nf_conntrack_netlink.c

linux 内核源代码

/* Connection tracking via netlink socket. Allows for user space
 * protocol helpers and general trouble making from userspace.
 *
 * (C) 2001 by Jay Schulist <[e-mail]>
 * (C) 2002-2006 by Harald Welte <[e-mail]>
 * (C) 2003 by Patrick Mchardy <[e-mail]>
 * (C) 2005-2007 by Pablo Neira Ayuso <[e-mail]>
 *
 * Initial connection tracking 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/init.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/timer.h>
#include <linux/skbuff.h>
#include <linux/errno.h>
#include <linux/netlink.h>
#include <linux/spinlock.h>
#include <linux/interrupt.h>
#include <linux/notifier.h>

#include <linux/netfilter.h>
#include <net/netlink.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/nf_conntrack_expect.h>
#include <net/netfilter/nf_conntrack_helper.h>
#include <net/netfilter/nf_conntrack_l3proto.h>
#include <net/netfilter/nf_conntrack_l4proto.h>
#include <net/netfilter/nf_conntrack_tuple.h>
#ifdef CONFIG_NF_NAT_NEEDED
#include <net/netfilter/nf_nat_core.h>
#include <net/netfilter/nf_nat_protocol.h>
#endif

#include <linux/netfilter/nfnetlink.h>
#include <linux/netfilter/nfnetlink_conntrack.h>

MODULE_LICENSE("GPL");

static char __initdata version[] = "0.93";

static inline int
ctnetlink_dump_tuples_proto(struct sk_buff *skb,
			    const struct nf_conntrack_tuple *tuple,
			    struct nf_conntrack_l4proto *l4proto)
{
	int ret = 0;
	struct nlattr *nest_parms;

	nest_parms = nla_nest_start(skb, CTA_TUPLE_PROTO | NLA_F_NESTED);
	if (!nest_parms)
		goto nla_put_failure;
	NLA_PUT(skb, CTA_PROTO_NUM, sizeof(u_int8_t), &tuple->dst.protonum);

	if (likely(l4proto->tuple_to_nlattr))
		ret = l4proto->tuple_to_nlattr(skb, tuple);

	nla_nest_end(skb, nest_parms);

	return ret;

nla_put_failure:
	return -1;
}

static inline int
ctnetlink_dump_tuples_ip(struct sk_buff *skb,
			 const struct nf_conntrack_tuple *tuple,
			 struct nf_conntrack_l3proto *l3proto)
{
	int ret = 0;
	struct nlattr *nest_parms;

	nest_parms = nla_nest_start(skb, CTA_TUPLE_IP | NLA_F_NESTED);
	if (!nest_parms)
		goto nla_put_failure;

	if (likely(l3proto->tuple_to_nlattr))
		ret = l3proto->tuple_to_nlattr(skb, tuple);

	nla_nest_end(skb, nest_parms);

	return ret;

nla_put_failure:
	return -1;
}

static inline int
ctnetlink_dump_tuples(struct sk_buff *skb,
		      const struct nf_conntrack_tuple *tuple)
{
	int ret;
	struct nf_conntrack_l3proto *l3proto;
	struct nf_conntrack_l4proto *l4proto;

	l3proto = nf_ct_l3proto_find_get(tuple->src.l3num);
	ret = ctnetlink_dump_tuples_ip(skb, tuple, l3proto);
	nf_ct_l3proto_put(l3proto);

	if (unlikely(ret < 0))
		return ret;

	l4proto = nf_ct_l4proto_find_get(tuple->src.l3num, tuple->dst.protonum);
	ret = ctnetlink_dump_tuples_proto(skb, tuple, l4proto);
	nf_ct_l4proto_put(l4proto);

	return ret;
}

static inline int
ctnetlink_dump_status(struct sk_buff *skb, const struct nf_conn *ct)
{
	__be32 status = htonl((u_int32_t) ct->status);
	NLA_PUT(skb, CTA_STATUS, sizeof(status), &status);
	return 0;

nla_put_failure:
	return -1;
}

static inline int
ctnetlink_dump_timeout(struct sk_buff *skb, const struct nf_conn *ct)
{
	long timeout_l = ct->timeout.expires - jiffies;
	__be32 timeout;

	if (timeout_l < 0)
		timeout = 0;
	else
		timeout = htonl(timeout_l / HZ);

	NLA_PUT(skb, CTA_TIMEOUT, sizeof(timeout), &timeout);
	return 0;

nla_put_failure:
	return -1;
}

static inline int
ctnetlink_dump_protoinfo(struct sk_buff *skb, const struct nf_conn *ct)
{
	struct nf_conntrack_l4proto *l4proto = nf_ct_l4proto_find_get(ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num, ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum);
	struct nlattr *nest_proto;
	int ret;

	if (!l4proto->to_nlattr) {
		nf_ct_l4proto_put(l4proto);
		return 0;
	}

	nest_proto = nla_nest_start(skb, CTA_PROTOINFO | NLA_F_NESTED);
	if (!nest_proto)
		goto nla_put_failure;

	ret = l4proto->to_nlattr(skb, nest_proto, ct);

	nf_ct_l4proto_put(l4proto);

	nla_nest_end(skb, nest_proto);

	return ret;

nla_put_failure:
	nf_ct_l4proto_put(l4proto);
	return -1;
}

static inline int
ctnetlink_dump_helpinfo(struct sk_buff *skb, const struct nf_conn *ct)
{
	struct nlattr *nest_helper;
	const struct nf_conn_help *help = nfct_help(ct);
	struct nf_conntrack_helper *helper;

	if (!help)
		return 0;

	rcu_read_lock();
	helper = rcu_dereference(help->helper);
	if (!helper)
		goto out;

	nest_helper = nla_nest_start(skb, CTA_HELP | NLA_F_NESTED);
	if (!nest_helper)
		goto nla_put_failure;
	NLA_PUT(skb, CTA_HELP_NAME, strlen(helper->name), helper->name);

	if (helper->to_nlattr)
		helper->to_nlattr(skb, ct);

	nla_nest_end(skb, nest_helper);
out:
	rcu_read_unlock();
	return 0;

nla_put_failure:
	rcu_read_unlock();
	return -1;
}

#ifdef CONFIG_NF_CT_ACCT
static inline int
ctnetlink_dump_counters(struct sk_buff *skb, const struct nf_conn *ct,
			enum ip_conntrack_dir dir)
{
	enum ctattr_type type = dir ? CTA_COUNTERS_REPLY: CTA_COUNTERS_ORIG;
	struct nlattr *nest_count;
	__be32 tmp;

	nest_count = nla_nest_start(skb, type | NLA_F_NESTED);
	if (!nest_count)
		goto nla_put_failure;

	tmp = htonl(ct->counters[dir].packets);
	NLA_PUT(skb, CTA_COUNTERS32_PACKETS, sizeof(u_int32_t), &tmp);

	tmp = htonl(ct->counters[dir].bytes);
	NLA_PUT(skb, CTA_COUNTERS32_BYTES, sizeof(u_int32_t), &tmp);

	nla_nest_end(skb, nest_count);

	return 0;

nla_put_failure:
	return -1;
}
#else
#define ctnetlink_dump_counters(a, b, c) (0)
#endif

#ifdef CONFIG_NF_CONNTRACK_MARK
static inline int
ctnetlink_dump_mark(struct sk_buff *skb, const struct nf_conn *ct)
{
	__be32 mark = htonl(ct->mark);

	NLA_PUT(skb, CTA_MARK, sizeof(u_int32_t), &mark);
	return 0;

nla_put_failure:
	return -1;
}
#else
#define ctnetlink_dump_mark(a, b) (0)
#endif

static inline int
ctnetlink_dump_id(struct sk_buff *skb, const struct nf_conn *ct)
{
	__be32 id = htonl((unsigned long)ct);
	NLA_PUT(skb, CTA_ID, sizeof(u_int32_t), &id);
	return 0;

nla_put_failure:
	return -1;
}

static inline int
ctnetlink_dump_use(struct sk_buff *skb, const struct nf_conn *ct)
{
	__be32 use = htonl(atomic_read(&ct->ct_general.use));

	NLA_PUT(skb, CTA_USE, sizeof(u_int32_t), &use);
	return 0;

nla_put_failure:
	return -1;
}

#define tuple(ct, dir) (&(ct)->tuplehash[dir].tuple)

static int
ctnetlink_fill_info(struct sk_buff *skb, u32 pid, u32 seq,
		    int event, int nowait,
		    const struct nf_conn *ct)
{
	struct nlmsghdr *nlh;
	struct nfgenmsg *nfmsg;
	struct nlattr *nest_parms;
	unsigned char *b = skb_tail_pointer(skb);

	event |= NFNL_SUBSYS_CTNETLINK << 8;
	nlh    = NLMSG_PUT(skb, pid, seq, event, sizeof(struct nfgenmsg));
	nfmsg  = NLMSG_DATA(nlh);

	nlh->nlmsg_flags    = (nowait && pid) ? NLM_F_MULTI : 0;
	nfmsg->nfgen_family =
		ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num;
	nfmsg->version      = NFNETLINK_V0;
	nfmsg->res_id	    = 0;

	nest_parms = nla_nest_start(skb, CTA_TUPLE_ORIG | NLA_F_NESTED);
	if (!nest_parms)
		goto nla_put_failure;
	if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_ORIGINAL)) < 0)
		goto nla_put_failure;
	nla_nest_end(skb, nest_parms);

	nest_parms = nla_nest_start(skb, CTA_TUPLE_REPLY | NLA_F_NESTED);
	if (!nest_parms)
		goto nla_put_failure;
	if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_REPLY)) < 0)
		goto nla_put_failure;
	nla_nest_end(skb, nest_parms);

	if (ctnetlink_dump_status(skb, ct) < 0 ||
	    ctnetlink_dump_timeout(skb, ct) < 0 ||
	    ctnetlink_dump_counters(skb, ct, IP_CT_DIR_ORIGINAL) < 0 ||
	    ctnetlink_dump_counters(skb, ct, IP_CT_DIR_REPLY) < 0 ||
	    ctnetlink_dump_protoinfo(skb, ct) < 0 ||
	    ctnetlink_dump_helpinfo(skb, ct) < 0 ||
	    ctnetlink_dump_mark(skb, ct) < 0 ||
	    ctnetlink_dump_id(skb, ct) < 0 ||
	    ctnetlink_dump_use(skb, ct) < 0)
		goto nla_put_failure;

	nlh->nlmsg_len = skb_tail_pointer(skb) - b;
	return skb->len;

nlmsg_failure:
nla_put_failure:
	nlmsg_trim(skb, b);
	return -1;
}

#ifdef CONFIG_NF_CONNTRACK_EVENTS
static int ctnetlink_conntrack_event(struct notifier_block *this,
				     unsigned long events, void *ptr)
{
	struct nlmsghdr *nlh;
	struct nfgenmsg *nfmsg;
	struct nlattr *nest_parms;
	struct nf_conn *ct = (struct nf_conn *)ptr;
	struct sk_buff *skb;
	unsigned int type;
	sk_buff_data_t b;
	unsigned int flags = 0, group;

	/* ignore our fake conntrack entry */
	if (ct == &nf_conntrack_untracked)
		return NOTIFY_DONE;

	if (events & IPCT_DESTROY) {
		type = IPCTNL_MSG_CT_DELETE;
		group = NFNLGRP_CONNTRACK_DESTROY;
	} else  if (events & (IPCT_NEW | IPCT_RELATED)) {
		type = IPCTNL_MSG_CT_NEW;
		flags = NLM_F_CREATE|NLM_F_EXCL;
		group = NFNLGRP_CONNTRACK_NEW;
	} else  if (events & (IPCT_STATUS | IPCT_PROTOINFO)) {
		type = IPCTNL_MSG_CT_NEW;
		group = NFNLGRP_CONNTRACK_UPDATE;
	} else
		return NOTIFY_DONE;

	if (!nfnetlink_has_listeners(group))
		return NOTIFY_DONE;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_ATOMIC);
	if (!skb)
		return NOTIFY_DONE;

	b = skb->tail;

	type |= NFNL_SUBSYS_CTNETLINK << 8;
	nlh   = NLMSG_PUT(skb, 0, 0, type, sizeof(struct nfgenmsg));
	nfmsg = NLMSG_DATA(nlh);

	nlh->nlmsg_flags    = flags;
	nfmsg->nfgen_family = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num;
	nfmsg->version	= NFNETLINK_V0;
	nfmsg->res_id	= 0;

	nest_parms = nla_nest_start(skb, CTA_TUPLE_ORIG | NLA_F_NESTED);
	if (!nest_parms)
		goto nla_put_failure;
	if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_ORIGINAL)) < 0)
		goto nla_put_failure;
	nla_nest_end(skb, nest_parms);

	nest_parms = nla_nest_start(skb, CTA_TUPLE_REPLY | NLA_F_NESTED);
	if (!nest_parms)
		goto nla_put_failure;
	if (ctnetlink_dump_tuples(skb, tuple(ct, IP_CT_DIR_REPLY)) < 0)
		goto nla_put_failure;
	nla_nest_end(skb, nest_parms);

	if (events & IPCT_DESTROY) {
		if (ctnetlink_dump_counters(skb, ct, IP_CT_DIR_ORIGINAL) < 0 ||
		    ctnetlink_dump_counters(skb, ct, IP_CT_DIR_REPLY) < 0)
			goto nla_put_failure;
	} else {
		if (ctnetlink_dump_status(skb, ct) < 0)
			goto nla_put_failure;

		if (ctnetlink_dump_timeout(skb, ct) < 0)
			goto nla_put_failure;

		if (events & IPCT_PROTOINFO
		    && ctnetlink_dump_protoinfo(skb, ct) < 0)
			goto nla_put_failure;

		if ((events & IPCT_HELPER || nfct_help(ct))
		    && ctnetlink_dump_helpinfo(skb, ct) < 0)
			goto nla_put_failure;

#ifdef CONFIG_NF_CONNTRACK_MARK
		if ((events & IPCT_MARK || ct->mark)
		    && ctnetlink_dump_mark(skb, ct) < 0)
			goto nla_put_failure;
#endif

		if (events & IPCT_COUNTER_FILLING &&
		    (ctnetlink_dump_counters(skb, ct, IP_CT_DIR_ORIGINAL) < 0 ||
		     ctnetlink_dump_counters(skb, ct, IP_CT_DIR_REPLY) < 0))
			goto nla_put_failure;
	}

	nlh->nlmsg_len = skb->tail - b;
	nfnetlink_send(skb, 0, group, 0);
	return NOTIFY_DONE;

nlmsg_failure:
nla_put_failure:
	kfree_skb(skb);
	return NOTIFY_DONE;
}
#endif /* CONFIG_NF_CONNTRACK_EVENTS */

static int ctnetlink_done(struct netlink_callback *cb)
{
	if (cb->args[1])
		nf_ct_put((struct nf_conn *)cb->args[1]);
	return 0;
}

#define L3PROTO(ct) ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num

static int
ctnetlink_dump_table(struct sk_buff *skb, struct netlink_callback *cb)
{
	struct nf_conn *ct, *last;
	struct nf_conntrack_tuple_hash *h;
	struct hlist_node *n;
	struct nfgenmsg *nfmsg = NLMSG_DATA(cb->nlh);
	u_int8_t l3proto = nfmsg->nfgen_family;

	read_lock_bh(&nf_conntrack_lock);
	last = (struct nf_conn *)cb->args[1];
	for (; cb->args[0] < nf_conntrack_htable_size; cb->args[0]++) {
restart:
		hlist_for_each_entry(h, n, &nf_conntrack_hash[cb->args[0]],
				     hnode) {
			if (NF_CT_DIRECTION(h) != IP_CT_DIR_ORIGINAL)
				continue;
			ct = nf_ct_tuplehash_to_ctrack(h);
			/* Dump entries of a given L3 protocol number.
			 * If it is not specified, ie. l3proto == 0,
			 * then dump everything. */
			if (l3proto && L3PROTO(ct) != l3proto)
				continue;
			if (cb->args[1]) {
				if (ct != last)
					continue;
				cb->args[1] = 0;
			}
			if (ctnetlink_fill_info(skb, NETLINK_CB(cb->skb).pid,
						cb->nlh->nlmsg_seq,
						IPCTNL_MSG_CT_NEW,
						1, ct) < 0) {
				nf_conntrack_get(&ct->ct_general);
				cb->args[1] = (unsigned long)ct;
				goto out;
			}
#ifdef CONFIG_NF_CT_ACCT
			if (NFNL_MSG_TYPE(cb->nlh->nlmsg_type) ==
						IPCTNL_MSG_CT_GET_CTRZERO)
				memset(&ct->counters, 0, sizeof(ct->counters));
#endif
		}
		if (cb->args[1]) {
			cb->args[1] = 0;
			goto restart;
		}
	}
out:
	read_unlock_bh(&nf_conntrack_lock);
	if (last)
		nf_ct_put(last);

	return skb->len;
}

static inline int
ctnetlink_parse_tuple_ip(struct nlattr *attr, struct nf_conntrack_tuple *tuple)
{
	struct nlattr *tb[CTA_IP_MAX+1];
	struct nf_conntrack_l3proto *l3proto;
	int ret = 0;

	nla_parse_nested(tb, CTA_IP_MAX, attr, NULL);

	l3proto = nf_ct_l3proto_find_get(tuple->src.l3num);

	if (likely(l3proto->nlattr_to_tuple)) {
		ret = nla_validate_nested(attr, CTA_IP_MAX,
					  l3proto->nla_policy);
		if (ret == 0)
			ret = l3proto->nlattr_to_tuple(tb, tuple);
	}

	nf_ct_l3proto_put(l3proto);

	return ret;
}

static const struct nla_policy proto_nla_policy[CTA_PROTO_MAX+1] = {
	[CTA_PROTO_NUM]	= { .type = NLA_U8 },
};

static inline int
ctnetlink_parse_tuple_proto(struct nlattr *attr,
			    struct nf_conntrack_tuple *tuple)
{
	struct nlattr *tb[CTA_PROTO_MAX+1];
	struct nf_conntrack_l4proto *l4proto;
	int ret = 0;

	ret = nla_parse_nested(tb, CTA_PROTO_MAX, attr, proto_nla_policy);
	if (ret < 0)
		return ret;

	if (!tb[CTA_PROTO_NUM])
		return -EINVAL;
	tuple->dst.protonum = *(u_int8_t *)nla_data(tb[CTA_PROTO_NUM]);

	l4proto = nf_ct_l4proto_find_get(tuple->src.l3num, tuple->dst.protonum);

	if (likely(l4proto->nlattr_to_tuple)) {
		ret = nla_validate_nested(attr, CTA_PROTO_MAX,
					  l4proto->nla_policy);
		if (ret == 0)
			ret = l4proto->nlattr_to_tuple(tb, tuple);
	}

	nf_ct_l4proto_put(l4proto);

	return ret;
}

static inline int
ctnetlink_parse_tuple(struct nlattr *cda[], struct nf_conntrack_tuple *tuple,
		      enum ctattr_tuple type, u_int8_t l3num)
{
	struct nlattr *tb[CTA_TUPLE_MAX+1];
	int err;

	memset(tuple, 0, sizeof(*tuple));

	nla_parse_nested(tb, CTA_TUPLE_MAX, cda[type], NULL);