[deliverable/linux.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
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)
{
	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);
	}
}

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,
				u_int16_t subsys_id)
{
	struct sk_buff *oskb = skb;
	struct net *net = sock_net(skb->sk);
	const struct nfnetlink_subsystem *ss;
	const struct nfnl_callback *nc;
	static LIST_HEAD(err_list);
	u32 status;
	int err;

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

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

	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(skb);
		}
	}

	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 (nlmsg_len(nlh) < sizeof(struct nfgenmsg) ||
		    skb->len < nlh->nlmsg_len) {
			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);
			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));
			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, 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) {
				status |= NFNL_BATCH_REPLAY;
				goto next;
			}
		}
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);
				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;
		}
next:
		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(oskb);
		nfnl_err_reset(&err_list);
		nfnl_unlock(subsys_id);
		kfree_skb(skb);
		goto replay;
	} else if (status == NFNL_BATCH_DONE) {
		ss->commit(oskb);
	} else {
		ss->abort(oskb);
	}

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

static void nfnetlink_rcv(struct sk_buff *skb)
{
	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	u_int16_t res_id;
	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);
		/* 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);
	} 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("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
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>
67ca3966	6	* (C) 2005,2007 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 HW	24	#include <linux/skbuff.h>
f9e815b3 HW	25	#include <asm/uaccess.h>
f9e815b3 HW	26	#include <net/sock.h>
f9e815b3 HW	27	#include <linux/init.h>
f9e815b3	28
573ce260	29	#include <net/netlink.h>
f9e815b3 HW	30	#include <linux/netfilter/nfnetlink.h>
	31
	32	MODULE_LICENSE("GPL");
4fdb3bb7 HW	33	MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
4fdb3bb7 HW	34	MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
f9e815b3 HW	35
	36	static char __initdata nfversion[] = "0.30";
	37
c14b78e7 PNA	38	static struct {
	39	struct mutex mutex;
	40	const struct nfnetlink_subsystem __rcu *subsys;
	41	} table[NFNL_SUBSYS_COUNT];
f9e815b3	42
03292745 PNA	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,
97840cb6 PNA	50	[NFNLGRP_NFTABLES] = NFNL_SUBSYS_NFTABLES,
97840cb6 PNA	51	[NFNLGRP_ACCT_QUOTA] = NFNL_SUBSYS_ACCT,
03292745 PNA	52	};
03292745 PNA	53
c14b78e7	54	void nfnl_lock(__u8 subsys_id)
f9e815b3	55	{
c14b78e7	56	mutex_lock(&table[subsys_id].mutex);
f9e815b3	57	}
e6a7d3c0	58	EXPORT_SYMBOL_GPL(nfnl_lock);
f9e815b3	59
c14b78e7	60	void nfnl_unlock(__u8 subsys_id)
a3c5029c	61	{
c14b78e7	62	mutex_unlock(&table[subsys_id].mutex);
f9e815b3	63	}
e6a7d3c0	64	EXPORT_SYMBOL_GPL(nfnl_unlock);
f9e815b3	65
0eb5db7a	66	#ifdef CONFIG_PROVE_LOCKING
875e0829	67	bool lockdep_nfnl_is_held(u8 subsys_id)
0eb5db7a PM	68	{
	69	return lockdep_is_held(&table[subsys_id].mutex);
	70	}
	71	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	72	#endif
	73
7c8d4cb4	74	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	75	{
c14b78e7 PNA	76	nfnl_lock(n->subsys_id);
	77	if (table[n->subsys_id].subsys) {
	78	nfnl_unlock(n->subsys_id);
0ab43f84 HW	79	return -EBUSY;
0ab43f84 HW	80	}
c14b78e7 PNA	81	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	82	nfnl_unlock(n->subsys_id);
f9e815b3 HW	83
	84	return 0;
	85	}
f4bc177f	86	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	87
7c8d4cb4	88	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	89	{
c14b78e7 PNA	90	nfnl_lock(n->subsys_id);
	91	table[n->subsys_id].subsys = NULL;
	92	nfnl_unlock(n->subsys_id);
6b75e3e8	93	synchronize_rcu();
f9e815b3 HW	94	return 0;
f9e815b3 HW	95	}
f4bc177f	96	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	97
7c8d4cb4	98	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
f9e815b3 HW	99	{
	100	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	101
ac0f1d98	102	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	103	return NULL;
f9e815b3 HW	104
c14b78e7	105	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	106	}
f9e815b3 HW	107
7c8d4cb4 PM	108	static inline const struct nfnl_callback *
7c8d4cb4 PM	109	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
f9e815b3 HW	110	{
f9e815b3 HW	111	u_int8_t cb_id = NFNL_MSG_TYPE(type);
601e68e1	112
67ca3966	113	if (cb_id >= ss->cb_count)
f9e815b3	114	return NULL;
f9e815b3 HW	115
	116	return &ss->cb[cb_id];
	117	}
	118
cd8c20b6	119	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	120	{
cd8c20b6	121	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	122	}
	123	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	124
3ab1f683 PM	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
ec464e5d	132	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	133	unsigned int group, int echo, gfp_t flags)
f9e815b3	134	{
ec464e5d	135	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	136	}
f4bc177f	137	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	138
ec464e5d	139	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	140	{
ec464e5d	141	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	142	}
	143	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	144
ec464e5d PM	145	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	146	int flags)
f9e815b3	147	{
ec464e5d	148	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	149	}
f4bc177f	150	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	151
f9e815b3 HW	152	/* Process one complete nfnetlink message. */
1d00a4eb	153	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	154	{
cd8c20b6	155	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	156	const struct nfnl_callback *nc;
7c8d4cb4 PM	157	const struct nfnetlink_subsystem *ss;
1d00a4eb	158	int type, err;
f9e815b3	159
f9e815b3	160	/* All the messages must at least contain nfgenmsg */
573ce260	161	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	162	return 0;
f9e815b3 HW	163
f9e815b3 HW	164	type = nlh->nlmsg_type;
e6a7d3c0	165	replay:
6b75e3e8	166	rcu_read_lock();
f9e815b3	167	ss = nfnetlink_get_subsys(type);
0ab43f84	168	if (!ss) {
95a5afca	169	#ifdef CONFIG_MODULES
6b75e3e8	170	rcu_read_unlock();
37d2e7a2	171	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	172	rcu_read_lock();
37d2e7a2	173	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	174	if (!ss)
0ab43f84 HW	175	#endif
6b75e3e8 ED	176	{
6b75e3e8 ED	177	rcu_read_unlock();
1d00a4eb	178	return -EINVAL;
6b75e3e8	179	}
0ab43f84	180	}
f9e815b3 HW	181
f9e815b3 HW	182	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	183	if (!nc) {
6b75e3e8 ED	184	rcu_read_unlock();
1d00a4eb	185	return -EINVAL;
6b75e3e8	186	}
f9e815b3	187
f9e815b3	188	{
573ce260	189	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
e3730578	190	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
f49c857f PNA	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;
c14b78e7	194	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f PNA	195
	196	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	197	attr, attrlen, ss->cb[cb_id].policy);
4009e188 TB	198	if (err < 0) {
4009e188 TB	199	rcu_read_unlock();
f49c857f	200	return err;
4009e188	201	}
601e68e1	202
6b75e3e8 ED	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();
c14b78e7 PNA	209	nfnl_lock(subsys_id);
c14b78e7 PNA	210	if (rcu_dereference_protected(table[subsys_id].subsys,
9df9e783	211	lockdep_is_held(&table[subsys_id].mutex)) != ss \|\|
6b75e3e8 ED	212	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	213	err = -EAGAIN;
59560a38	214	else if (nc->call)
6b75e3e8 ED	215	err = nc->call(net->nfnl, skb, nlh,
6b75e3e8 ED	216	(const struct nlattr **)cda);
59560a38 TB	217	else
59560a38 TB	218	err = -EINVAL;
c14b78e7	219	nfnl_unlock(subsys_id);
6b75e3e8	220	}
e6a7d3c0 PNA	221	if (err == -EAGAIN)
	222	goto replay;
	223	return err;
f9e815b3	224	}
f9e815b3 HW	225	}
f9e815b3 HW	226
cbb8125e PNA	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
6742b9e3 PNA	272	enum {
	273	NFNL_BATCH_FAILURE = (1 << 0),
	274	NFNL_BATCH_DONE = (1 << 1),
	275	NFNL_BATCH_REPLAY = (1 << 2),
	276	};
	277
0628b123 PNA	278	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	279	u_int16_t subsys_id)
	280	{
0f816232	281	struct sk_buff *oskb = skb;
0628b123 PNA	282	struct net *net = sock_net(skb->sk);
	283	const struct nfnetlink_subsystem *ss;
	284	const struct nfnl_callback *nc;
cbb8125e	285	static LIST_HEAD(err_list);
6742b9e3	286	u32 status;
0628b123 PNA	287	int err;
	288
	289	if (subsys_id >= NFNL_SUBSYS_COUNT)
	290	return netlink_ack(skb, nlh, -EINVAL);
	291	replay:
6742b9e3 PNA	292	status = 0;
6742b9e3 PNA	293
0f816232 DJ	294	skb = netlink_skb_clone(oskb, GFP_KERNEL);
0f816232 DJ	295	if (!skb)
0628b123 PNA	296	return netlink_ack(oskb, nlh, -ENOMEM);
0628b123 PNA	297
0628b123 PNA	298	nfnl_lock(subsys_id);
	299	ss = rcu_dereference_protected(table[subsys_id].subsys,
	300	lockdep_is_held(&table[subsys_id].mutex));
	301	if (!ss) {
	302	#ifdef CONFIG_MODULES
	303	nfnl_unlock(subsys_id);
	304	request_module("nfnetlink-subsys-%d", subsys_id);
	305	nfnl_lock(subsys_id);
	306	ss = rcu_dereference_protected(table[subsys_id].subsys,
	307	lockdep_is_held(&table[subsys_id].mutex));
	308	if (!ss)
	309	#endif
	310	{
	311	nfnl_unlock(subsys_id);
ecd15dd7	312	netlink_ack(skb, nlh, -EOPNOTSUPP);
0f816232	313	return kfree_skb(skb);
0628b123 PNA	314	}
	315	}
	316
	317	if (!ss->commit \|\| !ss->abort) {
	318	nfnl_unlock(subsys_id);
ecd15dd7 DF	319	netlink_ack(skb, nlh, -EOPNOTSUPP);
ecd15dd7 DF	320	return kfree_skb(skb);
0628b123 PNA	321	}
	322
	323	while (skb->len >= nlmsg_total_size(0)) {
	324	int msglen, type;
	325
	326	nlh = nlmsg_hdr(skb);
	327	err = 0;
	328
9ea2aa8b PNA	329	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg) \|\|
9ea2aa8b PNA	330	skb->len < nlh->nlmsg_len) {
0628b123 PNA	331	err = -EINVAL;
	332	goto ack;
	333	}
	334
	335	/* Only requests are handled by the kernel */
	336	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	337	err = -EINVAL;
	338	goto ack;
	339	}
	340
	341	type = nlh->nlmsg_type;
	342	if (type == NFNL_MSG_BATCH_BEGIN) {
	343	/* Malformed: Batch begin twice */
cbb8125e	344	nfnl_err_reset(&err_list);
6742b9e3	345	status \|= NFNL_BATCH_FAILURE;
0628b123 PNA	346	goto done;
0628b123 PNA	347	} else if (type == NFNL_MSG_BATCH_END) {
6742b9e3	348	status \|= NFNL_BATCH_DONE;
0628b123 PNA	349	goto done;
	350	} else if (type < NLMSG_MIN_TYPE) {
	351	err = -EINVAL;
	352	goto ack;
	353	}
	354
	355	/* We only accept a batch with messages for the same
	356	* subsystem.
	357	*/
	358	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	359	err = -EINVAL;
	360	goto ack;
	361	}
	362
	363	nc = nfnetlink_find_client(type, ss);
	364	if (!nc) {
	365	err = -EINVAL;
	366	goto ack;
	367	}
	368
	369	{
	370	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	371	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	372	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	373	struct nlattr attr = (void )nlh + min_len;
	374	int attrlen = nlh->nlmsg_len - min_len;
	375
	376	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	377	attr, attrlen, ss->cb[cb_id].policy);
	378	if (err < 0)
	379	goto ack;
	380
	381	if (nc->call_batch) {
633c9a84	382	err = nc->call_batch(net, net->nfnl, skb, nlh,
0628b123 PNA	383	(const struct nlattr **)cda);
	384	}
	385
	386	/* The lock was released to autoload some module, we
	387	* have to abort and start from scratch using the
	388	* original skb.
	389	*/
	390	if (err == -EAGAIN) {
6742b9e3 PNA	391	status \|= NFNL_BATCH_REPLAY;
6742b9e3 PNA	392	goto next;
0628b123 PNA	393	}
	394	}
	395	ack:
	396	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
cbb8125e PNA	397	/* Errors are delivered once the full batch has been
	398	* processed, this avoids that the same error is
	399	* reported several times when replaying the batch.
	400	*/
	401	if (nfnl_err_add(&err_list, nlh, err) < 0) {
	402	/* We failed to enqueue an error, reset the
	403	* list of errors and send OOM to userspace
	404	* pointing to the batch header.
	405	*/
	406	nfnl_err_reset(&err_list);
	407	netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM);
6742b9e3	408	status \|= NFNL_BATCH_FAILURE;
cbb8125e PNA	409	goto done;
cbb8125e PNA	410	}
0628b123 PNA	411	/* We don't stop processing the batch on errors, thus,
	412	* userspace gets all the errors that the batch
	413	* triggers.
	414	*/
0628b123	415	if (err)
6742b9e3	416	status \|= NFNL_BATCH_FAILURE;
0628b123	417	}
6742b9e3	418	next:
0628b123 PNA	419	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	420	if (msglen > skb->len)
	421	msglen = skb->len;
	422	skb_pull(skb, msglen);
	423	}
	424	done:
6742b9e3 PNA	425	if (status & NFNL_BATCH_REPLAY) {
	426	ss->abort(oskb);
	427	nfnl_err_reset(&err_list);
	428	nfnl_unlock(subsys_id);
	429	kfree_skb(skb);
	430	goto replay;
	431	} else if (status == NFNL_BATCH_DONE) {
fc04733a	432	ss->commit(oskb);
6742b9e3	433	} else {
fc04733a	434	ss->abort(oskb);
6742b9e3	435	}
0628b123	436
cbb8125e	437	nfnl_err_deliver(&err_list, oskb);
0628b123	438	nfnl_unlock(subsys_id);
0f816232	439	kfree_skb(skb);
0628b123 PNA	440	}
0628b123 PNA	441
cd40b7d3	442	static void nfnetlink_rcv(struct sk_buff *skb)
f9e815b3	443	{
0628b123	444	struct nlmsghdr *nlh = nlmsg_hdr(skb);
a9de9777	445	u_int16_t res_id;
0628b123 PNA	446	int msglen;
0628b123 PNA	447
0628b123 PNA	448	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	449	skb->len < nlh->nlmsg_len)
	450	return;
	451
90f62cf3	452	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
cdbe7c2d JB	453	netlink_ack(skb, nlh, -EPERM);
	454	return;
	455	}
	456
0628b123 PNA	457	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	458	struct nfgenmsg *nfgenmsg;
	459
	460	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	461	if (msglen > skb->len)
	462	msglen = skb->len;
	463
	464	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	465	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	466	return;
	467
	468	nfgenmsg = nlmsg_data(nlh);
	469	skb_pull(skb, msglen);
a9de9777 PNA	470	/* Work around old nft using host byte order */
	471	if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
	472	res_id = NFNL_SUBSYS_NFTABLES;
	473	else
	474	res_id = ntohs(nfgenmsg->res_id);
	475	nfnetlink_rcv_batch(skb, nlh, res_id);
0628b123 PNA	476	} else {
	477	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	478	}
f9e815b3 HW	479	}
f9e815b3 HW	480
03292745	481	#ifdef CONFIG_MODULES
023e2cfa	482	static int nfnetlink_bind(struct net *net, int group)
03292745 PNA	483	{
03292745 PNA	484	const struct nfnetlink_subsystem *ss;
97840cb6 PNA	485	int type;
	486
	487	if (group <= NFNLGRP_NONE \|\| group > NFNLGRP_MAX)
62924af2	488	return 0;
97840cb6 PNA	489
97840cb6 PNA	490	type = nfnl_group2type[group];
03292745 PNA	491
03292745 PNA	492	rcu_read_lock();
dbc3617f	493	ss = nfnetlink_get_subsys(type << 8);
03292745	494	rcu_read_unlock();
bfe4bc71 RGB	495	if (!ss)
bfe4bc71 RGB	496	request_module("nfnetlink-subsys-%d", type);
4f520900	497	return 0;
03292745 PNA	498	}
	499	#endif
	500
cd8c20b6	501	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	502	{
cd8c20b6	503	struct sock *nfnl;
a31f2d17 PNA	504	struct netlink_kernel_cfg cfg = {
	505	.groups = NFNLGRP_MAX,
	506	.input = nfnetlink_rcv,
03292745 PNA	507	#ifdef CONFIG_MODULES
	508	.bind = nfnetlink_bind,
	509	#endif
a31f2d17	510	};
cd8c20b6	511
9f00d977	512	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	513	if (!nfnl)
	514	return -ENOMEM;
	515	net->nfnl_stash = nfnl;
cf778b00	516	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	517	return 0;
f9e815b3 HW	518	}
f9e815b3 HW	519
cd8c20b6	520	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	521	{
cd8c20b6	522	struct net *net;
f9e815b3	523
cd8c20b6	524	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	525	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	526	synchronize_net();
	527	list_for_each_entry(net, net_exit_list, exit_list)
	528	netlink_kernel_release(net->nfnl_stash);
	529	}
f9e815b3	530
cd8c20b6 AD	531	static struct pernet_operations nfnetlink_net_ops = {
	532	.init = nfnetlink_net_init,
	533	.exit_batch = nfnetlink_net_exit_batch,
	534	};
	535
	536	static int __init nfnetlink_init(void)
	537	{
c14b78e7 PNA	538	int i;
c14b78e7 PNA	539
97840cb6 PNA	540	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
	541	BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);
	542
c14b78e7 PNA	543	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	544	mutex_init(&table[i].mutex);
	545
654d0fbd	546	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
cd8c20b6	547	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	548	}
f9e815b3 HW	549
cd8c20b6 AD	550	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	551	{
654d0fbd	552	pr_info("Removing netfilter NETLINK layer.\n");
cd8c20b6 AD	553	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	554	}
f9e815b3 HW	555	module_init(nfnetlink_init);
f9e815b3 HW	556	module_exit(nfnetlink_exit);