[mirror_ubuntu-artful-kernel.git] / net / ipv4 / netfilter / nf_nat_l3proto_ipv4.c

/*
 * (C) 1999-2001 Paul `Rusty' Russell
 * (C) 2002-2006 Netfilter Core Team <coreteam@netfilter.org>
 * (C) 2011 Patrick McHardy <kaber@trash.net>
 *
 * 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/module.h>
#include <linux/skbuff.h>
#include <linux/ip.h>
#include <linux/icmp.h>
#include <linux/netfilter.h>
#include <linux/netfilter_ipv4.h>
#include <net/secure_seq.h>
#include <net/checksum.h>
#include <net/route.h>
#include <net/ip.h>

#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_nat_core.h>
#include <net/netfilter/nf_nat_l3proto.h>
#include <net/netfilter/nf_nat_l4proto.h>

static const struct nf_nat_l3proto nf_nat_l3proto_ipv4;

#ifdef CONFIG_XFRM
static void nf_nat_ipv4_decode_session(struct sk_buff *skb,
				       const struct nf_conn *ct,
				       enum ip_conntrack_dir dir,
				       unsigned long statusbit,
				       struct flowi *fl)
{
	const struct nf_conntrack_tuple *t = &ct->tuplehash[dir].tuple;
	struct flowi4 *fl4 = &fl->u.ip4;

	if (ct->status & statusbit) {
		fl4->daddr = t->dst.u3.ip;
		if (t->dst.protonum == IPPROTO_TCP ||
		    t->dst.protonum == IPPROTO_UDP ||
		    t->dst.protonum == IPPROTO_UDPLITE ||
		    t->dst.protonum == IPPROTO_DCCP ||
		    t->dst.protonum == IPPROTO_SCTP)
			fl4->fl4_dport = t->dst.u.all;
	}

	statusbit ^= IPS_NAT_MASK;

	if (ct->status & statusbit) {
		fl4->saddr = t->src.u3.ip;
		if (t->dst.protonum == IPPROTO_TCP ||
		    t->dst.protonum == IPPROTO_UDP ||
		    t->dst.protonum == IPPROTO_UDPLITE ||
		    t->dst.protonum == IPPROTO_DCCP ||
		    t->dst.protonum == IPPROTO_SCTP)
			fl4->fl4_sport = t->src.u.all;
	}
}
#endif /* CONFIG_XFRM */

static bool nf_nat_ipv4_in_range(const struct nf_conntrack_tuple *t,
				 const struct nf_nat_range *range)
{
	return ntohl(t->src.u3.ip) >= ntohl(range->min_addr.ip) &&
	       ntohl(t->src.u3.ip) <= ntohl(range->max_addr.ip);
}

static u32 nf_nat_ipv4_secure_port(const struct nf_conntrack_tuple *t,
				   __be16 dport)
{
	return secure_ipv4_port_ephemeral(t->src.u3.ip, t->dst.u3.ip, dport);
}

static bool nf_nat_ipv4_manip_pkt(struct sk_buff *skb,
				  unsigned int iphdroff,
				  const struct nf_nat_l4proto *l4proto,
				  const struct nf_conntrack_tuple *target,
				  enum nf_nat_manip_type maniptype)
{
	struct iphdr *iph;
	unsigned int hdroff;

	if (!skb_make_writable(skb, iphdroff + sizeof(*iph)))
		return false;

	iph = (void *)skb->data + iphdroff;
	hdroff = iphdroff + iph->ihl * 4;

	if (!l4proto->manip_pkt(skb, &nf_nat_l3proto_ipv4, iphdroff, hdroff,
				target, maniptype))
		return false;
	iph = (void *)skb->data + iphdroff;

	if (maniptype == NF_NAT_MANIP_SRC) {
		csum_replace4(&iph->check, iph->saddr, target->src.u3.ip);
		iph->saddr = target->src.u3.ip;
	} else {
		csum_replace4(&iph->check, iph->daddr, target->dst.u3.ip);
		iph->daddr = target->dst.u3.ip;
	}
	return true;
}

static void nf_nat_ipv4_csum_update(struct sk_buff *skb,
				    unsigned int iphdroff, __sum16 *check,
				    const struct nf_conntrack_tuple *t,
				    enum nf_nat_manip_type maniptype)
{
	struct iphdr *iph = (struct iphdr *)(skb->data + iphdroff);
	__be32 oldip, newip;

	if (maniptype == NF_NAT_MANIP_SRC) {
		oldip = iph->saddr;
		newip = t->src.u3.ip;
	} else {
		oldip = iph->daddr;
		newip = t->dst.u3.ip;
	}
	inet_proto_csum_replace4(check, skb, oldip, newip, true);
}

static void nf_nat_ipv4_csum_recalc(struct sk_buff *skb,
				    u8 proto, void *data, __sum16 *check,
				    int datalen, int oldlen)
{
	if (skb->ip_summed != CHECKSUM_PARTIAL) {
		const struct iphdr *iph = ip_hdr(skb);

		skb->ip_summed = CHECKSUM_PARTIAL;
		skb->csum_start = skb_headroom(skb) + skb_network_offset(skb) +
			ip_hdrlen(skb);
		skb->csum_offset = (void *)check - data;
		*check = ~csum_tcpudp_magic(iph->saddr, iph->daddr, datalen,
					    proto, 0);
	} else
		inet_proto_csum_replace2(check, skb,
					 htons(oldlen), htons(datalen), true);
}

#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
static int nf_nat_ipv4_nlattr_to_range(struct nlattr *tb[],
				       struct nf_nat_range *range)
{
	if (tb[CTA_NAT_V4_MINIP]) {
		range->min_addr.ip = nla_get_be32(tb[CTA_NAT_V4_MINIP]);
		range->flags |= NF_NAT_RANGE_MAP_IPS;
	}

	if (tb[CTA_NAT_V4_MAXIP])
		range->max_addr.ip = nla_get_be32(tb[CTA_NAT_V4_MAXIP]);
	else
		range->max_addr.ip = range->min_addr.ip;

	return 0;
}
#endif

static const struct nf_nat_l3proto nf_nat_l3proto_ipv4 = {
	.l3proto		= NFPROTO_IPV4,
	.in_range		= nf_nat_ipv4_in_range,
	.secure_port		= nf_nat_ipv4_secure_port,
	.manip_pkt		= nf_nat_ipv4_manip_pkt,
	.csum_update		= nf_nat_ipv4_csum_update,
	.csum_recalc		= nf_nat_ipv4_csum_recalc,
#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
	.nlattr_to_range	= nf_nat_ipv4_nlattr_to_range,
#endif
#ifdef CONFIG_XFRM
	.decode_session		= nf_nat_ipv4_decode_session,
#endif
};

int nf_nat_icmp_reply_translation(struct sk_buff *skb,
				  struct nf_conn *ct,
				  enum ip_conntrack_info ctinfo,
				  unsigned int hooknum)
{
	struct {
		struct icmphdr	icmp;
		struct iphdr	ip;
	} *inside;
	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	enum nf_nat_manip_type manip = HOOK2MANIP(hooknum);
	unsigned int hdrlen = ip_hdrlen(skb);
	const struct nf_nat_l4proto *l4proto;
	struct nf_conntrack_tuple target;
	unsigned long statusbit;

	NF_CT_ASSERT(ctinfo == IP_CT_RELATED || ctinfo == IP_CT_RELATED_REPLY);

	if (!skb_make_writable(skb, hdrlen + sizeof(*inside)))
		return 0;
	if (nf_ip_checksum(skb, hooknum, hdrlen, 0))
		return 0;

	inside = (void *)skb->data + hdrlen;
	if (inside->icmp.type == ICMP_REDIRECT) {
		if ((ct->status & IPS_NAT_DONE_MASK) != IPS_NAT_DONE_MASK)
			return 0;
		if (ct->status & IPS_NAT_MASK)
			return 0;
	}

	if (manip == NF_NAT_MANIP_SRC)
		statusbit = IPS_SRC_NAT;
	else
		statusbit = IPS_DST_NAT;

	/* Invert if this is reply direction */
	if (dir == IP_CT_DIR_REPLY)
		statusbit ^= IPS_NAT_MASK;

	if (!(ct->status & statusbit))
		return 1;

	l4proto = __nf_nat_l4proto_find(NFPROTO_IPV4, inside->ip.protocol);
	if (!nf_nat_ipv4_manip_pkt(skb, hdrlen + sizeof(inside->icmp),
				   l4proto, &ct->tuplehash[!dir].tuple, !manip))
		return 0;

	if (skb->ip_summed != CHECKSUM_PARTIAL) {
		/* Reloading "inside" here since manip_pkt may reallocate */
		inside = (void *)skb->data + hdrlen;
		inside->icmp.checksum = 0;
		inside->icmp.checksum =
			csum_fold(skb_checksum(skb, hdrlen,
					       skb->len - hdrlen, 0));
	}

	/* Change outer to look like the reply to an incoming packet */
	nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);
	l4proto = __nf_nat_l4proto_find(NFPROTO_IPV4, 0);
	if (!nf_nat_ipv4_manip_pkt(skb, 0, l4proto, &target, manip))
		return 0;

	return 1;
}
EXPORT_SYMBOL_GPL(nf_nat_icmp_reply_translation);

unsigned int
nf_nat_ipv4_fn(void *priv, struct sk_buff *skb,
	       const struct nf_hook_state *state,
	       unsigned int (*do_chain)(void *priv,
					struct sk_buff *skb,
					const struct nf_hook_state *state,
					struct nf_conn *ct))
{
	struct nf_conn *ct;
	enum ip_conntrack_info ctinfo;
	struct nf_conn_nat *nat;
	/* maniptype == SRC for postrouting. */
	enum nf_nat_manip_type maniptype = HOOK2MANIP(state->hook);

	ct = nf_ct_get(skb, &ctinfo);
	/* Can't track?  It's not due to stress, or conntrack would
	 * have dropped it.  Hence it's the user's responsibilty to
	 * packet filter it out, or implement conntrack/NAT for that
	 * protocol. 8) --RR
	 */
	if (!ct)
		return NF_ACCEPT;

	nat = nf_ct_nat_ext_add(ct);
	if (nat == NULL)
		return NF_ACCEPT;

	switch (ctinfo) {
	case IP_CT_RELATED:
	case IP_CT_RELATED_REPLY:
		if (ip_hdr(skb)->protocol == IPPROTO_ICMP) {
			if (!nf_nat_icmp_reply_translation(skb, ct, ctinfo,
							   state->hook))
				return NF_DROP;
			else
				return NF_ACCEPT;
		}
		/* Fall thru... (Only ICMPs can be IP_CT_IS_REPLY) */
	case IP_CT_NEW:
		/* Seen it before?  This can happen for loopback, retrans,
		 * or local packets.
		 */
		if (!nf_nat_initialized(ct, maniptype)) {
			unsigned int ret;

			ret = do_chain(priv, skb, state, ct);
			if (ret != NF_ACCEPT)
				return ret;

			if (nf_nat_initialized(ct, HOOK2MANIP(state->hook)))
				break;

			ret = nf_nat_alloc_null_binding(ct, state->hook);
			if (ret != NF_ACCEPT)
				return ret;
		} else {
			pr_debug("Already setup manip %s for ct %p\n",
				 maniptype == NF_NAT_MANIP_SRC ? "SRC" : "DST",
				 ct);
			if (nf_nat_oif_changed(state->hook, ctinfo, nat,
					       state->out))
				goto oif_changed;
		}
		break;

	default:
		/* ESTABLISHED */
		NF_CT_ASSERT(ctinfo == IP_CT_ESTABLISHED ||
			     ctinfo == IP_CT_ESTABLISHED_REPLY);
		if (nf_nat_oif_changed(state->hook, ctinfo, nat, state->out))
			goto oif_changed;
	}

	return nf_nat_packet(ct, ctinfo, state->hook, skb);

oif_changed:
	nf_ct_kill_acct(ct, ctinfo, skb);
	return NF_DROP;
}
EXPORT_SYMBOL_GPL(nf_nat_ipv4_fn);

unsigned int
nf_nat_ipv4_in(void *priv, struct sk_buff *skb,
	       const struct nf_hook_state *state,
	       unsigned int (*do_chain)(void *priv,
					 struct sk_buff *skb,
					 const struct nf_hook_state *state,
					 struct nf_conn *ct))
{
	unsigned int ret;
	__be32 daddr = ip_hdr(skb)->daddr;

	ret = nf_nat_ipv4_fn(priv, skb, state, do_chain);
	if (ret != NF_DROP && ret != NF_STOLEN &&
	    daddr != ip_hdr(skb)->daddr)
		skb_dst_drop(skb);

	return ret;
}
EXPORT_SYMBOL_GPL(nf_nat_ipv4_in);

unsigned int
nf_nat_ipv4_out(void *priv, struct sk_buff *skb,
		const struct nf_hook_state *state,
		unsigned int (*do_chain)(void *priv,
					  struct sk_buff *skb,
					  const struct nf_hook_state *state,
					  struct nf_conn *ct))
{
#ifdef CONFIG_XFRM
	const struct nf_conn *ct;
	enum ip_conntrack_info ctinfo;
	int err;
#endif
	unsigned int ret;

	/* root is playing with raw sockets. */
	if (skb->len < sizeof(struct iphdr) ||
	    ip_hdrlen(skb) < sizeof(struct iphdr))
		return NF_ACCEPT;

	ret = nf_nat_ipv4_fn(priv, skb, state, do_chain);
#ifdef CONFIG_XFRM
	if (ret != NF_DROP && ret != NF_STOLEN &&
	    !(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
	    (ct = nf_ct_get(skb, &ctinfo)) != NULL) {
		enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);

		if ((ct->tuplehash[dir].tuple.src.u3.ip !=
		     ct->tuplehash[!dir].tuple.dst.u3.ip) ||
		    (ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMP &&
		     ct->tuplehash[dir].tuple.src.u.all !=
		     ct->tuplehash[!dir].tuple.dst.u.all)) {
			err = nf_xfrm_me_harder(state->net, skb, AF_INET);
			if (err < 0)
				ret = NF_DROP_ERR(err);
		}
	}
#endif
	return ret;
}
EXPORT_SYMBOL_GPL(nf_nat_ipv4_out);

unsigned int
nf_nat_ipv4_local_fn(void *priv, struct sk_buff *skb,
		     const struct nf_hook_state *state,
		     unsigned int (*do_chain)(void *priv,
					       struct sk_buff *skb,
					       const struct nf_hook_state *state,
					       struct nf_conn *ct))
{
	const struct nf_conn *ct;
	enum ip_conntrack_info ctinfo;
	unsigned int ret;
	int err;

	/* root is playing with raw sockets. */
	if (skb->len < sizeof(struct iphdr) ||
	    ip_hdrlen(skb) < sizeof(struct iphdr))
		return NF_ACCEPT;

	ret = nf_nat_ipv4_fn(priv, skb, state, do_chain);
	if (ret != NF_DROP && ret != NF_STOLEN &&
	    (ct = nf_ct_get(skb, &ctinfo)) != NULL) {
		enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);

		if (ct->tuplehash[dir].tuple.dst.u3.ip !=
		    ct->tuplehash[!dir].tuple.src.u3.ip) {
			err = ip_route_me_harder(state->net, skb, RTN_UNSPEC);
			if (err < 0)
				ret = NF_DROP_ERR(err);
		}
#ifdef CONFIG_XFRM
		else if (!(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
			 ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMP &&
			 ct->tuplehash[dir].tuple.dst.u.all !=
			 ct->tuplehash[!dir].tuple.src.u.all) {
			err = nf_xfrm_me_harder(state->net, skb, AF_INET);
			if (err < 0)
				ret = NF_DROP_ERR(err);
		}
#endif
	}
	return ret;
}
EXPORT_SYMBOL_GPL(nf_nat_ipv4_local_fn);

static int __init nf_nat_l3proto_ipv4_init(void)
{
	int err;

	err = nf_nat_l4proto_register(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
	if (err < 0)
		goto err1;
	err = nf_nat_l3proto_register(&nf_nat_l3proto_ipv4);
	if (err < 0)
		goto err2;
	return err;

err2:
	nf_nat_l4proto_unregister(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
err1:
	return err;
}

static void __exit nf_nat_l3proto_ipv4_exit(void)
{
	nf_nat_l3proto_unregister(&nf_nat_l3proto_ipv4);
	nf_nat_l4proto_unregister(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
}

MODULE_LICENSE("GPL");
MODULE_ALIAS("nf-nat-" __stringify(AF_INET));

module_init(nf_nat_l3proto_ipv4_init);
module_exit(nf_nat_l3proto_ipv4_exit);
Commit	Line	Data
c7232c99 PM	1	/*
	2	* (C) 1999-2001 Paul `Rusty' Russell
	3	* (C) 2002-2006 Netfilter Core Team <coreteam@netfilter.org>
	4	* (C) 2011 Patrick McHardy <kaber@trash.net>
	5	*
	6	* This program is free software; you can redistribute it and/or modify
	7	* it under the terms of the GNU General Public License version 2 as
	8	* published by the Free Software Foundation.
	9	*/
	10
	11	#include <linux/types.h>
	12	#include <linux/module.h>
	13	#include <linux/skbuff.h>
	14	#include <linux/ip.h>
	15	#include <linux/icmp.h>
	16	#include <linux/netfilter.h>
	17	#include <linux/netfilter_ipv4.h>
	18	#include <net/secure_seq.h>
	19	#include <net/checksum.h>
	20	#include <net/route.h>
	21	#include <net/ip.h>
	22
	23	#include <net/netfilter/nf_conntrack_core.h>
	24	#include <net/netfilter/nf_conntrack.h>
	25	#include <net/netfilter/nf_nat_core.h>
	26	#include <net/netfilter/nf_nat_l3proto.h>
	27	#include <net/netfilter/nf_nat_l4proto.h>
	28
	29	static const struct nf_nat_l3proto nf_nat_l3proto_ipv4;
	30
	31	#ifdef CONFIG_XFRM
	32	static void nf_nat_ipv4_decode_session(struct sk_buff *skb,
	33	const struct nf_conn *ct,
	34	enum ip_conntrack_dir dir,
	35	unsigned long statusbit,
	36	struct flowi *fl)
	37	{
	38	const struct nf_conntrack_tuple *t = &ct->tuplehash[dir].tuple;
	39	struct flowi4 *fl4 = &fl->u.ip4;
	40
	41	if (ct->status & statusbit) {
	42	fl4->daddr = t->dst.u3.ip;
	43	if (t->dst.protonum == IPPROTO_TCP \|\|
	44	t->dst.protonum == IPPROTO_UDP \|\|
	45	t->dst.protonum == IPPROTO_UDPLITE \|\|
	46	t->dst.protonum == IPPROTO_DCCP \|\|
	47	t->dst.protonum == IPPROTO_SCTP)
	48	fl4->fl4_dport = t->dst.u.all;
	49	}
	50
	51	statusbit ^= IPS_NAT_MASK;
	52
	53	if (ct->status & statusbit) {
	54	fl4->saddr = t->src.u3.ip;
	55	if (t->dst.protonum == IPPROTO_TCP \|\|
	56	t->dst.protonum == IPPROTO_UDP \|\|
	57	t->dst.protonum == IPPROTO_UDPLITE \|\|
	58	t->dst.protonum == IPPROTO_DCCP \|\|
	59	t->dst.protonum == IPPROTO_SCTP)
	60	fl4->fl4_sport = t->src.u.all;
	61	}
	62	}
	63	#endif /* CONFIG_XFRM */
	64
65	static bool nf_nat_ipv4_in_range(const struct nf_conntrack_tuple *t,
66	const struct nf_nat_range *range)
67	{
68	return ntohl(t->src.u3.ip) >= ntohl(range->min_addr.ip) &&
69	ntohl(t->src.u3.ip) <= ntohl(range->max_addr.ip);
70	}
71
72	static u32 nf_nat_ipv4_secure_port(const struct nf_conntrack_tuple *t,
73	__be16 dport)
74	{
75	return secure_ipv4_port_ephemeral(t->src.u3.ip, t->dst.u3.ip, dport);
76	}
77
78	static bool nf_nat_ipv4_manip_pkt(struct sk_buff *skb,
79	unsigned int iphdroff,
80	const struct nf_nat_l4proto *l4proto,
81	const struct nf_conntrack_tuple *target,
82	enum nf_nat_manip_type maniptype)
83	{
84	struct iphdr *iph;
85	unsigned int hdroff;
86
87	if (!skb_make_writable(skb, iphdroff + sizeof(*iph)))
88	return false;
89
90	iph = (void *)skb->data + iphdroff;
91	hdroff = iphdroff + iph->ihl * 4;
92
93	if (!l4proto->manip_pkt(skb, &nf_nat_l3proto_ipv4, iphdroff, hdroff,
94	target, maniptype))
95	return false;
96	iph = (void *)skb->data + iphdroff;
97
98	if (maniptype == NF_NAT_MANIP_SRC) {
99	csum_replace4(&iph->check, iph->saddr, target->src.u3.ip);
100	iph->saddr = target->src.u3.ip;
101	} else {
102	csum_replace4(&iph->check, iph->daddr, target->dst.u3.ip);
103	iph->daddr = target->dst.u3.ip;
104	}
105	return true;
106	}
107
108	static void nf_nat_ipv4_csum_update(struct sk_buff *skb,
109	unsigned int iphdroff, __sum16 *check,
110	const struct nf_conntrack_tuple *t,
111	enum nf_nat_manip_type maniptype)
112	{
113	struct iphdr iph = (struct iphdr )(skb->data + iphdroff);
114	__be32 oldip, newip;
115
116	if (maniptype == NF_NAT_MANIP_SRC) {
117	oldip = iph->saddr;
118	newip = t->src.u3.ip;
119	} else {
120	oldip = iph->daddr;
121	newip = t->dst.u3.ip;
122	}
4b048d6d	123	inet_proto_csum_replace4(check, skb, oldip, newip, true);
c7232c99 PM	124	}
	125
	126	static void nf_nat_ipv4_csum_recalc(struct sk_buff *skb,
	127	u8 proto, void data, __sum16 check,
	128	int datalen, int oldlen)
	129	{
c7232c99	130	if (skb->ip_summed != CHECKSUM_PARTIAL) {
26461905 JR	131	const struct iphdr *iph = ip_hdr(skb);
	132
	133	skb->ip_summed = CHECKSUM_PARTIAL;
	134	skb->csum_start = skb_headroom(skb) + skb_network_offset(skb) +
	135	ip_hdrlen(skb);
	136	skb->csum_offset = (void *)check - data;
	137	*check = ~csum_tcpudp_magic(iph->saddr, iph->daddr, datalen,
	138	proto, 0);
c7232c99 PM	139	} else
c7232c99 PM	140	inet_proto_csum_replace2(check, skb,
4b048d6d	141	htons(oldlen), htons(datalen), true);
c7232c99 PM	142	}
c7232c99 PM	143
24de3d37	144	#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
c7232c99 PM	145	static int nf_nat_ipv4_nlattr_to_range(struct nlattr *tb[],
	146	struct nf_nat_range *range)
	147	{
	148	if (tb[CTA_NAT_V4_MINIP]) {
	149	range->min_addr.ip = nla_get_be32(tb[CTA_NAT_V4_MINIP]);
	150	range->flags \|= NF_NAT_RANGE_MAP_IPS;
	151	}
	152
	153	if (tb[CTA_NAT_V4_MAXIP])
	154	range->max_addr.ip = nla_get_be32(tb[CTA_NAT_V4_MAXIP]);
	155	else
	156	range->max_addr.ip = range->min_addr.ip;
	157
	158	return 0;
	159	}
24de3d37	160	#endif
c7232c99 PM	161
	162	static const struct nf_nat_l3proto nf_nat_l3proto_ipv4 = {
	163	.l3proto = NFPROTO_IPV4,
	164	.in_range = nf_nat_ipv4_in_range,
	165	.secure_port = nf_nat_ipv4_secure_port,
	166	.manip_pkt = nf_nat_ipv4_manip_pkt,
	167	.csum_update = nf_nat_ipv4_csum_update,
	168	.csum_recalc = nf_nat_ipv4_csum_recalc,
24de3d37	169	#if IS_ENABLED(CONFIG_NF_CT_NETLINK)
c7232c99	170	.nlattr_to_range = nf_nat_ipv4_nlattr_to_range,
24de3d37	171	#endif
c7232c99 PM	172	#ifdef CONFIG_XFRM
	173	.decode_session = nf_nat_ipv4_decode_session,
	174	#endif
	175	};
	176
	177	int nf_nat_icmp_reply_translation(struct sk_buff *skb,
	178	struct nf_conn *ct,
	179	enum ip_conntrack_info ctinfo,
	180	unsigned int hooknum)
	181	{
	182	struct {
	183	struct icmphdr icmp;
	184	struct iphdr ip;
	185	} *inside;
	186	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	187	enum nf_nat_manip_type manip = HOOK2MANIP(hooknum);
	188	unsigned int hdrlen = ip_hdrlen(skb);
	189	const struct nf_nat_l4proto *l4proto;
	190	struct nf_conntrack_tuple target;
	191	unsigned long statusbit;
	192
	193	NF_CT_ASSERT(ctinfo == IP_CT_RELATED \|\| ctinfo == IP_CT_RELATED_REPLY);
	194
	195	if (!skb_make_writable(skb, hdrlen + sizeof(*inside)))
	196	return 0;
	197	if (nf_ip_checksum(skb, hooknum, hdrlen, 0))
	198	return 0;
	199
	200	inside = (void *)skb->data + hdrlen;
	201	if (inside->icmp.type == ICMP_REDIRECT) {
	202	if ((ct->status & IPS_NAT_DONE_MASK) != IPS_NAT_DONE_MASK)
	203	return 0;
	204	if (ct->status & IPS_NAT_MASK)
	205	return 0;
	206	}
	207
	208	if (manip == NF_NAT_MANIP_SRC)
	209	statusbit = IPS_SRC_NAT;
	210	else
	211	statusbit = IPS_DST_NAT;
	212
	213	/* Invert if this is reply direction */
	214	if (dir == IP_CT_DIR_REPLY)
	215	statusbit ^= IPS_NAT_MASK;
	216
	217	if (!(ct->status & statusbit))
	218	return 1;
	219
	220	l4proto = __nf_nat_l4proto_find(NFPROTO_IPV4, inside->ip.protocol);
	221	if (!nf_nat_ipv4_manip_pkt(skb, hdrlen + sizeof(inside->icmp),
	222	l4proto, &ct->tuplehash[!dir].tuple, !manip))
	223	return 0;
	224
	225	if (skb->ip_summed != CHECKSUM_PARTIAL) {
	226	/* Reloading "inside" here since manip_pkt may reallocate */
	227	inside = (void *)skb->data + hdrlen;
	228	inside->icmp.checksum = 0;
	229	inside->icmp.checksum =
	230	csum_fold(skb_checksum(skb, hdrlen,
	231	skb->len - hdrlen, 0));
	232	}
	233
	234	/* Change outer to look like the reply to an incoming packet */
	235	nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);
236	l4proto = __nf_nat_l4proto_find(NFPROTO_IPV4, 0);
237	if (!nf_nat_ipv4_manip_pkt(skb, 0, l4proto, &target, manip))
238	return 0;
239
240	return 1;
241	}
242	EXPORT_SYMBOL_GPL(nf_nat_icmp_reply_translation);
243
30766f4c	244	unsigned int
06198b34	245	nf_nat_ipv4_fn(void priv, struct sk_buff skb,
d7cf4081	246	const struct nf_hook_state *state,
06198b34	247	unsigned int (do_chain)(void priv,
30766f4c	248	struct sk_buff *skb,
d7cf4081	249	const struct nf_hook_state *state,
30766f4c PNA	250	struct nf_conn *ct))
	251	{
	252	struct nf_conn *ct;
	253	enum ip_conntrack_info ctinfo;
	254	struct nf_conn_nat *nat;
	255	/* maniptype == SRC for postrouting. */
082a758f	256	enum nf_nat_manip_type maniptype = HOOK2MANIP(state->hook);
30766f4c	257
30766f4c PNA	258	ct = nf_ct_get(skb, &ctinfo);
	259	/* Can't track? It's not due to stress, or conntrack would
	260	* have dropped it. Hence it's the user's responsibilty to
	261	* packet filter it out, or implement conntrack/NAT for that
	262	* protocol. 8) --RR
	263	*/
	264	if (!ct)
	265	return NF_ACCEPT;
	266
30766f4c PNA	267	nat = nf_ct_nat_ext_add(ct);
	268	if (nat == NULL)
	269	return NF_ACCEPT;
	270
	271	switch (ctinfo) {
	272	case IP_CT_RELATED:
	273	case IP_CT_RELATED_REPLY:
	274	if (ip_hdr(skb)->protocol == IPPROTO_ICMP) {
	275	if (!nf_nat_icmp_reply_translation(skb, ct, ctinfo,
082a758f	276	state->hook))
30766f4c PNA	277	return NF_DROP;
	278	else
	279	return NF_ACCEPT;
	280	}
	281	/* Fall thru... (Only ICMPs can be IP_CT_IS_REPLY) */
	282	case IP_CT_NEW:
	283	/* Seen it before? This can happen for loopback, retrans,
	284	* or local packets.
	285	*/
	286	if (!nf_nat_initialized(ct, maniptype)) {
	287	unsigned int ret;
	288
06198b34	289	ret = do_chain(priv, skb, state, ct);
30766f4c PNA	290	if (ret != NF_ACCEPT)
	291	return ret;
	292
082a758f	293	if (nf_nat_initialized(ct, HOOK2MANIP(state->hook)))
30766f4c PNA	294	break;
30766f4c PNA	295
082a758f	296	ret = nf_nat_alloc_null_binding(ct, state->hook);
30766f4c PNA	297	if (ret != NF_ACCEPT)
	298	return ret;
	299	} else {
	300	pr_debug("Already setup manip %s for ct %p\n",
	301	maniptype == NF_NAT_MANIP_SRC ? "SRC" : "DST",
	302	ct);
082a758f	303	if (nf_nat_oif_changed(state->hook, ctinfo, nat,
d7cf4081	304	state->out))
30766f4c PNA	305	goto oif_changed;
	306	}
	307	break;
	308
	309	default:
	310	/* ESTABLISHED */
	311	NF_CT_ASSERT(ctinfo == IP_CT_ESTABLISHED \|\|
	312	ctinfo == IP_CT_ESTABLISHED_REPLY);
082a758f	313	if (nf_nat_oif_changed(state->hook, ctinfo, nat, state->out))
30766f4c PNA	314	goto oif_changed;
	315	}
	316
082a758f	317	return nf_nat_packet(ct, ctinfo, state->hook, skb);
30766f4c PNA	318
	319	oif_changed:
	320	nf_ct_kill_acct(ct, ctinfo, skb);
	321	return NF_DROP;
	322	}
	323	EXPORT_SYMBOL_GPL(nf_nat_ipv4_fn);
	324
	325	unsigned int
06198b34	326	nf_nat_ipv4_in(void priv, struct sk_buff skb,
d7cf4081	327	const struct nf_hook_state *state,
06198b34	328	unsigned int (do_chain)(void priv,
30766f4c	329	struct sk_buff *skb,
d7cf4081	330	const struct nf_hook_state *state,
30766f4c PNA	331	struct nf_conn *ct))
	332	{
	333	unsigned int ret;
	334	__be32 daddr = ip_hdr(skb)->daddr;
	335
06198b34	336	ret = nf_nat_ipv4_fn(priv, skb, state, do_chain);
30766f4c PNA	337	if (ret != NF_DROP && ret != NF_STOLEN &&
	338	daddr != ip_hdr(skb)->daddr)
	339	skb_dst_drop(skb);
	340
	341	return ret;
	342	}
	343	EXPORT_SYMBOL_GPL(nf_nat_ipv4_in);
	344
	345	unsigned int
06198b34	346	nf_nat_ipv4_out(void priv, struct sk_buff skb,
d7cf4081	347	const struct nf_hook_state *state,
06198b34	348	unsigned int (do_chain)(void priv,
30766f4c	349	struct sk_buff *skb,
d7cf4081	350	const struct nf_hook_state *state,
30766f4c PNA	351	struct nf_conn *ct))
	352	{
	353	#ifdef CONFIG_XFRM
	354	const struct nf_conn *ct;
	355	enum ip_conntrack_info ctinfo;
	356	int err;
	357	#endif
	358	unsigned int ret;
	359
	360	/* root is playing with raw sockets. */
	361	if (skb->len < sizeof(struct iphdr) \|\|
	362	ip_hdrlen(skb) < sizeof(struct iphdr))
	363	return NF_ACCEPT;
	364
06198b34	365	ret = nf_nat_ipv4_fn(priv, skb, state, do_chain);
30766f4c PNA	366	#ifdef CONFIG_XFRM
	367	if (ret != NF_DROP && ret != NF_STOLEN &&
	368	!(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
	369	(ct = nf_ct_get(skb, &ctinfo)) != NULL) {
	370	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	371
	372	if ((ct->tuplehash[dir].tuple.src.u3.ip !=
	373	ct->tuplehash[!dir].tuple.dst.u3.ip) \|\|
	374	(ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMP &&
	375	ct->tuplehash[dir].tuple.src.u.all !=
	376	ct->tuplehash[!dir].tuple.dst.u.all)) {
c7af6483	377	err = nf_xfrm_me_harder(state->net, skb, AF_INET);
30766f4c PNA	378	if (err < 0)
	379	ret = NF_DROP_ERR(err);
	380	}
	381	}
	382	#endif
	383	return ret;
	384	}
	385	EXPORT_SYMBOL_GPL(nf_nat_ipv4_out);
	386
	387	unsigned int
06198b34	388	nf_nat_ipv4_local_fn(void priv, struct sk_buff skb,
d7cf4081	389	const struct nf_hook_state *state,
06198b34	390	unsigned int (do_chain)(void priv,
30766f4c	391	struct sk_buff *skb,
d7cf4081	392	const struct nf_hook_state *state,
30766f4c PNA	393	struct nf_conn *ct))
	394	{
	395	const struct nf_conn *ct;
	396	enum ip_conntrack_info ctinfo;
	397	unsigned int ret;
	398	int err;
	399
	400	/* root is playing with raw sockets. */
	401	if (skb->len < sizeof(struct iphdr) \|\|
	402	ip_hdrlen(skb) < sizeof(struct iphdr))
	403	return NF_ACCEPT;
	404
06198b34	405	ret = nf_nat_ipv4_fn(priv, skb, state, do_chain);
30766f4c PNA	406	if (ret != NF_DROP && ret != NF_STOLEN &&
	407	(ct = nf_ct_get(skb, &ctinfo)) != NULL) {
	408	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	409
	410	if (ct->tuplehash[dir].tuple.dst.u3.ip !=
	411	ct->tuplehash[!dir].tuple.src.u3.ip) {
e45f5066	412	err = ip_route_me_harder(state->net, skb, RTN_UNSPEC);
30766f4c PNA	413	if (err < 0)
	414	ret = NF_DROP_ERR(err);
	415	}
	416	#ifdef CONFIG_XFRM
	417	else if (!(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
	418	ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMP &&
	419	ct->tuplehash[dir].tuple.dst.u.all !=
	420	ct->tuplehash[!dir].tuple.src.u.all) {
c7af6483	421	err = nf_xfrm_me_harder(state->net, skb, AF_INET);
30766f4c PNA	422	if (err < 0)
	423	ret = NF_DROP_ERR(err);
	424	}
	425	#endif
	426	}
	427	return ret;
	428	}
	429	EXPORT_SYMBOL_GPL(nf_nat_ipv4_local_fn);
	430
c7232c99 PM	431	static int __init nf_nat_l3proto_ipv4_init(void)
	432	{
	433	int err;
	434
	435	err = nf_nat_l4proto_register(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
	436	if (err < 0)
	437	goto err1;
	438	err = nf_nat_l3proto_register(&nf_nat_l3proto_ipv4);
	439	if (err < 0)
	440	goto err2;
	441	return err;
	442
	443	err2:
	444	nf_nat_l4proto_unregister(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
	445	err1:
	446	return err;
	447	}
	448
	449	static void __exit nf_nat_l3proto_ipv4_exit(void)
	450	{
	451	nf_nat_l3proto_unregister(&nf_nat_l3proto_ipv4);
	452	nf_nat_l4proto_unregister(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
	453	}
	454
	455	MODULE_LICENSE("GPL");
	456	MODULE_ALIAS("nf-nat-" __stringify(AF_INET));
	457
	458	module_init(nf_nat_l3proto_ipv4_init);
	459	module_exit(nf_nat_l3proto_ipv4_exit);