[deliverable/linux.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)
{
	__be32 _addrs[2], *ap;
	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));
}

static int ipv4_print_conntrack(struct seq_file *s,
				const struct nf_conn *conntrack)
{
	return 0;
}

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

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

	if (skb)
		ip_send_check(ip_hdr(skb));

	return skb;
}

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

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

	/* This is where we call the helper: as the packet goes out. */
	ct = nf_ct_get(*pskb, &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(pskb, skb_network_offset(*pskb) + ip_hdrlen(*pskb),
			    ct, ctinfo);
}

static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
					  struct sk_buff **pskb,
					  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 ((*pskb)->nfct)
		return NF_ACCEPT;

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

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

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

/* Connection tracking may drop packets, but never alters them, so
   make it the first hook. */
static struct nf_hook_ops ipv4_conntrack_ops[] = {
	{
		.hook		= ipv4_conntrack_defrag,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_PRE_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv4_conntrack_in,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_PRE_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK,
	},
	{
		.hook           = ipv4_conntrack_defrag,
		.owner          = THIS_MODULE,
		.pf             = PF_INET,
		.hooknum        = NF_IP_LOCAL_OUT,
		.priority       = NF_IP_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv4_conntrack_local,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_LOCAL_OUT,
		.priority	= NF_IP_PRI_CONNTRACK,
	},
	{
		.hook		= ipv4_conntrack_help,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
	},
	{
		.hook		= ipv4_conntrack_help,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
	},
	{
		.hook		= ipv4_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM,
	},
	{
		.hook		= ipv4_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_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)
{
	struct inet_sock *inet = inet_sk(sk);
	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_nfattr(struct sk_buff *skb,
				const struct nf_conntrack_tuple *tuple)
{
	NFA_PUT(skb, CTA_IP_V4_SRC, sizeof(u_int32_t),
		&tuple->src.u3.ip);
	NFA_PUT(skb, CTA_IP_V4_DST, sizeof(u_int32_t),
		&tuple->dst.u3.ip);
	return 0;

nfattr_failure:
	return -1;
}

static const size_t cta_min_ip[CTA_IP_MAX] = {
	[CTA_IP_V4_SRC-1]       = sizeof(u_int32_t),
	[CTA_IP_V4_DST-1]       = sizeof(u_int32_t),
};

static int ipv4_nfattr_to_tuple(struct nfattr *tb[],
				struct nf_conntrack_tuple *t)
{
	if (!tb[CTA_IP_V4_SRC-1] || !tb[CTA_IP_V4_DST-1])
		return -EINVAL;

	if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip))
		return -EINVAL;

	t->src.u3.ip = *(__be32 *)NFA_DATA(tb[CTA_IP_V4_SRC-1]);
	t->dst.u3.ip = *(__be32 *)NFA_DATA(tb[CTA_IP_V4_DST-1]);

	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,
	.print_conntrack = ipv4_print_conntrack,
	.get_l4proto	 = ipv4_get_l4proto,
#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)
	.tuple_to_nfattr = ipv4_tuple_to_nfattr,
	.nfattr_to_tuple = ipv4_nfattr_to_tuple,
#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_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	{
bff9a89b	30	__be32 _addrs[2], *ap;
9fb9cbb1 YK	31	ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr),
	32	sizeof(u_int32_t) * 2, _addrs);
	33	if (ap == NULL)
	34	return 0;
	35
	36	tuple->src.u3.ip = ap[0];
	37	tuple->dst.u3.ip = ap[1];
	38
	39	return 1;
	40	}
	41
	42	static int ipv4_invert_tuple(struct nf_conntrack_tuple *tuple,
	43	const struct nf_conntrack_tuple *orig)
	44	{
	45	tuple->src.u3.ip = orig->dst.u3.ip;
	46	tuple->dst.u3.ip = orig->src.u3.ip;
	47
	48	return 1;
	49	}
	50
	51	static int ipv4_print_tuple(struct seq_file *s,
	52	const struct nf_conntrack_tuple *tuple)
	53	{
	54	return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ",
e905a9ed	55	NIPQUAD(tuple->src.u3.ip),
9fb9cbb1 YK	56	NIPQUAD(tuple->dst.u3.ip));
	57	}
	58
	59	static int ipv4_print_conntrack(struct seq_file *s,
	60	const struct nf_conn *conntrack)
	61	{
	62	return 0;
	63	}
	64
	65	/* Returns new sk_buff, or NULL */
	66	static struct sk_buff *
	67	nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user)
	68	{
	69	skb_orphan(skb);
	70
e905a9ed YH	71	local_bh_disable();
	72	skb = ip_defrag(skb, user);
	73	local_bh_enable();
9fb9cbb1	74
e905a9ed	75	if (skb)
eddc9ec5	76	ip_send_check(ip_hdr(skb));
9fb9cbb1	77
e905a9ed	78	return skb;
9fb9cbb1 YK	79	}
9fb9cbb1 YK	80
ffc30690 YK	81	static int ipv4_get_l4proto(const struct sk_buff *skb, unsigned int nhoff,
ffc30690 YK	82	unsigned int dataoff, u_int8_t protonum)
9fb9cbb1	83	{
ffc30690 YK	84	struct iphdr _iph, *iph;
	85
	86	iph = skb_header_pointer(skb, nhoff, sizeof(_iph), &_iph);
	87	if (iph == NULL)
	88	return -NF_DROP;
	89
0fb96701 PM	90	/* Conntrack defragments packets, we might still see fragments
	91	* inside ICMP packets though. */
	92	if (iph->frag_off & htons(IP_OFFSET))
9fb9cbb1	93	return -NF_DROP;
9fb9cbb1	94
ffc30690 YK	95	*dataoff = nhoff + (iph->ihl << 2);
ffc30690 YK	96	*protonum = iph->protocol;
9fb9cbb1 YK	97
	98	return NF_ACCEPT;
	99	}
	100
9fb9cbb1 YK	101	static unsigned int ipv4_confirm(unsigned int hooknum,
	102	struct sk_buff **pskb,
	103	const struct net_device *in,
	104	const struct net_device *out,
	105	int (okfn)(struct sk_buff ))
	106	{
	107	/* We've seen it coming out the other side: confirm it */
	108	return nf_conntrack_confirm(pskb);
	109	}
	110
	111	static unsigned int ipv4_conntrack_help(unsigned int hooknum,
	112	struct sk_buff **pskb,
	113	const struct net_device *in,
	114	const struct net_device *out,
	115	int (okfn)(struct sk_buff ))
	116	{
	117	struct nf_conn *ct;
	118	enum ip_conntrack_info ctinfo;
dc808fe2	119	struct nf_conn_help *help;
3c158f7f	120	struct nf_conntrack_helper *helper;
9fb9cbb1 YK	121
	122	/* This is where we call the helper: as the packet goes out. */
	123	ct = nf_ct_get(*pskb, &ctinfo);
6442f1cf	124	if (!ct \|\| ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
dc808fe2 HW	125	return NF_ACCEPT;
	126
	127	help = nfct_help(ct);
3c158f7f	128	if (!help)
dc808fe2	129	return NF_ACCEPT;
3c158f7f PM	130	/* rcu_read_lock()ed by nf_hook_slow */
	131	helper = rcu_dereference(help->helper);
	132	if (!helper)
	133	return NF_ACCEPT;
	134	return helper->help(pskb, skb_network_offset(pskb) + ip_hdrlen(pskb),
	135	ct, ctinfo);
9fb9cbb1 YK	136	}
	137
	138	static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
	139	struct sk_buff **pskb,
	140	const struct net_device *in,
	141	const struct net_device *out,
	142	int (okfn)(struct sk_buff ))
	143	{
9fb9cbb1 YK	144	/* Previously seen (loopback)? Ignore. Do this before
	145	fragment check. */
	146	if ((*pskb)->nfct)
	147	return NF_ACCEPT;
9fb9cbb1 YK	148
9fb9cbb1 YK	149	/* Gather fragments. */
eddc9ec5	150	if (ip_hdr(*pskb)->frag_off & htons(IP_MF \| IP_OFFSET)) {
9fb9cbb1 YK	151	pskb = nf_ct_ipv4_gather_frags(pskb,
	152	hooknum == NF_IP_PRE_ROUTING ?
	153	IP_DEFRAG_CONNTRACK_IN :
	154	IP_DEFRAG_CONNTRACK_OUT);
	155	if (!*pskb)
	156	return NF_STOLEN;
	157	}
	158	return NF_ACCEPT;
	159	}
	160
9fb9cbb1 YK	161	static unsigned int ipv4_conntrack_in(unsigned int hooknum,
	162	struct sk_buff **pskb,
	163	const struct net_device *in,
	164	const struct net_device *out,
	165	int (okfn)(struct sk_buff ))
	166	{
	167	return nf_conntrack_in(PF_INET, hooknum, pskb);
	168	}
	169
	170	static unsigned int ipv4_conntrack_local(unsigned int hooknum,
e905a9ed YH	171	struct sk_buff **pskb,
	172	const struct net_device *in,
	173	const struct net_device *out,
	174	int (okfn)(struct sk_buff ))
9fb9cbb1 YK	175	{
	176	/* root is playing with raw sockets. */
	177	if ((*pskb)->len < sizeof(struct iphdr)
c9bdd4b5	178	\|\| ip_hdrlen(*pskb) < sizeof(struct iphdr)) {
9fb9cbb1 YK	179	if (net_ratelimit())
	180	printk("ipt_hook: happy cracking.\n");
	181	return NF_ACCEPT;
	182	}
	183	return nf_conntrack_in(PF_INET, hooknum, pskb);
	184	}
	185
	186	/* Connection tracking may drop packets, but never alters them, so
	187	make it the first hook. */
964ddaa1 PM	188	static struct nf_hook_ops ipv4_conntrack_ops[] = {
	189	{
	190	.hook = ipv4_conntrack_defrag,
	191	.owner = THIS_MODULE,
	192	.pf = PF_INET,
	193	.hooknum = NF_IP_PRE_ROUTING,
	194	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	195	},
	196	{
	197	.hook = ipv4_conntrack_in,
	198	.owner = THIS_MODULE,
	199	.pf = PF_INET,
	200	.hooknum = NF_IP_PRE_ROUTING,
	201	.priority = NF_IP_PRI_CONNTRACK,
	202	},
	203	{
	204	.hook = ipv4_conntrack_defrag,
	205	.owner = THIS_MODULE,
	206	.pf = PF_INET,
	207	.hooknum = NF_IP_LOCAL_OUT,
	208	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	209	},
	210	{
	211	.hook = ipv4_conntrack_local,
	212	.owner = THIS_MODULE,
	213	.pf = PF_INET,
	214	.hooknum = NF_IP_LOCAL_OUT,
	215	.priority = NF_IP_PRI_CONNTRACK,
	216	},
	217	{
	218	.hook = ipv4_conntrack_help,
	219	.owner = THIS_MODULE,
	220	.pf = PF_INET,
	221	.hooknum = NF_IP_POST_ROUTING,
	222	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	223	},
	224	{
	225	.hook = ipv4_conntrack_help,
	226	.owner = THIS_MODULE,
	227	.pf = PF_INET,
	228	.hooknum = NF_IP_LOCAL_IN,
	229	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	230	},
	231	{
	232	.hook = ipv4_confirm,
	233	.owner = THIS_MODULE,
	234	.pf = PF_INET,
	235	.hooknum = NF_IP_POST_ROUTING,
	236	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	237	},
	238	{
	239	.hook = ipv4_confirm,
	240	.owner = THIS_MODULE,
	241	.pf = PF_INET,
	242	.hooknum = NF_IP_LOCAL_IN,
	243	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	244	},
9fb9cbb1 YK	245	};
9fb9cbb1 YK	246
a999e683 PM	247	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	248	static int log_invalid_proto_min = 0;
	249	static int log_invalid_proto_max = 255;
	250
	251	static ctl_table ip_ct_sysctl_table[] = {
	252	{
	253	.ctl_name = NET_IPV4_NF_CONNTRACK_MAX,
	254	.procname = "ip_conntrack_max",
	255	.data = &nf_conntrack_max,
	256	.maxlen = sizeof(int),
	257	.mode = 0644,
	258	.proc_handler = &proc_dointvec,
	259	},
	260	{
	261	.ctl_name = NET_IPV4_NF_CONNTRACK_COUNT,
	262	.procname = "ip_conntrack_count",
	263	.data = &nf_conntrack_count,
	264	.maxlen = sizeof(int),
	265	.mode = 0444,
	266	.proc_handler = &proc_dointvec,
	267	},
	268	{
	269	.ctl_name = NET_IPV4_NF_CONNTRACK_BUCKETS,
	270	.procname = "ip_conntrack_buckets",
	271	.data = &nf_conntrack_htable_size,
	272	.maxlen = sizeof(unsigned int),
	273	.mode = 0444,
	274	.proc_handler = &proc_dointvec,
	275	},
	276	{
	277	.ctl_name = NET_IPV4_NF_CONNTRACK_CHECKSUM,
	278	.procname = "ip_conntrack_checksum",
	279	.data = &nf_conntrack_checksum,
	280	.maxlen = sizeof(int),
	281	.mode = 0644,
	282	.proc_handler = &proc_dointvec,
	283	},
	284	{
	285	.ctl_name = NET_IPV4_NF_CONNTRACK_LOG_INVALID,
	286	.procname = "ip_conntrack_log_invalid",
	287	.data = &nf_ct_log_invalid,
	288	.maxlen = sizeof(unsigned int),
	289	.mode = 0644,
	290	.proc_handler = &proc_dointvec_minmax,
	291	.strategy = &sysctl_intvec,
	292	.extra1 = &log_invalid_proto_min,
	293	.extra2 = &log_invalid_proto_max,
	294	},
	295	{
	296	.ctl_name = 0
	297	}
	298	};
	299	#endif /* CONFIG_SYSCTL && CONFIG_NF_CONNTRACK_PROC_COMPAT */
	300
9fb9cbb1 YK	301	/* Fast function for those who don't want to parse /proc (and I don't
	302	blame them). */
	303	/* Reversing the socket's dst/src point of view gives us the reply
	304	mapping. */
	305	static int
	306	getorigdst(struct sock sk, int optval, void __user user, int *len)
	307	{
	308	struct inet_sock *inet = inet_sk(sk);
	309	struct nf_conntrack_tuple_hash *h;
	310	struct nf_conntrack_tuple tuple;
e905a9ed	311
9fb9cbb1 YK	312	NF_CT_TUPLE_U_BLANK(&tuple);
	313	tuple.src.u3.ip = inet->rcv_saddr;
	314	tuple.src.u.tcp.port = inet->sport;
	315	tuple.dst.u3.ip = inet->daddr;
	316	tuple.dst.u.tcp.port = inet->dport;
	317	tuple.src.l3num = PF_INET;
	318	tuple.dst.protonum = IPPROTO_TCP;
	319
	320	/* We only do TCP at the moment: is there a better way? */
	321	if (strcmp(sk->sk_prot->name, "TCP")) {
0d53778e	322	pr_debug("SO_ORIGINAL_DST: Not a TCP socket\n");
9fb9cbb1 YK	323	return -ENOPROTOOPT;
	324	}
	325
	326	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
0d53778e PM	327	pr_debug("SO_ORIGINAL_DST: len %d not %Zu\n",
0d53778e PM	328	*len, sizeof(struct sockaddr_in));
9fb9cbb1 YK	329	return -EINVAL;
	330	}
	331
330f7db5	332	h = nf_conntrack_find_get(&tuple);
9fb9cbb1 YK	333	if (h) {
	334	struct sockaddr_in sin;
	335	struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
	336
	337	sin.sin_family = AF_INET;
	338	sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	339	.tuple.dst.u.tcp.port;
	340	sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	341	.tuple.dst.u3.ip;
6c813c3f	342	memset(sin.sin_zero, 0, sizeof(sin.sin_zero));
9fb9cbb1	343
0d53778e PM	344	pr_debug("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
0d53778e PM	345	NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
9fb9cbb1 YK	346	nf_ct_put(ct);
	347	if (copy_to_user(user, &sin, sizeof(sin)) != 0)
	348	return -EFAULT;
	349	else
	350	return 0;
	351	}
0d53778e PM	352	pr_debug("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
	353	NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
	354	NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
9fb9cbb1 YK	355	return -ENOENT;
	356	}
	357
e281db5c	358	#if defined(CONFIG_NF_CT_NETLINK) \|\| defined(CONFIG_NF_CT_NETLINK_MODULE)
c1d10adb PNA	359
	360	#include <linux/netfilter/nfnetlink.h>
	361	#include <linux/netfilter/nfnetlink_conntrack.h>
	362
	363	static int ipv4_tuple_to_nfattr(struct sk_buff *skb,
	364	const struct nf_conntrack_tuple *tuple)
	365	{
	366	NFA_PUT(skb, CTA_IP_V4_SRC, sizeof(u_int32_t),
	367	&tuple->src.u3.ip);
	368	NFA_PUT(skb, CTA_IP_V4_DST, sizeof(u_int32_t),
	369	&tuple->dst.u3.ip);
	370	return 0;
	371
	372	nfattr_failure:
	373	return -1;
	374	}
	375
	376	static const size_t cta_min_ip[CTA_IP_MAX] = {
	377	[CTA_IP_V4_SRC-1] = sizeof(u_int32_t),
	378	[CTA_IP_V4_DST-1] = sizeof(u_int32_t),
	379	};
	380
	381	static int ipv4_nfattr_to_tuple(struct nfattr *tb[],
	382	struct nf_conntrack_tuple *t)
	383	{
	384	if (!tb[CTA_IP_V4_SRC-1] \|\| !tb[CTA_IP_V4_DST-1])
	385	return -EINVAL;
	386
	387	if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip))
	388	return -EINVAL;
	389
bff9a89b PM	390	t->src.u3.ip = (__be32 )NFA_DATA(tb[CTA_IP_V4_SRC-1]);
bff9a89b PM	391	t->dst.u3.ip = (__be32 )NFA_DATA(tb[CTA_IP_V4_DST-1]);
c1d10adb PNA	392
	393	return 0;
	394	}
	395	#endif
	396
9fb9cbb1 YK	397	static struct nf_sockopt_ops so_getorigdst = {
	398	.pf = PF_INET,
	399	.get_optmin = SO_ORIGINAL_DST,
	400	.get_optmax = SO_ORIGINAL_DST+1,
	401	.get = &getorigdst,
16fcec35	402	.owner = THIS_MODULE,
9fb9cbb1 YK	403	};
9fb9cbb1 YK	404
61075af5	405	struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 __read_mostly = {
9fb9cbb1 YK	406	.l3proto = PF_INET,
	407	.name = "ipv4",
	408	.pkt_to_tuple = ipv4_pkt_to_tuple,
	409	.invert_tuple = ipv4_invert_tuple,
	410	.print_tuple = ipv4_print_tuple,
	411	.print_conntrack = ipv4_print_conntrack,
ffc30690	412	.get_l4proto = ipv4_get_l4proto,
e281db5c	413	#if defined(CONFIG_NF_CT_NETLINK) \|\| defined(CONFIG_NF_CT_NETLINK_MODULE)
c1d10adb PNA	414	.tuple_to_nfattr = ipv4_tuple_to_nfattr,
c1d10adb PNA	415	.nfattr_to_tuple = ipv4_nfattr_to_tuple,
a999e683 PM	416	#endif
	417	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	418	.ctl_table_path = nf_net_ipv4_netfilter_sysctl_path,
	419	.ctl_table = ip_ct_sysctl_table,
c1d10adb	420	#endif
9fb9cbb1 YK	421	.me = THIS_MODULE,
	422	};
	423
32292a7f	424	MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
d2483dde	425	MODULE_ALIAS("ip_conntrack");
32292a7f PM	426	MODULE_LICENSE("GPL");
	427
	428	static int __init nf_conntrack_l3proto_ipv4_init(void)
9fb9cbb1 YK	429	{
	430	int ret = 0;
	431
32292a7f	432	need_conntrack();
9fb9cbb1 YK	433
	434	ret = nf_register_sockopt(&so_getorigdst);
	435	if (ret < 0) {
	436	printk(KERN_ERR "Unable to register netfilter socket option\n");
32292a7f	437	return ret;
9fb9cbb1 YK	438	}
9fb9cbb1 YK	439
605dcad6	440	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	441	if (ret < 0) {
	442	printk("nf_conntrack_ipv4: can't register tcp.\n");
	443	goto cleanup_sockopt;
	444	}
	445
605dcad6	446	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
9fb9cbb1 YK	447	if (ret < 0) {
	448	printk("nf_conntrack_ipv4: can't register udp.\n");
	449	goto cleanup_tcp;
	450	}
	451
605dcad6	452	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
9fb9cbb1 YK	453	if (ret < 0) {
	454	printk("nf_conntrack_ipv4: can't register icmp.\n");
	455	goto cleanup_udp;
	456	}
	457
	458	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	459	if (ret < 0) {
	460	printk("nf_conntrack_ipv4: can't register ipv4\n");
	461	goto cleanup_icmp;
	462	}
	463
964ddaa1 PM	464	ret = nf_register_hooks(ipv4_conntrack_ops,
964ddaa1 PM	465	ARRAY_SIZE(ipv4_conntrack_ops));
9fb9cbb1	466	if (ret < 0) {
964ddaa1	467	printk("nf_conntrack_ipv4: can't register hooks.\n");
9fb9cbb1 YK	468	goto cleanup_ipv4;
9fb9cbb1 YK	469	}
e4bd8bce PM	470	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	471	ret = nf_conntrack_ipv4_compat_init();
	472	if (ret < 0)
	473	goto cleanup_hooks;
	474	#endif
9fb9cbb1	475	return ret;
e4bd8bce PM	476	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
e4bd8bce PM	477	cleanup_hooks:
e905a9ed	478	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
e4bd8bce	479	#endif
9fb9cbb1 YK	480	cleanup_ipv4:
	481	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	482	cleanup_icmp:
605dcad6	483	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
9fb9cbb1	484	cleanup_udp:
605dcad6	485	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
9fb9cbb1	486	cleanup_tcp:
605dcad6	487	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	488	cleanup_sockopt:
9fb9cbb1 YK	489	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	490	return ret;
	491	}
	492
65b4b4e8	493	static void __exit nf_conntrack_l3proto_ipv4_fini(void)
9fb9cbb1	494	{
32292a7f	495	synchronize_net();
e4bd8bce PM	496	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	497	nf_conntrack_ipv4_compat_fini();
	498	#endif
32292a7f PM	499	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
32292a7f PM	500	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
605dcad6 MJ	501	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	502	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	503	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
32292a7f	504	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	505	}
9fb9cbb1 YK	506
65b4b4e8 AM	507	module_init(nf_conntrack_l3proto_ipv4_init);
65b4b4e8 AM	508	module_exit(nf_conntrack_l3proto_ipv4_fini);
591e6206 PM	509
	510	void need_ipv4_conntrack(void)
	511	{
	512	return;
	513	}
	514	EXPORT_SYMBOL_GPL(need_ipv4_conntrack);