[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,
	[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)
{
	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);

	skb->sk = oskb->sk;

	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->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,
33d5a7b1	52	[NFNLGRP_NFTRACE] = NFNL_SUBSYS_NFTABLES,
03292745 PNA	53	};
03292745 PNA	54
c14b78e7	55	void nfnl_lock(__u8 subsys_id)
f9e815b3	56	{
c14b78e7	57	mutex_lock(&table[subsys_id].mutex);
f9e815b3	58	}
e6a7d3c0	59	EXPORT_SYMBOL_GPL(nfnl_lock);
f9e815b3	60
c14b78e7	61	void nfnl_unlock(__u8 subsys_id)
a3c5029c	62	{
c14b78e7	63	mutex_unlock(&table[subsys_id].mutex);
f9e815b3	64	}
e6a7d3c0	65	EXPORT_SYMBOL_GPL(nfnl_unlock);
f9e815b3	66
0eb5db7a	67	#ifdef CONFIG_PROVE_LOCKING
875e0829	68	bool lockdep_nfnl_is_held(u8 subsys_id)
0eb5db7a PM	69	{
	70	return lockdep_is_held(&table[subsys_id].mutex);
	71	}
	72	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	73	#endif
	74
7c8d4cb4	75	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	76	{
c14b78e7 PNA	77	nfnl_lock(n->subsys_id);
	78	if (table[n->subsys_id].subsys) {
	79	nfnl_unlock(n->subsys_id);
0ab43f84 HW	80	return -EBUSY;
0ab43f84 HW	81	}
c14b78e7 PNA	82	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	83	nfnl_unlock(n->subsys_id);
f9e815b3 HW	84
	85	return 0;
	86	}
f4bc177f	87	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	88
7c8d4cb4	89	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	90	{
c14b78e7 PNA	91	nfnl_lock(n->subsys_id);
	92	table[n->subsys_id].subsys = NULL;
	93	nfnl_unlock(n->subsys_id);
6b75e3e8	94	synchronize_rcu();
f9e815b3 HW	95	return 0;
f9e815b3 HW	96	}
f4bc177f	97	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	98
7c8d4cb4	99	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
f9e815b3 HW	100	{
	101	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	102
ac0f1d98	103	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	104	return NULL;
f9e815b3 HW	105
c14b78e7	106	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	107	}
f9e815b3 HW	108
7c8d4cb4 PM	109	static inline const struct nfnl_callback *
7c8d4cb4 PM	110	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
f9e815b3 HW	111	{
f9e815b3 HW	112	u_int8_t cb_id = NFNL_MSG_TYPE(type);
601e68e1	113
67ca3966	114	if (cb_id >= ss->cb_count)
f9e815b3	115	return NULL;
f9e815b3 HW	116
	117	return &ss->cb[cb_id];
	118	}
	119
cd8c20b6	120	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	121	{
cd8c20b6	122	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	123	}
	124	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	125
3ab1f683 PM	126	struct sk_buff nfnetlink_alloc_skb(struct net net, unsigned int size,
	127	u32 dst_portid, gfp_t gfp_mask)
	128	{
	129	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
	130	}
	131	EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);
	132
ec464e5d	133	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	134	unsigned int group, int echo, gfp_t flags)
f9e815b3	135	{
ec464e5d	136	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	137	}
f4bc177f	138	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	139
ec464e5d	140	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	141	{
ec464e5d	142	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	143	}
	144	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	145
ec464e5d PM	146	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	147	int flags)
f9e815b3	148	{
ec464e5d	149	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	150	}
f4bc177f	151	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	152
f9e815b3 HW	153	/* Process one complete nfnetlink message. */
1d00a4eb	154	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	155	{
cd8c20b6	156	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	157	const struct nfnl_callback *nc;
7c8d4cb4 PM	158	const struct nfnetlink_subsystem *ss;
1d00a4eb	159	int type, err;
f9e815b3	160
f9e815b3	161	/* All the messages must at least contain nfgenmsg */
573ce260	162	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	163	return 0;
f9e815b3 HW	164
f9e815b3 HW	165	type = nlh->nlmsg_type;
e6a7d3c0	166	replay:
6b75e3e8	167	rcu_read_lock();
f9e815b3	168	ss = nfnetlink_get_subsys(type);
0ab43f84	169	if (!ss) {
95a5afca	170	#ifdef CONFIG_MODULES
6b75e3e8	171	rcu_read_unlock();
37d2e7a2	172	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	173	rcu_read_lock();
37d2e7a2	174	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	175	if (!ss)
0ab43f84 HW	176	#endif
6b75e3e8 ED	177	{
6b75e3e8 ED	178	rcu_read_unlock();
1d00a4eb	179	return -EINVAL;
6b75e3e8	180	}
0ab43f84	181	}
f9e815b3 HW	182
f9e815b3 HW	183	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	184	if (!nc) {
6b75e3e8 ED	185	rcu_read_unlock();
1d00a4eb	186	return -EINVAL;
6b75e3e8	187	}
f9e815b3	188
f9e815b3	189	{
573ce260	190	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
e3730578	191	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
f49c857f PNA	192	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	193	struct nlattr attr = (void )nlh + min_len;
	194	int attrlen = nlh->nlmsg_len - min_len;
c14b78e7	195	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f PNA	196
	197	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	198	attr, attrlen, ss->cb[cb_id].policy);
4009e188 TB	199	if (err < 0) {
4009e188 TB	200	rcu_read_unlock();
f49c857f	201	return err;
4009e188	202	}
601e68e1	203
6b75e3e8 ED	204	if (nc->call_rcu) {
	205	err = nc->call_rcu(net->nfnl, skb, nlh,
	206	(const struct nlattr **)cda);
	207	rcu_read_unlock();
	208	} else {
	209	rcu_read_unlock();
c14b78e7 PNA	210	nfnl_lock(subsys_id);
c14b78e7 PNA	211	if (rcu_dereference_protected(table[subsys_id].subsys,
9df9e783	212	lockdep_is_held(&table[subsys_id].mutex)) != ss \|\|
6b75e3e8 ED	213	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	214	err = -EAGAIN;
59560a38	215	else if (nc->call)
6b75e3e8 ED	216	err = nc->call(net->nfnl, skb, nlh,
6b75e3e8 ED	217	(const struct nlattr **)cda);
59560a38 TB	218	else
59560a38 TB	219	err = -EINVAL;
c14b78e7	220	nfnl_unlock(subsys_id);
6b75e3e8	221	}
e6a7d3c0 PNA	222	if (err == -EAGAIN)
	223	goto replay;
	224	return err;
f9e815b3	225	}
f9e815b3 HW	226	}
f9e815b3 HW	227
cbb8125e PNA	228	struct nfnl_err {
	229	struct list_head head;
	230	struct nlmsghdr *nlh;
	231	int err;
	232	};
	233
	234	static int nfnl_err_add(struct list_head list, struct nlmsghdr nlh, int err)
	235	{
	236	struct nfnl_err *nfnl_err;
	237
	238	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	239	if (nfnl_err == NULL)
	240	return -ENOMEM;
	241
	242	nfnl_err->nlh = nlh;
	243	nfnl_err->err = err;
	244	list_add_tail(&nfnl_err->head, list);
	245
	246	return 0;
	247	}
	248
	249	static void nfnl_err_del(struct nfnl_err *nfnl_err)
	250	{
	251	list_del(&nfnl_err->head);
	252	kfree(nfnl_err);
	253	}
	254
	255	static void nfnl_err_reset(struct list_head *err_list)
	256	{
	257	struct nfnl_err nfnl_err, next;
	258
	259	list_for_each_entry_safe(nfnl_err, next, err_list, head)
	260	nfnl_err_del(nfnl_err);
	261	}
	262
	263	static void nfnl_err_deliver(struct list_head err_list, struct sk_buff skb)
	264	{
	265	struct nfnl_err nfnl_err, next;
	266
	267	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
	268	netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
	269	nfnl_err_del(nfnl_err);
	270	}
	271	}
	272
6742b9e3 PNA	273	enum {
	274	NFNL_BATCH_FAILURE = (1 << 0),
	275	NFNL_BATCH_DONE = (1 << 1),
	276	NFNL_BATCH_REPLAY = (1 << 2),
	277	};
	278
0628b123 PNA	279	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	280	u_int16_t subsys_id)
	281	{
0f816232	282	struct sk_buff *oskb = skb;
0628b123 PNA	283	struct net *net = sock_net(skb->sk);
	284	const struct nfnetlink_subsystem *ss;
	285	const struct nfnl_callback *nc;
cbb8125e	286	static LIST_HEAD(err_list);
6742b9e3	287	u32 status;
0628b123 PNA	288	int err;
	289
	290	if (subsys_id >= NFNL_SUBSYS_COUNT)
	291	return netlink_ack(skb, nlh, -EINVAL);
	292	replay:
6742b9e3 PNA	293	status = 0;
6742b9e3 PNA	294
0f816232 DJ	295	skb = netlink_skb_clone(oskb, GFP_KERNEL);
0f816232 DJ	296	if (!skb)
0628b123 PNA	297	return netlink_ack(oskb, nlh, -ENOMEM);
0628b123 PNA	298
0f816232	299	skb->sk = oskb->sk;
0628b123 PNA	300
	301	nfnl_lock(subsys_id);
	302	ss = rcu_dereference_protected(table[subsys_id].subsys,
	303	lockdep_is_held(&table[subsys_id].mutex));
	304	if (!ss) {
	305	#ifdef CONFIG_MODULES
	306	nfnl_unlock(subsys_id);
	307	request_module("nfnetlink-subsys-%d", subsys_id);
	308	nfnl_lock(subsys_id);
	309	ss = rcu_dereference_protected(table[subsys_id].subsys,
	310	lockdep_is_held(&table[subsys_id].mutex));
	311	if (!ss)
	312	#endif
	313	{
	314	nfnl_unlock(subsys_id);
ecd15dd7	315	netlink_ack(skb, nlh, -EOPNOTSUPP);
0f816232	316	return kfree_skb(skb);
0628b123 PNA	317	}
	318	}
	319
	320	if (!ss->commit \|\| !ss->abort) {
	321	nfnl_unlock(subsys_id);
ecd15dd7 DF	322	netlink_ack(skb, nlh, -EOPNOTSUPP);
ecd15dd7 DF	323	return kfree_skb(skb);
0628b123 PNA	324	}
	325
	326	while (skb->len >= nlmsg_total_size(0)) {
	327	int msglen, type;
	328
	329	nlh = nlmsg_hdr(skb);
	330	err = 0;
	331
9ea2aa8b PNA	332	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg) \|\|
9ea2aa8b PNA	333	skb->len < nlh->nlmsg_len) {
0628b123 PNA	334	err = -EINVAL;
	335	goto ack;
	336	}
	337
	338	/* Only requests are handled by the kernel */
	339	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	340	err = -EINVAL;
	341	goto ack;
	342	}
	343
	344	type = nlh->nlmsg_type;
	345	if (type == NFNL_MSG_BATCH_BEGIN) {
	346	/* Malformed: Batch begin twice */
cbb8125e	347	nfnl_err_reset(&err_list);
6742b9e3	348	status \|= NFNL_BATCH_FAILURE;
0628b123 PNA	349	goto done;
0628b123 PNA	350	} else if (type == NFNL_MSG_BATCH_END) {
6742b9e3	351	status \|= NFNL_BATCH_DONE;
0628b123 PNA	352	goto done;
	353	} else if (type < NLMSG_MIN_TYPE) {
	354	err = -EINVAL;
	355	goto ack;
	356	}
	357
	358	/* We only accept a batch with messages for the same
	359	* subsystem.
	360	*/
	361	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	362	err = -EINVAL;
	363	goto ack;
	364	}
	365
	366	nc = nfnetlink_find_client(type, ss);
	367	if (!nc) {
	368	err = -EINVAL;
	369	goto ack;
	370	}
	371
	372	{
	373	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	374	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	375	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	376	struct nlattr attr = (void )nlh + min_len;
	377	int attrlen = nlh->nlmsg_len - min_len;
	378
	379	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	380	attr, attrlen, ss->cb[cb_id].policy);
	381	if (err < 0)
	382	goto ack;
	383
	384	if (nc->call_batch) {
	385	err = nc->call_batch(net->nfnl, skb, nlh,
	386	(const struct nlattr **)cda);
	387	}
	388
	389	/* The lock was released to autoload some module, we
	390	* have to abort and start from scratch using the
	391	* original skb.
	392	*/
	393	if (err == -EAGAIN) {
6742b9e3 PNA	394	status \|= NFNL_BATCH_REPLAY;
6742b9e3 PNA	395	goto next;
0628b123 PNA	396	}
	397	}
	398	ack:
	399	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
cbb8125e PNA	400	/* Errors are delivered once the full batch has been
	401	* processed, this avoids that the same error is
	402	* reported several times when replaying the batch.
	403	*/
	404	if (nfnl_err_add(&err_list, nlh, err) < 0) {
	405	/* We failed to enqueue an error, reset the
	406	* list of errors and send OOM to userspace
	407	* pointing to the batch header.
	408	*/
	409	nfnl_err_reset(&err_list);
	410	netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM);
6742b9e3	411	status \|= NFNL_BATCH_FAILURE;
cbb8125e PNA	412	goto done;
cbb8125e PNA	413	}
0628b123 PNA	414	/* We don't stop processing the batch on errors, thus,
	415	* userspace gets all the errors that the batch
	416	* triggers.
	417	*/
0628b123	418	if (err)
6742b9e3	419	status \|= NFNL_BATCH_FAILURE;
0628b123	420	}
6742b9e3	421	next:
0628b123 PNA	422	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	423	if (msglen > skb->len)
	424	msglen = skb->len;
	425	skb_pull(skb, msglen);
	426	}
	427	done:
6742b9e3 PNA	428	if (status & NFNL_BATCH_REPLAY) {
	429	ss->abort(oskb);
	430	nfnl_err_reset(&err_list);
	431	nfnl_unlock(subsys_id);
	432	kfree_skb(skb);
	433	goto replay;
	434	} else if (status == NFNL_BATCH_DONE) {
fc04733a	435	ss->commit(oskb);
6742b9e3	436	} else {
fc04733a	437	ss->abort(oskb);
6742b9e3	438	}
0628b123	439
cbb8125e	440	nfnl_err_deliver(&err_list, oskb);
0628b123	441	nfnl_unlock(subsys_id);
0f816232	442	kfree_skb(skb);
0628b123 PNA	443	}
0628b123 PNA	444
cd40b7d3	445	static void nfnetlink_rcv(struct sk_buff *skb)
f9e815b3	446	{
0628b123	447	struct nlmsghdr *nlh = nlmsg_hdr(skb);
a9de9777	448	u_int16_t res_id;
0628b123 PNA	449	int msglen;
0628b123 PNA	450
0628b123 PNA	451	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	452	skb->len < nlh->nlmsg_len)
	453	return;
	454
90f62cf3	455	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
cdbe7c2d JB	456	netlink_ack(skb, nlh, -EPERM);
	457	return;
	458	}
	459
0628b123 PNA	460	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	461	struct nfgenmsg *nfgenmsg;
	462
	463	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	464	if (msglen > skb->len)
	465	msglen = skb->len;
	466
	467	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	468	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	469	return;
	470
	471	nfgenmsg = nlmsg_data(nlh);
	472	skb_pull(skb, msglen);
a9de9777 PNA	473	/* Work around old nft using host byte order */
	474	if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
	475	res_id = NFNL_SUBSYS_NFTABLES;
	476	else
	477	res_id = ntohs(nfgenmsg->res_id);
	478	nfnetlink_rcv_batch(skb, nlh, res_id);
0628b123 PNA	479	} else {
	480	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	481	}
f9e815b3 HW	482	}
f9e815b3 HW	483
03292745	484	#ifdef CONFIG_MODULES
023e2cfa	485	static int nfnetlink_bind(struct net *net, int group)
03292745 PNA	486	{
03292745 PNA	487	const struct nfnetlink_subsystem *ss;
97840cb6 PNA	488	int type;
	489
	490	if (group <= NFNLGRP_NONE \|\| group > NFNLGRP_MAX)
62924af2	491	return 0;
97840cb6 PNA	492
97840cb6 PNA	493	type = nfnl_group2type[group];
03292745 PNA	494
03292745 PNA	495	rcu_read_lock();
dbc3617f	496	ss = nfnetlink_get_subsys(type << 8);
03292745	497	rcu_read_unlock();
bfe4bc71 RGB	498	if (!ss)
bfe4bc71 RGB	499	request_module("nfnetlink-subsys-%d", type);
4f520900	500	return 0;
03292745 PNA	501	}
	502	#endif
	503
cd8c20b6	504	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	505	{
cd8c20b6	506	struct sock *nfnl;
a31f2d17 PNA	507	struct netlink_kernel_cfg cfg = {
	508	.groups = NFNLGRP_MAX,
	509	.input = nfnetlink_rcv,
03292745 PNA	510	#ifdef CONFIG_MODULES
	511	.bind = nfnetlink_bind,
	512	#endif
a31f2d17	513	};
cd8c20b6	514
9f00d977	515	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	516	if (!nfnl)
	517	return -ENOMEM;
	518	net->nfnl_stash = nfnl;
cf778b00	519	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	520	return 0;
f9e815b3 HW	521	}
f9e815b3 HW	522
cd8c20b6	523	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	524	{
cd8c20b6	525	struct net *net;
f9e815b3	526
cd8c20b6	527	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	528	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	529	synchronize_net();
	530	list_for_each_entry(net, net_exit_list, exit_list)
	531	netlink_kernel_release(net->nfnl_stash);
	532	}
f9e815b3	533
cd8c20b6 AD	534	static struct pernet_operations nfnetlink_net_ops = {
	535	.init = nfnetlink_net_init,
	536	.exit_batch = nfnetlink_net_exit_batch,
	537	};
	538
	539	static int __init nfnetlink_init(void)
	540	{
c14b78e7 PNA	541	int i;
c14b78e7 PNA	542
97840cb6 PNA	543	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
	544	BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);
	545
c14b78e7 PNA	546	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	547	mutex_init(&table[i].mutex);
	548
654d0fbd	549	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
cd8c20b6	550	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	551	}
f9e815b3 HW	552
cd8c20b6 AD	553	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	554	{
654d0fbd	555	pr_info("Removing netfilter NETLINK layer.\n");
cd8c20b6 AD	556	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	557	}
f9e815b3 HW	558	module_init(nfnetlink_init);
f9e815b3 HW	559	module_exit(nfnetlink_exit);