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

/* ip_nat_helper.c - generic support functions for NAT helpers
 *
 * (C) 2000-2002 Harald Welte <laforge@netfilter.org>
 * (C) 2003-2004 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.
 *
 * 	14 Jan 2002 Harald Welte <laforge@gnumonks.org>:
 *		- add support for SACK adjustment
 *	14 Mar 2002 Harald Welte <laforge@gnumonks.org>:
 *		- merge SACK support into newnat API
 *	16 Aug 2002 Brian J. Murrell <netfilter@interlinx.bc.ca>:
 *		- make ip_nat_resize_packet more generic (TCP and UDP)
 *		- add ip_nat_mangle_udp_packet
 */
#include <linux/module.h>
#include <linux/kmod.h>
#include <linux/types.h>
#include <linux/timer.h>
#include <linux/skbuff.h>
#include <linux/netfilter_ipv4.h>
#include <net/checksum.h>
#include <net/icmp.h>
#include <net/ip.h>
#include <net/tcp.h>
#include <net/udp.h>

#include <linux/netfilter_ipv4/ip_conntrack.h>
#include <linux/netfilter_ipv4/ip_conntrack_helper.h>
#include <linux/netfilter_ipv4/ip_nat.h>
#include <linux/netfilter_ipv4/ip_nat_protocol.h>
#include <linux/netfilter_ipv4/ip_nat_core.h>
#include <linux/netfilter_ipv4/ip_nat_helper.h>

#if 0
#define DEBUGP printk
#define DUMP_OFFSET(x)	printk("offset_before=%d, offset_after=%d, correction_pos=%u\n", x->offset_before, x->offset_after, x->correction_pos);
#else
#define DEBUGP(format, args...)
#define DUMP_OFFSET(x)
#endif

static DEFINE_SPINLOCK(ip_nat_seqofs_lock);

/* Setup TCP sequence correction given this change at this sequence */
static inline void
adjust_tcp_sequence(u32 seq,
		    int sizediff,
		    struct ip_conntrack *ct,
		    enum ip_conntrack_info ctinfo)
{
	int dir;
	struct ip_nat_seq *this_way, *other_way;

	DEBUGP("ip_nat_resize_packet: old_size = %u, new_size = %u\n",
		(*skb)->len, new_size);

	dir = CTINFO2DIR(ctinfo);

	this_way = &ct->nat.info.seq[dir];
	other_way = &ct->nat.info.seq[!dir];

	DEBUGP("ip_nat_resize_packet: Seq_offset before: ");
	DUMP_OFFSET(this_way);

	spin_lock_bh(&ip_nat_seqofs_lock);

	/* SYN adjust. If it's uninitialized, or this is after last
	 * correction, record it: we don't handle more than one
	 * adjustment in the window, but do deal with common case of a
	 * retransmit */
	if (this_way->offset_before == this_way->offset_after
	    || before(this_way->correction_pos, seq)) {
		    this_way->correction_pos = seq;
		    this_way->offset_before = this_way->offset_after;
		    this_way->offset_after += sizediff;
	}
	spin_unlock_bh(&ip_nat_seqofs_lock);

	DEBUGP("ip_nat_resize_packet: Seq_offset after: ");
	DUMP_OFFSET(this_way);
}

/* Frobs data inside this packet, which is linear. */
static void mangle_contents(struct sk_buff *skb,
			    unsigned int dataoff,
			    unsigned int match_offset,
			    unsigned int match_len,
			    const char *rep_buffer,
			    unsigned int rep_len)
{
	unsigned char *data;

	BUG_ON(skb_is_nonlinear(skb));
	data = (unsigned char *)skb->nh.iph + dataoff;

	/* move post-replacement */
	memmove(data + match_offset + rep_len,
		data + match_offset + match_len,
		skb->tail - (data + match_offset + match_len));

	/* insert data from buffer */
	memcpy(data + match_offset, rep_buffer, rep_len);

	/* update skb info */
	if (rep_len > match_len) {
		DEBUGP("ip_nat_mangle_packet: Extending packet by "
			"%u from %u bytes\n", rep_len - match_len,
		       skb->len);
		skb_put(skb, rep_len - match_len);
	} else {
		DEBUGP("ip_nat_mangle_packet: Shrinking packet from "
			"%u from %u bytes\n", match_len - rep_len,
		       skb->len);
		__skb_trim(skb, skb->len + rep_len - match_len);
	}

	/* fix IP hdr checksum information */
	skb->nh.iph->tot_len = htons(skb->len);
	ip_send_check(skb->nh.iph);
}

/* Unusual, but possible case. */
static int enlarge_skb(struct sk_buff **pskb, unsigned int extra)
{
	struct sk_buff *nskb;

	if ((*pskb)->len + extra > 65535)
		return 0;

	nskb = skb_copy_expand(*pskb, skb_headroom(*pskb), extra, GFP_ATOMIC);
	if (!nskb)
		return 0;

	/* Transfer socket to new skb. */
	if ((*pskb)->sk)
		skb_set_owner_w(nskb, (*pskb)->sk);
	kfree_skb(*pskb);
	*pskb = nskb;
	return 1;
}

/* Generic function for mangling variable-length address changes inside
 * NATed TCP connections (like the PORT XXX,XXX,XXX,XXX,XXX,XXX
 * command in FTP).
 *
 * Takes care about all the nasty sequence number changes, checksumming,
 * skb enlargement, ...
 *
 * */
int
ip_nat_mangle_tcp_packet(struct sk_buff **pskb,
			 struct ip_conntrack *ct,
			 enum ip_conntrack_info ctinfo,
			 unsigned int match_offset,
			 unsigned int match_len,
			 const char *rep_buffer,
			 unsigned int rep_len)
{
	struct iphdr *iph;
	struct tcphdr *tcph;
	int oldlen, datalen;

	if (!skb_make_writable(pskb, (*pskb)->len))
		return 0;

	if (rep_len > match_len
	    && rep_len - match_len > skb_tailroom(*pskb)
	    && !enlarge_skb(pskb, rep_len - match_len))
		return 0;

	SKB_LINEAR_ASSERT(*pskb);

	iph = (*pskb)->nh.iph;
	tcph = (void *)iph + iph->ihl*4;

	oldlen = (*pskb)->len - iph->ihl*4;
	mangle_contents(*pskb, iph->ihl*4 + tcph->doff*4,
			match_offset, match_len, rep_buffer, rep_len);

	datalen = (*pskb)->len - iph->ihl*4;
	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
		tcph->check = 0;
		tcph->check = tcp_v4_check(datalen,
					   iph->saddr, iph->daddr,
					   csum_partial((char *)tcph,
							datalen, 0));
	} else
		nf_proto_csum_replace2(&tcph->check, *pskb,
					htons(oldlen), htons(datalen), 1);

	if (rep_len != match_len) {
		set_bit(IPS_SEQ_ADJUST_BIT, &ct->status);
		adjust_tcp_sequence(ntohl(tcph->seq),
				    (int)rep_len - (int)match_len,
				    ct, ctinfo);
		/* Tell TCP window tracking about seq change */
		ip_conntrack_tcp_update(*pskb, ct, CTINFO2DIR(ctinfo));
	}
	return 1;
}
EXPORT_SYMBOL(ip_nat_mangle_tcp_packet);

/* Generic function for mangling variable-length address changes inside
 * NATed UDP connections (like the CONNECT DATA XXXXX MESG XXXXX INDEX XXXXX
 * command in the Amanda protocol)
 *
 * Takes care about all the nasty sequence number changes, checksumming,
 * skb enlargement, ...
 *
 * XXX - This function could be merged with ip_nat_mangle_tcp_packet which
 *       should be fairly easy to do.
 */
int
ip_nat_mangle_udp_packet(struct sk_buff **pskb,
			 struct ip_conntrack *ct,
			 enum ip_conntrack_info ctinfo,
			 unsigned int match_offset,
			 unsigned int match_len,
			 const char *rep_buffer,
			 unsigned int rep_len)
{
	struct iphdr *iph;
	struct udphdr *udph;
	int datalen, oldlen;

	/* UDP helpers might accidentally mangle the wrong packet */
	iph = (*pskb)->nh.iph;
	if ((*pskb)->len < iph->ihl*4 + sizeof(*udph) +
			       match_offset + match_len)
		return 0;

	if (!skb_make_writable(pskb, (*pskb)->len))
		return 0;

	if (rep_len > match_len
	    && rep_len - match_len > skb_tailroom(*pskb)
	    && !enlarge_skb(pskb, rep_len - match_len))
		return 0;

	iph = (*pskb)->nh.iph;
	udph = (void *)iph + iph->ihl*4;

	oldlen = (*pskb)->len - iph->ihl*4;
	mangle_contents(*pskb, iph->ihl*4 + sizeof(*udph),
			match_offset, match_len, rep_buffer, rep_len);

	/* update the length of the UDP packet */
	datalen = (*pskb)->len - iph->ihl*4;
	udph->len = htons(datalen);

	/* fix udp checksum if udp checksum was previously calculated */
	if (!udph->check && (*pskb)->ip_summed != CHECKSUM_PARTIAL)
		return 1;

	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
		udph->check = 0;
		udph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
						datalen, IPPROTO_UDP,
						csum_partial((char *)udph,
							     datalen, 0));
		if (!udph->check)
			udph->check = CSUM_MANGLED_0;
	} else
		nf_proto_csum_replace2(&udph->check, *pskb,
					htons(oldlen), htons(datalen), 1);
	return 1;
}
EXPORT_SYMBOL(ip_nat_mangle_udp_packet);

/* Adjust one found SACK option including checksum correction */
static void
sack_adjust(struct sk_buff *skb,
	    struct tcphdr *tcph,
	    unsigned int sackoff,
	    unsigned int sackend,
	    struct ip_nat_seq *natseq)
{
	while (sackoff < sackend) {
		struct tcp_sack_block_wire *sack;
		__be32 new_start_seq, new_end_seq;

		sack = (void *)skb->data + sackoff;
		if (after(ntohl(sack->start_seq) - natseq->offset_before,
			  natseq->correction_pos))
			new_start_seq = htonl(ntohl(sack->start_seq)
					- natseq->offset_after);
		else
			new_start_seq = htonl(ntohl(sack->start_seq)
					- natseq->offset_before);

		if (after(ntohl(sack->end_seq) - natseq->offset_before,
			  natseq->correction_pos))
			new_end_seq = htonl(ntohl(sack->end_seq)
				      - natseq->offset_after);
		else
			new_end_seq = htonl(ntohl(sack->end_seq)
				      - natseq->offset_before);

		DEBUGP("sack_adjust: start_seq: %d->%d, end_seq: %d->%d\n",
			ntohl(sack->start_seq), new_start_seq,
			ntohl(sack->end_seq), new_end_seq);

		nf_proto_csum_replace4(&tcph->check, skb,
					sack->start_seq, new_start_seq, 0);
		nf_proto_csum_replace4(&tcph->check, skb,
					sack->end_seq, new_end_seq, 0);
		sack->start_seq = new_start_seq;
		sack->end_seq = new_end_seq;
		sackoff += sizeof(*sack);
	}
}

/* TCP SACK sequence number adjustment */
static inline unsigned int
ip_nat_sack_adjust(struct sk_buff **pskb,
		   struct tcphdr *tcph,
		   struct ip_conntrack *ct,
		   enum ip_conntrack_info ctinfo)
{
	unsigned int dir, optoff, optend;

	optoff = ip_hdrlen(*pskb) + sizeof(struct tcphdr);
	optend = ip_hdrlen(*pskb) + tcph->doff * 4;

	if (!skb_make_writable(pskb, optend))
		return 0;

	dir = CTINFO2DIR(ctinfo);

	while (optoff < optend) {
		/* Usually: option, length. */
		unsigned char *op = (*pskb)->data + optoff;

		switch (op[0]) {
		case TCPOPT_EOL:
			return 1;
		case TCPOPT_NOP:
			optoff++;
			continue;
		default:
			/* no partial options */
			if (optoff + 1 == optend
			    || optoff + op[1] > optend
			    || op[1] < 2)
				return 0;
			if (op[0] == TCPOPT_SACK
			    && op[1] >= 2+TCPOLEN_SACK_PERBLOCK
			    && ((op[1] - 2) % TCPOLEN_SACK_PERBLOCK) == 0)
				sack_adjust(*pskb, tcph, optoff+2,
					    optoff+op[1],
					    &ct->nat.info.seq[!dir]);
			optoff += op[1];
		}
	}
	return 1;
}

/* TCP sequence number adjustment.  Returns 1 on success, 0 on failure */
int
ip_nat_seq_adjust(struct sk_buff **pskb,
		  struct ip_conntrack *ct,
		  enum ip_conntrack_info ctinfo)
{
	struct tcphdr *tcph;
	int dir;
	__be32 newseq, newack;
	struct ip_nat_seq *this_way, *other_way;

	dir = CTINFO2DIR(ctinfo);

	this_way = &ct->nat.info.seq[dir];
	other_way = &ct->nat.info.seq[!dir];

	if (!skb_make_writable(pskb, ip_hdrlen(*pskb) + sizeof(*tcph)))
		return 0;

	tcph = (void *)(*pskb)->data + ip_hdrlen(*pskb);
	if (after(ntohl(tcph->seq), this_way->correction_pos))
		newseq = htonl(ntohl(tcph->seq) + this_way->offset_after);
	else
		newseq = htonl(ntohl(tcph->seq) + this_way->offset_before);

	if (after(ntohl(tcph->ack_seq) - other_way->offset_before,
		  other_way->correction_pos))
		newack = htonl(ntohl(tcph->ack_seq) - other_way->offset_after);
	else
		newack = htonl(ntohl(tcph->ack_seq) - other_way->offset_before);

	nf_proto_csum_replace4(&tcph->check, *pskb, tcph->seq, newseq, 0);
	nf_proto_csum_replace4(&tcph->check, *pskb, tcph->ack_seq, newack, 0);

	DEBUGP("Adjusting sequence number from %u->%u, ack from %u->%u\n",
		ntohl(tcph->seq), ntohl(newseq), ntohl(tcph->ack_seq),
		ntohl(newack));

	tcph->seq = newseq;
	tcph->ack_seq = newack;

	if (!ip_nat_sack_adjust(pskb, tcph, ct, ctinfo))
		return 0;

	ip_conntrack_tcp_update(*pskb, ct, dir);

	return 1;
}
EXPORT_SYMBOL(ip_nat_seq_adjust);

/* Setup NAT on this expected conntrack so it follows master. */
/* If we fail to get a free NAT slot, we'll get dropped on confirm */
void ip_nat_follow_master(struct ip_conntrack *ct,
			  struct ip_conntrack_expect *exp)
{
	struct ip_nat_range range;

	/* This must be a fresh one. */
	BUG_ON(ct->status & IPS_NAT_DONE_MASK);

	/* Change src to where master sends to */
	range.flags = IP_NAT_RANGE_MAP_IPS;
	range.min_ip = range.max_ip
		= ct->master->tuplehash[!exp->dir].tuple.dst.ip;
	/* hook doesn't matter, but it has to do source manip */
	ip_nat_setup_info(ct, &range, NF_IP_POST_ROUTING);

	/* For DST manip, map port here to where it's expected. */
	range.flags = (IP_NAT_RANGE_MAP_IPS | IP_NAT_RANGE_PROTO_SPECIFIED);
	range.min = range.max = exp->saved_proto;
	range.min_ip = range.max_ip
		= ct->master->tuplehash[!exp->dir].tuple.src.ip;
	/* hook doesn't matter, but it has to do destination manip */
	ip_nat_setup_info(ct, &range, NF_IP_PRE_ROUTING);
}
EXPORT_SYMBOL(ip_nat_follow_master);
Commit	Line	Data
e905a9ed	1	/* ip_nat_helper.c - generic support functions for NAT helpers
1da177e4 LT	2	*
	3	* (C) 2000-2002 Harald Welte <laforge@netfilter.org>
	4	* (C) 2003-2004 Netfilter Core Team <coreteam@netfilter.org>
	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	* 14 Jan 2002 Harald Welte <laforge@gnumonks.org>:
e905a9ed	11	* - add support for SACK adjustment
1da177e4 LT	12	* 14 Mar 2002 Harald Welte <laforge@gnumonks.org>:
	13	* - merge SACK support into newnat API
	14	* 16 Aug 2002 Brian J. Murrell <netfilter@interlinx.bc.ca>:
	15	* - make ip_nat_resize_packet more generic (TCP and UDP)
	16	* - add ip_nat_mangle_udp_packet
	17	*/
1da177e4 LT	18	#include <linux/module.h>
	19	#include <linux/kmod.h>
	20	#include <linux/types.h>
	21	#include <linux/timer.h>
	22	#include <linux/skbuff.h>
	23	#include <linux/netfilter_ipv4.h>
	24	#include <net/checksum.h>
	25	#include <net/icmp.h>
	26	#include <net/ip.h>
	27	#include <net/tcp.h>
	28	#include <net/udp.h>
	29
1da177e4 LT	30	#include <linux/netfilter_ipv4/ip_conntrack.h>
	31	#include <linux/netfilter_ipv4/ip_conntrack_helper.h>
	32	#include <linux/netfilter_ipv4/ip_nat.h>
	33	#include <linux/netfilter_ipv4/ip_nat_protocol.h>
	34	#include <linux/netfilter_ipv4/ip_nat_core.h>
	35	#include <linux/netfilter_ipv4/ip_nat_helper.h>
1da177e4 LT	36
	37	#if 0
	38	#define DEBUGP printk
	39	#define DUMP_OFFSET(x) printk("offset_before=%d, offset_after=%d, correction_pos=%u\n", x->offset_before, x->offset_after, x->correction_pos);
	40	#else
	41	#define DEBUGP(format, args...)
	42	#define DUMP_OFFSET(x)
	43	#endif
	44
e45b1be8	45	static DEFINE_SPINLOCK(ip_nat_seqofs_lock);
1da177e4 LT	46
1da177e4 LT	47	/* Setup TCP sequence correction given this change at this sequence */
e905a9ed	48	static inline void
1da177e4 LT	49	adjust_tcp_sequence(u32 seq,
1da177e4 LT	50	int sizediff,
e905a9ed	51	struct ip_conntrack *ct,
1da177e4 LT	52	enum ip_conntrack_info ctinfo)
	53	{
	54	int dir;
	55	struct ip_nat_seq this_way, other_way;
	56
	57	DEBUGP("ip_nat_resize_packet: old_size = %u, new_size = %u\n",
	58	(*skb)->len, new_size);
	59
	60	dir = CTINFO2DIR(ctinfo);
	61
	62	this_way = &ct->nat.info.seq[dir];
	63	other_way = &ct->nat.info.seq[!dir];
	64
	65	DEBUGP("ip_nat_resize_packet: Seq_offset before: ");
	66	DUMP_OFFSET(this_way);
	67
e45b1be8	68	spin_lock_bh(&ip_nat_seqofs_lock);
1da177e4 LT	69
	70	/* SYN adjust. If it's uninitialized, or this is after last
	71	* correction, record it: we don't handle more than one
	72	* adjustment in the window, but do deal with common case of a
	73	* retransmit */
	74	if (this_way->offset_before == this_way->offset_after
	75	\|\| before(this_way->correction_pos, seq)) {
	76	this_way->correction_pos = seq;
	77	this_way->offset_before = this_way->offset_after;
	78	this_way->offset_after += sizediff;
	79	}
e45b1be8	80	spin_unlock_bh(&ip_nat_seqofs_lock);
1da177e4 LT	81
	82	DEBUGP("ip_nat_resize_packet: Seq_offset after: ");
	83	DUMP_OFFSET(this_way);
	84	}
	85
	86	/* Frobs data inside this packet, which is linear. */
	87	static void mangle_contents(struct sk_buff *skb,
	88	unsigned int dataoff,
	89	unsigned int match_offset,
	90	unsigned int match_len,
	91	const char *rep_buffer,
	92	unsigned int rep_len)
	93	{
	94	unsigned char *data;
	95
	96	BUG_ON(skb_is_nonlinear(skb));
	97	data = (unsigned char *)skb->nh.iph + dataoff;
	98
	99	/* move post-replacement */
	100	memmove(data + match_offset + rep_len,
	101	data + match_offset + match_len,
	102	skb->tail - (data + match_offset + match_len));
	103
	104	/* insert data from buffer */
	105	memcpy(data + match_offset, rep_buffer, rep_len);
	106
	107	/* update skb info */
	108	if (rep_len > match_len) {
	109	DEBUGP("ip_nat_mangle_packet: Extending packet by "
	110	"%u from %u bytes\n", rep_len - match_len,
	111	skb->len);
	112	skb_put(skb, rep_len - match_len);
	113	} else {
	114	DEBUGP("ip_nat_mangle_packet: Shrinking packet from "
	115	"%u from %u bytes\n", match_len - rep_len,
	116	skb->len);
	117	__skb_trim(skb, skb->len + rep_len - match_len);
	118	}
	119
	120	/* fix IP hdr checksum information */
	121	skb->nh.iph->tot_len = htons(skb->len);
	122	ip_send_check(skb->nh.iph);
	123	}
	124
	125	/* Unusual, but possible case. */
	126	static int enlarge_skb(struct sk_buff **pskb, unsigned int extra)
	127	{
	128	struct sk_buff *nskb;
	129
	130	if ((*pskb)->len + extra > 65535)
	131	return 0;
	132
	133	nskb = skb_copy_expand(pskb, skb_headroom(pskb), extra, GFP_ATOMIC);
	134	if (!nskb)
	135	return 0;
	136
	137	/* Transfer socket to new skb. */
	138	if ((*pskb)->sk)
	139	skb_set_owner_w(nskb, (*pskb)->sk);
1da177e4 LT	140	kfree_skb(*pskb);
	141	*pskb = nskb;
	142	return 1;
	143	}
	144
	145	/* Generic function for mangling variable-length address changes inside
	146	* NATed TCP connections (like the PORT XXX,XXX,XXX,XXX,XXX,XXX
	147	* command in FTP).
	148	*
	149	* Takes care about all the nasty sequence number changes, checksumming,
	150	* skb enlargement, ...
	151	*
	152	* */
e905a9ed	153	int
1da177e4 LT	154	ip_nat_mangle_tcp_packet(struct sk_buff **pskb,
	155	struct ip_conntrack *ct,
	156	enum ip_conntrack_info ctinfo,
	157	unsigned int match_offset,
	158	unsigned int match_len,
	159	const char *rep_buffer,
	160	unsigned int rep_len)
	161	{
	162	struct iphdr *iph;
	163	struct tcphdr *tcph;
4cf411de	164	int oldlen, datalen;
1da177e4	165
089af26c	166	if (!skb_make_writable(pskb, (*pskb)->len))
1da177e4 LT	167	return 0;
	168
	169	if (rep_len > match_len
	170	&& rep_len - match_len > skb_tailroom(*pskb)
	171	&& !enlarge_skb(pskb, rep_len - match_len))
	172	return 0;
	173
	174	SKB_LINEAR_ASSERT(*pskb);
	175
	176	iph = (*pskb)->nh.iph;
	177	tcph = (void )iph + iph->ihl4;
	178
4cf411de	179	oldlen = (pskb)->len - iph->ihl4;
1da177e4 LT	180	mangle_contents(pskb, iph->ihl4 + tcph->doff*4,
	181	match_offset, match_len, rep_buffer, rep_len);
	182
	183	datalen = (pskb)->len - iph->ihl4;
4cf411de PM	184	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
4cf411de PM	185	tcph->check = 0;
ba7808ea	186	tcph->check = tcp_v4_check(datalen,
4cf411de PM	187	iph->saddr, iph->daddr,
4cf411de PM	188	csum_partial((char *)tcph,
e905a9ed	189	datalen, 0));
4cf411de	190	} else
43bc0ca7 AV	191	nf_proto_csum_replace2(&tcph->check, *pskb,
43bc0ca7 AV	192	htons(oldlen), htons(datalen), 1);
1da177e4 LT	193
	194	if (rep_len != match_len) {
	195	set_bit(IPS_SEQ_ADJUST_BIT, &ct->status);
	196	adjust_tcp_sequence(ntohl(tcph->seq),
	197	(int)rep_len - (int)match_len,
	198	ct, ctinfo);
	199	/* Tell TCP window tracking about seq change */
	200	ip_conntrack_tcp_update(*pskb, ct, CTINFO2DIR(ctinfo));
	201	}
	202	return 1;
	203	}
188bab3a	204	EXPORT_SYMBOL(ip_nat_mangle_tcp_packet);
e905a9ed	205
1da177e4 LT	206	/* Generic function for mangling variable-length address changes inside
	207	* NATed UDP connections (like the CONNECT DATA XXXXX MESG XXXXX INDEX XXXXX
	208	* command in the Amanda protocol)
	209	*
	210	* Takes care about all the nasty sequence number changes, checksumming,
	211	* skb enlargement, ...
	212	*
	213	* XXX - This function could be merged with ip_nat_mangle_tcp_packet which
	214	* should be fairly easy to do.
	215	*/
e905a9ed	216	int
1da177e4 LT	217	ip_nat_mangle_udp_packet(struct sk_buff **pskb,
	218	struct ip_conntrack *ct,
	219	enum ip_conntrack_info ctinfo,
	220	unsigned int match_offset,
	221	unsigned int match_len,
	222	const char *rep_buffer,
	223	unsigned int rep_len)
	224	{
	225	struct iphdr *iph;
	226	struct udphdr *udph;
4cf411de	227	int datalen, oldlen;
1da177e4 LT	228
	229	/* UDP helpers might accidentally mangle the wrong packet */
	230	iph = (*pskb)->nh.iph;
e905a9ed YH	231	if ((pskb)->len < iph->ihl4 + sizeof(*udph) +
e905a9ed YH	232	match_offset + match_len)
1da177e4 LT	233	return 0;
1da177e4 LT	234
089af26c	235	if (!skb_make_writable(pskb, (*pskb)->len))
1da177e4 LT	236	return 0;
	237
	238	if (rep_len > match_len
	239	&& rep_len - match_len > skb_tailroom(*pskb)
	240	&& !enlarge_skb(pskb, rep_len - match_len))
	241	return 0;
	242
	243	iph = (*pskb)->nh.iph;
	244	udph = (void )iph + iph->ihl4;
4cf411de PM	245
4cf411de PM	246	oldlen = (pskb)->len - iph->ihl4;
1da177e4 LT	247	mangle_contents(pskb, iph->ihl4 + sizeof(*udph),
	248	match_offset, match_len, rep_buffer, rep_len);
	249
	250	/* update the length of the UDP packet */
4cf411de PM	251	datalen = (pskb)->len - iph->ihl4;
4cf411de PM	252	udph->len = htons(datalen);
1da177e4 LT	253
1da177e4 LT	254	/* fix udp checksum if udp checksum was previously calculated */
4cf411de PM	255	if (!udph->check && (*pskb)->ip_summed != CHECKSUM_PARTIAL)
	256	return 1;
	257
	258	if ((*pskb)->ip_summed != CHECKSUM_PARTIAL) {
1da177e4 LT	259	udph->check = 0;
1da177e4 LT	260	udph->check = csum_tcpudp_magic(iph->saddr, iph->daddr,
e905a9ed YH	261	datalen, IPPROTO_UDP,
	262	csum_partial((char *)udph,
	263	datalen, 0));
4cf411de	264	if (!udph->check)
f6ab0288	265	udph->check = CSUM_MANGLED_0;
4cf411de	266	} else
43bc0ca7 AV	267	nf_proto_csum_replace2(&udph->check, *pskb,
43bc0ca7 AV	268	htons(oldlen), htons(datalen), 1);
1da177e4 LT	269	return 1;
1da177e4 LT	270	}
188bab3a	271	EXPORT_SYMBOL(ip_nat_mangle_udp_packet);
1da177e4 LT	272
	273	/* Adjust one found SACK option including checksum correction */
	274	static void
	275	sack_adjust(struct sk_buff *skb,
e905a9ed	276	struct tcphdr *tcph,
1da177e4 LT	277	unsigned int sackoff,
	278	unsigned int sackend,
	279	struct ip_nat_seq *natseq)
	280	{
	281	while (sackoff < sackend) {
269bd27e	282	struct tcp_sack_block_wire *sack;
a76b11dd	283	__be32 new_start_seq, new_end_seq;
1da177e4 LT	284
	285	sack = (void *)skb->data + sackoff;
	286	if (after(ntohl(sack->start_seq) - natseq->offset_before,
	287	natseq->correction_pos))
a76b11dd AV	288	new_start_seq = htonl(ntohl(sack->start_seq)
a76b11dd AV	289	- natseq->offset_after);
1da177e4	290	else
a76b11dd AV	291	new_start_seq = htonl(ntohl(sack->start_seq)
a76b11dd AV	292	- natseq->offset_before);
1da177e4 LT	293
	294	if (after(ntohl(sack->end_seq) - natseq->offset_before,
	295	natseq->correction_pos))
a76b11dd AV	296	new_end_seq = htonl(ntohl(sack->end_seq)
a76b11dd AV	297	- natseq->offset_after);
1da177e4	298	else
a76b11dd AV	299	new_end_seq = htonl(ntohl(sack->end_seq)
a76b11dd AV	300	- natseq->offset_before);
1da177e4 LT	301
	302	DEBUGP("sack_adjust: start_seq: %d->%d, end_seq: %d->%d\n",
	303	ntohl(sack->start_seq), new_start_seq,
	304	ntohl(sack->end_seq), new_end_seq);
	305
43bc0ca7 AV	306	nf_proto_csum_replace4(&tcph->check, skb,
	307	sack->start_seq, new_start_seq, 0);
	308	nf_proto_csum_replace4(&tcph->check, skb,
	309	sack->end_seq, new_end_seq, 0);
1da177e4 LT	310	sack->start_seq = new_start_seq;
	311	sack->end_seq = new_end_seq;
	312	sackoff += sizeof(*sack);
	313	}
	314	}
	315
	316	/* TCP SACK sequence number adjustment */
	317	static inline unsigned int
	318	ip_nat_sack_adjust(struct sk_buff **pskb,
	319	struct tcphdr *tcph,
	320	struct ip_conntrack *ct,
	321	enum ip_conntrack_info ctinfo)
	322	{
	323	unsigned int dir, optoff, optend;
	324
c9bdd4b5 ACM	325	optoff = ip_hdrlen(*pskb) + sizeof(struct tcphdr);
c9bdd4b5 ACM	326	optend = ip_hdrlen(pskb) + tcph->doff 4;
1da177e4	327
089af26c	328	if (!skb_make_writable(pskb, optend))
1da177e4 LT	329	return 0;
	330
	331	dir = CTINFO2DIR(ctinfo);
	332
	333	while (optoff < optend) {
	334	/* Usually: option, length. */
	335	unsigned char op = (pskb)->data + optoff;
	336
	337	switch (op[0]) {
	338	case TCPOPT_EOL:
	339	return 1;
	340	case TCPOPT_NOP:
	341	optoff++;
	342	continue;
	343	default:
	344	/* no partial options */
	345	if (optoff + 1 == optend
	346	\|\| optoff + op[1] > optend
	347	\|\| op[1] < 2)
	348	return 0;
	349	if (op[0] == TCPOPT_SACK
	350	&& op[1] >= 2+TCPOLEN_SACK_PERBLOCK
	351	&& ((op[1] - 2) % TCPOLEN_SACK_PERBLOCK) == 0)
	352	sack_adjust(*pskb, tcph, optoff+2,
	353	optoff+op[1],
	354	&ct->nat.info.seq[!dir]);
	355	optoff += op[1];
	356	}
	357	}
	358	return 1;
	359	}
	360
	361	/* TCP sequence number adjustment. Returns 1 on success, 0 on failure */
	362	int
e905a9ed YH	363	ip_nat_seq_adjust(struct sk_buff **pskb,
e905a9ed YH	364	struct ip_conntrack *ct,
1da177e4 LT	365	enum ip_conntrack_info ctinfo)
	366	{
	367	struct tcphdr *tcph;
a76b11dd AV	368	int dir;
a76b11dd AV	369	__be32 newseq, newack;
e905a9ed	370	struct ip_nat_seq this_way, other_way;
1da177e4 LT	371
	372	dir = CTINFO2DIR(ctinfo);
	373
	374	this_way = &ct->nat.info.seq[dir];
	375	other_way = &ct->nat.info.seq[!dir];
	376
c9bdd4b5	377	if (!skb_make_writable(pskb, ip_hdrlen(pskb) + sizeof(tcph)))
1da177e4 LT	378	return 0;
1da177e4 LT	379
c9bdd4b5	380	tcph = (void )(pskb)->data + ip_hdrlen(*pskb);
1da177e4	381	if (after(ntohl(tcph->seq), this_way->correction_pos))
a76b11dd	382	newseq = htonl(ntohl(tcph->seq) + this_way->offset_after);
1da177e4	383	else
a76b11dd	384	newseq = htonl(ntohl(tcph->seq) + this_way->offset_before);
1da177e4 LT	385
	386	if (after(ntohl(tcph->ack_seq) - other_way->offset_before,
	387	other_way->correction_pos))
a76b11dd	388	newack = htonl(ntohl(tcph->ack_seq) - other_way->offset_after);
1da177e4	389	else
a76b11dd	390	newack = htonl(ntohl(tcph->ack_seq) - other_way->offset_before);
1da177e4	391
43bc0ca7 AV	392	nf_proto_csum_replace4(&tcph->check, *pskb, tcph->seq, newseq, 0);
43bc0ca7 AV	393	nf_proto_csum_replace4(&tcph->check, *pskb, tcph->ack_seq, newack, 0);
1da177e4 LT	394
	395	DEBUGP("Adjusting sequence number from %u->%u, ack from %u->%u\n",
	396	ntohl(tcph->seq), ntohl(newseq), ntohl(tcph->ack_seq),
	397	ntohl(newack));
	398
	399	tcph->seq = newseq;
	400	tcph->ack_seq = newack;
	401
	402	if (!ip_nat_sack_adjust(pskb, tcph, ct, ctinfo))
	403	return 0;
	404
	405	ip_conntrack_tcp_update(*pskb, ct, dir);
	406
	407	return 1;
	408	}
188bab3a	409	EXPORT_SYMBOL(ip_nat_seq_adjust);
1da177e4 LT	410
	411	/* Setup NAT on this expected conntrack so it follows master. */
	412	/* If we fail to get a free NAT slot, we'll get dropped on confirm */
	413	void ip_nat_follow_master(struct ip_conntrack *ct,
	414	struct ip_conntrack_expect *exp)
	415	{
	416	struct ip_nat_range range;
	417
	418	/* This must be a fresh one. */
	419	BUG_ON(ct->status & IPS_NAT_DONE_MASK);
	420
	421	/* Change src to where master sends to */
	422	range.flags = IP_NAT_RANGE_MAP_IPS;
	423	range.min_ip = range.max_ip
	424	= ct->master->tuplehash[!exp->dir].tuple.dst.ip;
	425	/* hook doesn't matter, but it has to do source manip */
	426	ip_nat_setup_info(ct, &range, NF_IP_POST_ROUTING);
	427
	428	/* For DST manip, map port here to where it's expected. */
	429	range.flags = (IP_NAT_RANGE_MAP_IPS \| IP_NAT_RANGE_PROTO_SPECIFIED);
	430	range.min = range.max = exp->saved_proto;
	431	range.min_ip = range.max_ip
	432	= ct->master->tuplehash[!exp->dir].tuple.src.ip;
	433	/* hook doesn't matter, but it has to do destination manip */
	434	ip_nat_setup_info(ct, &range, NF_IP_PRE_ROUTING);
	435	}
188bab3a	436	EXPORT_SYMBOL(ip_nat_follow_master);