = 0;
 		return 0;
 	}
 	if (addr->sa_family != AF_NETLINK)
@@ -399,7 +644,7 @@ static int netlink_connect(struct socket *sock, struct sockaddr *addr,
 	if (err == 0) {
 		sk->sk_state	= NETLINK_CONNECTED;
 		nlk->dst_pid 	= nladdr->nl_pid;
-		nlk->dst_groups = nladdr->nl_groups;
+		nlk->dst_group  = ffs(nladdr->nl_groups);
 	}
 
 	return err;
@@ -408,7 +653,7 @@ static int netlink_connect(struct socket *sock, struct sockaddr *addr,
 static int netlink_getname(struct socket *sock, struct sockaddr *addr, int *addr_len, int peer)
 {
 	struct sock *sk = sock->sk;
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 	struct sockaddr_nl *nladdr=(struct sockaddr_nl *)addr;
 	
 	nladdr->nl_family = AF_NETLINK;
@@ -417,10 +662,10 @@ static int netlink_getname(struct socket *sock, struct sockaddr *addr, int *addr
 
 	if (peer) {
 		nladdr->nl_pid = nlk->dst_pid;
-		nladdr->nl_groups = nlk->dst_groups;
+		nladdr->nl_groups = netlink_group_mask(nlk->dst_group);
 	} else {
 		nladdr->nl_pid = nlk->pid;
-		nladdr->nl_groups = nlk->groups;
+		nladdr->nl_groups = nlk->groups ? nlk->groups[0] : 0;
 	}
 	return 0;
 }
@@ -433,11 +678,11 @@ static void netlink_overrun(struct sock *sk)
 	}
 }
 
-struct sock *netlink_getsockbypid(struct sock *ssk, u32 pid)
+static struct sock *netlink_getsockbypid(struct sock *ssk, u32 pid)
 {
 	int protocol = ssk->sk_protocol;
 	struct sock *sock;
-	struct netlink_opt *nlk;
+	struct netlink_sock *nlk;
 
 	sock = netlink_lookup(protocol, pid);
 	if (!sock)
@@ -456,14 +701,13 @@ struct sock *netlink_getsockbypid(struct sock *ssk, u32 pid)
 
 struct sock *netlink_getsockbyfilp(struct file *filp)
 {
-	struct inode *inode = filp->f_dentry->d_inode;
-	struct socket *socket;
+	struct inode *inode = filp->f_path.dentry->d_inode;
 	struct sock *sock;
 
-	if (!inode->i_sock || !(socket = SOCKET_I(inode)))
+	if (!S_ISSOCK(inode->i_mode))
 		return ERR_PTR(-ENOTSOCK);
 
-	sock = socket->sk;
+	sock = SOCKET_I(inode)->sk;
 	if (sock->sk_family != AF_NETLINK)
 		return ERR_PTR(-EINVAL);
 
@@ -481,21 +725,18 @@ struct sock *netlink_getsockbyfilp(struct file *filp)
  * 0: continue
  * 1: repeat lookup - reference dropped while waiting for socket memory.
  */
-int netlink_attachskb(struct sock *sk, struct sk_buff *skb, int nonblock, long timeo)
+int netlink_attachskb(struct sock *sk, struct sk_buff *skb, int nonblock,
+		long timeo, struct sock *ssk)
 {
-	struct netlink_opt *nlk;
+	struct netlink_sock *nlk;
 
 	nlk = nlk_sk(sk);
 
-#ifdef NL_EMULATE_DEV
-	if (nlk->handler)
-		return 0;
-#endif
 	if (atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf ||
 	    test_bit(0, &nlk->state)) {
 		DECLARE_WAITQUEUE(wait, current);
 		if (!timeo) {
-			if (!nlk->pid)
+			if (!ssk || nlk_sk(ssk)->pid == 0)
 				netlink_overrun(sk);
 			sock_put(sk);
 			kfree_skb(skb);
@@ -520,26 +761,14 @@ int netlink_attachskb(struct sock *sk, struct sk_buff *skb, int nonblock, long t
 		}
 		return 1;
 	}
-	skb_orphan(skb);
 	skb_set_owner_r(skb, sk);
 	return 0;
 }
 
 int netlink_sendskb(struct sock *sk, struct sk_buff *skb, int protocol)
 {
-	struct netlink_opt *nlk;
 	int len = skb->len;
 
-	nlk = nlk_sk(sk);
-#ifdef NL_EMULATE_DEV
-	if (nlk->handler) {
-		skb_orphan(skb);
-		len = nlk->handler(protocol, skb);
-		sock_put(sk);
-		return len;
-	}
-#endif
-
 	skb_queue_tail(&sk->sk_receive_queue, skb);
 	sk->sk_data_ready(sk, len);
 	sock_put(sk);
@@ -552,21 +781,29 @@ void netlink_detachskb(struct sock *sk, struct sk_buff *skb)
 	sock_put(sk);
 }
 
-static inline void netlink_trim(struct sk_buff *skb, int allocation)
+static inline struct sk_buff *netlink_trim(struct sk_buff *skb,
+					   gfp_t allocation)
 {
-	int delta = skb->end - skb->tail;
+	int delta;
 
-	/* If the packet is charged to a socket, the modification
-	 * of truesize below is illegal and will corrupt socket
-	 * buffer accounting state.
-	 */
-	BUG_ON(skb->list != NULL);
+	skb_orphan(skb);
 
+	delta = skb->end - skb->tail;
 	if (delta * 2 < skb->truesize)
-		return;
-	if (pskb_expand_head(skb, 0, -delta, allocation))
-		return;
-	skb->truesize -= delta;
+		return skb;
+
+	if (skb_shared(skb)) {
+		struct sk_buff *nskb = skb_clone(skb, allocation);
+		if (!nskb)
+			return skb;
+		kfree_skb(skb);
+		skb = nskb;
+	}
+
+	if (!pskb_expand_head(skb, 0, -delta, allocation))
+		skb->truesize -= delta;
+
+	return skb;
 }
 
 int netlink_unicast(struct sock *ssk, struct sk_buff *skb, u32 pid, int nonblock)
@@ -575,7 +812,7 @@ int netlink_unicast(struct sock *ssk, struct sk_buff *skb, u32 pid, int nonblock
 	int err;
 	long timeo;
 
-	netlink_trim(skb, gfp_any());
+	skb = netlink_trim(skb, gfp_any());
 
 	timeo = sock_sndtimeo(ssk, nonblock);
 retry:
@@ -584,7 +821,7 @@ retry:
 		kfree_skb(skb);
 		return PTR_ERR(sk);
 	}
-	err = netlink_attachskb(sk, skb, nonblock, timeo);
+	err = netlink_attachskb(sk, skb, nonblock, timeo, ssk);
 	if (err == 1)
 		goto retry;
 	if (err)
@@ -593,117 +830,277 @@ retry:
 	return netlink_sendskb(sk, skb, ssk->sk_protocol);
 }
 
+int netlink_has_listeners(struct sock *sk, unsigned int group)
+{
+	int res = 0;
+
+	BUG_ON(!(nlk_sk(sk)->flags & NETLINK_KERNEL_SOCKET));
+	if (group - 1 < nl_table[sk->sk_protocol].groups)
+		res = test_bit(group - 1, nl_table[sk->sk_protocol].listeners);
+	return res;
+}
+EXPORT_SYMBOL_GPL(netlink_has_listeners);
+
 static __inline__ int netlink_broadcast_deliver(struct sock *sk, struct sk_buff *skb)
 {
-	struct netlink_opt *nlk = nlk_sk(sk);
-#ifdef NL_EMULATE_DEV
-	if (nlk->handler) {
-		skb_orphan(skb);
-		nlk->handler(sk->sk_protocol, skb);
-		return 0;
-	} else
-#endif
+	struct netlink_sock *nlk = nlk_sk(sk);
+
 	if (atomic_read(&sk->sk_rmem_alloc) <= sk->sk_rcvbuf &&
 	    !test_bit(0, &nlk->state)) {
-                skb_orphan(skb);
 		skb_set_owner_r(skb, sk);
 		skb_queue_tail(&sk->sk_receive_queue, skb);
 		sk->sk_data_ready(sk, skb->len);
-		return 0;
+		return atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf;
 	}
 	return -1;
 }
 
+struct netlink_broadcast_data {
+	struct sock *exclude_sk;
+	u32 pid;
+	u32 group;
+	int failure;
+	int congested;
+	int delivered;
+	gfp_t allocation;
+	struct sk_buff *skb, *skb2;
+};
+
+static inline int do_one_broadcast(struct sock *sk,
+				   struct netlink_broadcast_data *p)
+{
+	struct netlink_sock *nlk = nlk_sk(sk);
+	int val;
+
+	if (p->exclude_sk == sk)
+		goto out;
+
+	if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups ||
+	    !test_bit(p->group - 1, nlk->groups))
+		goto out;
+
+	if (p->failure) {
+		netlink_overrun(sk);
+		goto out;
+	}
+
+	sock_hold(sk);
+	if (p->skb2 == NULL) {
+		if (skb_shared(p->skb)) {
+			p->skb2 = skb_clone(p->skb, p->allocation);
+		} else {
+			p->skb2 = skb_get(p->skb);
+			/*
+			 * skb ownership may have been set when
+			 * delivered to a previous socket.
+			 */
+			skb_orphan(p->skb2);
+		}
+	}
+	if (p->skb2 == NULL) {
+		netlink_overrun(sk);
+		/* Clone failed. Notify ALL listeners. */
+		p->failure = 1;
+	} else if ((val = netlink_broadcast_deliver(sk, p->skb2)) < 0) {
+		netlink_overrun(sk);
+	} else {
+		p->congested |= val;
+		p->delivered = 1;
+		p->skb2 = NULL;
+	}
+	sock_put(sk);
+
+out:
+	return 0;
+}
+
 int netlink_broadcast(struct sock *ssk, struct sk_buff *skb, u32 pid,
-		      u32 group, int allocation)
+		      u32 group, gfp_t allocation)
 {
-	struct sock *sk;
+	struct netlink_broadcast_data info;
 	struct hlist_node *node;
-	struct sk_buff *skb2 = NULL;
-	int protocol = ssk->sk_protocol;
-	int failure = 0, delivered = 0;
+	struct sock *sk;
+
+	skb = netlink_trim(skb, allocation);
 
-	netlink_trim(skb, allocation);
+	info.exclude_sk = ssk;
+	info.pid = pid;
+	info.group = group;
+	info.failure = 0;
+	info.congested = 0;
+	info.delivered = 0;
+	info.allocation = allocation;
+	info.skb = skb;
+	info.skb2 = NULL;
 
 	/* While we sleep in clone, do not allow to change socket list */
 
 	netlink_lock_table();
 
-	sk_for_each(sk, node, &nl_table[protocol]) {
-		struct netlink_opt *nlk = nlk_sk(sk);
-
-		if (ssk == sk)
-			continue;
-
-		if (nlk->pid == pid || !(nlk->groups & group))
-			continue;
-
-		if (failure) {
-			netlink_overrun(sk);
-			continue;
-		}
+	sk_for_each_bound(sk, node, &nl_table[ssk->sk_protocol].mc_list)
+		do_one_broadcast(sk, &info);
 
-		sock_hold(sk);
-		if (skb2 == NULL) {
-			if (atomic_read(&skb->users) != 1) {
-				skb2 = skb_clone(skb, allocation);
-			} else {
-				skb2 = skb;
-				atomic_inc(&skb->users);
-			}
-		}
-		if (skb2 == NULL) {
-			netlink_overrun(sk);
-			/* Clone failed. Notify ALL listeners. */
-			failure = 1;
-		} else if (netlink_broadcast_deliver(sk, skb2)) {
-			netlink_overrun(sk);
-		} else {
-			delivered = 1;
-			skb2 = NULL;
-		}
-		sock_put(sk);
-	}
+	kfree_skb(skb);
 
 	netlink_unlock_table();
 
-	if (skb2)
-		kfree_skb(skb2);
-	kfree_skb(skb);
+	if (info.skb2)
+		kfree_skb(info.skb2);
 
-	if (delivered)
+	if (info.delivered) {
+		if (info.congested && (allocation & __GFP_WAIT))
+			yield();
 		return 0;
-	if (failure)
+	}
+	if (info.failure)
 		return -ENOBUFS;
 	return -ESRCH;
 }
 
+struct netlink_set_err_data {
+	struct sock *exclude_sk;
+	u32 pid;
+	u32 group;
+	int code;
+};
+
+static inline int do_one_set_err(struct sock *sk,
+				 struct netlink_set_err_data *p)
+{
+	struct netlink_sock *nlk = nlk_sk(sk);
+
+	if (sk == p->exclude_sk)
+		goto out;
+
+	if (nlk->pid == p->pid || p->group - 1 >= nlk->ngroups ||
+	    !test_bit(p->group - 1, nlk->groups))
+		goto out;
+
+	sk->sk_err = p->code;
+	sk->sk_error_report(sk);
+out:
+	return 0;
+}
+
 void netlink_set_err(struct sock *ssk, u32 pid, u32 group, int code)
 {
-	struct sock *sk;
+	struct netlink_set_err_data info;
 	struct hlist_node *node;
-	int protocol = ssk->sk_protocol;
+	struct sock *sk;
+
+	info.exclude_sk = ssk;
+	info.pid = pid;
+	info.group = group;
+	info.code = code;
 
 	read_lock(&nl_table_lock);
-	sk_for_each(sk, node, &nl_table[protocol]) {
-		struct netlink_opt *nlk = nlk_sk(sk);
-		if (ssk == sk)
-			continue;
 
-		if (nlk->pid == pid || !(nlk->groups & group))
-			continue;
+	sk_for_each_bound(sk, node, &nl_table[ssk->sk_protocol].mc_list)
+		do_one_set_err(sk, &info);
 
-		sk->sk_err = code;
-		sk->sk_error_report(sk);
-	}
 	read_unlock(&nl_table_lock);
 }
 
+static int netlink_setsockopt(struct socket *sock, int level, int optname,
+                              char __user *optval, int optlen)
+{
+	struct sock *sk = sock->sk;
+	struct netlink_sock *nlk = nlk_sk(sk);
+	int val = 0, err;
+
+	if (level != SOL_NETLINK)
+		return -ENOPROTOOPT;
+
+	if (optlen >= sizeof(int) &&
+	    get_user(val, (int __user *)optval))
+		return -EFAULT;
+
+	switch (optname) {
+	case NETLINK_PKTINFO:
+		if (val)
+			nlk->flags |= NETLINK_RECV_PKTINFO;
+		else
+			nlk->flags &= ~NETLINK_RECV_PKTINFO;
+		err = 0;
+		break;
+	case NETLINK_ADD_MEMBERSHIP:
+	case NETLINK_DROP_MEMBERSHIP: {
+		unsigned int subscriptions;
+		int old, new = optname == NETLINK_ADD_MEMBERSHIP ? 1 : 0;
+
+		if (!netlink_capable(sock, NL_NONROOT_RECV))
+			return -EPERM;
+		if (nlk->groups == NULL) {
+			err = netlink_alloc_groups(sk);
+			if (err)
+				return err;
+		}
+		if (!val || val - 1 >= nlk->ngroups)
+			return -EINVAL;
+		netlink_table_grab();
+		old = test_bit(val - 1, nlk->groups);
+		subscriptions = nlk->subscriptions - old + new;
+		if (new)
+			__set_bit(val - 1, nlk->groups);
+		else
+			__clear_bit(val - 1, nlk->groups);
+		netlink_update_subscriptions(sk, subscriptions);
+		netlink_update_listeners(sk);
+		netlink_table_ungrab();
+		err = 0;
+		break;
+	}
+	default:
+		err = -ENOPROTOOPT;
+	}
+	return err;
+}
+
+static int netlink_getsockopt(struct socket *sock, int level, int optname,
+                              char __user *optval, int __user *optlen)
+{
+	struct sock *sk = sock->sk;
+	struct netlink_sock *nlk = nlk_sk(sk);
+	int len, val, err;
+
+	if (level != SOL_NETLINK)
+		return -ENOPROTOOPT;
+
+	if (get_user(len, optlen))
+		return -EFAULT;
+	if (len < 0)
+		return -EINVAL;
+
+	switch (optname) {
+	case NETLINK_PKTINFO:
+		if (len < sizeof(int))
+			return -EINVAL;
+		len = sizeof(int);
+		val = nlk->flags & NETLINK_RECV_PKTINFO ? 1 : 0;
+		if (put_user(len, optlen) ||
+		    put_user(val, optval))
+			return -EFAULT;
+		err = 0;
+		break;
+	default:
+		err = -ENOPROTOOPT;
+	}
+	return err;
+}
+
+static void netlink_cmsg_recv_pktinfo(struct msghdr *msg, struct sk_buff *skb)
+{
+	struct nl_pktinfo info;
+
+	info.group = NETLINK_CB(skb).dst_group;
+	put_cmsg(msg, SOL_NETLINK, NETLINK_PKTINFO, sizeof(info), &info);
+}
+
 static inline void netlink_rcv_wake(struct sock *sk)
 {
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 
-	if (!skb_queue_len(&sk->sk_receive_queue))
+	if (skb_queue_empty(&sk->sk_receive_queue))
 		clear_bit(0, &nlk->state);
 	if (!test_bit(0, &nlk->state))
 		wake_up_interruptible(&nlk->wait);
@@ -714,10 +1111,10 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock,
 {
 	struct sock_iocb *siocb = kiocb_to_siocb(kiocb);
 	struct sock *sk = sock->sk;
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 	struct sockaddr_nl *addr=msg->msg_name;
 	u32 dst_pid;
-	u32 dst_groups;
+	u32 dst_group;
 	struct sk_buff *skb;
 	int err;
 	struct scm_cookie scm;
@@ -735,12 +1132,12 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock,
 		if (addr->nl_family != AF_NETLINK)
 			return -EINVAL;
 		dst_pid = addr->nl_pid;
-		dst_groups = addr->nl_groups;
-		if (dst_groups && !netlink_capable(sock, NL_NONROOT_SEND))
+		dst_group = ffs(addr->nl_groups);
+		if (dst_group && !netlink_capable(sock, NL_NONROOT_SEND))
 			return -EPERM;
 	} else {
 		dst_pid = nlk->dst_pid;
-		dst_groups = nlk->dst_groups;
+		dst_group = nlk->dst_group;
 	}
 
 	if (!nlk->pid) {
@@ -758,9 +1155,9 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock,
 		goto out;
 
 	NETLINK_CB(skb).pid	= nlk->pid;
-	NETLINK_CB(skb).groups	= nlk->groups;
-	NETLINK_CB(skb).dst_pid = dst_pid;
-	NETLINK_CB(skb).dst_groups = dst_groups;
+	NETLINK_CB(skb).dst_group = dst_group;
+	NETLINK_CB(skb).loginuid = audit_get_loginuid(current->audit_context);
+	selinux_get_task_sid(current, &(NETLINK_CB(skb).sid));
 	memcpy(NETLINK_CREDS(skb), &siocb->scm->creds, sizeof(struct ucred));
 
 	/* What can I do? Netlink is asynchronous, so that
@@ -781,9 +1178,9 @@ static int netlink_sendmsg(struct kiocb *kiocb, struct socket *sock,
 		goto out;
 	}
 
-	if (dst_groups) {
+	if (dst_group) {
 		atomic_inc(&skb->users);
-		netlink_broadcast(sk, skb, dst_pid, dst_groups, GFP_KERNEL);
+		netlink_broadcast(sk, skb, dst_pid, dst_group, GFP_KERNEL);
 	}
 	err = netlink_unicast(sk, skb, dst_pid, msg->msg_flags&MSG_DONTWAIT);
 
@@ -798,7 +1195,7 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock,
 	struct sock_iocb *siocb = kiocb_to_siocb(kiocb);
 	struct scm_cookie scm;
 	struct sock *sk = sock->sk;
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 	int noblock = flags&MSG_DONTWAIT;
 	size_t copied;
 	struct sk_buff *skb;
@@ -829,10 +1226,13 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock,
 		addr->nl_family = AF_NETLINK;
 		addr->nl_pad    = 0;
 		addr->nl_pid	= NETLINK_CB(skb).pid;
-		addr->nl_groups	= NETLINK_CB(skb).dst_groups;
+		addr->nl_groups	= netlink_group_mask(NETLINK_CB(skb).dst_group);
 		msg->msg_namelen = sizeof(*addr);
 	}
 
+	if (nlk->flags & NETLINK_RECV_PKTINFO)
+		netlink_cmsg_recv_pktinfo(msg, skb);
+
 	if (NULL == siocb->scm) {
 		memset(&scm, 0, sizeof(scm));
 		siocb->scm = &scm;
@@ -852,7 +1252,7 @@ out:
 
 static void netlink_data_ready(struct sock *sk, int len)
 {
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 
 	if (nlk->data_ready)
 		nlk->data_ready(sk, len);
@@ -866,10 +1266,16 @@ static void netlink_data_ready(struct sock *sk, int len)
  */
 
 struct sock *
-netlink_kernel_create(int unit, void (*input)(struct sock *sk, int len))
+netlink_kernel_create(int unit, unsigned int groups,
+                      void (*input)(struct sock *sk, int len),
+                      struct module *module)
 {
 	struct socket *sock;
 	struct sock *sk;
+	struct netlink_sock *nlk;
+	unsigned long *listeners = NULL;
+
+	BUG_ON(!nl_table);
 
 	if (unit<0 || unit>=MAX_LINKS)
 		return NULL;
@@ -877,23 +1283,46 @@ netlink_kernel_create(int unit, void (*input)(struct sock *sk, int len))
 	if (sock_create_lite(PF_NETLINK, SOCK_DGRAM, unit, &sock))
 		return NULL;
 
-	if (netlink_create(sock, unit) < 0) {
-		sock_release(sock);
-		return NULL;
-	}
+	if (__netlink_create(sock, unit) < 0)
+		goto out_sock_release;
+
+	if (groups < 32)
+		groups = 32;
+
+	listeners = kzalloc(NLGRPSZ(groups), GFP_KERNEL);
+	if (!listeners)
+		goto out_sock_release;
+
 	sk = sock->sk;
 	sk->sk_data_ready = netlink_data_ready;
 	if (input)
 		nlk_sk(sk)->data_ready = input;
 
-	netlink_insert(sk, 0);
+	if (netlink_insert(sk, 0))
+		goto out_sock_release;
+
+	nlk = nlk_sk(sk);
+	nlk->flags |= NETLINK_KERNEL_SOCKET;
+
+	netlink_table_grab();
+	nl_table[unit].groups = groups;
+	nl_table[unit].listeners = listeners;
+	nl_table[unit].module = module;
+	nl_table[unit].registered = 1;
+	netlink_table_ungrab();
+
 	return sk;
+
+out_sock_release:
+	kfree(listeners);
+	sock_release(sock);
+	return NULL;
 }
 
-void netlink_set_nonroot(int protocol, unsigned flags)
+void netlink_set_nonroot(int protocol, unsigned int flags)
 { 
-	if ((unsigned)protocol < MAX_LINKS) 
-		nl_nonroot[protocol] = flags;
+	if ((unsigned int)protocol < MAX_LINKS) 
+		nl_table[protocol].nl_nonroot = flags;
 } 
 
 static void netlink_destroy_callback(struct netlink_callback *cb)
@@ -910,23 +1339,22 @@ static void netlink_destroy_callback(struct netlink_callback *cb)
 
 static int netlink_dump(struct sock *sk)
 {
-	struct netlink_opt *nlk = nlk_sk(sk);
+	struct netlink_sock *nlk = nlk_sk(sk);
 	struct netlink_callback *cb;
 	struct sk_buff *skb;
 	struct nlmsghdr *nlh;
-	int len;
+	int len, err = -ENOBUFS;
 	
 	skb = sock_rmalloc(sk, NLMSG_GOODSIZE, 0, GFP_KERNEL);
 	if (!skb)
-		return -ENOBUFS;
+		goto errout;
 
 	spin_lock(&nlk->cb_lock);
 
 	cb = nlk->cb;
 	if (cb == NULL) {
-		spin_unlock(&nlk->cb_lock);
-		kfree_skb(skb);
-		return -EINVAL;
+		err = -EINVAL;
+		goto errout_skb;
 	}
 
 	len = cb->dump(skb, cb);
@@ -938,19 +1366,28 @@ static int netlink_dump(struct sock *sk)
 		return 0;
 	}
 
-	nlh = __nlmsg_put(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, NLMSG_DONE, sizeof(int));
-	nlh->nlmsg_flags |= NLM_F_MULTI;
-	memcpy(NLMSG_DATA(nlh), &len, sizeof(len));
+	nlh = nlmsg_put_answer(skb, cb, NLMSG_DONE, sizeof(len), NLM_F_MULTI);
+	if (!nlh)
+		goto errout_skb;
+
+	memcpy(nlmsg_data(nlh), &len, sizeof(len));
+
 	skb_queue_tail(&sk->sk_receive_queue, skb);
 	sk->sk_data_ready(sk, skb->len);
 
-	cb->done(cb);
+	if (cb->done)
+		cb->done(cb);
 	nlk->cb = NULL;
 	spin_unlock(&nlk->cb_lock);
 
 	netlink_destroy_callback(cb);
-	sock_put(sk);
 	return 0;
+
+errout_skb:
+	spin_unlock(&nlk->cb_lock);
+	kfree_skb(skb);
+errout:
+	return err;
 }
 
 int netlink_dump_start(struct sock *ssk, struct sk_buff *skb,
@@ -960,13 +1397,12 @@ int netlink_dump_start(struct sock *ssk, struct sk_buff *skb,
 {
 	struct netlink_callback *cb;
 	struct sock *sk;
-	struct netlink_opt *nlk;
+	struct netlink_sock *nlk;
 
-	cb = kmalloc(sizeof(*cb), GFP_KERNEL);
+	cb = kzalloc(sizeof(*cb), GFP_KERNEL);
 	if (cb == NULL)
 		return -ENOBUFS;
 
-	memset(cb, 0, sizeof(*cb));
 	cb->dump = dump;
 	cb->done = done;
 	cb->nlh = nlh;
@@ -991,6 +1427,7 @@ int netlink_dump_start(struct sock *ssk, struct sk_buff *skb,
 	spin_unlock(&nlk->cb_lock);
 
 	netlink_dump(sk);
+	sock_put(sk);
 	return 0;
 }
 
@@ -999,14 +1436,13 @@ void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err)
 	struct sk_buff *skb;
 	struct nlmsghdr *rep;
 	struct nlmsgerr *errmsg;
-	int size;
+	size_t payload = sizeof(*errmsg);
 
-	if (err == 0)
-		size = NLMSG_SPACE(sizeof(struct nlmsgerr));
-	else
-		size = NLMSG_SPACE(4 + NLMSG_ALIGN(nlh->nlmsg_len));
+	/* error messages get the original request appened */
+	if (err)
+		payload += nlmsg_len(nlh);
 
-	skb = alloc_skb(size, GFP_KERNEL);
+	skb = nlmsg_new(payload, GFP_KERNEL);
 	if (!skb) {
 		struct sock *sk;
 
@@ -1021,84 +1457,157 @@ void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err)
 	}
 
 	rep = __nlmsg_put(skb, NETLINK_CB(in_skb).pid, nlh->nlmsg_seq,
-			  NLMSG_ERROR, sizeof(struct nlmsgerr));
-	errmsg = NLMSG_DATA(rep);
+			  NLMSG_ERROR, sizeof(struct nlmsgerr), 0);
+	errmsg = nlmsg_data(rep);
 	errmsg->error = err;
-	memcpy(&errmsg->msg, nlh, err ? nlh->nlmsg_len : sizeof(struct nlmsghdr));
+	memcpy(&errmsg->msg, nlh, err ? nlh->nlmsg_len : sizeof(*nlh));
 	netlink_unicast(in_skb->sk, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT);
 }
 
+static int netlink_rcv_skb(struct sk_buff *skb, int (*cb)(struct sk_buff *,
+						     struct nlmsghdr *, int *))
+{
+	struct nlmsghdr *nlh;
+	int err;
 
-#ifdef NL_EMULATE_DEV
+	while (skb->len >= nlmsg_total_size(0)) {
+		nlh = (struct nlmsghdr *) skb->data;
 
-static rwlock_t nl_emu_lock = RW_LOCK_UNLOCKED;
+		if (nlh->nlmsg_len < NLMSG_HDRLEN || skb->len < nlh->nlmsg_len)
+			return 0;
+
+		if (cb(skb, nlh, &err) < 0) {
+			/* Not an error, but we have to interrupt processing
+			 * here. Note: that in this case we do not pull
+			 * message from skb, it will be processed later.
+			 */
+			if (err == 0)
+				return -1;
+			netlink_ack(skb, nlh, err);
+		} else if (nlh->nlmsg_flags & NLM_F_ACK)
+			netlink_ack(skb, nlh, 0);
+
+		netlink_queue_skip(nlh, skb);
+	}
 
-/*
- *	Backward compatibility.
- */	
- 
-int netlink_attach(int unit, int (*function)(int, struct sk_buff *skb))
-{
-	struct sock *sk = netlink_kernel_create(unit, NULL);
-	if (sk == NULL)
-		return -ENOBUFS;
-	nlk_sk(sk)->handler = function;
-	write_lock_bh(&nl_emu_lock);
-	netlink_kernel[unit] = sk->sk_socket;
-	write_unlock_bh(&nl_emu_lock);
 	return 0;
 }
 
-void netlink_detach(int unit)
+/**
+ * nelink_run_queue - Process netlink receive queue.
+ * @sk: Netlink socket containing the queue
+ * @qlen: Place to store queue length upon entry
+ * @cb: Callback function invoked for each netlink message found
+ *
+ * Processes as much as there was in the queue upon entry and invokes
+ * a callback function for each netlink message found. The callback
+ * function may refuse a message by returning a negative error code
+ * but setting the error pointer to 0 in which case this function
+ * returns with a qlen != 0.
+ *
+ * qlen must be initialized to 0 before the initial entry, afterwards
+ * the function may be called repeatedly until qlen reaches 0.
+ */
+void netlink_run_queue(struct sock *sk, unsigned int *qlen,
+		       int (*cb)(struct sk_buff *, struct nlmsghdr *, int *))
 {
-	struct socket *sock;
+	struct sk_buff *skb;
 
-	write_lock_bh(&nl_emu_lock);
-	sock = netlink_kernel[unit];
-	netlink_kernel[unit] = NULL;
-	write_unlock_bh(&nl_emu_lock);
+	if (!*qlen || *qlen > skb_queue_len(&sk->sk_receive_queue))
+		*qlen = skb_queue_len(&sk->sk_receive_queue);
+
+	for (; *qlen; (*qlen)--) {
+		skb = skb_dequeue(&sk->sk_receive_queue);
+		if (netlink_rcv_skb(skb, cb)) {
+			if (skb->len)
+				skb_queue_head(&sk->sk_receive_queue, skb);
+			else {
+				kfree_skb(skb);
+				(*qlen)--;
+			}
+			break;
+		}
 
-	sock_release(sock);
+		kfree_skb(skb);
+	}
 }
 
-int netlink_post(int unit, struct sk_buff *skb)
+/**
+ * netlink_queue_skip - Skip netlink message while processing queue.
+ * @nlh: Netlink message to be skipped
+ * @skb: Socket buffer containing the netlink messages.
+ *
+ * Pulls the given netlink message off the socket buffer so the next
+ * call to netlink_queue_run() will not reconsider the message.
+ */
+void netlink_queue_skip(struct nlmsghdr *nlh, struct sk_buff *skb)
 {
-	struct socket *sock;
+	int msglen = NLMSG_ALIGN(nlh->nlmsg_len);
 
-	read_lock(&nl_emu_lock);
-	sock = netlink_kernel[unit];
-	if (sock) {
-		struct sock *sk = sock->sk;
-		memset(skb->cb, 0, sizeof(skb->cb));
-		sock_hold(sk);
-		read_unlock(&nl_emu_lock);
+	if (msglen > skb->len)
+		msglen = skb->len;
 
-		netlink_broadcast(sk, skb, 0, ~0, GFP_ATOMIC);
+	skb_pull(skb, msglen);
+}
 
-		sock_put(sk);
-		return 0;
+/**
+ * nlmsg_notify - send a notification netlink message
+ * @sk: netlink socket to use
+ * @skb: notification message
+ * @pid: destination netlink pid for reports or 0
+ * @group: destination multicast group or 0
+ * @report: 1 to report back, 0 to disable
+ * @flags: allocation flags
+ */
+int nlmsg_notify(struct sock *sk, struct sk_buff *skb, u32 pid,
+		 unsigned int group, int report, gfp_t flags)
+{
+	int err = 0;
+
+	if (group) {
+		int exclude_pid = 0;
+
+		if (report) {
+			atomic_inc(&skb->users);
+			exclude_pid = pid;
+		}
+
+		/* errors reported via destination sk->sk_err */
+		nlmsg_multicast(sk, skb, exclude_pid, group, flags);
 	}
-	read_unlock(&nl_emu_lock);
-	return -EUNATCH;
-}
 
-#endif
+	if (report)
+		err = nlmsg_unicast(sk, skb, pid);
+
+	return err;
+}
 
 #ifdef CONFIG_PROC_FS
+struct nl_seq_iter {
+	int link;
+	int hash_idx;
+};
+
 static struct sock *netlink_seq_socket_idx(struct seq_file *seq, loff_t pos)
 {
-	long i;
+	struct nl_seq_iter *iter = seq->private;
+	int i, j;
 	struct sock *s;
 	struct hlist_node *node;
 	loff_t off = 0;
 
 	for (i=0; i<MAX_LINKS; i++) {
-		sk_for_each(s, node, &nl_table[i]) {
-			if (off == pos) {
-				seq->private = (void *) i;
-				return s;
+		struct nl_pid_hash *hash = &nl_table[i].hash;
+
+		for (j = 0; j <= hash->mask; j++) {
+			sk_for_each(s, node, &hash->table[j]) {
+				if (off == pos) {
+					iter->link = i;
+					iter->hash_idx = j;
+					return s;
+				}
+				++off;
 			}
-			++off;
 		}
 	}
 	return NULL;
@@ -1113,6 +1622,8 @@ static void *netlink_seq_start(struct seq_file *seq, loff_t *pos)
 static void *netlink_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 {
 	struct sock *s;
+	struct nl_seq_iter *iter;
+	int i, j;
 
 	++*pos;
 
@@ -1120,18 +1631,29 @@ static void *netlink_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 		return netlink_seq_socket_idx(seq, 0);
 		
 	s = sk_next(v);
-	if (!s) {
-		long i = (long)seq->private;
+	if (s)
+		return s;
+
+	iter = seq->private;
+	i = iter->link;
+	j = iter->hash_idx + 1;
 
-		while (++i < MAX_LINKS) {
-			s = sk_head(&nl_table[i]);
+	do {
+		struct nl_pid_hash *hash = &nl_table[i].hash;
+
+		for (; j <= hash->mask; j++) {
+			s = sk_head(&hash->table[j]);
 			if (s) {
-				seq->private = (void *) i;
-				break;
+				iter->link = i;
+				iter->hash_idx = j;
+				return s;
 			}
 		}
-	}
-	return s;
+
+		j = 0;
+	} while (++i < MAX_LINKS);
+
+	return NULL;
 }
 
 static void netlink_seq_stop(struct seq_file *seq, void *v)
@@ -1148,13 +1670,13 @@ static int netlink_seq_show(struct seq_file *seq, void *v)
 			 "Rmem     Wmem     Dump     Locks\n");
 	else {
 		struct sock *s = v;
-		struct netlink_opt *nlk = nlk_sk(s);
+		struct netlink_sock *nlk = nlk_sk(s);
 
 		seq_printf(seq, "%p %-3d %-6d %08x %-8d %-8d %p %d\n",
 			   s,
 			   s->sk_protocol,
 			   nlk->pid,
-			   nlk->groups,
+			   nlk->groups ? (u32)nlk->groups[0] : 0,
 			   atomic_read(&s->sk_rmem_alloc),
 			   atomic_read(&s->sk_wmem_alloc),
 			   nlk->cb,
@@ -1175,7 +1697,23 @@ static struct seq_operations netlink_seq_ops = {
 
 static int netlink_seq_open(struct inode *inode, struct file *file)
 {
-	return seq_open(file, &netlink_seq_ops);
+	struct seq_file *seq;
+	struct nl_seq_iter *iter;
+	int err;
+
+	iter = kzalloc(sizeof(*iter), GFP_KERNEL);
+	if (!iter)
+		return -ENOMEM;
+
+	err = seq_open(file, &netlink_seq_ops);
+	if (err) {
+		kfree(iter);
+		return err;
+	}
+
+	seq = file->private_data;
+	seq->private = iter;
+	return 0;
 }
 
 static struct file_operations netlink_seq_fops = {
@@ -1183,22 +1721,22 @@ static struct file_operations netlink_seq_fops = {
 	.open		= netlink_seq_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
-	.release	= seq_release,
+	.release	= seq_release_private,
 };
 
 #endif
 
 int netlink_register_notifier(struct notifier_block *nb)
 {
-	return notifier_chain_register(&netlink_chain, nb);
+	return atomic_notifier_chain_register(&netlink_chain, nb);
 }
 
 int netlink_unregister_notifier(struct notifier_block *nb)
 {
-	return notifier_chain_unregister(&netlink_chain, nb);
+	return atomic_notifier_chain_unregister(&netlink_chain, nb);
 }
                 
-static struct proto_ops netlink_ops = {
+static const struct proto_ops netlink_ops = {
 	.family =	PF_NETLINK,
 	.owner =	THIS_MODULE,
 	.release =	netlink_release,
@@ -1211,8 +1749,8 @@ static struct proto_ops netlink_ops = {
 	.ioctl =	sock_no_ioctl,
 	.listen =	sock_no_listen,
 	.shutdown =	sock_no_shutdown,
-	.setsockopt =	sock_no_setsockopt,
-	.getsockopt =	sock_no_getsockopt,
+	.setsockopt =	netlink_setsockopt,
+	.getsockopt =	netlink_getsockopt,
 	.sendmsg =	netlink_sendmsg,
 	.recvmsg =	netlink_recvmsg,
 	.mmap =		sock_no_mmap,
@@ -1228,34 +1766,64 @@ static struct net_proto_family netlink_family_ops = {
 static int __init netlink_proto_init(void)
 {
 	struct sk_buff *dummy_skb;
+	int i;
+	unsigned long max;
+	unsigned int order;
+	int err = proto_register(&netlink_proto, 0);
 
-	if (sizeof(struct netlink_skb_parms) > sizeof(dummy_skb->cb)) {
-		printk(KERN_CRIT "netlink_init: panic\n");
-		return -1;
+	if (err != 0)
+		goto out;
+
+	BUILD_BUG_ON(sizeof(struct netlink_skb_parms) > sizeof(dummy_skb->cb));
+
+	nl_table = kcalloc(MAX_LINKS, sizeof(*nl_table), GFP_KERNEL);
+	if (!nl_table)
+		goto panic;
+
+	if (num_physpages >= (128 * 1024))
+		max = num_physpages >> (21 - PAGE_SHIFT);
+	else
+		max = num_physpages >> (23 - PAGE_SHIFT);
+
+	order = get_bitmask_order(max) - 1 + PAGE_SHIFT;
+	max = (1UL << order) / sizeof(struct hlist_head);
+	order = get_bitmask_order(max > UINT_MAX ? UINT_MAX : max) - 1;
+
+	for (i = 0; i < MAX_LINKS; i++) {
+		struct nl_pid_hash *hash = &nl_table[i].hash;
+
+		hash->table = nl_pid_hash_alloc(1 * sizeof(*hash->table));
+		if (!hash->table) {
+			while (i-- > 0)
+				nl_pid_hash_free(nl_table[i].hash.table,
+						 1 * sizeof(*hash->table));
+			kfree(nl_table);
+			goto panic;
+		}
+		memset(hash->table, 0, 1 * sizeof(*hash->table));
+		hash->max_shift = order;
+		hash->shift = 0;
+		hash->mask = 0;
+		hash->rehash_time = jiffies;
 	}
+
 	sock_register(&netlink_family_ops);
 #ifdef CONFIG_PROC_FS
 	proc_net_fops_create("netlink", 0, &netlink_seq_fops);
 #endif
 	/* The netlink device handler may be needed early. */ 
 	rtnetlink_init();
-	return 0;
-}
-
-static void __exit netlink_proto_exit(void)
-{
-       sock_unregister(PF_NETLINK);
-       proc_net_remove("netlink");
+out:
+	return err;
+panic:
+	panic("netlink_init: Cannot allocate nl_table\n");
 }
 
 core_initcall(netlink_proto_init);
-module_exit(netlink_proto_exit);
-
-MODULE_LICENSE("GPL");
-
-MODULE_ALIAS_NETPROTO(PF_NETLINK);
 
 EXPORT_SYMBOL(netlink_ack);
+EXPORT_SYMBOL(netlink_run_queue);
+EXPORT_SYMBOL(netlink_queue_skip);
 EXPORT_SYMBOL(netlink_broadcast);
 EXPORT_SYMBOL(netlink_dump_start);
 EXPORT_SYMBOL(netlink_kernel_create);
@@ -1264,9 +1832,4 @@ EXPORT_SYMBOL(netlink_set_err);
 EXPORT_SYMBOL(netlink_set_nonroot);
 EXPORT_SYMBOL(netlink_unicast);
 EXPORT_SYMBOL(netlink_unregister_notifier);
-
-#if defined(CONFIG_NETLINK_DEV) || defined(CONFIG_NETLINK_DEV_MODULE)
-EXPORT_SYMBOL(netlink_attach);
-EXPORT_SYMBOL(netlink_detach);
-EXPORT_SYMBOL(netlink_post);
-#endif
+EXPORT_SYMBOL(nlmsg_notify);