[linux-2.6-block.git] / net / netfilter / nfnetlink.c

/* Netfilter messages via netlink socket. Allows for user space
 * protocol helpers and general trouble making from userspace.
 *
 * (C) 2001 by Jay Schulist <jschlst@samba.org>,
 * (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
 * (C) 2005,2007 by Pablo Neira Ayuso <pablo@netfilter.org>
 *
 * Initial netfilter messages via netlink development funded and
 * generally made possible by Network Robots, Inc. (www.networkrobots.com)
 *
 * Further development of this code funded by Astaro AG (http://www.astaro.com)
 *
 * This software may be used and distributed according to the terms
 * of the GNU General Public License, incorporated herein by reference.
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <asm/uaccess.h>
#include <net/sock.h>
#include <linux/init.h>

#include <net/netlink.h>
#include <linux/netfilter/nfnetlink.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);

static char __initdata nfversion[] = "0.30";

static struct {
	struct mutex				mutex;
	const struct nfnetlink_subsystem __rcu	*subsys;
} table[NFNL_SUBSYS_COUNT];

static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	[NFNLGRP_CONNTRACK_NEW]		= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_UPDATE]	= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_DESTROY]	= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_EXP_NEW]	= NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_CONNTRACK_EXP_UPDATE]	= NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_NFTABLES]		= NFNL_SUBSYS_NFTABLES,
	[NFNLGRP_ACCT_QUOTA]		= NFNL_SUBSYS_ACCT,
};

void nfnl_lock(__u8 subsys_id)
{
	mutex_lock(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(nfnl_lock);

void nfnl_unlock(__u8 subsys_id)
{
	mutex_unlock(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(nfnl_unlock);

#ifdef CONFIG_PROVE_LOCKING
int lockdep_nfnl_is_held(u8 subsys_id)
{
	return lockdep_is_held(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
#endif

int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
{
	nfnl_lock(n->subsys_id);
	if (table[n->subsys_id].subsys) {
		nfnl_unlock(n->subsys_id);
		return -EBUSY;
	}
	rcu_assign_pointer(table[n->subsys_id].subsys, n);
	nfnl_unlock(n->subsys_id);

	return 0;
}
EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);

int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
{
	nfnl_lock(n->subsys_id);
	table[n->subsys_id].subsys = NULL;
	nfnl_unlock(n->subsys_id);
	synchronize_rcu();
	return 0;
}
EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);

static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
{
	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return NULL;

	return rcu_dereference(table[subsys_id].subsys);
}

static inline const struct nfnl_callback *
nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
{
	u_int8_t cb_id = NFNL_MSG_TYPE(type);

	if (cb_id >= ss->cb_count)
		return NULL;

	return &ss->cb[cb_id];
}

int nfnetlink_has_listeners(struct net *net, unsigned int group)
{
	return netlink_has_listeners(net->nfnl, group);
}
EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);

struct sk_buff *nfnetlink_alloc_skb(struct net *net, unsigned int size,
				    u32 dst_portid, gfp_t gfp_mask)
{
	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
}
EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);

int nfnetlink_send(struct sk_buff *skb, struct net *net, u32 portid,
		   unsigned int group, int echo, gfp_t flags)
{
	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
}
EXPORT_SYMBOL_GPL(nfnetlink_send);

int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
{
	return netlink_set_err(net->nfnl, portid, group, error);
}
EXPORT_SYMBOL_GPL(nfnetlink_set_err);

int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid,
		      int flags)
{
	return netlink_unicast(net->nfnl, skb, portid, flags);
}
EXPORT_SYMBOL_GPL(nfnetlink_unicast);

/* Process one complete nfnetlink message. */
static int nfnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
{
	struct net *net = sock_net(skb->sk);
	const struct nfnl_callback *nc;
	const struct nfnetlink_subsystem *ss;
	int type, err;

	/* All the messages must at least contain nfgenmsg */
	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
		return 0;

	type = nlh->nlmsg_type;
replay:
	rcu_read_lock();
	ss = nfnetlink_get_subsys(type);
	if (!ss) {
#ifdef CONFIG_MODULES
		rcu_read_unlock();
		request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
		rcu_read_lock();
		ss = nfnetlink_get_subsys(type);
		if (!ss)
#endif
		{
			rcu_read_unlock();
			return -EINVAL;
		}
	}

	nc = nfnetlink_find_client(type, ss);
	if (!nc) {
		rcu_read_unlock();
		return -EINVAL;
	}

	{
		int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
		u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
		struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
		struct nlattr *attr = (void *)nlh + min_len;
		int attrlen = nlh->nlmsg_len - min_len;
		__u8 subsys_id = NFNL_SUBSYS_ID(type);

		err = nla_parse(cda, ss->cb[cb_id].attr_count,
				attr, attrlen, ss->cb[cb_id].policy);
		if (err < 0) {
			rcu_read_unlock();
			return err;
		}

		if (nc->call_rcu) {
			err = nc->call_rcu(net->nfnl, skb, nlh,
					   (const struct nlattr **)cda);
			rcu_read_unlock();
		} else {
			rcu_read_unlock();
			nfnl_lock(subsys_id);
			if (rcu_dereference_protected(table[subsys_id].subsys,
				lockdep_is_held(&table[subsys_id].mutex)) != ss ||
			    nfnetlink_find_client(type, ss) != nc)
				err = -EAGAIN;
			else if (nc->call)
				err = nc->call(net->nfnl, skb, nlh,
						   (const struct nlattr **)cda);
			else
				err = -EINVAL;
			nfnl_unlock(subsys_id);
		}
		if (err == -EAGAIN)
			goto replay;
		return err;
	}
}

struct nfnl_err {
	struct list_head	head;
	struct nlmsghdr		*nlh;
	int			err;
};

static int nfnl_err_add(struct list_head *list, struct nlmsghdr *nlh, int err)
{
	struct nfnl_err *nfnl_err;

	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	if (nfnl_err == NULL)
		return -ENOMEM;

	nfnl_err->nlh = nlh;
	nfnl_err->err = err;
	list_add_tail(&nfnl_err->head, list);

	return 0;
}

static void nfnl_err_del(struct nfnl_err *nfnl_err)
{
	list_del(&nfnl_err->head);
	kfree(nfnl_err);
}

static void nfnl_err_reset(struct list_head *err_list)
{
	struct nfnl_err *nfnl_err, *next;

	list_for_each_entry_safe(nfnl_err, next, err_list, head)
		nfnl_err_del(nfnl_err);
}

static void nfnl_err_deliver(struct list_head *err_list, struct sk_buff *skb)
{
	struct nfnl_err *nfnl_err, *next;

	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
		netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
		nfnl_err_del(nfnl_err);
	}
}

static void nfnetlink_rcv_batch(struct sk_buff *skb, struct nlmsghdr *nlh,
				u_int16_t subsys_id)
{
	struct sk_buff *nskb, *oskb = skb;
	struct net *net = sock_net(skb->sk);
	const struct nfnetlink_subsystem *ss;
	const struct nfnl_callback *nc;
	bool success = true, done = false;
	static LIST_HEAD(err_list);
	int err;

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return netlink_ack(skb, nlh, -EINVAL);
replay:
	nskb = netlink_skb_clone(oskb, GFP_KERNEL);
	if (!nskb)
		return netlink_ack(oskb, nlh, -ENOMEM);

	nskb->sk = oskb->sk;
	skb = nskb;

	nfnl_lock(subsys_id);
	ss = rcu_dereference_protected(table[subsys_id].subsys,
				       lockdep_is_held(&table[subsys_id].mutex));
	if (!ss) {
#ifdef CONFIG_MODULES
		nfnl_unlock(subsys_id);
		request_module("nfnetlink-subsys-%d", subsys_id);
		nfnl_lock(subsys_id);
		ss = rcu_dereference_protected(table[subsys_id].subsys,
					       lockdep_is_held(&table[subsys_id].mutex));
		if (!ss)
#endif
		{
			nfnl_unlock(subsys_id);
			netlink_ack(skb, nlh, -EOPNOTSUPP);
			return kfree_skb(nskb);
		}
	}

	if (!ss->commit || !ss->abort) {
		nfnl_unlock(subsys_id);
		netlink_ack(skb, nlh, -EOPNOTSUPP);
		return kfree_skb(skb);
	}

	while (skb->len >= nlmsg_total_size(0)) {
		int msglen, type;

		nlh = nlmsg_hdr(skb);
		err = 0;

		if (nlh->nlmsg_len < NLMSG_HDRLEN) {
			err = -EINVAL;
			goto ack;
		}

		/* Only requests are handled by the kernel */
		if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
			err = -EINVAL;
			goto ack;
		}

		type = nlh->nlmsg_type;
		if (type == NFNL_MSG_BATCH_BEGIN) {
			/* Malformed: Batch begin twice */
			nfnl_err_reset(&err_list);
			success = false;
			goto done;
		} else if (type == NFNL_MSG_BATCH_END) {
			done = true;
			goto done;
		} else if (type < NLMSG_MIN_TYPE) {
			err = -EINVAL;
			goto ack;
		}

		/* We only accept a batch with messages for the same
		 * subsystem.
		 */
		if (NFNL_SUBSYS_ID(type) != subsys_id) {
			err = -EINVAL;
			goto ack;
		}

		nc = nfnetlink_find_client(type, ss);
		if (!nc) {
			err = -EINVAL;
			goto ack;
		}

		{
			int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
			u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
			struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
			struct nlattr *attr = (void *)nlh + min_len;
			int attrlen = nlh->nlmsg_len - min_len;

			err = nla_parse(cda, ss->cb[cb_id].attr_count,
					attr, attrlen, ss->cb[cb_id].policy);
			if (err < 0)
				goto ack;

			if (nc->call_batch) {
				err = nc->call_batch(net->nfnl, skb, nlh,
						     (const struct nlattr **)cda);
			}

			/* The lock was released to autoload some module, we
			 * have to abort and start from scratch using the
			 * original skb.
			 */
			if (err == -EAGAIN) {
				nfnl_err_reset(&err_list);
				ss->abort(oskb);
				nfnl_unlock(subsys_id);
				kfree_skb(nskb);
				goto replay;
			}
		}
ack:
		if (nlh->nlmsg_flags & NLM_F_ACK || err) {
			/* Errors are delivered once the full batch has been
			 * processed, this avoids that the same error is
			 * reported several times when replaying the batch.
			 */
			if (nfnl_err_add(&err_list, nlh, err) < 0) {
				/* We failed to enqueue an error, reset the
				 * list of errors and send OOM to userspace
				 * pointing to the batch header.
				 */
				nfnl_err_reset(&err_list);
				netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM);
				success = false;
				goto done;
			}
			/* We don't stop processing the batch on errors, thus,
			 * userspace gets all the errors that the batch
			 * triggers.
			 */
			if (err)
				success = false;
		}

		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;
		skb_pull(skb, msglen);
	}
done:
	if (success && done)
		ss->commit(oskb);
	else
		ss->abort(oskb);

	nfnl_err_deliver(&err_list, oskb);
	nfnl_unlock(subsys_id);
	kfree_skb(nskb);
}

static void nfnetlink_rcv(struct sk_buff *skb)
{
	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	int msglen;

	if (nlh->nlmsg_len < NLMSG_HDRLEN ||
	    skb->len < nlh->nlmsg_len)
		return;

	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
		netlink_ack(skb, nlh, -EPERM);
		return;
	}

	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
		struct nfgenmsg *nfgenmsg;

		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;

		if (nlh->nlmsg_len < NLMSG_HDRLEN ||
		    skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
			return;

		nfgenmsg = nlmsg_data(nlh);
		skb_pull(skb, msglen);
		nfnetlink_rcv_batch(skb, nlh, nfgenmsg->res_id);
	} else {
		netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	}
}

#ifdef CONFIG_MODULES
static int nfnetlink_bind(struct net *net, int group)
{
	const struct nfnetlink_subsystem *ss;
	int type;

	if (group <= NFNLGRP_NONE || group > NFNLGRP_MAX)
		return -EINVAL;

	type = nfnl_group2type[group];

	rcu_read_lock();
	ss = nfnetlink_get_subsys(type);
	rcu_read_unlock();
	if (!ss)
		request_module("nfnetlink-subsys-%d", type);
	return 0;
}
#endif

static int __net_init nfnetlink_net_init(struct net *net)
{
	struct sock *nfnl;
	struct netlink_kernel_cfg cfg = {
		.groups	= NFNLGRP_MAX,
		.input	= nfnetlink_rcv,
#ifdef CONFIG_MODULES
		.bind	= nfnetlink_bind,
#endif
	};

	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
	if (!nfnl)
		return -ENOMEM;
	net->nfnl_stash = nfnl;
	rcu_assign_pointer(net->nfnl, nfnl);
	return 0;
}

static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
{
	struct net *net;

	list_for_each_entry(net, net_exit_list, exit_list)
		RCU_INIT_POINTER(net->nfnl, NULL);
	synchronize_net();
	list_for_each_entry(net, net_exit_list, exit_list)
		netlink_kernel_release(net->nfnl_stash);
}

static struct pernet_operations nfnetlink_net_ops = {
	.init		= nfnetlink_net_init,
	.exit_batch	= nfnetlink_net_exit_batch,
};

static int __init nfnetlink_init(void)
{
	int i;

	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
		BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);

	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
		mutex_init(&table[i].mutex);

	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
	return register_pernet_subsys(&nfnetlink_net_ops);
}

static void __exit nfnetlink_exit(void)
{
	pr_info("Removing netfilter NETLINK layer.\n");
	unregister_pernet_subsys(&nfnetlink_net_ops);
}
module_init(nfnetlink_init);
module_exit(nfnetlink_exit);
Commit	Line	Data
	1	/* Netfilter messages via netlink socket. Allows for user space
	2	* protocol helpers and general trouble making from userspace.
	3	*
	4	* (C) 2001 by Jay Schulist <jschlst@samba.org>,
	5	* (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
	6	* (C) 2005,2007 by Pablo Neira Ayuso <pablo@netfilter.org>
	7	*
	8	* Initial netfilter messages via netlink development funded and
	9	* generally made possible by Network Robots, Inc. (www.networkrobots.com)
	10	*
	11	* Further development of this code funded by Astaro AG (http://www.astaro.com)
	12	*
	13	* This software may be used and distributed according to the terms
	14	* of the GNU General Public License, incorporated herein by reference.
	15	*/
	16
	17	#include <linux/module.h>
	18	#include <linux/types.h>
	19	#include <linux/socket.h>
	20	#include <linux/kernel.h>
	21	#include <linux/string.h>
	22	#include <linux/sockios.h>
	23	#include <linux/net.h>
	24	#include <linux/skbuff.h>
	25	#include <asm/uaccess.h>
	26	#include <net/sock.h>
	27	#include <linux/init.h>
	28
	29	#include <net/netlink.h>
	30	#include <linux/netfilter/nfnetlink.h>
	31
	32	MODULE_LICENSE("GPL");
	33	MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
	34	MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
	35
	36	static char __initdata nfversion[] = "0.30";
	37
	38	static struct {
	39	struct mutex mutex;
	40	const struct nfnetlink_subsystem __rcu *subsys;
	41	} table[NFNL_SUBSYS_COUNT];
	42
	43	static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	44	[NFNLGRP_CONNTRACK_NEW] = NFNL_SUBSYS_CTNETLINK,
	45	[NFNLGRP_CONNTRACK_UPDATE] = NFNL_SUBSYS_CTNETLINK,
	46	[NFNLGRP_CONNTRACK_DESTROY] = NFNL_SUBSYS_CTNETLINK,
	47	[NFNLGRP_CONNTRACK_EXP_NEW] = NFNL_SUBSYS_CTNETLINK_EXP,
	48	[NFNLGRP_CONNTRACK_EXP_UPDATE] = NFNL_SUBSYS_CTNETLINK_EXP,
	49	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
	50	[NFNLGRP_NFTABLES] = NFNL_SUBSYS_NFTABLES,
	51	[NFNLGRP_ACCT_QUOTA] = NFNL_SUBSYS_ACCT,
	52	};
	53
	54	void nfnl_lock(__u8 subsys_id)
	55	{
	56	mutex_lock(&table[subsys_id].mutex);
	57	}
	58	EXPORT_SYMBOL_GPL(nfnl_lock);
	59
	60	void nfnl_unlock(__u8 subsys_id)
	61	{
	62	mutex_unlock(&table[subsys_id].mutex);
	63	}
	64	EXPORT_SYMBOL_GPL(nfnl_unlock);
	65
	66	#ifdef CONFIG_PROVE_LOCKING
	67	int lockdep_nfnl_is_held(u8 subsys_id)
	68	{
	69	return lockdep_is_held(&table[subsys_id].mutex);
	70	}
	71	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	72	#endif
	73
	74	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
	75	{
	76	nfnl_lock(n->subsys_id);
	77	if (table[n->subsys_id].subsys) {
	78	nfnl_unlock(n->subsys_id);
	79	return -EBUSY;
	80	}
	81	rcu_assign_pointer(table[n->subsys_id].subsys, n);
	82	nfnl_unlock(n->subsys_id);
	83
	84	return 0;
	85	}
	86	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
	87
	88	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
	89	{
	90	nfnl_lock(n->subsys_id);
	91	table[n->subsys_id].subsys = NULL;
	92	nfnl_unlock(n->subsys_id);
	93	synchronize_rcu();
	94	return 0;
	95	}
	96	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
	97
	98	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
	99	{
	100	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	101
	102	if (subsys_id >= NFNL_SUBSYS_COUNT)
	103	return NULL;
	104
	105	return rcu_dereference(table[subsys_id].subsys);
	106	}
	107
	108	static inline const struct nfnl_callback *
	109	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
	110	{
	111	u_int8_t cb_id = NFNL_MSG_TYPE(type);
	112
	113	if (cb_id >= ss->cb_count)
	114	return NULL;
	115
	116	return &ss->cb[cb_id];
	117	}
	118
	119	int nfnetlink_has_listeners(struct net *net, unsigned int group)
	120	{
	121	return netlink_has_listeners(net->nfnl, group);
	122	}
	123	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	124
	125	struct sk_buff nfnetlink_alloc_skb(struct net net, unsigned int size,
	126	u32 dst_portid, gfp_t gfp_mask)
	127	{
	128	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
	129	}
	130	EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);
	131
	132	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
	133	unsigned int group, int echo, gfp_t flags)
	134	{
	135	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
	136	}
	137	EXPORT_SYMBOL_GPL(nfnetlink_send);
	138
	139	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
	140	{
	141	return netlink_set_err(net->nfnl, portid, group, error);
	142	}
	143	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	144
	145	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
	146	int flags)
	147	{
	148	return netlink_unicast(net->nfnl, skb, portid, flags);
	149	}
	150	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
	151
	152	/* Process one complete nfnetlink message. */
	153	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
	154	{
	155	struct net *net = sock_net(skb->sk);
	156	const struct nfnl_callback *nc;
	157	const struct nfnetlink_subsystem *ss;
	158	int type, err;
	159
	160	/* All the messages must at least contain nfgenmsg */
	161	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
	162	return 0;
	163
	164	type = nlh->nlmsg_type;
	165	replay:
	166	rcu_read_lock();
	167	ss = nfnetlink_get_subsys(type);
	168	if (!ss) {
	169	#ifdef CONFIG_MODULES
	170	rcu_read_unlock();
	171	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
	172	rcu_read_lock();
	173	ss = nfnetlink_get_subsys(type);
	174	if (!ss)
	175	#endif
	176	{
	177	rcu_read_unlock();
	178	return -EINVAL;
	179	}
	180	}
	181
	182	nc = nfnetlink_find_client(type, ss);
	183	if (!nc) {
	184	rcu_read_unlock();
	185	return -EINVAL;
	186	}
	187
	188	{
	189	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	190	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	191	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	192	struct nlattr attr = (void )nlh + min_len;
	193	int attrlen = nlh->nlmsg_len - min_len;
	194	__u8 subsys_id = NFNL_SUBSYS_ID(type);
	195
	196	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	197	attr, attrlen, ss->cb[cb_id].policy);
	198	if (err < 0) {
	199	rcu_read_unlock();
	200	return err;
	201	}
	202
	203	if (nc->call_rcu) {
	204	err = nc->call_rcu(net->nfnl, skb, nlh,
	205	(const struct nlattr **)cda);
	206	rcu_read_unlock();
	207	} else {
	208	rcu_read_unlock();
	209	nfnl_lock(subsys_id);
	210	if (rcu_dereference_protected(table[subsys_id].subsys,
	211	lockdep_is_held(&table[subsys_id].mutex)) != ss \|\|
	212	nfnetlink_find_client(type, ss) != nc)
	213	err = -EAGAIN;
	214	else if (nc->call)
	215	err = nc->call(net->nfnl, skb, nlh,
	216	(const struct nlattr **)cda);
	217	else
	218	err = -EINVAL;
	219	nfnl_unlock(subsys_id);
	220	}
	221	if (err == -EAGAIN)
	222	goto replay;
	223	return err;
	224	}
	225	}
	226
	227	struct nfnl_err {
	228	struct list_head head;
	229	struct nlmsghdr *nlh;
	230	int err;
	231	};
	232
	233	static int nfnl_err_add(struct list_head list, struct nlmsghdr nlh, int err)
	234	{
	235	struct nfnl_err *nfnl_err;
	236
	237	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	238	if (nfnl_err == NULL)
	239	return -ENOMEM;
	240
	241	nfnl_err->nlh = nlh;
	242	nfnl_err->err = err;
	243	list_add_tail(&nfnl_err->head, list);
	244
	245	return 0;
	246	}
	247
	248	static void nfnl_err_del(struct nfnl_err *nfnl_err)
	249	{
	250	list_del(&nfnl_err->head);
	251	kfree(nfnl_err);
	252	}
	253
	254	static void nfnl_err_reset(struct list_head *err_list)
	255	{
	256	struct nfnl_err nfnl_err, next;
	257
	258	list_for_each_entry_safe(nfnl_err, next, err_list, head)
	259	nfnl_err_del(nfnl_err);
	260	}
	261
	262	static void nfnl_err_deliver(struct list_head err_list, struct sk_buff skb)
	263	{
	264	struct nfnl_err nfnl_err, next;
	265
	266	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
	267	netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
	268	nfnl_err_del(nfnl_err);
	269	}
	270	}
	271
	272	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	273	u_int16_t subsys_id)
	274	{
	275	struct sk_buff nskb, oskb = skb;
	276	struct net *net = sock_net(skb->sk);
	277	const struct nfnetlink_subsystem *ss;
	278	const struct nfnl_callback *nc;
	279	bool success = true, done = false;
	280	static LIST_HEAD(err_list);
	281	int err;
	282
	283	if (subsys_id >= NFNL_SUBSYS_COUNT)
	284	return netlink_ack(skb, nlh, -EINVAL);
	285	replay:
	286	nskb = netlink_skb_clone(oskb, GFP_KERNEL);
	287	if (!nskb)
	288	return netlink_ack(oskb, nlh, -ENOMEM);
	289
	290	nskb->sk = oskb->sk;
	291	skb = nskb;
	292
	293	nfnl_lock(subsys_id);
	294	ss = rcu_dereference_protected(table[subsys_id].subsys,
	295	lockdep_is_held(&table[subsys_id].mutex));
	296	if (!ss) {
	297	#ifdef CONFIG_MODULES
	298	nfnl_unlock(subsys_id);
	299	request_module("nfnetlink-subsys-%d", subsys_id);
	300	nfnl_lock(subsys_id);
	301	ss = rcu_dereference_protected(table[subsys_id].subsys,
	302	lockdep_is_held(&table[subsys_id].mutex));
	303	if (!ss)
	304	#endif
	305	{
	306	nfnl_unlock(subsys_id);
	307	netlink_ack(skb, nlh, -EOPNOTSUPP);
	308	return kfree_skb(nskb);
	309	}
	310	}
	311
	312	if (!ss->commit \|\| !ss->abort) {
	313	nfnl_unlock(subsys_id);
	314	netlink_ack(skb, nlh, -EOPNOTSUPP);
	315	return kfree_skb(skb);
	316	}
	317
	318	while (skb->len >= nlmsg_total_size(0)) {
	319	int msglen, type;
	320
	321	nlh = nlmsg_hdr(skb);
	322	err = 0;
	323
	324	if (nlh->nlmsg_len < NLMSG_HDRLEN) {
	325	err = -EINVAL;
	326	goto ack;
	327	}
	328
	329	/* Only requests are handled by the kernel */
	330	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	331	err = -EINVAL;
	332	goto ack;
	333	}
	334
	335	type = nlh->nlmsg_type;
	336	if (type == NFNL_MSG_BATCH_BEGIN) {
	337	/* Malformed: Batch begin twice */
	338	nfnl_err_reset(&err_list);
	339	success = false;
	340	goto done;
	341	} else if (type == NFNL_MSG_BATCH_END) {
	342	done = true;
	343	goto done;
	344	} else if (type < NLMSG_MIN_TYPE) {
	345	err = -EINVAL;
	346	goto ack;
	347	}
	348
	349	/* We only accept a batch with messages for the same
	350	* subsystem.
	351	*/
	352	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	353	err = -EINVAL;
	354	goto ack;
	355	}
	356
	357	nc = nfnetlink_find_client(type, ss);
	358	if (!nc) {
	359	err = -EINVAL;
	360	goto ack;
	361	}
	362
	363	{
	364	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	365	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	366	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	367	struct nlattr attr = (void )nlh + min_len;
	368	int attrlen = nlh->nlmsg_len - min_len;
	369
	370	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	371	attr, attrlen, ss->cb[cb_id].policy);
	372	if (err < 0)
	373	goto ack;
	374
	375	if (nc->call_batch) {
	376	err = nc->call_batch(net->nfnl, skb, nlh,
	377	(const struct nlattr **)cda);
	378	}
	379
	380	/* The lock was released to autoload some module, we
	381	* have to abort and start from scratch using the
	382	* original skb.
	383	*/
	384	if (err == -EAGAIN) {
	385	nfnl_err_reset(&err_list);
	386	ss->abort(oskb);
	387	nfnl_unlock(subsys_id);
	388	kfree_skb(nskb);
	389	goto replay;
	390	}
	391	}
	392	ack:
	393	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
	394	/* Errors are delivered once the full batch has been
	395	* processed, this avoids that the same error is
	396	* reported several times when replaying the batch.
	397	*/
	398	if (nfnl_err_add(&err_list, nlh, err) < 0) {
	399	/* We failed to enqueue an error, reset the
	400	* list of errors and send OOM to userspace
	401	* pointing to the batch header.
	402	*/
	403	nfnl_err_reset(&err_list);
	404	netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM);
	405	success = false;
	406	goto done;
	407	}
	408	/* We don't stop processing the batch on errors, thus,
	409	* userspace gets all the errors that the batch
	410	* triggers.
	411	*/
	412	if (err)
	413	success = false;
	414	}
	415
	416	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	417	if (msglen > skb->len)
	418	msglen = skb->len;
	419	skb_pull(skb, msglen);
	420	}
	421	done:
	422	if (success && done)
	423	ss->commit(oskb);
	424	else
	425	ss->abort(oskb);
	426
	427	nfnl_err_deliver(&err_list, oskb);
	428	nfnl_unlock(subsys_id);
	429	kfree_skb(nskb);
	430	}
	431
	432	static void nfnetlink_rcv(struct sk_buff *skb)
	433	{
	434	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	435	int msglen;
	436
	437	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	438	skb->len < nlh->nlmsg_len)
	439	return;
	440
	441	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
	442	netlink_ack(skb, nlh, -EPERM);
	443	return;
	444	}
	445
	446	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	447	struct nfgenmsg *nfgenmsg;
	448
	449	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	450	if (msglen > skb->len)
	451	msglen = skb->len;
	452
	453	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	454	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	455	return;
	456
	457	nfgenmsg = nlmsg_data(nlh);
	458	skb_pull(skb, msglen);
	459	nfnetlink_rcv_batch(skb, nlh, nfgenmsg->res_id);
	460	} else {
	461	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	462	}
	463	}
	464
	465	#ifdef CONFIG_MODULES
	466	static int nfnetlink_bind(struct net *net, int group)
	467	{
	468	const struct nfnetlink_subsystem *ss;
	469	int type;
	470
	471	if (group <= NFNLGRP_NONE \|\| group > NFNLGRP_MAX)
	472	return -EINVAL;
	473
	474	type = nfnl_group2type[group];
	475
	476	rcu_read_lock();
	477	ss = nfnetlink_get_subsys(type);
	478	rcu_read_unlock();
	479	if (!ss)
	480	request_module("nfnetlink-subsys-%d", type);
	481	return 0;
	482	}
	483	#endif
	484
	485	static int __net_init nfnetlink_net_init(struct net *net)
	486	{
	487	struct sock *nfnl;
	488	struct netlink_kernel_cfg cfg = {
	489	.groups = NFNLGRP_MAX,
	490	.input = nfnetlink_rcv,
	491	#ifdef CONFIG_MODULES
	492	.bind = nfnetlink_bind,
	493	#endif
	494	};
	495
	496	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
	497	if (!nfnl)
	498	return -ENOMEM;
	499	net->nfnl_stash = nfnl;
	500	rcu_assign_pointer(net->nfnl, nfnl);
	501	return 0;
	502	}
	503
	504	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
	505	{
	506	struct net *net;
	507
	508	list_for_each_entry(net, net_exit_list, exit_list)
	509	RCU_INIT_POINTER(net->nfnl, NULL);
	510	synchronize_net();
	511	list_for_each_entry(net, net_exit_list, exit_list)
	512	netlink_kernel_release(net->nfnl_stash);
	513	}
	514
	515	static struct pernet_operations nfnetlink_net_ops = {
	516	.init = nfnetlink_net_init,
	517	.exit_batch = nfnetlink_net_exit_batch,
	518	};
	519
	520	static int __init nfnetlink_init(void)
	521	{
	522	int i;
	523
	524	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
	525	BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);
	526
	527	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	528	mutex_init(&table[i].mutex);
	529
	530	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
	531	return register_pernet_subsys(&nfnetlink_net_ops);
	532	}
	533
	534	static void __exit nfnetlink_exit(void)
	535	{
	536	pr_info("Removing netfilter NETLINK layer.\n");
	537	unregister_pernet_subsys(&nfnetlink_net_ops);
	538	}
	539	module_init(nfnetlink_init);
	540	module_exit(nfnetlink_exit);