[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>
#include <net/netfilter/nf_nat_helper.h>

int (*nf_nat_seq_adjust_hook)(struct sk_buff *skb,
			      struct nf_conn *ct,
			      enum ip_conntrack_info ctinfo);
EXPORT_SYMBOL_GPL(nf_nat_seq_adjust_hook);

static bool 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 false;

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

	return true;
}

static bool 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 true;
}

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 *))
{
	struct nf_conn *ct;
	enum ip_conntrack_info ctinfo;
	const struct nf_conn_help *help;
	const struct nf_conntrack_helper *helper;
	unsigned int ret;

	/* 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)
		goto out;

	help = nfct_help(ct);
	if (!help)
		goto out;

	/* rcu_read_lock()ed by nf_hook_slow */
	helper = rcu_dereference(help->helper);
	if (!helper)
		goto out;

	ret = helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
			   ct, ctinfo);
	if (ret != NF_ACCEPT)
		return ret;

	if (test_bit(IPS_SEQ_ADJUST_BIT, &ct->status)) {
		typeof(nf_nat_seq_adjust_hook) seq_adjust;

		seq_adjust = rcu_dereference(nf_nat_seq_adjust_hook);
		if (!seq_adjust || !seq_adjust(skb, ct, ctinfo))
			return NF_DROP;
	}
out:
	/* We've seen it coming out the other side: confirm it */
	return nf_conntrack_confirm(skb);
}

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(dev_net(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(dev_net(out), 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_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		= &init_net.ct.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		= &init_net.ct.sysctl_checksum,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_LOG_INVALID,
		.procname	= "ip_conntrack_log_invalid",
		.data		= &init_net.ct.sysctl_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;

	memset(&tuple, 0, sizeof(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(sock_net(sk), &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>
dd13b010 PM	26	#include <net/netfilter/nf_nat_helper.h>
	27
	28	int (nf_nat_seq_adjust_hook)(struct sk_buff skb,
	29	struct nf_conn *ct,
	30	enum ip_conntrack_info ctinfo);
	31	EXPORT_SYMBOL_GPL(nf_nat_seq_adjust_hook);
9fb9cbb1	32
8ce8439a JE	33	static bool ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
8ce8439a JE	34	struct nf_conntrack_tuple *tuple)
9fb9cbb1	35	{
32948588 JE	36	const __be32 *ap;
32948588 JE	37	__be32 _addrs[2];
9fb9cbb1 YK	38	ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr),
	39	sizeof(u_int32_t) * 2, _addrs);
	40	if (ap == NULL)
8ce8439a	41	return false;
9fb9cbb1 YK	42
	43	tuple->src.u3.ip = ap[0];
	44	tuple->dst.u3.ip = ap[1];
	45
8ce8439a	46	return true;
9fb9cbb1 YK	47	}
9fb9cbb1 YK	48
8ce8439a JE	49	static bool ipv4_invert_tuple(struct nf_conntrack_tuple *tuple,
8ce8439a JE	50	const struct nf_conntrack_tuple *orig)
9fb9cbb1 YK	51	{
	52	tuple->src.u3.ip = orig->dst.u3.ip;
	53	tuple->dst.u3.ip = orig->src.u3.ip;
	54
8ce8439a	55	return true;
9fb9cbb1 YK	56	}
	57
	58	static int ipv4_print_tuple(struct seq_file *s,
	59	const struct nf_conntrack_tuple *tuple)
	60	{
	61	return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ",
e905a9ed	62	NIPQUAD(tuple->src.u3.ip),
9fb9cbb1 YK	63	NIPQUAD(tuple->dst.u3.ip));
	64	}
	65
9fb9cbb1	66	/* Returns new sk_buff, or NULL */
776c729e	67	static int nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user)
9fb9cbb1	68	{
776c729e HX	69	int err;
776c729e HX	70
9fb9cbb1 YK	71	skb_orphan(skb);
9fb9cbb1 YK	72
e905a9ed	73	local_bh_disable();
776c729e	74	err = ip_defrag(skb, user);
e905a9ed	75	local_bh_enable();
9fb9cbb1	76
776c729e	77	if (!err)
eddc9ec5	78	ip_send_check(ip_hdr(skb));
9fb9cbb1	79
776c729e	80	return err;
9fb9cbb1 YK	81	}
9fb9cbb1 YK	82
ffc30690 YK	83	static int ipv4_get_l4proto(const struct sk_buff *skb, unsigned int nhoff,
ffc30690 YK	84	unsigned int dataoff, u_int8_t protonum)
9fb9cbb1	85	{
32948588 JE	86	const struct iphdr *iph;
32948588 JE	87	struct iphdr _iph;
ffc30690 YK	88
	89	iph = skb_header_pointer(skb, nhoff, sizeof(_iph), &_iph);
	90	if (iph == NULL)
	91	return -NF_DROP;
	92
0fb96701 PM	93	/* Conntrack defragments packets, we might still see fragments
	94	* inside ICMP packets though. */
	95	if (iph->frag_off & htons(IP_OFFSET))
9fb9cbb1	96	return -NF_DROP;
9fb9cbb1	97
ffc30690 YK	98	*dataoff = nhoff + (iph->ihl << 2);
ffc30690 YK	99	*protonum = iph->protocol;
9fb9cbb1 YK	100
	101	return NF_ACCEPT;
	102	}
	103
9fb9cbb1	104	static unsigned int ipv4_confirm(unsigned int hooknum,
3db05fea	105	struct sk_buff *skb,
9fb9cbb1 YK	106	const struct net_device *in,
	107	const struct net_device *out,
	108	int (okfn)(struct sk_buff ))
9fb9cbb1 YK	109	{
	110	struct nf_conn *ct;
	111	enum ip_conntrack_info ctinfo;
32948588 JE	112	const struct nf_conn_help *help;
32948588 JE	113	const struct nf_conntrack_helper *helper;
dd13b010	114	unsigned int ret;
9fb9cbb1 YK	115
9fb9cbb1 YK	116	/* This is where we call the helper: as the packet goes out. */
3db05fea	117	ct = nf_ct_get(skb, &ctinfo);
6442f1cf	118	if (!ct \|\| ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
dd13b010	119	goto out;
dc808fe2 HW	120
dc808fe2 HW	121	help = nfct_help(ct);
3c158f7f	122	if (!help)
dd13b010 PM	123	goto out;
dd13b010 PM	124
3c158f7f PM	125	/* rcu_read_lock()ed by nf_hook_slow */
	126	helper = rcu_dereference(help->helper);
	127	if (!helper)
dd13b010 PM	128	goto out;
	129
	130	ret = helper->help(skb, skb_network_offset(skb) + ip_hdrlen(skb),
	131	ct, ctinfo);
	132	if (ret != NF_ACCEPT)
	133	return ret;
	134
	135	if (test_bit(IPS_SEQ_ADJUST_BIT, &ct->status)) {
	136	typeof(nf_nat_seq_adjust_hook) seq_adjust;
	137
	138	seq_adjust = rcu_dereference(nf_nat_seq_adjust_hook);
	139	if (!seq_adjust \|\| !seq_adjust(skb, ct, ctinfo))
	140	return NF_DROP;
	141	}
	142	out:
	143	/* We've seen it coming out the other side: confirm it */
	144	return nf_conntrack_confirm(skb);
9fb9cbb1 YK	145	}
	146
	147	static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
3db05fea	148	struct sk_buff *skb,
9fb9cbb1 YK	149	const struct net_device *in,
	150	const struct net_device *out,
	151	int (okfn)(struct sk_buff ))
	152	{
9fb9cbb1 YK	153	/* Previously seen (loopback)? Ignore. Do this before
9fb9cbb1 YK	154	fragment check. */
3db05fea	155	if (skb->nfct)
9fb9cbb1	156	return NF_ACCEPT;
9fb9cbb1 YK	157
9fb9cbb1 YK	158	/* Gather fragments. */
3db05fea HX	159	if (ip_hdr(skb)->frag_off & htons(IP_MF \| IP_OFFSET)) {
3db05fea HX	160	if (nf_ct_ipv4_gather_frags(skb,
6e23ae2a	161	hooknum == NF_INET_PRE_ROUTING ?
776c729e HX	162	IP_DEFRAG_CONNTRACK_IN :
776c729e HX	163	IP_DEFRAG_CONNTRACK_OUT))
9fb9cbb1 YK	164	return NF_STOLEN;
	165	}
	166	return NF_ACCEPT;
	167	}
	168
9fb9cbb1	169	static unsigned int ipv4_conntrack_in(unsigned int hooknum,
3db05fea	170	struct sk_buff *skb,
9fb9cbb1 YK	171	const struct net_device *in,
	172	const struct net_device *out,
	173	int (okfn)(struct sk_buff ))
	174	{
a702a65f	175	return nf_conntrack_in(dev_net(in), PF_INET, hooknum, skb);
9fb9cbb1 YK	176	}
	177
	178	static unsigned int ipv4_conntrack_local(unsigned int hooknum,
3db05fea	179	struct sk_buff *skb,
e905a9ed YH	180	const struct net_device *in,
	181	const struct net_device *out,
	182	int (okfn)(struct sk_buff ))
9fb9cbb1 YK	183	{
9fb9cbb1 YK	184	/* root is playing with raw sockets. */
3db05fea HX	185	if (skb->len < sizeof(struct iphdr) \|\|
3db05fea HX	186	ip_hdrlen(skb) < sizeof(struct iphdr)) {
9fb9cbb1 YK	187	if (net_ratelimit())
	188	printk("ipt_hook: happy cracking.\n");
	189	return NF_ACCEPT;
	190	}
a702a65f	191	return nf_conntrack_in(dev_net(out), PF_INET, hooknum, skb);
9fb9cbb1 YK	192	}
	193
	194	/* Connection tracking may drop packets, but never alters them, so
	195	make it the first hook. */
1999414a	196	static struct nf_hook_ops ipv4_conntrack_ops[] __read_mostly = {
964ddaa1 PM	197	{
	198	.hook = ipv4_conntrack_defrag,
	199	.owner = THIS_MODULE,
	200	.pf = PF_INET,
6e23ae2a	201	.hooknum = NF_INET_PRE_ROUTING,
964ddaa1 PM	202	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	203	},
	204	{
	205	.hook = ipv4_conntrack_in,
	206	.owner = THIS_MODULE,
	207	.pf = PF_INET,
6e23ae2a	208	.hooknum = NF_INET_PRE_ROUTING,
964ddaa1 PM	209	.priority = NF_IP_PRI_CONNTRACK,
	210	},
	211	{
	212	.hook = ipv4_conntrack_defrag,
	213	.owner = THIS_MODULE,
	214	.pf = PF_INET,
6e23ae2a	215	.hooknum = NF_INET_LOCAL_OUT,
964ddaa1 PM	216	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	217	},
	218	{
	219	.hook = ipv4_conntrack_local,
	220	.owner = THIS_MODULE,
	221	.pf = PF_INET,
6e23ae2a	222	.hooknum = NF_INET_LOCAL_OUT,
964ddaa1 PM	223	.priority = NF_IP_PRI_CONNTRACK,
964ddaa1 PM	224	},
964ddaa1 PM	225	{
	226	.hook = ipv4_confirm,
	227	.owner = THIS_MODULE,
	228	.pf = PF_INET,
6e23ae2a	229	.hooknum = NF_INET_POST_ROUTING,
964ddaa1 PM	230	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	231	},
	232	{
	233	.hook = ipv4_confirm,
	234	.owner = THIS_MODULE,
	235	.pf = PF_INET,
6e23ae2a	236	.hooknum = NF_INET_LOCAL_IN,
964ddaa1 PM	237	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
964ddaa1 PM	238	},
9fb9cbb1 YK	239	};
9fb9cbb1 YK	240
a999e683 PM	241	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	242	static int log_invalid_proto_min = 0;
	243	static int log_invalid_proto_max = 255;
	244
	245	static ctl_table ip_ct_sysctl_table[] = {
	246	{
	247	.ctl_name = NET_IPV4_NF_CONNTRACK_MAX,
	248	.procname = "ip_conntrack_max",
	249	.data = &nf_conntrack_max,
	250	.maxlen = sizeof(int),
	251	.mode = 0644,
	252	.proc_handler = &proc_dointvec,
	253	},
	254	{
	255	.ctl_name = NET_IPV4_NF_CONNTRACK_COUNT,
	256	.procname = "ip_conntrack_count",
49ac8713	257	.data = &init_net.ct.count,
a999e683 PM	258	.maxlen = sizeof(int),
	259	.mode = 0444,
	260	.proc_handler = &proc_dointvec,
	261	},
	262	{
	263	.ctl_name = NET_IPV4_NF_CONNTRACK_BUCKETS,
	264	.procname = "ip_conntrack_buckets",
	265	.data = &nf_conntrack_htable_size,
	266	.maxlen = sizeof(unsigned int),
	267	.mode = 0444,
	268	.proc_handler = &proc_dointvec,
	269	},
	270	{
	271	.ctl_name = NET_IPV4_NF_CONNTRACK_CHECKSUM,
	272	.procname = "ip_conntrack_checksum",
c04d0552	273	.data = &init_net.ct.sysctl_checksum,
a999e683 PM	274	.maxlen = sizeof(int),
	275	.mode = 0644,
	276	.proc_handler = &proc_dointvec,
	277	},
	278	{
	279	.ctl_name = NET_IPV4_NF_CONNTRACK_LOG_INVALID,
	280	.procname = "ip_conntrack_log_invalid",
c2a2c7e0	281	.data = &init_net.ct.sysctl_log_invalid,
a999e683 PM	282	.maxlen = sizeof(unsigned int),
	283	.mode = 0644,
	284	.proc_handler = &proc_dointvec_minmax,
	285	.strategy = &sysctl_intvec,
	286	.extra1 = &log_invalid_proto_min,
	287	.extra2 = &log_invalid_proto_max,
	288	},
	289	{
	290	.ctl_name = 0
	291	}
	292	};
	293	#endif /* CONFIG_SYSCTL && CONFIG_NF_CONNTRACK_PROC_COMPAT */
	294
9fb9cbb1 YK	295	/* Fast function for those who don't want to parse /proc (and I don't
	296	blame them). */
	297	/* Reversing the socket's dst/src point of view gives us the reply
	298	mapping. */
	299	static int
	300	getorigdst(struct sock sk, int optval, void __user user, int *len)
	301	{
32948588 JE	302	const struct inet_sock *inet = inet_sk(sk);
32948588 JE	303	const struct nf_conntrack_tuple_hash *h;
9fb9cbb1	304	struct nf_conntrack_tuple tuple;
e905a9ed	305
443a70d5	306	memset(&tuple, 0, sizeof(tuple));
9fb9cbb1 YK	307	tuple.src.u3.ip = inet->rcv_saddr;
	308	tuple.src.u.tcp.port = inet->sport;
	309	tuple.dst.u3.ip = inet->daddr;
	310	tuple.dst.u.tcp.port = inet->dport;
	311	tuple.src.l3num = PF_INET;
	312	tuple.dst.protonum = IPPROTO_TCP;
	313
	314	/* We only do TCP at the moment: is there a better way? */
	315	if (strcmp(sk->sk_prot->name, "TCP")) {
0d53778e	316	pr_debug("SO_ORIGINAL_DST: Not a TCP socket\n");
9fb9cbb1 YK	317	return -ENOPROTOOPT;
	318	}
	319
	320	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
0d53778e PM	321	pr_debug("SO_ORIGINAL_DST: len %d not %Zu\n",
0d53778e PM	322	*len, sizeof(struct sockaddr_in));
9fb9cbb1 YK	323	return -EINVAL;
	324	}
	325
400dad39	326	h = nf_conntrack_find_get(sock_net(sk), &tuple);
9fb9cbb1 YK	327	if (h) {
	328	struct sockaddr_in sin;
	329	struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
	330
	331	sin.sin_family = AF_INET;
	332	sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	333	.tuple.dst.u.tcp.port;
	334	sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	335	.tuple.dst.u3.ip;
6c813c3f	336	memset(sin.sin_zero, 0, sizeof(sin.sin_zero));
9fb9cbb1	337
0d53778e PM	338	pr_debug("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
0d53778e PM	339	NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
9fb9cbb1 YK	340	nf_ct_put(ct);
	341	if (copy_to_user(user, &sin, sizeof(sin)) != 0)
	342	return -EFAULT;
	343	else
	344	return 0;
	345	}
0d53778e PM	346	pr_debug("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
	347	NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
	348	NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
9fb9cbb1 YK	349	return -ENOENT;
	350	}
	351
e281db5c	352	#if defined(CONFIG_NF_CT_NETLINK) \|\| defined(CONFIG_NF_CT_NETLINK_MODULE)
c1d10adb PNA	353
	354	#include <linux/netfilter/nfnetlink.h>
	355	#include <linux/netfilter/nfnetlink_conntrack.h>
	356
fdf70832	357	static int ipv4_tuple_to_nlattr(struct sk_buff *skb,
c1d10adb PNA	358	const struct nf_conntrack_tuple *tuple)
c1d10adb PNA	359	{
77236b6e PM	360	NLA_PUT_BE32(skb, CTA_IP_V4_SRC, tuple->src.u3.ip);
77236b6e PM	361	NLA_PUT_BE32(skb, CTA_IP_V4_DST, tuple->dst.u3.ip);
c1d10adb PNA	362	return 0;
c1d10adb PNA	363
df6fb868	364	nla_put_failure:
c1d10adb PNA	365	return -1;
	366	}
	367
f73e924c PM	368	static const struct nla_policy ipv4_nla_policy[CTA_IP_MAX+1] = {
	369	[CTA_IP_V4_SRC] = { .type = NLA_U32 },
	370	[CTA_IP_V4_DST] = { .type = NLA_U32 },
c1d10adb PNA	371	};
c1d10adb PNA	372
fdf70832	373	static int ipv4_nlattr_to_tuple(struct nlattr *tb[],
c1d10adb PNA	374	struct nf_conntrack_tuple *t)
c1d10adb PNA	375	{
df6fb868	376	if (!tb[CTA_IP_V4_SRC] \|\| !tb[CTA_IP_V4_DST])
c1d10adb PNA	377	return -EINVAL;
c1d10adb PNA	378
77236b6e PM	379	t->src.u3.ip = nla_get_be32(tb[CTA_IP_V4_SRC]);
77236b6e PM	380	t->dst.u3.ip = nla_get_be32(tb[CTA_IP_V4_DST]);
c1d10adb PNA	381
	382	return 0;
	383	}
	384	#endif
	385
9fb9cbb1 YK	386	static struct nf_sockopt_ops so_getorigdst = {
	387	.pf = PF_INET,
	388	.get_optmin = SO_ORIGINAL_DST,
	389	.get_optmax = SO_ORIGINAL_DST+1,
	390	.get = &getorigdst,
16fcec35	391	.owner = THIS_MODULE,
9fb9cbb1 YK	392	};
9fb9cbb1 YK	393
61075af5	394	struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 __read_mostly = {
9fb9cbb1 YK	395	.l3proto = PF_INET,
	396	.name = "ipv4",
	397	.pkt_to_tuple = ipv4_pkt_to_tuple,
	398	.invert_tuple = ipv4_invert_tuple,
	399	.print_tuple = ipv4_print_tuple,
ffc30690	400	.get_l4proto = ipv4_get_l4proto,
e281db5c	401	#if defined(CONFIG_NF_CT_NETLINK) \|\| defined(CONFIG_NF_CT_NETLINK_MODULE)
fdf70832 PM	402	.tuple_to_nlattr = ipv4_tuple_to_nlattr,
fdf70832 PM	403	.nlattr_to_tuple = ipv4_nlattr_to_tuple,
f73e924c	404	.nla_policy = ipv4_nla_policy,
a999e683 PM	405	#endif
	406	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	407	.ctl_table_path = nf_net_ipv4_netfilter_sysctl_path,
	408	.ctl_table = ip_ct_sysctl_table,
c1d10adb	409	#endif
9fb9cbb1 YK	410	.me = THIS_MODULE,
	411	};
	412
fae718dd PM	413	module_param_call(hashsize, nf_conntrack_set_hashsize, param_get_uint,
	414	&nf_conntrack_htable_size, 0600);
	415
32292a7f	416	MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
d2483dde	417	MODULE_ALIAS("ip_conntrack");
32292a7f PM	418	MODULE_LICENSE("GPL");
	419
	420	static int __init nf_conntrack_l3proto_ipv4_init(void)
9fb9cbb1 YK	421	{
	422	int ret = 0;
	423
32292a7f	424	need_conntrack();
9fb9cbb1 YK	425
	426	ret = nf_register_sockopt(&so_getorigdst);
	427	if (ret < 0) {
	428	printk(KERN_ERR "Unable to register netfilter socket option\n");
32292a7f	429	return ret;
9fb9cbb1 YK	430	}
9fb9cbb1 YK	431
605dcad6	432	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	433	if (ret < 0) {
	434	printk("nf_conntrack_ipv4: can't register tcp.\n");
	435	goto cleanup_sockopt;
	436	}
	437
605dcad6	438	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
9fb9cbb1 YK	439	if (ret < 0) {
	440	printk("nf_conntrack_ipv4: can't register udp.\n");
	441	goto cleanup_tcp;
	442	}
	443
605dcad6	444	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
9fb9cbb1 YK	445	if (ret < 0) {
	446	printk("nf_conntrack_ipv4: can't register icmp.\n");
	447	goto cleanup_udp;
	448	}
	449
	450	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	451	if (ret < 0) {
	452	printk("nf_conntrack_ipv4: can't register ipv4\n");
	453	goto cleanup_icmp;
	454	}
	455
964ddaa1 PM	456	ret = nf_register_hooks(ipv4_conntrack_ops,
964ddaa1 PM	457	ARRAY_SIZE(ipv4_conntrack_ops));
9fb9cbb1	458	if (ret < 0) {
964ddaa1	459	printk("nf_conntrack_ipv4: can't register hooks.\n");
9fb9cbb1 YK	460	goto cleanup_ipv4;
9fb9cbb1 YK	461	}
e4bd8bce PM	462	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	463	ret = nf_conntrack_ipv4_compat_init();
	464	if (ret < 0)
	465	goto cleanup_hooks;
	466	#endif
9fb9cbb1	467	return ret;
e4bd8bce PM	468	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
e4bd8bce PM	469	cleanup_hooks:
e905a9ed	470	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
e4bd8bce	471	#endif
9fb9cbb1 YK	472	cleanup_ipv4:
	473	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	474	cleanup_icmp:
605dcad6	475	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
9fb9cbb1	476	cleanup_udp:
605dcad6	477	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
9fb9cbb1	478	cleanup_tcp:
605dcad6	479	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	480	cleanup_sockopt:
9fb9cbb1 YK	481	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	482	return ret;
	483	}
	484
65b4b4e8	485	static void __exit nf_conntrack_l3proto_ipv4_fini(void)
9fb9cbb1	486	{
32292a7f	487	synchronize_net();
e4bd8bce PM	488	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	489	nf_conntrack_ipv4_compat_fini();
	490	#endif
32292a7f PM	491	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
32292a7f PM	492	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
605dcad6 MJ	493	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	494	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	495	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
32292a7f	496	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	497	}
9fb9cbb1 YK	498
65b4b4e8 AM	499	module_init(nf_conntrack_l3proto_ipv4_init);
65b4b4e8 AM	500	module_exit(nf_conntrack_l3proto_ipv4_fini);
591e6206 PM	501
	502	void need_ipv4_conntrack(void)
	503	{
	504	return;
	505	}
	506	EXPORT_SYMBOL_GPL(need_ipv4_conntrack);