[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) 2005-2017 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 <linux/uaccess.h>
#include <net/sock.h>
#include <linux/init.h>
#include <linux/sched/signal.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);

#define nfnl_dereference_protected(id) \
	rcu_dereference_protected(table[(id)].subsys, \
				  lockdep_nfnl_is_held((id)))

#define NFNL_MAX_ATTR_COUNT	32

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,
	[NFNLGRP_NFTRACE]		= NFNL_SUBSYS_NFTABLES,
};

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
bool 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)
{
	u8 cb_id;

	/* Sanity-check attr_count size to avoid stack buffer overflow. */
	for (cb_id = 0; cb_id < n->cb_count; cb_id++)
		if (WARN_ON(n->cb[cb_id].attr_count > NFNL_MAX_ATTR_COUNT))
			return -EINVAL;

	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(u16 type)
{
	u8 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(u16 type, const struct nfnetlink_subsystem *ss)
{
	u8 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);

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 netlink_ext_ack *extack)
{
	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));
		u8 cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
		struct nlattr *cda[NFNL_MAX_ATTR_COUNT + 1];
		struct nlattr *attr = (void *)nlh + min_len;
		int attrlen = nlh->nlmsg_len - min_len;
		__u8 subsys_id = NFNL_SUBSYS_ID(type);

		/* Sanity-check NFNL_MAX_ATTR_COUNT */
		if (ss->cb[cb_id].attr_count > NFNL_MAX_ATTR_COUNT) {
			rcu_read_unlock();
			return -ENOMEM;
		}

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

		if (nc->call_rcu) {
			err = nc->call_rcu(net, net->nfnl, skb, nlh,
					   (const struct nlattr **)cda,
					   extack);
			rcu_read_unlock();
		} else {
			rcu_read_unlock();
			nfnl_lock(subsys_id);
			if (nfnl_dereference_protected(subsys_id) != ss ||
			    nfnetlink_find_client(type, ss) != nc)
				err = -EAGAIN;
			else if (nc->call)
				err = nc->call(net, net->nfnl, skb, nlh,
					       (const struct nlattr **)cda,
					       extack);
			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;
	struct netlink_ext_ack	extack;
};

static int nfnl_err_add(struct list_head *list, struct nlmsghdr *nlh, int err,
			const struct netlink_ext_ack *extack)
{
	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;
	nfnl_err->extack = *extack;
	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->extack);
		nfnl_err_del(nfnl_err);
	}
}

enum {
	NFNL_BATCH_FAILURE	= (1 << 0),
	NFNL_BATCH_DONE		= (1 << 1),
	NFNL_BATCH_REPLAY	= (1 << 2),
};

static void nfnetlink_rcv_batch(struct sk_buff *skb, struct nlmsghdr *nlh,
				u16 subsys_id, u32 genid)
{
	struct sk_buff *oskb = skb;
	struct net *net = sock_net(skb->sk);
	const struct nfnetlink_subsystem *ss;
	const struct nfnl_callback *nc;
	struct netlink_ext_ack extack;
	LIST_HEAD(err_list);
	u32 status;
	int err;

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return netlink_ack(skb, nlh, -EINVAL, NULL);
replay:
	status = 0;

	skb = netlink_skb_clone(oskb, GFP_KERNEL);
	if (!skb)
		return netlink_ack(oskb, nlh, -ENOMEM, NULL);

	nfnl_lock(subsys_id);
	ss = nfnl_dereference_protected(subsys_id);
	if (!ss) {
#ifdef CONFIG_MODULES
		nfnl_unlock(subsys_id);
		request_module("nfnetlink-subsys-%d", subsys_id);
		nfnl_lock(subsys_id);
		ss = nfnl_dereference_protected(subsys_id);
		if (!ss)
#endif
		{
			nfnl_unlock(subsys_id);
			netlink_ack(oskb, nlh, -EOPNOTSUPP, NULL);
			return kfree_skb(skb);
		}
	}

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

	if (!try_module_get(ss->owner)) {
		nfnl_unlock(subsys_id);
		netlink_ack(oskb, nlh, -EOPNOTSUPP, NULL);
		return kfree_skb(skb);
	}

	if (!ss->valid_genid(net, genid)) {
		module_put(ss->owner);
		nfnl_unlock(subsys_id);
		netlink_ack(oskb, nlh, -ERESTART, NULL);
		return kfree_skb(skb);
	}

	nfnl_unlock(subsys_id);

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

		if (fatal_signal_pending(current)) {
			nfnl_err_reset(&err_list);
			err = -EINTR;
			status = NFNL_BATCH_FAILURE;
			goto done;
		}

		memset(&extack, 0, sizeof(extack));
		nlh = nlmsg_hdr(skb);
		err = 0;

		if (nlh->nlmsg_len < NLMSG_HDRLEN ||
		    skb->len < nlh->nlmsg_len ||
		    nlmsg_len(nlh) < sizeof(struct nfgenmsg)) {
			nfnl_err_reset(&err_list);
			status |= NFNL_BATCH_FAILURE;
			goto done;
		}

		/* 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);
			status |= NFNL_BATCH_FAILURE;
			goto done;
		} else if (type == NFNL_MSG_BATCH_END) {
			status |= NFNL_BATCH_DONE;
			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));
			u8 cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
			struct nlattr *cda[NFNL_MAX_ATTR_COUNT + 1];
			struct nlattr *attr = (void *)nlh + min_len;
			int attrlen = nlh->nlmsg_len - min_len;

			/* Sanity-check NFTA_MAX_ATTR */
			if (ss->cb[cb_id].attr_count > NFNL_MAX_ATTR_COUNT) {
				err = -ENOMEM;
				goto ack;
			}

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

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

			/* The lock was released to autoload some module, we
			 * have to abort and start from scratch using the
			 * original skb.
			 */
			if (err == -EAGAIN) {
				status |= NFNL_BATCH_REPLAY;
				goto done;
			}
		}
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, &extack) < 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(oskb, nlmsg_hdr(oskb), -ENOMEM,
					    NULL);
				status |= NFNL_BATCH_FAILURE;
				goto done;
			}
			/* We don't stop processing the batch on errors, thus,
			 * userspace gets all the errors that the batch
			 * triggers.
			 */
			if (err)
				status |= NFNL_BATCH_FAILURE;
		}

		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;
		skb_pull(skb, msglen);
	}
done:
	if (status & NFNL_BATCH_REPLAY) {
		ss->abort(net, oskb);
		nfnl_err_reset(&err_list);
		kfree_skb(skb);
		module_put(ss->owner);
		goto replay;
	} else if (status == NFNL_BATCH_DONE) {
		err = ss->commit(net, oskb);
		if (err == -EAGAIN) {
			status |= NFNL_BATCH_REPLAY;
			goto done;
		} else if (err) {
			ss->abort(net, oskb);
			netlink_ack(oskb, nlmsg_hdr(oskb), err, NULL);
		}
	} else {
		ss->abort(net, oskb);
	}
	if (ss->cleanup)
		ss->cleanup(net);

	nfnl_err_deliver(&err_list, oskb);
	kfree_skb(skb);
	module_put(ss->owner);
}

static const struct nla_policy nfnl_batch_policy[NFNL_BATCH_MAX + 1] = {
	[NFNL_BATCH_GENID]	= { .type = NLA_U32 },
};

static void nfnetlink_rcv_skb_batch(struct sk_buff *skb, struct nlmsghdr *nlh)
{
	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	struct nlattr *attr = (void *)nlh + min_len;
	struct nlattr *cda[NFNL_BATCH_MAX + 1];
	int attrlen = nlh->nlmsg_len - min_len;
	struct nfgenmsg *nfgenmsg;
	int msglen, err;
	u32 gen_id = 0;
	u16 res_id;

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

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

	err = nla_parse_deprecated(cda, NFNL_BATCH_MAX, attr, attrlen,
				   nfnl_batch_policy, NULL);
	if (err < 0) {
		netlink_ack(skb, nlh, err, NULL);
		return;
	}
	if (cda[NFNL_BATCH_GENID])
		gen_id = ntohl(nla_get_be32(cda[NFNL_BATCH_GENID]));

	nfgenmsg = nlmsg_data(nlh);
	skb_pull(skb, msglen);
	/* Work around old nft using host byte order */
	if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
		res_id = NFNL_SUBSYS_NFTABLES;
	else
		res_id = ntohs(nfgenmsg->res_id);

	nfnetlink_rcv_batch(skb, nlh, res_id, gen_id);
}

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

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

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

	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN)
		nfnetlink_rcv_skb_batch(skb, nlh);
	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 0;

	type = nfnl_group2type[group];

	rcu_read_lock();
	ss = nfnetlink_get_subsys(type << 8);
	rcu_read_unlock();
	if (!ss)
		request_module_nowait("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);

	return register_pernet_subsys(&nfnetlink_net_ops);
}

static void __exit nfnetlink_exit(void)
{
	unregister_pernet_subsys(&nfnetlink_net_ops);
}
module_init(nfnetlink_init);
module_exit(nfnetlink_exit);
Commit	Line	Data
f9e815b3 HW	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>
8c4d4e8b	6	* (C) 2005-2017 by Pablo Neira Ayuso <pablo@netfilter.org>
f9e815b3 HW	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
f9e815b3 HW	17	#include <linux/module.h>
	18	#include <linux/types.h>
	19	#include <linux/socket.h>
	20	#include <linux/kernel.h>
f9e815b3 HW	21	#include <linux/string.h>
	22	#include <linux/sockios.h>
	23	#include <linux/net.h>
f9e815b3	24	#include <linux/skbuff.h>
7c0f6ba6	25	#include <linux/uaccess.h>
f9e815b3 HW	26	#include <net/sock.h>
f9e815b3 HW	27	#include <linux/init.h>
8a3d4c36	28	#include <linux/sched/signal.h>
f9e815b3	29
573ce260	30	#include <net/netlink.h>
f9e815b3 HW	31	#include <linux/netfilter/nfnetlink.h>
	32
	33	MODULE_LICENSE("GPL");
4fdb3bb7 HW	34	MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
4fdb3bb7 HW	35	MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
f9e815b3	36
9c55d3b5 FW	37	#define nfnl_dereference_protected(id) \
	38	rcu_dereference_protected(table[(id)].subsys, \
	39	lockdep_nfnl_is_held((id)))
	40
7b7744e2 KC	41	#define NFNL_MAX_ATTR_COUNT 32
7b7744e2 KC	42
c14b78e7 PNA	43	static struct {
	44	struct mutex mutex;
	45	const struct nfnetlink_subsystem __rcu *subsys;
	46	} table[NFNL_SUBSYS_COUNT];
f9e815b3	47
03292745 PNA	48	static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	49	[NFNLGRP_CONNTRACK_NEW] = NFNL_SUBSYS_CTNETLINK,
	50	[NFNLGRP_CONNTRACK_UPDATE] = NFNL_SUBSYS_CTNETLINK,
	51	[NFNLGRP_CONNTRACK_DESTROY] = NFNL_SUBSYS_CTNETLINK,
	52	[NFNLGRP_CONNTRACK_EXP_NEW] = NFNL_SUBSYS_CTNETLINK_EXP,
	53	[NFNLGRP_CONNTRACK_EXP_UPDATE] = NFNL_SUBSYS_CTNETLINK_EXP,
	54	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
97840cb6 PNA	55	[NFNLGRP_NFTABLES] = NFNL_SUBSYS_NFTABLES,
97840cb6 PNA	56	[NFNLGRP_ACCT_QUOTA] = NFNL_SUBSYS_ACCT,
33d5a7b1	57	[NFNLGRP_NFTRACE] = NFNL_SUBSYS_NFTABLES,
03292745 PNA	58	};
03292745 PNA	59
c14b78e7	60	void nfnl_lock(__u8 subsys_id)
f9e815b3	61	{
c14b78e7	62	mutex_lock(&table[subsys_id].mutex);
f9e815b3	63	}
e6a7d3c0	64	EXPORT_SYMBOL_GPL(nfnl_lock);
f9e815b3	65
c14b78e7	66	void nfnl_unlock(__u8 subsys_id)
a3c5029c	67	{
c14b78e7	68	mutex_unlock(&table[subsys_id].mutex);
f9e815b3	69	}
e6a7d3c0	70	EXPORT_SYMBOL_GPL(nfnl_unlock);
f9e815b3	71
0eb5db7a	72	#ifdef CONFIG_PROVE_LOCKING
875e0829	73	bool lockdep_nfnl_is_held(u8 subsys_id)
0eb5db7a PM	74	{
	75	return lockdep_is_held(&table[subsys_id].mutex);
	76	}
	77	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	78	#endif
	79
7c8d4cb4	80	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	81	{
7b7744e2 KC	82	u8 cb_id;
	83
	84	/* Sanity-check attr_count size to avoid stack buffer overflow. */
	85	for (cb_id = 0; cb_id < n->cb_count; cb_id++)
	86	if (WARN_ON(n->cb[cb_id].attr_count > NFNL_MAX_ATTR_COUNT))
	87	return -EINVAL;
	88
c14b78e7 PNA	89	nfnl_lock(n->subsys_id);
	90	if (table[n->subsys_id].subsys) {
	91	nfnl_unlock(n->subsys_id);
0ab43f84 HW	92	return -EBUSY;
0ab43f84 HW	93	}
c14b78e7 PNA	94	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	95	nfnl_unlock(n->subsys_id);
f9e815b3 HW	96
	97	return 0;
	98	}
f4bc177f	99	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	100
7c8d4cb4	101	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	102	{
c14b78e7 PNA	103	nfnl_lock(n->subsys_id);
	104	table[n->subsys_id].subsys = NULL;
	105	nfnl_unlock(n->subsys_id);
6b75e3e8	106	synchronize_rcu();
f9e815b3 HW	107	return 0;
f9e815b3 HW	108	}
f4bc177f	109	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	110
b745d035	111	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u16 type)
f9e815b3	112	{
b745d035	113	u8 subsys_id = NFNL_SUBSYS_ID(type);
f9e815b3	114
ac0f1d98	115	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	116	return NULL;
f9e815b3 HW	117
c14b78e7	118	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	119	}
f9e815b3 HW	120
7c8d4cb4	121	static inline const struct nfnl_callback *
b745d035	122	nfnetlink_find_client(u16 type, const struct nfnetlink_subsystem *ss)
f9e815b3	123	{
b745d035	124	u8 cb_id = NFNL_MSG_TYPE(type);
601e68e1	125
67ca3966	126	if (cb_id >= ss->cb_count)
f9e815b3	127	return NULL;
f9e815b3 HW	128
	129	return &ss->cb[cb_id];
	130	}
	131
cd8c20b6	132	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	133	{
cd8c20b6	134	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	135	}
	136	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	137
ec464e5d	138	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	139	unsigned int group, int echo, gfp_t flags)
f9e815b3	140	{
ec464e5d	141	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	142	}
f4bc177f	143	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	144
ec464e5d	145	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	146	{
ec464e5d	147	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	148	}
	149	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	150
ec464e5d PM	151	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	152	int flags)
f9e815b3	153	{
ec464e5d	154	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	155	}
f4bc177f	156	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	157
f9e815b3 HW	158	/* Process one complete nfnetlink message. */
2d4bc933 JB	159	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh,
2d4bc933 JB	160	struct netlink_ext_ack *extack)
f9e815b3	161	{
cd8c20b6	162	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	163	const struct nfnl_callback *nc;
7c8d4cb4 PM	164	const struct nfnetlink_subsystem *ss;
1d00a4eb	165	int type, err;
f9e815b3	166
f9e815b3	167	/* All the messages must at least contain nfgenmsg */
573ce260	168	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	169	return 0;
f9e815b3 HW	170
f9e815b3 HW	171	type = nlh->nlmsg_type;
e6a7d3c0	172	replay:
6b75e3e8	173	rcu_read_lock();
f9e815b3	174	ss = nfnetlink_get_subsys(type);
0ab43f84	175	if (!ss) {
95a5afca	176	#ifdef CONFIG_MODULES
6b75e3e8	177	rcu_read_unlock();
37d2e7a2	178	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	179	rcu_read_lock();
37d2e7a2	180	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	181	if (!ss)
0ab43f84 HW	182	#endif
6b75e3e8 ED	183	{
6b75e3e8 ED	184	rcu_read_unlock();
1d00a4eb	185	return -EINVAL;
6b75e3e8	186	}
0ab43f84	187	}
f9e815b3 HW	188
f9e815b3 HW	189	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	190	if (!nc) {
6b75e3e8 ED	191	rcu_read_unlock();
1d00a4eb	192	return -EINVAL;
6b75e3e8	193	}
f9e815b3	194
f9e815b3	195	{
573ce260	196	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
b745d035	197	u8 cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
7b7744e2	198	struct nlattr *cda[NFNL_MAX_ATTR_COUNT + 1];
f49c857f PNA	199	struct nlattr attr = (void )nlh + min_len;
f49c857f PNA	200	int attrlen = nlh->nlmsg_len - min_len;
c14b78e7	201	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f	202
7b7744e2 KC	203	/* Sanity-check NFNL_MAX_ATTR_COUNT */
	204	if (ss->cb[cb_id].attr_count > NFNL_MAX_ATTR_COUNT) {
	205	rcu_read_unlock();
	206	return -ENOMEM;
	207	}
	208
8cb08174 JB	209	err = nla_parse_deprecated(cda, ss->cb[cb_id].attr_count,
	210	attr, attrlen,
	211	ss->cb[cb_id].policy, extack);
4009e188 TB	212	if (err < 0) {
4009e188 TB	213	rcu_read_unlock();
f49c857f	214	return err;
4009e188	215	}
601e68e1	216
6b75e3e8	217	if (nc->call_rcu) {
7b8002a1	218	err = nc->call_rcu(net, net->nfnl, skb, nlh,
04ba724b PNA	219	(const struct nlattr **)cda,
04ba724b PNA	220	extack);
6b75e3e8 ED	221	rcu_read_unlock();
	222	} else {
	223	rcu_read_unlock();
c14b78e7	224	nfnl_lock(subsys_id);
9c55d3b5	225	if (nfnl_dereference_protected(subsys_id) != ss \|\|
6b75e3e8 ED	226	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	227	err = -EAGAIN;
59560a38	228	else if (nc->call)
7b8002a1	229	err = nc->call(net, net->nfnl, skb, nlh,
04ba724b PNA	230	(const struct nlattr **)cda,
04ba724b PNA	231	extack);
59560a38 TB	232	else
59560a38 TB	233	err = -EINVAL;
c14b78e7	234	nfnl_unlock(subsys_id);
6b75e3e8	235	}
e6a7d3c0 PNA	236	if (err == -EAGAIN)
	237	goto replay;
	238	return err;
f9e815b3	239	}
f9e815b3 HW	240	}
f9e815b3 HW	241
cbb8125e PNA	242	struct nfnl_err {
	243	struct list_head head;
	244	struct nlmsghdr *nlh;
	245	int err;
04ba724b	246	struct netlink_ext_ack extack;
cbb8125e PNA	247	};
cbb8125e PNA	248
04ba724b PNA	249	static int nfnl_err_add(struct list_head list, struct nlmsghdr nlh, int err,
04ba724b PNA	250	const struct netlink_ext_ack *extack)
cbb8125e PNA	251	{
	252	struct nfnl_err *nfnl_err;
	253
	254	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	255	if (nfnl_err == NULL)
	256	return -ENOMEM;
	257
	258	nfnl_err->nlh = nlh;
	259	nfnl_err->err = err;
04ba724b	260	nfnl_err->extack = *extack;
cbb8125e PNA	261	list_add_tail(&nfnl_err->head, list);
	262
	263	return 0;
	264	}
	265
	266	static void nfnl_err_del(struct nfnl_err *nfnl_err)
	267	{
	268	list_del(&nfnl_err->head);
	269	kfree(nfnl_err);
	270	}
	271
	272	static void nfnl_err_reset(struct list_head *err_list)
	273	{
	274	struct nfnl_err nfnl_err, next;
	275
	276	list_for_each_entry_safe(nfnl_err, next, err_list, head)
	277	nfnl_err_del(nfnl_err);
	278	}
	279
	280	static void nfnl_err_deliver(struct list_head err_list, struct sk_buff skb)
	281	{
	282	struct nfnl_err nfnl_err, next;
	283
	284	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
04ba724b PNA	285	netlink_ack(skb, nfnl_err->nlh, nfnl_err->err,
04ba724b PNA	286	&nfnl_err->extack);
cbb8125e PNA	287	nfnl_err_del(nfnl_err);
	288	}
	289	}
	290
6742b9e3 PNA	291	enum {
	292	NFNL_BATCH_FAILURE = (1 << 0),
	293	NFNL_BATCH_DONE = (1 << 1),
	294	NFNL_BATCH_REPLAY = (1 << 2),
	295	};
	296
0628b123	297	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
8c4d4e8b	298	u16 subsys_id, u32 genid)
0628b123	299	{
0f816232	300	struct sk_buff *oskb = skb;
0628b123 PNA	301	struct net *net = sock_net(skb->sk);
	302	const struct nfnetlink_subsystem *ss;
	303	const struct nfnl_callback *nc;
04ba724b	304	struct netlink_ext_ack extack;
4eba8b78	305	LIST_HEAD(err_list);
6742b9e3	306	u32 status;
0628b123 PNA	307	int err;
	308
	309	if (subsys_id >= NFNL_SUBSYS_COUNT)
2d4bc933	310	return netlink_ack(skb, nlh, -EINVAL, NULL);
0628b123	311	replay:
6742b9e3 PNA	312	status = 0;
6742b9e3 PNA	313
0f816232 DJ	314	skb = netlink_skb_clone(oskb, GFP_KERNEL);
0f816232 DJ	315	if (!skb)
2d4bc933	316	return netlink_ack(oskb, nlh, -ENOMEM, NULL);
0628b123	317
0628b123	318	nfnl_lock(subsys_id);
9c55d3b5	319	ss = nfnl_dereference_protected(subsys_id);
0628b123 PNA	320	if (!ss) {
	321	#ifdef CONFIG_MODULES
	322	nfnl_unlock(subsys_id);
	323	request_module("nfnetlink-subsys-%d", subsys_id);
	324	nfnl_lock(subsys_id);
9c55d3b5	325	ss = nfnl_dereference_protected(subsys_id);
0628b123 PNA	326	if (!ss)
	327	#endif
	328	{
	329	nfnl_unlock(subsys_id);
2d4bc933	330	netlink_ack(oskb, nlh, -EOPNOTSUPP, NULL);
0f816232	331	return kfree_skb(skb);
0628b123 PNA	332	}
	333	}
	334
ca2f18be	335	if (!ss->valid_genid \|\| !ss->commit \|\| !ss->abort) {
0628b123	336	nfnl_unlock(subsys_id);
2d4bc933	337	netlink_ack(oskb, nlh, -EOPNOTSUPP, NULL);
ecd15dd7	338	return kfree_skb(skb);
0628b123 PNA	339	}
0628b123 PNA	340
be2ab5b4 FW	341	if (!try_module_get(ss->owner)) {
	342	nfnl_unlock(subsys_id);
	343	netlink_ack(oskb, nlh, -EOPNOTSUPP, NULL);
	344	return kfree_skb(skb);
	345	}
	346
ca2f18be	347	if (!ss->valid_genid(net, genid)) {
be2ab5b4	348	module_put(ss->owner);
8c4d4e8b	349	nfnl_unlock(subsys_id);
2d4bc933	350	netlink_ack(oskb, nlh, -ERESTART, NULL);
8c4d4e8b PNA	351	return kfree_skb(skb);
	352	}
	353
f102d66b FW	354	nfnl_unlock(subsys_id);
f102d66b FW	355
0628b123 PNA	356	while (skb->len >= nlmsg_total_size(0)) {
	357	int msglen, type;
	358
8a3d4c36 FW	359	if (fatal_signal_pending(current)) {
	360	nfnl_err_reset(&err_list);
	361	err = -EINTR;
	362	status = NFNL_BATCH_FAILURE;
	363	goto done;
	364	}
	365
04ba724b	366	memset(&extack, 0, sizeof(extack));
0628b123 PNA	367	nlh = nlmsg_hdr(skb);
	368	err = 0;
	369
c58d6c93 PT	370	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	371	skb->len < nlh->nlmsg_len \|\|
	372	nlmsg_len(nlh) < sizeof(struct nfgenmsg)) {
	373	nfnl_err_reset(&err_list);
	374	status \|= NFNL_BATCH_FAILURE;
	375	goto done;
0628b123 PNA	376	}
	377
	378	/* Only requests are handled by the kernel */
	379	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	380	err = -EINVAL;
	381	goto ack;
	382	}
	383
	384	type = nlh->nlmsg_type;
	385	if (type == NFNL_MSG_BATCH_BEGIN) {
	386	/* Malformed: Batch begin twice */
cbb8125e	387	nfnl_err_reset(&err_list);
6742b9e3	388	status \|= NFNL_BATCH_FAILURE;
0628b123 PNA	389	goto done;
0628b123 PNA	390	} else if (type == NFNL_MSG_BATCH_END) {
6742b9e3	391	status \|= NFNL_BATCH_DONE;
0628b123 PNA	392	goto done;
	393	} else if (type < NLMSG_MIN_TYPE) {
	394	err = -EINVAL;
	395	goto ack;
	396	}
	397
	398	/* We only accept a batch with messages for the same
	399	* subsystem.
	400	*/
	401	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	402	err = -EINVAL;
	403	goto ack;
	404	}
	405
	406	nc = nfnetlink_find_client(type, ss);
	407	if (!nc) {
	408	err = -EINVAL;
	409	goto ack;
	410	}
	411
	412	{
	413	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
b745d035	414	u8 cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
7b7744e2	415	struct nlattr *cda[NFNL_MAX_ATTR_COUNT + 1];
0628b123 PNA	416	struct nlattr attr = (void )nlh + min_len;
	417	int attrlen = nlh->nlmsg_len - min_len;
	418
7b7744e2 KC	419	/* Sanity-check NFTA_MAX_ATTR */
	420	if (ss->cb[cb_id].attr_count > NFNL_MAX_ATTR_COUNT) {
	421	err = -ENOMEM;
	422	goto ack;
	423	}
	424
8cb08174 JB	425	err = nla_parse_deprecated(cda,
	426	ss->cb[cb_id].attr_count,
	427	attr, attrlen,
	428	ss->cb[cb_id].policy, NULL);
0628b123 PNA	429	if (err < 0)
	430	goto ack;
	431
	432	if (nc->call_batch) {
633c9a84	433	err = nc->call_batch(net, net->nfnl, skb, nlh,
04ba724b PNA	434	(const struct nlattr **)cda,
04ba724b PNA	435	&extack);
0628b123 PNA	436	}
	437
	438	/* The lock was released to autoload some module, we
	439	* have to abort and start from scratch using the
	440	* original skb.
	441	*/
	442	if (err == -EAGAIN) {
6742b9e3	443	status \|= NFNL_BATCH_REPLAY;
71ad00c5	444	goto done;
0628b123 PNA	445	}
	446	}
	447	ack:
	448	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
cbb8125e PNA	449	/* Errors are delivered once the full batch has been
	450	* processed, this avoids that the same error is
	451	* reported several times when replaying the batch.
	452	*/
04ba724b	453	if (nfnl_err_add(&err_list, nlh, err, &extack) < 0) {
cbb8125e PNA	454	/* We failed to enqueue an error, reset the
	455	* list of errors and send OOM to userspace
	456	* pointing to the batch header.
	457	*/
	458	nfnl_err_reset(&err_list);
2d4bc933 JB	459	netlink_ack(oskb, nlmsg_hdr(oskb), -ENOMEM,
2d4bc933 JB	460	NULL);
6742b9e3	461	status \|= NFNL_BATCH_FAILURE;
cbb8125e PNA	462	goto done;
cbb8125e PNA	463	}
0628b123 PNA	464	/* We don't stop processing the batch on errors, thus,
	465	* userspace gets all the errors that the batch
	466	* triggers.
	467	*/
0628b123	468	if (err)
6742b9e3	469	status \|= NFNL_BATCH_FAILURE;
0628b123	470	}
71ad00c5	471
0628b123 PNA	472	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	473	if (msglen > skb->len)
	474	msglen = skb->len;
	475	skb_pull(skb, msglen);
	476	}
	477	done:
6742b9e3	478	if (status & NFNL_BATCH_REPLAY) {
f102d66b	479	ss->abort(net, oskb);
6742b9e3	480	nfnl_err_reset(&err_list);
6742b9e3	481	kfree_skb(skb);
be2ab5b4	482	module_put(ss->owner);
6742b9e3 PNA	483	goto replay;
6742b9e3 PNA	484	} else if (status == NFNL_BATCH_DONE) {
00308791 FW	485	err = ss->commit(net, oskb);
	486	if (err == -EAGAIN) {
	487	status \|= NFNL_BATCH_REPLAY;
	488	goto done;
	489	} else if (err) {
	490	ss->abort(net, oskb);
	491	netlink_ack(oskb, nlmsg_hdr(oskb), err, NULL);
	492	}
6742b9e3	493	} else {
5913beaf	494	ss->abort(net, oskb);
6742b9e3	495	}
a654de8f PNA	496	if (ss->cleanup)
a654de8f PNA	497	ss->cleanup(net);
0628b123	498
cbb8125e	499	nfnl_err_deliver(&err_list, oskb);
0f816232	500	kfree_skb(skb);
be2ab5b4	501	module_put(ss->owner);
0628b123 PNA	502	}
0628b123 PNA	503
8c4d4e8b PNA	504	static const struct nla_policy nfnl_batch_policy[NFNL_BATCH_MAX + 1] = {
	505	[NFNL_BATCH_GENID] = { .type = NLA_U32 },
	506	};
	507
48656835	508	static void nfnetlink_rcv_skb_batch(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	509	{
8c4d4e8b PNA	510	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	511	struct nlattr attr = (void )nlh + min_len;
	512	struct nlattr *cda[NFNL_BATCH_MAX + 1];
	513	int attrlen = nlh->nlmsg_len - min_len;
48656835	514	struct nfgenmsg *nfgenmsg;
8c4d4e8b PNA	515	int msglen, err;
8c4d4e8b PNA	516	u32 gen_id = 0;
b745d035	517	u16 res_id;
0628b123	518
48656835 PNA	519	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	520	if (msglen > skb->len)
	521	msglen = skb->len;
	522
f55ce7b0	523	if (skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
48656835 PNA	524	return;
48656835 PNA	525
8cb08174 JB	526	err = nla_parse_deprecated(cda, NFNL_BATCH_MAX, attr, attrlen,
8cb08174 JB	527	nfnl_batch_policy, NULL);
8c4d4e8b	528	if (err < 0) {
2d4bc933	529	netlink_ack(skb, nlh, err, NULL);
8c4d4e8b PNA	530	return;
	531	}
	532	if (cda[NFNL_BATCH_GENID])
	533	gen_id = ntohl(nla_get_be32(cda[NFNL_BATCH_GENID]));
	534
48656835 PNA	535	nfgenmsg = nlmsg_data(nlh);
	536	skb_pull(skb, msglen);
	537	/* Work around old nft using host byte order */
	538	if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
	539	res_id = NFNL_SUBSYS_NFTABLES;
	540	else
	541	res_id = ntohs(nfgenmsg->res_id);
	542
8c4d4e8b	543	nfnetlink_rcv_batch(skb, nlh, res_id, gen_id);
48656835 PNA	544	}
	545
	546	static void nfnetlink_rcv(struct sk_buff *skb)
	547	{
	548	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	549
f55ce7b0 MJ	550	if (skb->len < NLMSG_HDRLEN \|\|
f55ce7b0 MJ	551	nlh->nlmsg_len < NLMSG_HDRLEN \|\|
0628b123 PNA	552	skb->len < nlh->nlmsg_len)
	553	return;
	554
90f62cf3	555	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
2d4bc933	556	netlink_ack(skb, nlh, -EPERM, NULL);
cdbe7c2d JB	557	return;
	558	}
	559
48656835 PNA	560	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN)
	561	nfnetlink_rcv_skb_batch(skb, nlh);
	562	else
d4ef3835	563	netlink_rcv_skb(skb, nfnetlink_rcv_msg);
f9e815b3 HW	564	}
f9e815b3 HW	565
03292745	566	#ifdef CONFIG_MODULES
023e2cfa	567	static int nfnetlink_bind(struct net *net, int group)
03292745 PNA	568	{
03292745 PNA	569	const struct nfnetlink_subsystem *ss;
97840cb6 PNA	570	int type;
	571
	572	if (group <= NFNLGRP_NONE \|\| group > NFNLGRP_MAX)
62924af2	573	return 0;
97840cb6 PNA	574
97840cb6 PNA	575	type = nfnl_group2type[group];
03292745 PNA	576
03292745 PNA	577	rcu_read_lock();
dbc3617f	578	ss = nfnetlink_get_subsys(type << 8);
03292745	579	rcu_read_unlock();
bfe4bc71	580	if (!ss)
1b0890cd	581	request_module_nowait("nfnetlink-subsys-%d", type);
4f520900	582	return 0;
03292745 PNA	583	}
	584	#endif
	585
cd8c20b6	586	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	587	{
cd8c20b6	588	struct sock *nfnl;
a31f2d17 PNA	589	struct netlink_kernel_cfg cfg = {
	590	.groups = NFNLGRP_MAX,
	591	.input = nfnetlink_rcv,
03292745 PNA	592	#ifdef CONFIG_MODULES
	593	.bind = nfnetlink_bind,
	594	#endif
a31f2d17	595	};
cd8c20b6	596
9f00d977	597	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	598	if (!nfnl)
	599	return -ENOMEM;
	600	net->nfnl_stash = nfnl;
cf778b00	601	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	602	return 0;
f9e815b3 HW	603	}
f9e815b3 HW	604
cd8c20b6	605	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	606	{
cd8c20b6	607	struct net *net;
f9e815b3	608
cd8c20b6	609	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	610	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	611	synchronize_net();
	612	list_for_each_entry(net, net_exit_list, exit_list)
	613	netlink_kernel_release(net->nfnl_stash);
	614	}
f9e815b3	615
cd8c20b6 AD	616	static struct pernet_operations nfnetlink_net_ops = {
	617	.init = nfnetlink_net_init,
	618	.exit_batch = nfnetlink_net_exit_batch,
	619	};
	620
	621	static int __init nfnetlink_init(void)
	622	{
c14b78e7 PNA	623	int i;
c14b78e7 PNA	624
97840cb6 PNA	625	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
	626	BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);
	627
c14b78e7 PNA	628	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	629	mutex_init(&table[i].mutex);
	630
cd8c20b6	631	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	632	}
f9e815b3 HW	633
cd8c20b6 AD	634	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	635	{
cd8c20b6 AD	636	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	637	}
f9e815b3 HW	638	module_init(nfnetlink_init);
f9e815b3 HW	639	module_exit(nfnetlink_exit);