[GitHub/mt8127/android_kernel_alcatel_ttab.git] / net / netfilter / nf_nat_helper.c

/* nf_nat_helper.c - generic support functions for NAT helpers
 *
 * (C) 2000-2002 Harald Welte <laforge@netfilter.org>
 * (C) 2003-2006 Netfilter Core Team <coreteam@netfilter.org>
 * (C) 2007-2012 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/module.h>
#include <linux/gfp.h>
#include <linux/types.h>
#include <linux/skbuff.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <net/tcp.h>

#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_helper.h>
#include <net/netfilter/nf_conntrack_ecache.h>
#include <net/netfilter/nf_conntrack_expect.h>
#include <net/netfilter/nf_nat.h>
#include <net/netfilter/nf_nat_l3proto.h>
#include <net/netfilter/nf_nat_l4proto.h>
#include <net/netfilter/nf_nat_core.h>
#include <net/netfilter/nf_nat_helper.h>

#define DUMP_OFFSET(x) \
	pr_debug("offset_before=%d, offset_after=%d, correction_pos=%u\n", \
		 x->offset_before, x->offset_after, x->correction_pos);

static DEFINE_SPINLOCK(nf_nat_seqofs_lock);

/* Setup TCP sequence correction given this change at this sequence */
static inline void
adjust_tcp_sequence(u32 seq,
		    int sizediff,
		    struct nf_conn *ct,
		    enum ip_conntrack_info ctinfo)
{
	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	struct nf_conn_nat *nat = nfct_nat(ct);
	struct nf_nat_seq *this_way = &nat->seq[dir];

	pr_debug("adjust_tcp_sequence: seq = %u, sizediff = %d\n",
		 seq, sizediff);

	pr_debug("adjust_tcp_sequence: Seq_offset before: ");
	DUMP_OFFSET(this_way);

	spin_lock_bh(&nf_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(&nf_nat_seqofs_lock);

	pr_debug("adjust_tcp_sequence: Seq_offset after: ");
	DUMP_OFFSET(this_way);
}

/* Get the offset value, for conntrack */
s16 nf_nat_get_offset(const struct nf_conn *ct,
		      enum ip_conntrack_dir dir,
		      u32 seq)
{
	struct nf_conn_nat *nat = nfct_nat(ct);
	struct nf_nat_seq *this_way;
	s16 offset;

	if (!nat)
		return 0;

	this_way = &nat->seq[dir];
	spin_lock_bh(&nf_nat_seqofs_lock);
	offset = after(seq, this_way->correction_pos)
		 ? this_way->offset_after : this_way->offset_before;
	spin_unlock_bh(&nf_nat_seqofs_lock);

	return offset;
}

/* 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 = skb_network_header(skb) + dataoff;

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

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

	/* update skb info */
	if (rep_len > match_len) {
		pr_debug("nf_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 {
		pr_debug("nf_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);
	}

	if (nf_ct_l3num((struct nf_conn *)skb->nfct) == NFPROTO_IPV4) {
		/* fix IP hdr checksum information */
		ip_hdr(skb)->tot_len = htons(skb->len);
		ip_send_check(ip_hdr(skb));
	} else
		ipv6_hdr(skb)->payload_len =
			htons(skb->len - sizeof(struct ipv6hdr));
}

/* Unusual, but possible case. */
static int enlarge_skb(struct sk_buff *skb, unsigned int extra)
{
	if (skb->len + extra > 65535)
		return 0;

	if (pskb_expand_head(skb, 0, extra - skb_tailroom(skb), GFP_ATOMIC))
		return 0;

	return 1;
}

void nf_nat_set_seq_adjust(struct nf_conn *ct, enum ip_conntrack_info ctinfo,
			   __be32 seq, s16 off)
{
	if (!off)
		return;
	set_bit(IPS_SEQ_ADJUST_BIT, &ct->status);
	adjust_tcp_sequence(ntohl(seq), off, ct, ctinfo);
	nf_conntrack_event_cache(IPCT_NATSEQADJ, ct);
}
EXPORT_SYMBOL_GPL(nf_nat_set_seq_adjust);

void nf_nat_tcp_seq_adjust(struct sk_buff *skb, struct nf_conn *ct,
			   u32 ctinfo, int off)
{
	const struct tcphdr *th;

	if (nf_ct_protonum(ct) != IPPROTO_TCP)
		return;

	th = (struct tcphdr *)(skb_network_header(skb)+ ip_hdrlen(skb));
	nf_nat_set_seq_adjust(ct, ctinfo, th->seq, off);
}
EXPORT_SYMBOL_GPL(nf_nat_tcp_seq_adjust);

/* 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 __nf_nat_mangle_tcp_packet(struct sk_buff *skb,
			       struct nf_conn *ct,
			       enum ip_conntrack_info ctinfo,
			       unsigned int protoff,
			       unsigned int match_offset,
			       unsigned int match_len,
			       const char *rep_buffer,
			       unsigned int rep_len, bool adjust)
{
	const struct nf_nat_l3proto *l3proto;
	struct tcphdr *tcph;
	int oldlen, datalen;

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

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

	SKB_LINEAR_ASSERT(skb);

	tcph = (void *)skb->data + protoff;

	oldlen = skb->len - protoff;
	mangle_contents(skb, protoff + tcph->doff*4,
			match_offset, match_len, rep_buffer, rep_len);

	datalen = skb->len - protoff;

	l3proto = __nf_nat_l3proto_find(nf_ct_l3num(ct));
	l3proto->csum_recalc(skb, IPPROTO_TCP, tcph, &tcph->check,
			     datalen, oldlen);

	if (adjust && rep_len != match_len)
		nf_nat_set_seq_adjust(ct, ctinfo, tcph->seq,
				      (int)rep_len - (int)match_len);

	return 1;
}
EXPORT_SYMBOL(__nf_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 nf_nat_mangle_tcp_packet which
 *       should be fairly easy to do.
 */
int
nf_nat_mangle_udp_packet(struct sk_buff *skb,
			 struct nf_conn *ct,
			 enum ip_conntrack_info ctinfo,
			 unsigned int protoff,
			 unsigned int match_offset,
			 unsigned int match_len,
			 const char *rep_buffer,
			 unsigned int rep_len)
{
	const struct nf_nat_l3proto *l3proto;
	struct udphdr *udph;
	int datalen, oldlen;

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

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

	udph = (void *)skb->data + protoff;

	oldlen = skb->len - protoff;
	mangle_contents(skb, protoff + sizeof(*udph),
			match_offset, match_len, rep_buffer, rep_len);

	/* update the length of the UDP packet */
	datalen = skb->len - protoff;
	udph->len = htons(datalen);

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

	l3proto = __nf_nat_l3proto_find(nf_ct_l3num(ct));
	l3proto->csum_recalc(skb, IPPROTO_UDP, udph, &udph->check,
			     datalen, oldlen);

	return 1;
}
EXPORT_SYMBOL(nf_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 nf_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);

		pr_debug("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);

		inet_proto_csum_replace4(&tcph->check, skb,
					 sack->start_seq, new_start_seq, 0);
		inet_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
nf_nat_sack_adjust(struct sk_buff *skb,
		   unsigned int protoff,
		   struct tcphdr *tcph,
		   struct nf_conn *ct,
		   enum ip_conntrack_info ctinfo)
{
	unsigned int dir, optoff, optend;
	struct nf_conn_nat *nat = nfct_nat(ct);

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

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

	dir = CTINFO2DIR(ctinfo);

	while (optoff < optend) {
		/* Usually: option, length. */
		unsigned char *op = skb->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(skb, tcph, optoff+2,
					    optoff+op[1], &nat->seq[!dir]);
			optoff += op[1];
		}
	}
	return 1;
}

/* TCP sequence number adjustment.  Returns 1 on success, 0 on failure */
int
nf_nat_seq_adjust(struct sk_buff *skb,
		  struct nf_conn *ct,
		  enum ip_conntrack_info ctinfo,
		  unsigned int protoff)
{
	struct tcphdr *tcph;
	int dir;
	__be32 newseq, newack;
	s16 seqoff, ackoff;
	struct nf_conn_nat *nat = nfct_nat(ct);
	struct nf_nat_seq *this_way, *other_way;

	dir = CTINFO2DIR(ctinfo);

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

	if (!skb_make_writable(skb, protoff + sizeof(*tcph)))
		return 0;

	tcph = (void *)skb->data + protoff;
	if (after(ntohl(tcph->seq), this_way->correction_pos))
		seqoff = this_way->offset_after;
	else
		seqoff = this_way->offset_before;

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

	newseq = htonl(ntohl(tcph->seq) + seqoff);
	newack = htonl(ntohl(tcph->ack_seq) - ackoff);

	inet_proto_csum_replace4(&tcph->check, skb, tcph->seq, newseq, 0);
	inet_proto_csum_replace4(&tcph->check, skb, tcph->ack_seq, newack, 0);

	pr_debug("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;

	return nf_nat_sack_adjust(skb, protoff, tcph, ct, ctinfo);
}

/* 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 nf_nat_follow_master(struct nf_conn *ct,
			  struct nf_conntrack_expect *exp)
{
	struct nf_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 = NF_NAT_RANGE_MAP_IPS;
	range.min_addr = range.max_addr
		= ct->master->tuplehash[!exp->dir].tuple.dst.u3;
	nf_nat_setup_info(ct, &range, NF_NAT_MANIP_SRC);

	/* For DST manip, map port here to where it's expected. */
	range.flags = (NF_NAT_RANGE_MAP_IPS | NF_NAT_RANGE_PROTO_SPECIFIED);
	range.min_proto = range.max_proto = exp->saved_proto;
	range.min_addr = range.max_addr
		= ct->master->tuplehash[!exp->dir].tuple.src.u3;
	nf_nat_setup_info(ct, &range, NF_NAT_MANIP_DST);
}
EXPORT_SYMBOL(nf_nat_follow_master);
Commit	Line	Data
c7232c99	1	/* nf_nat_helper.c - generic support functions for NAT helpers
5b1158e9 JK	2	*
	3	* (C) 2000-2002 Harald Welte <laforge@netfilter.org>
	4	* (C) 2003-2006 Netfilter Core Team <coreteam@netfilter.org>
f229f6ce	5	* (C) 2007-2012 Patrick McHardy <kaber@trash.net>
5b1158e9 JK	6	*
	7	* This program is free software; you can redistribute it and/or modify
	8	* it under the terms of the GNU General Public License version 2 as
	9	* published by the Free Software Foundation.
	10	*/
	11	#include <linux/module.h>
5a0e3ad6	12	#include <linux/gfp.h>
5b1158e9	13	#include <linux/types.h>
5b1158e9 JK	14	#include <linux/skbuff.h>
	15	#include <linux/tcp.h>
	16	#include <linux/udp.h>
5b1158e9 JK	17	#include <net/tcp.h>
5b1158e9 JK	18
5b1158e9 JK	19	#include <net/netfilter/nf_conntrack.h>
5b1158e9 JK	20	#include <net/netfilter/nf_conntrack_helper.h>
13eae15a	21	#include <net/netfilter/nf_conntrack_ecache.h>
5b1158e9 JK	22	#include <net/netfilter/nf_conntrack_expect.h>
5b1158e9 JK	23	#include <net/netfilter/nf_nat.h>
c7232c99 PM	24	#include <net/netfilter/nf_nat_l3proto.h>
c7232c99 PM	25	#include <net/netfilter/nf_nat_l4proto.h>
5b1158e9 JK	26	#include <net/netfilter/nf_nat_core.h>
	27	#include <net/netfilter/nf_nat_helper.h>
	28
0d53778e PM	29	#define DUMP_OFFSET(x) \
	30	pr_debug("offset_before=%d, offset_after=%d, correction_pos=%u\n", \
	31	x->offset_before, x->offset_after, x->correction_pos);
5b1158e9 JK	32
	33	static DEFINE_SPINLOCK(nf_nat_seqofs_lock);
	34
	35	/* Setup TCP sequence correction given this change at this sequence */
	36	static inline void
	37	adjust_tcp_sequence(u32 seq,
	38	int sizediff,
	39	struct nf_conn *ct,
	40	enum ip_conntrack_info ctinfo)
	41	{
76ac8940	42	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
5b1158e9	43	struct nf_conn_nat *nat = nfct_nat(ct);
76ac8940	44	struct nf_nat_seq *this_way = &nat->seq[dir];
5b1158e9	45
76ac8940 HE	46	pr_debug("adjust_tcp_sequence: seq = %u, sizediff = %d\n",
76ac8940 HE	47	seq, sizediff);
5b1158e9	48
76ac8940	49	pr_debug("adjust_tcp_sequence: Seq_offset before: ");
5b1158e9 JK	50	DUMP_OFFSET(this_way);
	51
	52	spin_lock_bh(&nf_nat_seqofs_lock);
	53
	54	/* SYN adjust. If it's uninitialized, or this is after last
	55	* correction, record it: we don't handle more than one
	56	* adjustment in the window, but do deal with common case of a
	57	* retransmit */
	58	if (this_way->offset_before == this_way->offset_after \|\|
	59	before(this_way->correction_pos, seq)) {
76ac8940 HE	60	this_way->correction_pos = seq;
	61	this_way->offset_before = this_way->offset_after;
	62	this_way->offset_after += sizediff;
5b1158e9 JK	63	}
	64	spin_unlock_bh(&nf_nat_seqofs_lock);
	65
76ac8940	66	pr_debug("adjust_tcp_sequence: Seq_offset after: ");
5b1158e9 JK	67	DUMP_OFFSET(this_way);
	68	}
	69
f9dd09c7 JK	70	/* Get the offset value, for conntrack */
	71	s16 nf_nat_get_offset(const struct nf_conn *ct,
	72	enum ip_conntrack_dir dir,
	73	u32 seq)
	74	{
	75	struct nf_conn_nat *nat = nfct_nat(ct);
	76	struct nf_nat_seq *this_way;
	77	s16 offset;
	78
	79	if (!nat)
	80	return 0;
	81
	82	this_way = &nat->seq[dir];
	83	spin_lock_bh(&nf_nat_seqofs_lock);
	84	offset = after(seq, this_way->correction_pos)
	85	? this_way->offset_after : this_way->offset_before;
	86	spin_unlock_bh(&nf_nat_seqofs_lock);
	87
	88	return offset;
	89	}
f9dd09c7	90
5b1158e9 JK	91	/* Frobs data inside this packet, which is linear. */
	92	static void mangle_contents(struct sk_buff *skb,
	93	unsigned int dataoff,
	94	unsigned int match_offset,
	95	unsigned int match_len,
	96	const char *rep_buffer,
	97	unsigned int rep_len)
	98	{
	99	unsigned char *data;
	100
	101	BUG_ON(skb_is_nonlinear(skb));
eddc9ec5	102	data = skb_network_header(skb) + dataoff;
5b1158e9 JK	103
	104	/* move post-replacement */
	105	memmove(data + match_offset + rep_len,
	106	data + match_offset + match_len,
27a884dc ACM	107	skb->tail - (skb->network_header + dataoff +
27a884dc ACM	108	match_offset + match_len));
5b1158e9 JK	109
	110	/* insert data from buffer */
	111	memcpy(data + match_offset, rep_buffer, rep_len);
	112
	113	/* update skb info */
	114	if (rep_len > match_len) {
0d53778e PM	115	pr_debug("nf_nat_mangle_packet: Extending packet by "
0d53778e PM	116	"%u from %u bytes\n", rep_len - match_len, skb->len);
5b1158e9 JK	117	skb_put(skb, rep_len - match_len);
5b1158e9 JK	118	} else {
0d53778e PM	119	pr_debug("nf_nat_mangle_packet: Shrinking packet from "
0d53778e PM	120	"%u from %u bytes\n", match_len - rep_len, skb->len);
5b1158e9 JK	121	__skb_trim(skb, skb->len + rep_len - match_len);
	122	}
	123
c7232c99 PM	124	if (nf_ct_l3num((struct nf_conn *)skb->nfct) == NFPROTO_IPV4) {
	125	/* fix IP hdr checksum information */
	126	ip_hdr(skb)->tot_len = htons(skb->len);
	127	ip_send_check(ip_hdr(skb));
	128	} else
	129	ipv6_hdr(skb)->payload_len =
	130	htons(skb->len - sizeof(struct ipv6hdr));
5b1158e9 JK	131	}
	132
	133	/* Unusual, but possible case. */
3db05fea	134	static int enlarge_skb(struct sk_buff *skb, unsigned int extra)
5b1158e9	135	{
3db05fea	136	if (skb->len + extra > 65535)
5b1158e9 JK	137	return 0;
5b1158e9 JK	138
3db05fea	139	if (pskb_expand_head(skb, 0, extra - skb_tailroom(skb), GFP_ATOMIC))
5b1158e9 JK	140	return 0;
5b1158e9 JK	141
5b1158e9 JK	142	return 1;
	143	}
	144
010c0b9f PM	145	void nf_nat_set_seq_adjust(struct nf_conn *ct, enum ip_conntrack_info ctinfo,
	146	__be32 seq, s16 off)
	147	{
	148	if (!off)
	149	return;
	150	set_bit(IPS_SEQ_ADJUST_BIT, &ct->status);
	151	adjust_tcp_sequence(ntohl(seq), off, ct, ctinfo);
	152	nf_conntrack_event_cache(IPCT_NATSEQADJ, ct);
	153	}
	154	EXPORT_SYMBOL_GPL(nf_nat_set_seq_adjust);
	155
8c88f87c PNA	156	void nf_nat_tcp_seq_adjust(struct sk_buff skb, struct nf_conn ct,
	157	u32 ctinfo, int off)
	158	{
	159	const struct tcphdr *th;
	160
	161	if (nf_ct_protonum(ct) != IPPROTO_TCP)
	162	return;
	163
	164	th = (struct tcphdr *)(skb_network_header(skb)+ ip_hdrlen(skb));
	165	nf_nat_set_seq_adjust(ct, ctinfo, th->seq, off);
	166	}
	167	EXPORT_SYMBOL_GPL(nf_nat_tcp_seq_adjust);
	168
5b1158e9 JK	169	/* Generic function for mangling variable-length address changes inside
	170	* NATed TCP connections (like the PORT XXX,XXX,XXX,XXX,XXX,XXX
	171	* command in FTP).
	172	*
	173	* Takes care about all the nasty sequence number changes, checksumming,
	174	* skb enlargement, ...
	175	*
	176	* */
010c0b9f PM	177	int __nf_nat_mangle_tcp_packet(struct sk_buff *skb,
	178	struct nf_conn *ct,
	179	enum ip_conntrack_info ctinfo,
051966c0	180	unsigned int protoff,
010c0b9f PM	181	unsigned int match_offset,
	182	unsigned int match_len,
	183	const char *rep_buffer,
	184	unsigned int rep_len, bool adjust)
5b1158e9	185	{
c7232c99	186	const struct nf_nat_l3proto *l3proto;
5b1158e9 JK	187	struct tcphdr *tcph;
	188	int oldlen, datalen;
	189
3db05fea	190	if (!skb_make_writable(skb, skb->len))
5b1158e9 JK	191	return 0;
	192
	193	if (rep_len > match_len &&
3db05fea HX	194	rep_len - match_len > skb_tailroom(skb) &&
3db05fea HX	195	!enlarge_skb(skb, rep_len - match_len))
5b1158e9 JK	196	return 0;
5b1158e9 JK	197
3db05fea	198	SKB_LINEAR_ASSERT(skb);
5b1158e9	199
c7232c99	200	tcph = (void *)skb->data + protoff;
5b1158e9	201
c7232c99 PM	202	oldlen = skb->len - protoff;
c7232c99 PM	203	mangle_contents(skb, protoff + tcph->doff*4,
5b1158e9 JK	204	match_offset, match_len, rep_buffer, rep_len);
5b1158e9 JK	205
c7232c99 PM	206	datalen = skb->len - protoff;
	207
	208	l3proto = __nf_nat_l3proto_find(nf_ct_l3num(ct));
	209	l3proto->csum_recalc(skb, IPPROTO_TCP, tcph, &tcph->check,
	210	datalen, oldlen);
5b1158e9	211
010c0b9f PM	212	if (adjust && rep_len != match_len)
	213	nf_nat_set_seq_adjust(ct, ctinfo, tcph->seq,
	214	(int)rep_len - (int)match_len);
	215
5b1158e9 JK	216	return 1;
5b1158e9 JK	217	}
010c0b9f	218	EXPORT_SYMBOL(__nf_nat_mangle_tcp_packet);
5b1158e9 JK	219
	220	/* Generic function for mangling variable-length address changes inside
	221	* NATed UDP connections (like the CONNECT DATA XXXXX MESG XXXXX INDEX XXXXX
	222	* command in the Amanda protocol)
	223	*
	224	* Takes care about all the nasty sequence number changes, checksumming,
	225	* skb enlargement, ...
	226	*
	227	* XXX - This function could be merged with nf_nat_mangle_tcp_packet which
	228	* should be fairly easy to do.
	229	*/
	230	int
3db05fea	231	nf_nat_mangle_udp_packet(struct sk_buff *skb,
5b1158e9 JK	232	struct nf_conn *ct,
5b1158e9 JK	233	enum ip_conntrack_info ctinfo,
051966c0	234	unsigned int protoff,
5b1158e9 JK	235	unsigned int match_offset,
	236	unsigned int match_len,
	237	const char *rep_buffer,
	238	unsigned int rep_len)
	239	{
c7232c99	240	const struct nf_nat_l3proto *l3proto;
5b1158e9 JK	241	struct udphdr *udph;
	242	int datalen, oldlen;
	243
3db05fea	244	if (!skb_make_writable(skb, skb->len))
5b1158e9 JK	245	return 0;
	246
	247	if (rep_len > match_len &&
3db05fea HX	248	rep_len - match_len > skb_tailroom(skb) &&
3db05fea HX	249	!enlarge_skb(skb, rep_len - match_len))
5b1158e9 JK	250	return 0;
5b1158e9 JK	251
c7232c99	252	udph = (void *)skb->data + protoff;
5b1158e9	253
c7232c99 PM	254	oldlen = skb->len - protoff;
c7232c99 PM	255	mangle_contents(skb, protoff + sizeof(*udph),
5b1158e9 JK	256	match_offset, match_len, rep_buffer, rep_len);
	257
	258	/* update the length of the UDP packet */
c7232c99	259	datalen = skb->len - protoff;
5b1158e9 JK	260	udph->len = htons(datalen);
	261
	262	/* fix udp checksum if udp checksum was previously calculated */
3db05fea	263	if (!udph->check && skb->ip_summed != CHECKSUM_PARTIAL)
5b1158e9 JK	264	return 1;
5b1158e9 JK	265
c7232c99 PM	266	l3proto = __nf_nat_l3proto_find(nf_ct_l3num(ct));
	267	l3proto->csum_recalc(skb, IPPROTO_UDP, udph, &udph->check,
	268	datalen, oldlen);
5b1158e9 JK	269
	270	return 1;
	271	}
	272	EXPORT_SYMBOL(nf_nat_mangle_udp_packet);
	273
	274	/* Adjust one found SACK option including checksum correction */
	275	static void
	276	sack_adjust(struct sk_buff *skb,
	277	struct tcphdr *tcph,
	278	unsigned int sackoff,
	279	unsigned int sackend,
	280	struct nf_nat_seq *natseq)
	281	{
	282	while (sackoff < sackend) {
	283	struct tcp_sack_block_wire *sack;
	284	__be32 new_start_seq, new_end_seq;
	285
	286	sack = (void *)skb->data + sackoff;
	287	if (after(ntohl(sack->start_seq) - natseq->offset_before,
	288	natseq->correction_pos))
	289	new_start_seq = htonl(ntohl(sack->start_seq)
	290	- natseq->offset_after);
	291	else
	292	new_start_seq = htonl(ntohl(sack->start_seq)
	293	- natseq->offset_before);
	294
	295	if (after(ntohl(sack->end_seq) - natseq->offset_before,
	296	natseq->correction_pos))
	297	new_end_seq = htonl(ntohl(sack->end_seq)
	298	- natseq->offset_after);
	299	else
	300	new_end_seq = htonl(ntohl(sack->end_seq)
	301	- natseq->offset_before);
	302
0d53778e PM	303	pr_debug("sack_adjust: start_seq: %d->%d, end_seq: %d->%d\n",
	304	ntohl(sack->start_seq), new_start_seq,
	305	ntohl(sack->end_seq), new_end_seq);
5b1158e9	306
be0ea7d5 PM	307	inet_proto_csum_replace4(&tcph->check, skb,
	308	sack->start_seq, new_start_seq, 0);
	309	inet_proto_csum_replace4(&tcph->check, skb,
	310	sack->end_seq, new_end_seq, 0);
5b1158e9 JK	311	sack->start_seq = new_start_seq;
	312	sack->end_seq = new_end_seq;
	313	sackoff += sizeof(*sack);
	314	}
	315	}
	316
	317	/* TCP SACK sequence number adjustment */
	318	static inline unsigned int
3db05fea	319	nf_nat_sack_adjust(struct sk_buff *skb,
c7232c99	320	unsigned int protoff,
5b1158e9 JK	321	struct tcphdr *tcph,
	322	struct nf_conn *ct,
	323	enum ip_conntrack_info ctinfo)
	324	{
	325	unsigned int dir, optoff, optend;
	326	struct nf_conn_nat *nat = nfct_nat(ct);
	327
c7232c99 PM	328	optoff = protoff + sizeof(struct tcphdr);
c7232c99 PM	329	optend = protoff + tcph->doff * 4;
5b1158e9	330
3db05fea	331	if (!skb_make_writable(skb, optend))
5b1158e9 JK	332	return 0;
	333
	334	dir = CTINFO2DIR(ctinfo);
	335
	336	while (optoff < optend) {
	337	/* Usually: option, length. */
3db05fea	338	unsigned char *op = skb->data + optoff;
5b1158e9 JK	339
	340	switch (op[0]) {
	341	case TCPOPT_EOL:
	342	return 1;
	343	case TCPOPT_NOP:
	344	optoff++;
	345	continue;
	346	default:
	347	/* no partial options */
	348	if (optoff + 1 == optend \|\|
	349	optoff + op[1] > optend \|\|
	350	op[1] < 2)
	351	return 0;
	352	if (op[0] == TCPOPT_SACK &&
	353	op[1] >= 2+TCPOLEN_SACK_PERBLOCK &&
	354	((op[1] - 2) % TCPOLEN_SACK_PERBLOCK) == 0)
3db05fea	355	sack_adjust(skb, tcph, optoff+2,
b6b84d4a	356	optoff+op[1], &nat->seq[!dir]);
5b1158e9 JK	357	optoff += op[1];
	358	}
	359	}
	360	return 1;
	361	}
	362
	363	/* TCP sequence number adjustment. Returns 1 on success, 0 on failure */
	364	int
3db05fea	365	nf_nat_seq_adjust(struct sk_buff *skb,
5b1158e9	366	struct nf_conn *ct,
051966c0 PM	367	enum ip_conntrack_info ctinfo,
051966c0 PM	368	unsigned int protoff)
5b1158e9 JK	369	{
	370	struct tcphdr *tcph;
	371	int dir;
	372	__be32 newseq, newack;
a3a9f79e	373	s16 seqoff, ackoff;
5b1158e9 JK	374	struct nf_conn_nat *nat = nfct_nat(ct);
	375	struct nf_nat_seq this_way, other_way;
	376
	377	dir = CTINFO2DIR(ctinfo);
	378
b6b84d4a YK	379	this_way = &nat->seq[dir];
b6b84d4a YK	380	other_way = &nat->seq[!dir];
5b1158e9	381
051966c0	382	if (!skb_make_writable(skb, protoff + sizeof(*tcph)))
5b1158e9 JK	383	return 0;
5b1158e9 JK	384
051966c0	385	tcph = (void *)skb->data + protoff;
5b1158e9	386	if (after(ntohl(tcph->seq), this_way->correction_pos))
a3a9f79e	387	seqoff = this_way->offset_after;
5b1158e9	388	else
a3a9f79e	389	seqoff = this_way->offset_before;
5b1158e9 JK	390
	391	if (after(ntohl(tcph->ack_seq) - other_way->offset_before,
	392	other_way->correction_pos))
a3a9f79e	393	ackoff = other_way->offset_after;
5b1158e9	394	else
a3a9f79e PM	395	ackoff = other_way->offset_before;
	396
	397	newseq = htonl(ntohl(tcph->seq) + seqoff);
	398	newack = htonl(ntohl(tcph->ack_seq) - ackoff);
5b1158e9	399
be0ea7d5 PM	400	inet_proto_csum_replace4(&tcph->check, skb, tcph->seq, newseq, 0);
be0ea7d5 PM	401	inet_proto_csum_replace4(&tcph->check, skb, tcph->ack_seq, newack, 0);
5b1158e9	402
0d53778e PM	403	pr_debug("Adjusting sequence number from %u->%u, ack from %u->%u\n",
	404	ntohl(tcph->seq), ntohl(newseq), ntohl(tcph->ack_seq),
	405	ntohl(newack));
5b1158e9 JK	406
	407	tcph->seq = newseq;
	408	tcph->ack_seq = newack;
	409
c7232c99	410	return nf_nat_sack_adjust(skb, protoff, tcph, ct, ctinfo);
5b1158e9	411	}
5b1158e9 JK	412
	413	/* Setup NAT on this expected conntrack so it follows master. */
	414	/* If we fail to get a free NAT slot, we'll get dropped on confirm */
	415	void nf_nat_follow_master(struct nf_conn *ct,
	416	struct nf_conntrack_expect *exp)
	417	{
c7232c99	418	struct nf_nat_range range;
5b1158e9 JK	419
	420	/* This must be a fresh one. */
	421	BUG_ON(ct->status & IPS_NAT_DONE_MASK);
	422
	423	/* Change src to where master sends to */
cbc9f2f4	424	range.flags = NF_NAT_RANGE_MAP_IPS;
c7232c99 PM	425	range.min_addr = range.max_addr
c7232c99 PM	426	= ct->master->tuplehash[!exp->dir].tuple.dst.u3;
cbc9f2f4	427	nf_nat_setup_info(ct, &range, NF_NAT_MANIP_SRC);
5b1158e9 JK	428
5b1158e9 JK	429	/* For DST manip, map port here to where it's expected. */
cbc9f2f4	430	range.flags = (NF_NAT_RANGE_MAP_IPS \| NF_NAT_RANGE_PROTO_SPECIFIED);
c7232c99 PM	431	range.min_proto = range.max_proto = exp->saved_proto;
	432	range.min_addr = range.max_addr
	433	= ct->master->tuplehash[!exp->dir].tuple.src.u3;
cbc9f2f4	434	nf_nat_setup_info(ct, &range, NF_NAT_MANIP_DST);
5b1158e9 JK	435	}
5b1158e9 JK	436	EXPORT_SYMBOL(nf_nat_follow_master);