[net-next-2.6.git] / net / ipv4 / netfilter / nf_conntrack_l3proto_ipv4.c

/* (C) 1999-2001 Paul `Rusty' Russell
 * (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 */

#include <linux/types.h>
#include <linux/ip.h>
#include <linux/netfilter.h>
#include <linux/module.h>
#include <linux/skbuff.h>
#include <linux/icmp.h>
#include <linux/sysctl.h>
#include <net/route.h>
#include <net/ip.h>

#include <linux/netfilter_ipv4.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_helper.h>
#include <net/netfilter/nf_conntrack_l4proto.h>
#include <net/netfilter/nf_conntrack_l3proto.h>
#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/ipv4/nf_conntrack_ipv4.h>

static int ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
			     struct nf_conntrack_tuple *tuple)
{
	const __be32 *ap;
	__be32 _addrs[2];
	ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr),
				sizeof(u_int32_t) * 2, _addrs);
	if (ap == NULL)
		return 0;

	tuple->src.u3.ip = ap[0];
	tuple->dst.u3.ip = ap[1];

	return 1;
}

static int ipv4_invert_tuple(struct nf_conntrack_tuple *tuple,
			   const struct nf_conntrack_tuple *orig)
{
	tuple->src.u3.ip = orig->dst.u3.ip;
	tuple->dst.u3.ip = orig->src.u3.ip;

	return 1;
}

static int ipv4_print_tuple(struct seq_file *s,
			    const struct nf_conntrack_tuple *tuple)
{
	return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ",
			  NIPQUAD(tuple->src.u3.ip),
			  NIPQUAD(tuple->dst.u3.ip));
}

/* Returns new sk_buff, or NULL */
static int nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user)
{
	int err;

	skb_orphan(skb);

	local_bh_disable();
	err = ip_defrag(skb, user);
	local_bh_enable();

	if (!err)
		ip_send_check(ip_hdr(skb));

	return err;
}

static int ipv4_get_l4proto(const struct sk_buff *skb, unsigned int nhoff,
			    unsigned int *dataoff, u_int8_t *protonum)
{
	const struct iphdr *iph;
	struct iphdr _iph;

	iph = skb_header_pointer(skb, nhoff, sizeof(_iph), &_iph);
	if (iph == NULL)
		return -NF_DROP;

	/* Conntrack defragments packets, we might still see fragments
	 * inside ICMP packets though. */
	if (iph->frag_off & htons(IP_OFFSET))
		return -NF_DROP;

	*dataoff = nhoff + (iph->ihl << 2);
	*protonum = iph->protocol;

	return NF_ACCEPT;
}

static unsigned int ipv4_confirm(unsigned int hooknum,
				 struct sk_buff *skb,
				 const struct net_device *in,
				 const struct net_device *out,
				 int (*okfn)(struct sk_buff *))
{
	/* We've seen it coming out the other side: confirm it */
	return nf_conntrack_confirm(skb);
}

static unsigned int ipv4_conntrack_help(unsigned int hooknum,
				      struct sk_buff *skb,
				      const struct net_device *in,
				      const struct net_device *out,
				      int (*okfn)(struct sk_buff *))
{
	struct nf_conn *ct;
	enum ip_conntrack_info ctinfo;
	const struct nf_conn_help *help;
	const struct nf_conntrack_helper *helper;

	/* This is where we call the helper: as the packet goes out. */
	ct = nf_ct_get(skb, &ctinfo);
	if (!ct || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
		return NF_ACCEPT;

	help = nfct_help(ct);
	if (!help)
		return NF_ACCEPT;
	/* rcu_read_lock()ed by nf_hook_slow */
	helper = rcu_dereference(help->helper);
	if (!helper)
		return NF_ACCEPT;
	return helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
			    ct, ctinfo);
}

static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
					  struct sk_buff *skb,
					  const struct net_device *in,
					  const struct net_device *out,
					  int (*okfn)(struct sk_buff *))
{
	/* Previously seen (loopback)?  Ignore.  Do this before
	   fragment check. */
	if (skb->nfct)
		return NF_ACCEPT;

	/* Gather fragments. */
	if (ip_hdr(skb)->frag_off & htons(IP_MF | IP_OFFSET)) {
		if (nf_ct_ipv4_gather_frags(skb,
					    hooknum == NF_INET_PRE_ROUTING ?
					    IP_DEFRAG_CONNTRACK_IN :
					    IP_DEFRAG_CONNTRACK_OUT))
			return NF_STOLEN;
	}
	return NF_ACCEPT;
}

static unsigned int ipv4_conntrack_in(unsigned int hooknum,
				      struct sk_buff *skb,
				      const struct net_device *in,
				      const struct net_device *out,
				      int (*okfn)(struct sk_buff *))
{
	return nf_conntrack_in(PF_INET, hooknum, skb);
}

static unsigned int ipv4_conntrack_local(unsigned int hooknum,
					 struct sk_buff *skb,
					 const struct net_device *in,
					 const struct net_device *out,
					 int (*okfn)(struct sk_buff *))
{
	/* root is playing with raw sockets. */
	if (skb->len < sizeof(struct iphdr) ||
	    ip_hdrlen(skb) < sizeof(struct iphdr)) {
		if (net_ratelimit())
			printk("ipt_hook: happy cracking.\n");
		return NF_ACCEPT;
	}
	return nf_conntrack_in(PF_INET, hooknum, skb);
}

/* Connection tracking may drop packets, but never alters them, so
   make it the first hook. */
static struct nf_hook_ops ipv4_conntrack_ops[] __read_mostly = {
	{
		.hook		= ipv4_conntrack_defrag,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_PRE_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv4_conntrack_in,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_PRE_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK,
	},
	{
		.hook           = ipv4_conntrack_defrag,
		.owner          = THIS_MODULE,
		.pf             = PF_INET,
		.hooknum        = NF_INET_LOCAL_OUT,
		.priority       = NF_IP_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv4_conntrack_local,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_LOCAL_OUT,
		.priority	= NF_IP_PRI_CONNTRACK,
	},
	{
		.hook		= ipv4_conntrack_help,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
	},
	{
		.hook		= ipv4_conntrack_help,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
	},
	{
		.hook		= ipv4_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM,
	},
	{
		.hook		= ipv4_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_INET_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM,
	},
};

#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
static int log_invalid_proto_min = 0;
static int log_invalid_proto_max = 255;

static ctl_table ip_ct_sysctl_table[] = {
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_MAX,
		.procname	= "ip_conntrack_max",
		.data		= &nf_conntrack_max,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_COUNT,
		.procname	= "ip_conntrack_count",
		.data		= &nf_conntrack_count,
		.maxlen		= sizeof(int),
		.mode		= 0444,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_BUCKETS,
		.procname	= "ip_conntrack_buckets",
		.data		= &nf_conntrack_htable_size,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0444,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_CHECKSUM,
		.procname	= "ip_conntrack_checksum",
		.data		= &nf_conntrack_checksum,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_LOG_INVALID,
		.procname	= "ip_conntrack_log_invalid",
		.data		= &nf_ct_log_invalid,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec_minmax,
		.strategy	= &sysctl_intvec,
		.extra1		= &log_invalid_proto_min,
		.extra2		= &log_invalid_proto_max,
	},
	{
		.ctl_name	= 0
	}
};
#endif /* CONFIG_SYSCTL && CONFIG_NF_CONNTRACK_PROC_COMPAT */

/* Fast function for those who don't want to parse /proc (and I don't
   blame them). */
/* Reversing the socket's dst/src point of view gives us the reply
   mapping. */
static int
getorigdst(struct sock *sk, int optval, void __user *user, int *len)
{
	const struct inet_sock *inet = inet_sk(sk);
	const struct nf_conntrack_tuple_hash *h;
	struct nf_conntrack_tuple tuple;

	NF_CT_TUPLE_U_BLANK(&tuple);
	tuple.src.u3.ip = inet->rcv_saddr;
	tuple.src.u.tcp.port = inet->sport;
	tuple.dst.u3.ip = inet->daddr;
	tuple.dst.u.tcp.port = inet->dport;
	tuple.src.l3num = PF_INET;
	tuple.dst.protonum = IPPROTO_TCP;

	/* We only do TCP at the moment: is there a better way? */
	if (strcmp(sk->sk_prot->name, "TCP")) {
		pr_debug("SO_ORIGINAL_DST: Not a TCP socket\n");
		return -ENOPROTOOPT;
	}

	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
		pr_debug("SO_ORIGINAL_DST: len %d not %Zu\n",
			 *len, sizeof(struct sockaddr_in));
		return -EINVAL;
	}

	h = nf_conntrack_find_get(&tuple);
	if (h) {
		struct sockaddr_in sin;
		struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);

		sin.sin_family = AF_INET;
		sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
			.tuple.dst.u.tcp.port;
		sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
			.tuple.dst.u3.ip;
		memset(sin.sin_zero, 0, sizeof(sin.sin_zero));

		pr_debug("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
			 NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
		nf_ct_put(ct);
		if (copy_to_user(user, &sin, sizeof(sin)) != 0)
			return -EFAULT;
		else
			return 0;
	}
	pr_debug("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
		 NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
		 NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
	return -ENOENT;
}

#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)

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

static int ipv4_tuple_to_nlattr(struct sk_buff *skb,
				const struct nf_conntrack_tuple *tuple)
{
	NLA_PUT_BE32(skb, CTA_IP_V4_SRC, tuple->src.u3.ip);
	NLA_PUT_BE32(skb, CTA_IP_V4_DST, tuple->dst.u3.ip);
	return 0;

nla_put_failure:
	return -1;
}

static const struct nla_policy ipv4_nla_policy[CTA_IP_MAX+1] = {
	[CTA_IP_V4_SRC]	= { .type = NLA_U32 },
	[CTA_IP_V4_DST]	= { .type = NLA_U32 },
};

static int ipv4_nlattr_to_tuple(struct nlattr *tb[],
				struct nf_conntrack_tuple *t)
{
	if (!tb[CTA_IP_V4_SRC] || !tb[CTA_IP_V4_DST])
		return -EINVAL;

	t->src.u3.ip = nla_get_be32(tb[CTA_IP_V4_SRC]);
	t->dst.u3.ip = nla_get_be32(tb[CTA_IP_V4_DST]);

	return 0;
}
#endif

static struct nf_sockopt_ops so_getorigdst = {
	.pf		= PF_INET,
	.get_optmin	= SO_ORIGINAL_DST,
	.get_optmax	= SO_ORIGINAL_DST+1,
	.get		= &getorigdst,
	.owner		= THIS_MODULE,
};

struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 __read_mostly = {
	.l3proto	 = PF_INET,
	.name		 = "ipv4",
	.pkt_to_tuple	 = ipv4_pkt_to_tuple,
	.invert_tuple	 = ipv4_invert_tuple,
	.print_tuple	 = ipv4_print_tuple,
	.get_l4proto	 = ipv4_get_l4proto,
#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)
	.tuple_to_nlattr = ipv4_tuple_to_nlattr,
	.nlattr_to_tuple = ipv4_nlattr_to_tuple,
	.nla_policy	 = ipv4_nla_policy,
#endif
#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	.ctl_table_path  = nf_net_ipv4_netfilter_sysctl_path,
	.ctl_table	 = ip_ct_sysctl_table,
#endif
	.me		 = THIS_MODULE,
};

module_param_call(hashsize, nf_conntrack_set_hashsize, param_get_uint,
		  &nf_conntrack_htable_size, 0600);

MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
MODULE_ALIAS("ip_conntrack");
MODULE_LICENSE("GPL");

static int __init nf_conntrack_l3proto_ipv4_init(void)
{
	int ret = 0;

	need_conntrack();

	ret = nf_register_sockopt(&so_getorigdst);
	if (ret < 0) {
		printk(KERN_ERR "Unable to register netfilter socket option\n");
		return ret;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register tcp.\n");
		goto cleanup_sockopt;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register udp.\n");
		goto cleanup_tcp;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register icmp.\n");
		goto cleanup_udp;
	}

	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register ipv4\n");
		goto cleanup_icmp;
	}

	ret = nf_register_hooks(ipv4_conntrack_ops,
				ARRAY_SIZE(ipv4_conntrack_ops));
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register hooks.\n");
		goto cleanup_ipv4;
	}
#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	ret = nf_conntrack_ipv4_compat_init();
	if (ret < 0)
		goto cleanup_hooks;
#endif
	return ret;
#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
 cleanup_hooks:
	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
#endif
 cleanup_ipv4:
	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
 cleanup_icmp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
 cleanup_udp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
 cleanup_tcp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
 cleanup_sockopt:
	nf_unregister_sockopt(&so_getorigdst);
	return ret;
}

static void __exit nf_conntrack_l3proto_ipv4_fini(void)
{
	synchronize_net();
#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	nf_conntrack_ipv4_compat_fini();
#endif
	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
	nf_unregister_sockopt(&so_getorigdst);
}

module_init(nf_conntrack_l3proto_ipv4_init);
module_exit(nf_conntrack_l3proto_ipv4_fini);

void need_ipv4_conntrack(void)
{
	return;
}
EXPORT_SYMBOL_GPL(need_ipv4_conntrack);
Commit	Line	Data
9fb9cbb1 YK	1	/* (C) 1999-2001 Paul `Rusty' Russell
	2	* (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
	3	*
	4	* This program is free software; you can redistribute it and/or modify
	5	* it under the terms of the GNU General Public License version 2 as
	6	* published by the Free Software Foundation.
9fb9cbb1 YK	7	*/
9fb9cbb1 YK	8
9fb9cbb1 YK	9	#include <linux/types.h>
	10	#include <linux/ip.h>
	11	#include <linux/netfilter.h>
	12	#include <linux/module.h>
	13	#include <linux/skbuff.h>
	14	#include <linux/icmp.h>
	15	#include <linux/sysctl.h>
0ae2cfe7	16	#include <net/route.h>
9fb9cbb1 YK	17	#include <net/ip.h>
	18
	19	#include <linux/netfilter_ipv4.h>
	20	#include <net/netfilter/nf_conntrack.h>
	21	#include <net/netfilter/nf_conntrack_helper.h>
605dcad6	22	#include <net/netfilter/nf_conntrack_l4proto.h>
9fb9cbb1 YK	23	#include <net/netfilter/nf_conntrack_l3proto.h>
	24	#include <net/netfilter/nf_conntrack_core.h>
	25	#include <net/netfilter/ipv4/nf_conntrack_ipv4.h>
	26
9fb9cbb1 YK	27	static int ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
	28	struct nf_conntrack_tuple *tuple)
	29	{
32948588 JE	30	const __be32 *ap;
32948588 JE	31	__be32 _addrs[2];
9fb9cbb1 YK	32	ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr),
	33	sizeof(u_int32_t) * 2, _addrs);
	34	if (ap == NULL)
	35	return 0;
	36
	37	tuple->src.u3.ip = ap[0];
	38	tuple->dst.u3.ip = ap[1];
	39
	40	return 1;
	41	}
	42
	43	static int ipv4_invert_tuple(struct nf_conntrack_tuple *tuple,
	44	const struct nf_conntrack_tuple *orig)
	45	{
	46	tuple->src.u3.ip = orig->dst.u3.ip;
	47	tuple->dst.u3.ip = orig->src.u3.ip;
	48
	49	return 1;
	50	}
	51
	52	static int ipv4_print_tuple(struct seq_file *s,
	53	const struct nf_conntrack_tuple *tuple)
	54	{
	55	return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ",
e905a9ed	56	NIPQUAD(tuple->src.u3.ip),
9fb9cbb1 YK	57	NIPQUAD(tuple->dst.u3.ip));
	58	}
	59
9fb9cbb1	60	/* Returns new sk_buff, or NULL */
776c729e	61	static int nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user)
9fb9cbb1	62	{
776c729e HX	63	int err;
776c729e HX	64
9fb9cbb1 YK	65	skb_orphan(skb);
9fb9cbb1 YK	66
e905a9ed	67	local_bh_disable();
776c729e	68	err = ip_defrag(skb, user);
e905a9ed	69	local_bh_enable();
9fb9cbb1	70
776c729e	71	if (!err)
eddc9ec5	72	ip_send_check(ip_hdr(skb));
9fb9cbb1	73
776c729e	74	return err;
9fb9cbb1 YK	75	}
9fb9cbb1 YK	76
ffc30690 YK	77	static int ipv4_get_l4proto(const struct sk_buff *skb, unsigned int nhoff,
ffc30690 YK	78	unsigned int dataoff, u_int8_t protonum)
9fb9cbb1	79	{
32948588 JE	80	const struct iphdr *iph;
32948588 JE	81	struct iphdr _iph;
ffc30690 YK	82
	83	iph = skb_header_pointer(skb, nhoff, sizeof(_iph), &_iph);
	84	if (iph == NULL)
	85	return -NF_DROP;
	86
0fb96701 PM	87	/* Conntrack defragments packets, we might still see fragments
	88	* inside ICMP packets though. */
	89	if (iph->frag_off & htons(IP_OFFSET))
9fb9cbb1	90	return -NF_DROP;
9fb9cbb1	91
ffc30690 YK	92	*dataoff = nhoff + (iph->ihl << 2);
ffc30690 YK	93	*protonum = iph->protocol;
9fb9cbb1 YK	94
	95	return NF_ACCEPT;
	96	}
	97
9fb9cbb1	98	static unsigned int ipv4_confirm(unsigned int hooknum,
3db05fea	99	struct sk_buff *skb,
9fb9cbb1 YK	100	const struct net_device *in,
	101	const struct net_device *out,
	102	int (okfn)(struct sk_buff ))
	103	{
	104	/* We've seen it coming out the other side: confirm it */
3db05fea	105	return nf_conntrack_confirm(skb);
9fb9cbb1 YK	106	}
	107
	108	static unsigned int ipv4_conntrack_help(unsigned int hooknum,
3db05fea	109	struct sk_buff *skb,
9fb9cbb1 YK	110	const struct net_device *in,
	111	const struct net_device *out,
	112	int (okfn)(struct sk_buff ))
	113	{
	114	struct nf_conn *ct;
	115	enum ip_conntrack_info ctinfo;
32948588 JE	116	const struct nf_conn_help *help;
32948588 JE	117	const struct nf_conntrack_helper *helper;
9fb9cbb1 YK	118
9fb9cbb1 YK	119	/* This is where we call the helper: as the packet goes out. */
3db05fea	120	ct = nf_ct_get(skb, &ctinfo);
6442f1cf	121	if (!ct \|\| ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
dc808fe2 HW	122	return NF_ACCEPT;
	123
	124	help = nfct_help(ct);
3c158f7f	125	if (!help)
dc808fe2	126	return NF_ACCEPT;
3c158f7f PM	127	/* rcu_read_lock()ed by nf_hook_slow */
	128	helper = rcu_dereference(help->helper);
	129	if (!helper)
	130	return NF_ACCEPT;
3db05fea	131	return helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
3c158f7f	132	ct, ctinfo);
9fb9cbb1 YK	133	}
	134
	135	static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
3db05fea	136	struct sk_buff *skb,
9fb9cbb1 YK	137	const struct net_device *in,
	138	const struct net_device *out,
	139	int (okfn)(struct sk_buff ))
	140	{
9fb9cbb1 YK	141	/* Previously seen (loopback)? Ignore. Do this before
9fb9cbb1 YK	142	fragment check. */
3db05fea	143	if (skb->nfct)
9fb9cbb1	144	return NF_ACCEPT;
9fb9cbb1 YK	145
9fb9cbb1 YK	146	/* Gather fragments. */
3db05fea HX	147	if (ip_hdr(skb)->frag_off & htons(IP_MF \| IP_OFFSET)) {
3db05fea HX	148	if (nf_ct_ipv4_gather_frags(skb,
6e23ae2a	149	hooknum == NF_INET_PRE_ROUTING ?
776c729e HX	150	IP_DEFRAG_CONNTRACK_IN :
776c729e HX	151	IP_DEFRAG_CONNTRACK_OUT))
9fb9cbb1 YK	152	return NF_STOLEN;
	153	}
	154	return NF_ACCEPT;
	155	}
	156
9fb9cbb1	157	static unsigned int ipv4_conntrack_in(unsigned int hooknum,
3db05fea	158	struct sk_buff *skb,
9fb9cbb1 YK	159	const struct net_device *in,
	160	const struct net_device *out,
	161	int (okfn)(struct sk_buff ))
	162	{
3db05fea	163	return nf_conntrack_in(PF_INET, hooknum, skb);
9fb9cbb1 YK	164	}
	165
	166	static unsigned int ipv4_conntrack_local(unsigned int hooknum,
3db05fea	167	struct sk_buff *skb,
e905a9ed YH	168	const struct net_device *in,
	169	const struct net_device *out,
	170	int (okfn)(struct sk_buff ))
9fb9cbb1 YK	171	{
9fb9cbb1 YK	172	/* root is playing with raw sockets. */
3db05fea HX	173	if (skb->len < sizeof(struct iphdr) \|\|
3db05fea HX	174	ip_hdrlen(skb) < sizeof(struct iphdr)) {
9fb9cbb1 YK	175	if (net_ratelimit())
	176	printk("ipt_hook: happy cracking.\n");
	177	return NF_ACCEPT;
	178	}
3db05fea	179	return nf_conntrack_in(PF_INET, hooknum, skb);
9fb9cbb1 YK	180	}
	181
	182	/* Connection tracking may drop packets, but never alters them, so
	183	make it the first hook. */
1999414a	184	static struct nf_hook_ops ipv4_conntrack_ops[] __read_mostly = {
964ddaa1 PM	185	{
	186	.hook = ipv4_conntrack_defrag,
	187	.owner = THIS_MODULE,
	188	.pf = PF_INET,
6e23ae2a	189	.hooknum = NF_INET_PRE_ROUTING,
964ddaa1 PM	190	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	191	},
	192	{
	193	.hook = ipv4_conntrack_in,
	194	.owner = THIS_MODULE,
	195	.pf = PF_INET,
6e23ae2a	196	.hooknum = NF_INET_PRE_ROUTING,
964ddaa1 PM	197	.priority = NF_IP_PRI_CONNTRACK,
	198	},
	199	{
	200	.hook = ipv4_conntrack_defrag,
	201	.owner = THIS_MODULE,
	202	.pf = PF_INET,
6e23ae2a	203	.hooknum = NF_INET_LOCAL_OUT,
964ddaa1 PM	204	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	205	},
	206	{
	207	.hook = ipv4_conntrack_local,
	208	.owner = THIS_MODULE,
	209	.pf = PF_INET,
6e23ae2a	210	.hooknum = NF_INET_LOCAL_OUT,
964ddaa1 PM	211	.priority = NF_IP_PRI_CONNTRACK,
	212	},
	213	{
	214	.hook = ipv4_conntrack_help,
	215	.owner = THIS_MODULE,
	216	.pf = PF_INET,
6e23ae2a	217	.hooknum = NF_INET_POST_ROUTING,
964ddaa1 PM	218	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	219	},
	220	{
	221	.hook = ipv4_conntrack_help,
	222	.owner = THIS_MODULE,
	223	.pf = PF_INET,
6e23ae2a	224	.hooknum = NF_INET_LOCAL_IN,
964ddaa1 PM	225	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	226	},
	227	{
	228	.hook = ipv4_confirm,
	229	.owner = THIS_MODULE,
	230	.pf = PF_INET,
6e23ae2a	231	.hooknum = NF_INET_POST_ROUTING,
964ddaa1 PM	232	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	233	},
	234	{
	235	.hook = ipv4_confirm,
	236	.owner = THIS_MODULE,
	237	.pf = PF_INET,
6e23ae2a	238	.hooknum = NF_INET_LOCAL_IN,
964ddaa1 PM	239	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
964ddaa1 PM	240	},
9fb9cbb1 YK	241	};
9fb9cbb1 YK	242
a999e683 PM	243	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	244	static int log_invalid_proto_min = 0;
	245	static int log_invalid_proto_max = 255;
	246
	247	static ctl_table ip_ct_sysctl_table[] = {
	248	{
	249	.ctl_name = NET_IPV4_NF_CONNTRACK_MAX,
	250	.procname = "ip_conntrack_max",
	251	.data = &nf_conntrack_max,
	252	.maxlen = sizeof(int),
	253	.mode = 0644,
	254	.proc_handler = &proc_dointvec,
	255	},
	256	{
	257	.ctl_name = NET_IPV4_NF_CONNTRACK_COUNT,
	258	.procname = "ip_conntrack_count",
	259	.data = &nf_conntrack_count,
	260	.maxlen = sizeof(int),
	261	.mode = 0444,
	262	.proc_handler = &proc_dointvec,
	263	},
	264	{
	265	.ctl_name = NET_IPV4_NF_CONNTRACK_BUCKETS,
	266	.procname = "ip_conntrack_buckets",
	267	.data = &nf_conntrack_htable_size,
	268	.maxlen = sizeof(unsigned int),
	269	.mode = 0444,
	270	.proc_handler = &proc_dointvec,
	271	},
	272	{
	273	.ctl_name = NET_IPV4_NF_CONNTRACK_CHECKSUM,
	274	.procname = "ip_conntrack_checksum",
	275	.data = &nf_conntrack_checksum,
	276	.maxlen = sizeof(int),
	277	.mode = 0644,
	278	.proc_handler = &proc_dointvec,
	279	},
	280	{
	281	.ctl_name = NET_IPV4_NF_CONNTRACK_LOG_INVALID,
	282	.procname = "ip_conntrack_log_invalid",
	283	.data = &nf_ct_log_invalid,
	284	.maxlen = sizeof(unsigned int),
	285	.mode = 0644,
	286	.proc_handler = &proc_dointvec_minmax,
	287	.strategy = &sysctl_intvec,
	288	.extra1 = &log_invalid_proto_min,
	289	.extra2 = &log_invalid_proto_max,
	290	},
	291	{
	292	.ctl_name = 0
	293	}
	294	};
	295	#endif /* CONFIG_SYSCTL && CONFIG_NF_CONNTRACK_PROC_COMPAT */
	296
9fb9cbb1 YK	297	/* Fast function for those who don't want to parse /proc (and I don't
	298	blame them). */
	299	/* Reversing the socket's dst/src point of view gives us the reply
	300	mapping. */
	301	static int
	302	getorigdst(struct sock sk, int optval, void __user user, int *len)
	303	{
32948588 JE	304	const struct inet_sock *inet = inet_sk(sk);
32948588 JE	305	const struct nf_conntrack_tuple_hash *h;
9fb9cbb1	306	struct nf_conntrack_tuple tuple;
e905a9ed	307
9fb9cbb1 YK	308	NF_CT_TUPLE_U_BLANK(&tuple);
	309	tuple.src.u3.ip = inet->rcv_saddr;
	310	tuple.src.u.tcp.port = inet->sport;
	311	tuple.dst.u3.ip = inet->daddr;
	312	tuple.dst.u.tcp.port = inet->dport;
	313	tuple.src.l3num = PF_INET;
	314	tuple.dst.protonum = IPPROTO_TCP;
	315
	316	/* We only do TCP at the moment: is there a better way? */
	317	if (strcmp(sk->sk_prot->name, "TCP")) {
0d53778e	318	pr_debug("SO_ORIGINAL_DST: Not a TCP socket\n");
9fb9cbb1 YK	319	return -ENOPROTOOPT;
	320	}
	321
	322	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
0d53778e PM	323	pr_debug("SO_ORIGINAL_DST: len %d not %Zu\n",
0d53778e PM	324	*len, sizeof(struct sockaddr_in));
9fb9cbb1 YK	325	return -EINVAL;
	326	}
	327
330f7db5	328	h = nf_conntrack_find_get(&tuple);
9fb9cbb1 YK	329	if (h) {
	330	struct sockaddr_in sin;
	331	struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
	332
	333	sin.sin_family = AF_INET;
	334	sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	335	.tuple.dst.u.tcp.port;
	336	sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	337	.tuple.dst.u3.ip;
6c813c3f	338	memset(sin.sin_zero, 0, sizeof(sin.sin_zero));
9fb9cbb1	339
0d53778e PM	340	pr_debug("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
0d53778e PM	341	NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
9fb9cbb1 YK	342	nf_ct_put(ct);
	343	if (copy_to_user(user, &sin, sizeof(sin)) != 0)
	344	return -EFAULT;
	345	else
	346	return 0;
	347	}
0d53778e PM	348	pr_debug("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
	349	NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
	350	NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
9fb9cbb1 YK	351	return -ENOENT;
	352	}
	353
e281db5c	354	#if defined(CONFIG_NF_CT_NETLINK) \|\| defined(CONFIG_NF_CT_NETLINK_MODULE)
c1d10adb PNA	355
	356	#include <linux/netfilter/nfnetlink.h>
	357	#include <linux/netfilter/nfnetlink_conntrack.h>
	358
fdf70832	359	static int ipv4_tuple_to_nlattr(struct sk_buff *skb,
c1d10adb PNA	360	const struct nf_conntrack_tuple *tuple)
c1d10adb PNA	361	{
77236b6e PM	362	NLA_PUT_BE32(skb, CTA_IP_V4_SRC, tuple->src.u3.ip);
77236b6e PM	363	NLA_PUT_BE32(skb, CTA_IP_V4_DST, tuple->dst.u3.ip);
c1d10adb PNA	364	return 0;
c1d10adb PNA	365
df6fb868	366	nla_put_failure:
c1d10adb PNA	367	return -1;
	368	}
	369
f73e924c PM	370	static const struct nla_policy ipv4_nla_policy[CTA_IP_MAX+1] = {
	371	[CTA_IP_V4_SRC] = { .type = NLA_U32 },
	372	[CTA_IP_V4_DST] = { .type = NLA_U32 },
c1d10adb PNA	373	};
c1d10adb PNA	374
fdf70832	375	static int ipv4_nlattr_to_tuple(struct nlattr *tb[],
c1d10adb PNA	376	struct nf_conntrack_tuple *t)
c1d10adb PNA	377	{
df6fb868	378	if (!tb[CTA_IP_V4_SRC] \|\| !tb[CTA_IP_V4_DST])
c1d10adb PNA	379	return -EINVAL;
c1d10adb PNA	380
77236b6e PM	381	t->src.u3.ip = nla_get_be32(tb[CTA_IP_V4_SRC]);
77236b6e PM	382	t->dst.u3.ip = nla_get_be32(tb[CTA_IP_V4_DST]);
c1d10adb PNA	383
	384	return 0;
	385	}
	386	#endif
	387
9fb9cbb1 YK	388	static struct nf_sockopt_ops so_getorigdst = {
	389	.pf = PF_INET,
	390	.get_optmin = SO_ORIGINAL_DST,
	391	.get_optmax = SO_ORIGINAL_DST+1,
	392	.get = &getorigdst,
16fcec35	393	.owner = THIS_MODULE,
9fb9cbb1 YK	394	};
9fb9cbb1 YK	395
61075af5	396	struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 __read_mostly = {
9fb9cbb1 YK	397	.l3proto = PF_INET,
	398	.name = "ipv4",
	399	.pkt_to_tuple = ipv4_pkt_to_tuple,
	400	.invert_tuple = ipv4_invert_tuple,
	401	.print_tuple = ipv4_print_tuple,
ffc30690	402	.get_l4proto = ipv4_get_l4proto,
e281db5c	403	#if defined(CONFIG_NF_CT_NETLINK) \|\| defined(CONFIG_NF_CT_NETLINK_MODULE)
fdf70832 PM	404	.tuple_to_nlattr = ipv4_tuple_to_nlattr,
fdf70832 PM	405	.nlattr_to_tuple = ipv4_nlattr_to_tuple,
f73e924c	406	.nla_policy = ipv4_nla_policy,
a999e683 PM	407	#endif
	408	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	409	.ctl_table_path = nf_net_ipv4_netfilter_sysctl_path,
	410	.ctl_table = ip_ct_sysctl_table,
c1d10adb	411	#endif
9fb9cbb1 YK	412	.me = THIS_MODULE,
	413	};
	414
fae718dd PM	415	module_param_call(hashsize, nf_conntrack_set_hashsize, param_get_uint,
	416	&nf_conntrack_htable_size, 0600);
	417
32292a7f	418	MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
d2483dde	419	MODULE_ALIAS("ip_conntrack");
32292a7f PM	420	MODULE_LICENSE("GPL");
	421
	422	static int __init nf_conntrack_l3proto_ipv4_init(void)
9fb9cbb1 YK	423	{
	424	int ret = 0;
	425
32292a7f	426	need_conntrack();
9fb9cbb1 YK	427
	428	ret = nf_register_sockopt(&so_getorigdst);
	429	if (ret < 0) {
	430	printk(KERN_ERR "Unable to register netfilter socket option\n");
32292a7f	431	return ret;
9fb9cbb1 YK	432	}
9fb9cbb1 YK	433
605dcad6	434	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	435	if (ret < 0) {
	436	printk("nf_conntrack_ipv4: can't register tcp.\n");
	437	goto cleanup_sockopt;
	438	}
	439
605dcad6	440	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
9fb9cbb1 YK	441	if (ret < 0) {
	442	printk("nf_conntrack_ipv4: can't register udp.\n");
	443	goto cleanup_tcp;
	444	}
	445
605dcad6	446	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
9fb9cbb1 YK	447	if (ret < 0) {
	448	printk("nf_conntrack_ipv4: can't register icmp.\n");
	449	goto cleanup_udp;
	450	}
	451
	452	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	453	if (ret < 0) {
	454	printk("nf_conntrack_ipv4: can't register ipv4\n");
	455	goto cleanup_icmp;
	456	}
	457
964ddaa1 PM	458	ret = nf_register_hooks(ipv4_conntrack_ops,
964ddaa1 PM	459	ARRAY_SIZE(ipv4_conntrack_ops));
9fb9cbb1	460	if (ret < 0) {
964ddaa1	461	printk("nf_conntrack_ipv4: can't register hooks.\n");
9fb9cbb1 YK	462	goto cleanup_ipv4;
9fb9cbb1 YK	463	}
e4bd8bce PM	464	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	465	ret = nf_conntrack_ipv4_compat_init();
	466	if (ret < 0)
	467	goto cleanup_hooks;
	468	#endif
9fb9cbb1	469	return ret;
e4bd8bce PM	470	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
e4bd8bce PM	471	cleanup_hooks:
e905a9ed	472	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
e4bd8bce	473	#endif
9fb9cbb1 YK	474	cleanup_ipv4:
	475	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	476	cleanup_icmp:
605dcad6	477	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
9fb9cbb1	478	cleanup_udp:
605dcad6	479	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
9fb9cbb1	480	cleanup_tcp:
605dcad6	481	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	482	cleanup_sockopt:
9fb9cbb1 YK	483	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	484	return ret;
	485	}
	486
65b4b4e8	487	static void __exit nf_conntrack_l3proto_ipv4_fini(void)
9fb9cbb1	488	{
32292a7f	489	synchronize_net();
e4bd8bce PM	490	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	491	nf_conntrack_ipv4_compat_fini();
	492	#endif
32292a7f PM	493	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
32292a7f PM	494	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
605dcad6 MJ	495	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	496	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	497	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
32292a7f	498	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	499	}
9fb9cbb1 YK	500
65b4b4e8 AM	501	module_init(nf_conntrack_l3proto_ipv4_init);
65b4b4e8 AM	502	module_exit(nf_conntrack_l3proto_ipv4_fini);
591e6206 PM	503
	504	void need_ipv4_conntrack(void)
	505	{
	506	return;
	507	}
	508	EXPORT_SYMBOL_GPL(need_ipv4_conntrack);