nf_nat_core.c 21.3 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
/* NAT for netfilter; shared with compatibility layer. */

/* (C) 1999-2001 Paul `Rusty' Russell
 * (C) 2002-2006 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/module.h>
#include <linux/types.h>
#include <linux/timer.h>
#include <linux/skbuff.h>
15
#include <linux/gfp.h>
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
#include <net/checksum.h>
#include <net/icmp.h>
#include <net/ip.h>
#include <net/tcp.h>  /* For tcp_prot in getorigdst */
#include <linux/icmp.h>
#include <linux/udp.h>
#include <linux/jhash.h>

#include <linux/netfilter_ipv4.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/nf_nat.h>
#include <net/netfilter/nf_nat_protocol.h>
#include <net/netfilter/nf_nat_core.h>
#include <net/netfilter/nf_nat_helper.h>
#include <net/netfilter/nf_conntrack_helper.h>
#include <net/netfilter/nf_conntrack_l3proto.h>
33
#include <net/netfilter/nf_conntrack_zones.h>
34

35
static DEFINE_SPINLOCK(nf_nat_lock);
36

37
static struct nf_conntrack_l3proto *l3proto __read_mostly;
38
39

#define MAX_IP_NAT_PROTO 256
40
static const struct nf_nat_protocol __rcu *nf_nat_protos[MAX_IP_NAT_PROTO]
41
						__read_mostly;
42

43
static inline const struct nf_nat_protocol *
44
45
__nf_nat_proto_find(u_int8_t protonum)
{
46
	return rcu_dereference(nf_nat_protos[protonum]);
47
48
49
50
}

/* We keep an extra hash for each conntrack, for fast searching. */
static inline unsigned int
51
52
hash_by_src(const struct net *net, u16 zone,
	    const struct nf_conntrack_tuple *tuple)
53
{
54
55
	unsigned int hash;

56
	/* Original src, to ensure we map it consistently if poss. */
57
	hash = jhash_3words((__force u32)tuple->src.u3.ip,
58
			    (__force u32)tuple->src.u.all ^ zone,
59
			    tuple->dst.protonum, nf_conntrack_hash_rnd);
60
	return ((u64)hash * net->ipv4.nat_htable_size) >> 32;
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
}

/* Is this tuple already taken? (not by us) */
int
nf_nat_used_tuple(const struct nf_conntrack_tuple *tuple,
		  const struct nf_conn *ignored_conntrack)
{
	/* Conntrack tracking doesn't keep track of outgoing tuples; only
	   incoming ones.  NAT means they don't have a fixed mapping,
	   so we invert the tuple and look for the incoming reply.

	   We could keep a separate hash if this proves too slow. */
	struct nf_conntrack_tuple reply;

	nf_ct_invert_tuplepr(&reply, tuple);
	return nf_conntrack_tuple_taken(&reply, ignored_conntrack);
}
EXPORT_SYMBOL(nf_nat_used_tuple);

/* If we source map this tuple so reply looks like reply_tuple, will
 * that meet the constraints of range. */
static int
in_range(const struct nf_conntrack_tuple *tuple,
84
	 const struct nf_nat_ipv4_range *range)
85
{
86
	const struct nf_nat_protocol *proto;
87
	int ret = 0;
88
89
90

	/* If we are supposed to map IPs, then we must be in the
	   range specified, otherwise let this drag us onto a new src IP. */
91
	if (range->flags & NF_NAT_RANGE_MAP_IPS) {
92
93
94
95
96
		if (ntohl(tuple->src.u3.ip) < ntohl(range->min_ip) ||
		    ntohl(tuple->src.u3.ip) > ntohl(range->max_ip))
			return 0;
	}

97
98
	rcu_read_lock();
	proto = __nf_nat_proto_find(tuple->dst.protonum);
99
100
	if (!(range->flags & NF_NAT_RANGE_PROTO_SPECIFIED) ||
	    proto->in_range(tuple, NF_NAT_MANIP_SRC,
101
			    &range->min, &range->max))
102
103
		ret = 1;
	rcu_read_unlock();
104

105
	return ret;
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
}

static inline int
same_src(const struct nf_conn *ct,
	 const struct nf_conntrack_tuple *tuple)
{
	const struct nf_conntrack_tuple *t;

	t = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple;
	return (t->dst.protonum == tuple->dst.protonum &&
		t->src.u3.ip == tuple->src.u3.ip &&
		t->src.u.all == tuple->src.u.all);
}

/* Only called for SRC manip */
static int
122
find_appropriate_src(struct net *net, u16 zone,
123
		     const struct nf_conntrack_tuple *tuple,
124
		     struct nf_conntrack_tuple *result,
125
		     const struct nf_nat_ipv4_range *range)
126
{
127
	unsigned int h = hash_by_src(net, zone, tuple);
128
129
130
	const struct nf_conn_nat *nat;
	const struct nf_conn *ct;
	const struct hlist_node *n;
131

132
	rcu_read_lock();
133
	hlist_for_each_entry_rcu(nat, n, &net->ipv4.nat_bysource[h], bysource) {
134
		ct = nat->ct;
135
		if (same_src(ct, tuple) && nf_ct_zone(ct) == zone) {
136
137
138
139
140
141
			/* Copy source part from reply tuple. */
			nf_ct_invert_tuplepr(result,
				       &ct->tuplehash[IP_CT_DIR_REPLY].tuple);
			result->dst = tuple->dst;

			if (in_range(result, range)) {
142
				rcu_read_unlock();
143
144
145
146
				return 1;
			}
		}
	}
147
	rcu_read_unlock();
148
149
150
151
152
153
154
155
156
157
	return 0;
}

/* For [FUTURE] fragmentation handling, we want the least-used
   src-ip/dst-ip/proto triple.  Fairness doesn't come into it.  Thus
   if the range specifies 1.2.3.4 ports 10000-10005 and 1.2.3.5 ports
   1-65535, we don't do pro-rata allocation based on ports; we choose
   the ip with the lowest src-ip/dst-ip/proto usage.
*/
static void
158
find_best_ips_proto(u16 zone, struct nf_conntrack_tuple *tuple,
159
		    const struct nf_nat_ipv4_range *range,
160
161
162
163
164
165
166
167
		    const struct nf_conn *ct,
		    enum nf_nat_manip_type maniptype)
{
	__be32 *var_ipp;
	/* Host order */
	u_int32_t minip, maxip, j;

	/* No IP mapping?  Do nothing. */
168
	if (!(range->flags & NF_NAT_RANGE_MAP_IPS))
169
170
		return;

171
	if (maniptype == NF_NAT_MANIP_SRC)
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
		var_ipp = &tuple->src.u3.ip;
	else
		var_ipp = &tuple->dst.u3.ip;

	/* Fast path: only one choice. */
	if (range->min_ip == range->max_ip) {
		*var_ipp = range->min_ip;
		return;
	}

	/* Hashing source and destination IPs gives a fairly even
	 * spread in practice (if there are a small number of IPs
	 * involved, there usually aren't that many connections
	 * anyway).  The consistency means that servers see the same
	 * client coming from the same IP (some Internet Banking sites
	 * like this), even across reboots. */
	minip = ntohl(range->min_ip);
	maxip = ntohl(range->max_ip);
	j = jhash_2words((__force u32)tuple->src.u3.ip,
191
			 range->flags & NF_NAT_RANGE_PERSISTENT ?
192
				0 : (__force u32)tuple->dst.u3.ip ^ zone, 0);
193
194
	j = ((u64)j * (maxip - minip + 1)) >> 32;
	*var_ipp = htonl(minip + j);
195
196
}

197
198
199
/* Manipulate the tuple into the range given.  For NF_INET_POST_ROUTING,
 * we change the source to map into the range.  For NF_INET_PRE_ROUTING
 * and NF_INET_LOCAL_OUT, we change the destination to map into the
200
201
202
203
204
205
 * range.  It might not be possible to get a unique tuple, but we try.
 * At worst (or if we race), we will end up with a final duplicate in
 * __ip_conntrack_confirm and drop the packet. */
static void
get_unique_tuple(struct nf_conntrack_tuple *tuple,
		 const struct nf_conntrack_tuple *orig_tuple,
206
		 const struct nf_nat_ipv4_range *range,
207
208
209
		 struct nf_conn *ct,
		 enum nf_nat_manip_type maniptype)
{
210
	struct net *net = nf_ct_net(ct);
211
	const struct nf_nat_protocol *proto;
212
	u16 zone = nf_ct_zone(ct);
213
214
215
216
217
218
219
220

	/* 1) If this srcip/proto/src-proto-part is currently mapped,
	   and that same mapping gives a unique tuple within the given
	   range, use that.

	   This is only required for source (ie. NAT/masq) mappings.
	   So far, we don't do local source mappings, so multiple
	   manips not an issue.  */
221
222
	if (maniptype == NF_NAT_MANIP_SRC &&
	    !(range->flags & NF_NAT_RANGE_PROTO_RANDOM)) {
223
224
225
226
227
228
229
230
		/* try the original tuple first */
		if (in_range(orig_tuple, range)) {
			if (!nf_nat_used_tuple(orig_tuple, ct)) {
				*tuple = *orig_tuple;
				return;
			}
		} else if (find_appropriate_src(net, zone, orig_tuple, tuple,
			   range)) {
231
			pr_debug("get_unique_tuple: Found current src map\n");
232
233
			if (!nf_nat_used_tuple(tuple, ct))
				return;
234
235
236
237
238
239
		}
	}

	/* 2) Select the least-used IP/proto combination in the given
	   range. */
	*tuple = *orig_tuple;
240
	find_best_ips_proto(zone, tuple, range, ct, maniptype);
241
242
243
244

	/* 3) The per-protocol part of the manip is made to map into
	   the range to make a unique tuple. */

245
246
	rcu_read_lock();
	proto = __nf_nat_proto_find(orig_tuple->dst.protonum);
247
248

	/* Only bother mapping if it's not already in range and unique */
249
250
	if (!(range->flags & NF_NAT_RANGE_PROTO_RANDOM)) {
		if (range->flags & NF_NAT_RANGE_PROTO_SPECIFIED) {
251
252
253
254
255
256
257
258
259
			if (proto->in_range(tuple, maniptype, &range->min,
					    &range->max) &&
			    (range->min.all == range->max.all ||
			     !nf_nat_used_tuple(tuple, ct)))
				goto out;
		} else if (!nf_nat_used_tuple(tuple, ct)) {
			goto out;
		}
	}
260
261
262

	/* Last change: get protocol to try to obtain unique tuple. */
	proto->unique_tuple(tuple, range, maniptype, ct);
263
264
out:
	rcu_read_unlock();
265
266
267
268
}

unsigned int
nf_nat_setup_info(struct nf_conn *ct,
269
		  const struct nf_nat_ipv4_range *range,
270
		  enum nf_nat_manip_type maniptype)
271
{
272
	struct net *net = nf_ct_net(ct);
273
	struct nf_conntrack_tuple curr_tuple, new_tuple;
274
	struct nf_conn_nat *nat;
275

276
277
278
279
280
	/* nat helper or nfctnetlink also setup binding */
	nat = nfct_nat(ct);
	if (!nat) {
		nat = nf_ct_ext_add(ct, NF_CT_EXT_NAT, GFP_ATOMIC);
		if (nat == NULL) {
281
			pr_debug("failed to add NAT extension\n");
282
283
284
285
			return NF_ACCEPT;
		}
	}

286
287
	NF_CT_ASSERT(maniptype == NF_NAT_MANIP_SRC ||
		     maniptype == NF_NAT_MANIP_DST);
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
	BUG_ON(nf_nat_initialized(ct, maniptype));

	/* What we've got will look like inverse of reply. Normally
	   this is what is in the conntrack, except for prior
	   manipulations (future optimization: if num_manips == 0,
	   orig_tp =
	   conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple) */
	nf_ct_invert_tuplepr(&curr_tuple,
			     &ct->tuplehash[IP_CT_DIR_REPLY].tuple);

	get_unique_tuple(&new_tuple, &curr_tuple, range, ct, maniptype);

	if (!nf_ct_tuple_equal(&new_tuple, &curr_tuple)) {
		struct nf_conntrack_tuple reply;

		/* Alter conntrack table so will recognize replies. */
		nf_ct_invert_tuplepr(&reply, &new_tuple);
		nf_conntrack_alter_reply(ct, &reply);

		/* Non-atomic: we own this at the moment. */
308
		if (maniptype == NF_NAT_MANIP_SRC)
309
310
311
312
313
			ct->status |= IPS_SRC_NAT;
		else
			ct->status |= IPS_DST_NAT;
	}

314
	if (maniptype == NF_NAT_MANIP_SRC) {
315
316
		unsigned int srchash;

317
318
		srchash = hash_by_src(net, nf_ct_zone(ct),
				      &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple);
319
		spin_lock_bh(&nf_nat_lock);
320
		/* nf_conntrack_alter_reply might re-allocate extension area */
321
322
		nat = nfct_nat(ct);
		nat->ct = ct;
323
324
		hlist_add_head_rcu(&nat->bysource,
				   &net->ipv4.nat_bysource[srchash]);
325
		spin_unlock_bh(&nf_nat_lock);
326
327
328
	}

	/* It's done. */
329
	if (maniptype == NF_NAT_MANIP_DST)
330
		ct->status |= IPS_DST_NAT_DONE;
331
	else
332
		ct->status |= IPS_SRC_NAT_DONE;
333
334
335
336
337
338

	return NF_ACCEPT;
}
EXPORT_SYMBOL(nf_nat_setup_info);

/* Returns true if succeeded. */
339
static bool
340
manip_pkt(u_int16_t proto,
341
	  struct sk_buff *skb,
342
343
344
345
346
	  unsigned int iphdroff,
	  const struct nf_conntrack_tuple *target,
	  enum nf_nat_manip_type maniptype)
{
	struct iphdr *iph;
347
	const struct nf_nat_protocol *p;
348

349
	if (!skb_make_writable(skb, iphdroff + sizeof(*iph)))
350
		return false;
351

352
	iph = (void *)skb->data + iphdroff;
353
354

	/* Manipulate protcol part. */
355
356
357

	/* rcu_read_lock()ed by nf_hook_slow */
	p = __nf_nat_proto_find(proto);
358
	if (!p->manip_pkt(skb, iphdroff, target, maniptype))
359
		return false;
360

361
	iph = (void *)skb->data + iphdroff;
362

363
	if (maniptype == NF_NAT_MANIP_SRC) {
364
		csum_replace4(&iph->check, iph->saddr, target->src.u3.ip);
365
366
		iph->saddr = target->src.u3.ip;
	} else {
367
		csum_replace4(&iph->check, iph->daddr, target->dst.u3.ip);
368
369
		iph->daddr = target->dst.u3.ip;
	}
370
	return true;
371
372
373
374
375
376
}

/* Do packet manipulations according to nf_nat_setup_info. */
unsigned int nf_nat_packet(struct nf_conn *ct,
			   enum ip_conntrack_info ctinfo,
			   unsigned int hooknum,
377
			   struct sk_buff *skb)
378
379
380
381
382
{
	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	unsigned long statusbit;
	enum nf_nat_manip_type mtype = HOOK2MANIP(hooknum);

383
	if (mtype == NF_NAT_MANIP_SRC)
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
		statusbit = IPS_SRC_NAT;
	else
		statusbit = IPS_DST_NAT;

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

	/* Non-atomic: these bits don't change. */
	if (ct->status & statusbit) {
		struct nf_conntrack_tuple target;

		/* We are aiming to look like inverse of other direction. */
		nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);

399
		if (!manip_pkt(target.dst.protonum, skb, 0, &target, mtype))
400
401
402
403
404
405
406
407
408
409
			return NF_DROP;
	}
	return NF_ACCEPT;
}
EXPORT_SYMBOL_GPL(nf_nat_packet);

/* Dir is direction ICMP is coming from (opposite to packet it contains) */
int nf_nat_icmp_reply_translation(struct nf_conn *ct,
				  enum ip_conntrack_info ctinfo,
				  unsigned int hooknum,
410
				  struct sk_buff *skb)
411
412
413
414
415
{
	struct {
		struct icmphdr icmp;
		struct iphdr ip;
	} *inside;
416
	struct nf_conntrack_tuple target;
417
	int hdrlen = ip_hdrlen(skb);
418
419
420
421
	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
	unsigned long statusbit;
	enum nf_nat_manip_type manip = HOOK2MANIP(hooknum);

422
	if (!skb_make_writable(skb, hdrlen + sizeof(*inside)))
423
424
		return 0;

425
	inside = (void *)skb->data + hdrlen;
426
427
428

	/* We're actually going to mangle it beyond trivial checksum
	   adjustment, so make sure the current checksum is correct. */
429
	if (nf_ip_checksum(skb, hooknum, hdrlen, 0))
430
431
432
		return 0;

	/* Must be RELATED */
433
	NF_CT_ASSERT(skb->nfctinfo == IP_CT_RELATED ||
434
		     skb->nfctinfo == IP_CT_RELATED_REPLY);
435
436

	/* Redirects on non-null nats must be dropped, else they'll
437
438
	   start talking to each other without our translation, and be
	   confused... --RR */
439
440
441
442
443
444
445
446
447
	if (inside->icmp.type == ICMP_REDIRECT) {
		/* If NAT isn't finished, assume it and drop. */
		if ((ct->status & IPS_NAT_DONE_MASK) != IPS_NAT_DONE_MASK)
			return 0;

		if (ct->status & IPS_NAT_MASK)
			return 0;
	}

448
	if (manip == NF_NAT_MANIP_SRC)
449
450
451
452
453
454
455
456
457
458
459
		statusbit = IPS_SRC_NAT;
	else
		statusbit = IPS_DST_NAT;

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

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

460
	pr_debug("icmp_reply_translation: translating error %p manip %u "
461
		 "dir %s\n", skb, manip,
462
		 dir == IP_CT_DIR_ORIGINAL ? "ORIG" : "REPLY");
463
464
465
466
467
468

	/* Change inner back to look like incoming packet.  We do the
	   opposite manip on this hook to normal, because it might not
	   pass all hooks (locally-generated ICMP).  Consider incoming
	   packet: PREROUTING (DST manip), routing produces ICMP, goes
	   through POSTROUTING (which must correct the DST manip). */
469
470
	if (!manip_pkt(inside->ip.protocol, skb, hdrlen + sizeof(inside->icmp),
		       &ct->tuplehash[!dir].tuple, !manip))
471
472
		return 0;

473
	if (skb->ip_summed != CHECKSUM_PARTIAL) {
474
		/* Reloading "inside" here since manip_pkt inner. */
475
		inside = (void *)skb->data + hdrlen;
476
477
		inside->icmp.checksum = 0;
		inside->icmp.checksum =
478
479
			csum_fold(skb_checksum(skb, hdrlen,
					       skb->len - hdrlen, 0));
480
481
482
483
	}

	/* Change outer to look the reply to an incoming packet
	 * (proto 0 means don't invert per-proto part). */
484
485
486
	nf_ct_invert_tuplepr(&target, &ct->tuplehash[!dir].tuple);
	if (!manip_pkt(0, skb, 0, &target, manip))
		return 0;
487
488
489
490
491
492

	return 1;
}
EXPORT_SYMBOL_GPL(nf_nat_icmp_reply_translation);

/* Protocol registration. */
493
int nf_nat_protocol_register(const struct nf_nat_protocol *proto)
494
495
496
{
	int ret = 0;

497
	spin_lock_bh(&nf_nat_lock);
498
499
500
501
	if (rcu_dereference_protected(
			nf_nat_protos[proto->protonum],
			lockdep_is_held(&nf_nat_lock)
			) != &nf_nat_unknown_protocol) {
502
503
504
		ret = -EBUSY;
		goto out;
	}
505
	RCU_INIT_POINTER(nf_nat_protos[proto->protonum], proto);
506
 out:
507
	spin_unlock_bh(&nf_nat_lock);
508
509
510
511
	return ret;
}
EXPORT_SYMBOL(nf_nat_protocol_register);

Lucas De Marchi's avatar
Lucas De Marchi committed
512
/* No one stores the protocol anywhere; simply delete it. */
513
void nf_nat_protocol_unregister(const struct nf_nat_protocol *proto)
514
{
515
	spin_lock_bh(&nf_nat_lock);
516
	RCU_INIT_POINTER(nf_nat_protos[proto->protonum],
517
			   &nf_nat_unknown_protocol);
518
	spin_unlock_bh(&nf_nat_lock);
519
	synchronize_rcu();
520
521
522
}
EXPORT_SYMBOL(nf_nat_protocol_unregister);

Lucas De Marchi's avatar
Lucas De Marchi committed
523
/* No one using conntrack by the time this called. */
524
525
526
527
static void nf_nat_cleanup_conntrack(struct nf_conn *ct)
{
	struct nf_conn_nat *nat = nf_ct_ext_find(ct, NF_CT_EXT_NAT);

528
	if (nat == NULL || nat->ct == NULL)
529
530
		return;

531
	NF_CT_ASSERT(nat->ct->status & IPS_SRC_NAT_DONE);
532

533
	spin_lock_bh(&nf_nat_lock);
534
	hlist_del_rcu(&nat->bysource);
535
	spin_unlock_bh(&nf_nat_lock);
536
537
}

538
static void nf_nat_move_storage(void *new, void *old)
539
{
540
541
	struct nf_conn_nat *new_nat = new;
	struct nf_conn_nat *old_nat = old;
542
	struct nf_conn *ct = old_nat->ct;
543

544
	if (!ct || !(ct->status & IPS_SRC_NAT_DONE))
545
546
		return;

547
	spin_lock_bh(&nf_nat_lock);
548
	hlist_replace_rcu(&old_nat->bysource, &new_nat->bysource);
549
	spin_unlock_bh(&nf_nat_lock);
550
551
}

552
static struct nf_ct_ext_type nat_extend __read_mostly = {
553
554
555
556
557
558
	.len		= sizeof(struct nf_conn_nat),
	.align		= __alignof__(struct nf_conn_nat),
	.destroy	= nf_nat_cleanup_conntrack,
	.move		= nf_nat_move_storage,
	.id		= NF_CT_EXT_NAT,
	.flags		= NF_CT_EXT_F_PREALLOC,
559
560
};

561
562
563
564
565
566
567
568
569
570
571
572
#if defined(CONFIG_NF_CT_NETLINK) || defined(CONFIG_NF_CT_NETLINK_MODULE)

#include <linux/netfilter/nfnetlink.h>
#include <linux/netfilter/nfnetlink_conntrack.h>

static const struct nla_policy protonat_nla_policy[CTA_PROTONAT_MAX+1] = {
	[CTA_PROTONAT_PORT_MIN]	= { .type = NLA_U16 },
	[CTA_PROTONAT_PORT_MAX]	= { .type = NLA_U16 },
};

static int nfnetlink_parse_nat_proto(struct nlattr *attr,
				     const struct nf_conn *ct,
573
				     struct nf_nat_ipv4_range *range)
574
575
576
577
578
579
580
581
582
{
	struct nlattr *tb[CTA_PROTONAT_MAX+1];
	const struct nf_nat_protocol *npt;
	int err;

	err = nla_parse_nested(tb, CTA_PROTONAT_MAX, attr, protonat_nla_policy);
	if (err < 0)
		return err;

583
584
	rcu_read_lock();
	npt = __nf_nat_proto_find(nf_ct_protonum(ct));
585
586
	if (npt->nlattr_to_range)
		err = npt->nlattr_to_range(tb, range);
587
	rcu_read_unlock();
588
589
590
591
592
593
	return err;
}

static const struct nla_policy nat_nla_policy[CTA_NAT_MAX+1] = {
	[CTA_NAT_MINIP]		= { .type = NLA_U32 },
	[CTA_NAT_MAXIP]		= { .type = NLA_U32 },
594
	[CTA_NAT_PROTO]		= { .type = NLA_NESTED },
595
596
597
};

static int
598
nfnetlink_parse_nat(const struct nlattr *nat,
599
		    const struct nf_conn *ct, struct nf_nat_ipv4_range *range)
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
{
	struct nlattr *tb[CTA_NAT_MAX+1];
	int err;

	memset(range, 0, sizeof(*range));

	err = nla_parse_nested(tb, CTA_NAT_MAX, nat, nat_nla_policy);
	if (err < 0)
		return err;

	if (tb[CTA_NAT_MINIP])
		range->min_ip = nla_get_be32(tb[CTA_NAT_MINIP]);

	if (!tb[CTA_NAT_MAXIP])
		range->max_ip = range->min_ip;
	else
		range->max_ip = nla_get_be32(tb[CTA_NAT_MAXIP]);

	if (range->min_ip)
619
		range->flags |= NF_NAT_RANGE_MAP_IPS;
620
621
622
623
624
625
626
627
628
629
630
631
632
633

	if (!tb[CTA_NAT_PROTO])
		return 0;

	err = nfnetlink_parse_nat_proto(tb[CTA_NAT_PROTO], ct, range);
	if (err < 0)
		return err;

	return 0;
}

static int
nfnetlink_parse_nat_setup(struct nf_conn *ct,
			  enum nf_nat_manip_type manip,
634
			  const struct nlattr *attr)
635
{
636
	struct nf_nat_ipv4_range range;
637
638
639
640
641
642
643
644
645
646
647
648

	if (nfnetlink_parse_nat(attr, ct, &range) < 0)
		return -EINVAL;
	if (nf_nat_initialized(ct, manip))
		return -EEXIST;

	return nf_nat_setup_info(ct, &range, manip);
}
#else
static int
nfnetlink_parse_nat_setup(struct nf_conn *ct,
			  enum nf_nat_manip_type manip,
649
			  const struct nlattr *attr)
650
651
652
653
654
{
	return -EOPNOTSUPP;
}
#endif

655
656
static int __net_init nf_nat_net_init(struct net *net)
{
657
658
	/* Leave them the same for the moment. */
	net->ipv4.nat_htable_size = net->ct.htable_size;
659
	net->ipv4.nat_bysource = nf_ct_alloc_hashtable(&net->ipv4.nat_htable_size, 0);
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
	if (!net->ipv4.nat_bysource)
		return -ENOMEM;
	return 0;
}

/* Clear NAT section of all conntracks, in case we're loaded again. */
static int clean_nat(struct nf_conn *i, void *data)
{
	struct nf_conn_nat *nat = nfct_nat(i);

	if (!nat)
		return 0;
	memset(nat, 0, sizeof(*nat));
	i->status &= ~(IPS_NAT_MASK | IPS_NAT_DONE_MASK | IPS_SEQ_ADJUST);
	return 0;
}

static void __net_exit nf_nat_net_exit(struct net *net)
{
	nf_ct_iterate_cleanup(net, &clean_nat, NULL);
	synchronize_rcu();
681
	nf_ct_free_hashtable(net->ipv4.nat_bysource, net->ipv4.nat_htable_size);
682
683
684
685
686
687
688
}

static struct pernet_operations nf_nat_net_ops = {
	.init = nf_nat_net_init,
	.exit = nf_nat_net_exit,
};

689
690
691
692
693
static struct nf_ct_helper_expectfn follow_master_nat = {
	.name		= "nat-follow-master",
	.expectfn	= nf_nat_follow_master,
};

694
695
696
697
static struct nfq_ct_nat_hook nfq_ct_nat = {
	.seq_adjust	= nf_nat_tcp_seq_adjust,
};

698
699
700
static int __init nf_nat_init(void)
{
	size_t i;
701
702
	int ret;

703
704
	need_ipv4_conntrack();

705
706
707
708
709
	ret = nf_ct_extend_register(&nat_extend);
	if (ret < 0) {
		printk(KERN_ERR "nf_nat_core: Unable to register extension\n");
		return ret;
	}
710

711
712
	ret = register_pernet_subsys(&nf_nat_net_ops);
	if (ret < 0)
713
		goto cleanup_extend;
714
715

	/* Sew in builtin protocols. */
716
	spin_lock_bh(&nf_nat_lock);
717
	for (i = 0; i < MAX_IP_NAT_PROTO; i++)
718
719
720
721
		RCU_INIT_POINTER(nf_nat_protos[i], &nf_nat_unknown_protocol);
	RCU_INIT_POINTER(nf_nat_protos[IPPROTO_TCP], &nf_nat_protocol_tcp);
	RCU_INIT_POINTER(nf_nat_protos[IPPROTO_UDP], &nf_nat_protocol_udp);
	RCU_INIT_POINTER(nf_nat_protos[IPPROTO_ICMP], &nf_nat_protocol_icmp);
722
	spin_unlock_bh(&nf_nat_lock);
723
724

	/* Initialize fake conntrack so that NAT will skip it */
725
	nf_ct_untracked_status_or(IPS_NAT_DONE_MASK);
726
727

	l3proto = nf_ct_l3proto_find_get((u_int16_t)AF_INET);
728

729
730
	nf_ct_helper_expectfn_register(&follow_master_nat);

731
	BUG_ON(nf_nat_seq_adjust_hook != NULL);
732
	RCU_INIT_POINTER(nf_nat_seq_adjust_hook, nf_nat_seq_adjust);
733
	BUG_ON(nfnetlink_parse_nat_setup_hook != NULL);
734
	RCU_INIT_POINTER(nfnetlink_parse_nat_setup_hook,
735
			   nfnetlink_parse_nat_setup);
736
	BUG_ON(nf_ct_nat_offset != NULL);
737
	RCU_INIT_POINTER(nf_ct_nat_offset, nf_nat_get_offset);
738
	RCU_INIT_POINTER(nfq_ct_nat_hook, &nfq_ct_nat);
739
	return 0;
740
741
742
743

 cleanup_extend:
	nf_ct_extend_unregister(&nat_extend);
	return ret;
744
745
746
747
}

static void __exit nf_nat_cleanup(void)
{
748
	unregister_pernet_subsys(&nf_nat_net_ops);
749
	nf_ct_l3proto_put(l3proto);
750
	nf_ct_extend_unregister(&nat_extend);
751
	nf_ct_helper_expectfn_unregister(&follow_master_nat);
752
753
754
	RCU_INIT_POINTER(nf_nat_seq_adjust_hook, NULL);
	RCU_INIT_POINTER(nfnetlink_parse_nat_setup_hook, NULL);
	RCU_INIT_POINTER(nf_ct_nat_offset, NULL);
755
	RCU_INIT_POINTER(nfq_ct_nat_hook, NULL);
756
	synchronize_net();
757
758
759
}

MODULE_LICENSE("GPL");
760
MODULE_ALIAS("nf-nat-ipv4");
761
762
763

module_init(nf_nat_init);
module_exit(nf_nat_cleanup);