ip6_tables.c 58.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4
/*
 * Packet matching code.
 *
 * Copyright (C) 1999 Paul `Rusty' Russell & Michael J. Neuling
5
 * Copyright (C) 2000-2005 Netfilter Core Team <coreteam@netfilter.org>
L
Linus Torvalds 已提交
6 7 8 9 10
 *
 * 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.
 */
11
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
12
#include <linux/capability.h>
13
#include <linux/in.h>
L
Linus Torvalds 已提交
14 15 16 17 18
#include <linux/skbuff.h>
#include <linux/kmod.h>
#include <linux/vmalloc.h>
#include <linux/netdevice.h>
#include <linux/module.h>
R
Randy Dunlap 已提交
19
#include <linux/poison.h>
L
Linus Torvalds 已提交
20 21
#include <linux/icmpv6.h>
#include <net/ipv6.h>
22
#include <net/compat.h>
L
Linus Torvalds 已提交
23
#include <asm/uaccess.h>
I
Ingo Molnar 已提交
24
#include <linux/mutex.h>
L
Linus Torvalds 已提交
25
#include <linux/proc_fs.h>
26
#include <linux/err.h>
27
#include <linux/cpumask.h>
L
Linus Torvalds 已提交
28 29

#include <linux/netfilter_ipv6/ip6_tables.h>
30
#include <linux/netfilter/x_tables.h>
31
#include <net/netfilter/nf_log.h>
32
#include "../../netfilter/xt_repldata.h"
L
Linus Torvalds 已提交
33 34 35 36 37 38 39 40 41 42

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Netfilter Core Team <coreteam@netfilter.org>");
MODULE_DESCRIPTION("IPv6 packet filter");

/*#define DEBUG_IP_FIREWALL*/
/*#define DEBUG_ALLOW_ALL*/ /* Useful for remote debugging */
/*#define DEBUG_IP_FIREWALL_USER*/

#ifdef DEBUG_IP_FIREWALL
43
#define dprintf(format, args...) pr_info(format , ## args)
L
Linus Torvalds 已提交
44 45 46 47 48
#else
#define dprintf(format, args...)
#endif

#ifdef DEBUG_IP_FIREWALL_USER
49
#define duprintf(format, args...) pr_info(format , ## args)
L
Linus Torvalds 已提交
50 51 52 53 54 55 56 57 58
#else
#define duprintf(format, args...)
#endif

#ifdef CONFIG_NETFILTER_DEBUG
#define IP_NF_ASSERT(x)						\
do {								\
	if (!(x))						\
		printk("IP_NF_ASSERT: %s:%s:%u\n",		\
59
		       __func__, __FILE__, __LINE__);	\
L
Linus Torvalds 已提交
60 61 62 63 64 65 66 67 68 69 70
} while(0)
#else
#define IP_NF_ASSERT(x)
#endif

#if 0
/* All the better to debug you with... */
#define static
#define inline
#endif

71 72 73 74 75 76
void *ip6t_alloc_initial_table(const struct xt_table *info)
{
	return xt_alloc_initial_table(ip6t, IP6T);
}
EXPORT_SYMBOL_GPL(ip6t_alloc_initial_table);

77
/*
L
Linus Torvalds 已提交
78
   We keep a set of rules for each CPU, so we can avoid write-locking
79 80 81 82
   them in the softirq when updating the counters and therefore
   only need to read-lock in the softirq; doing a write_lock_bh() in user
   context stops packets coming through and allows user context to read
   the counters or update the rules.
L
Linus Torvalds 已提交
83 84 85 86

   Hence the start of any table is given by get_table() below.  */

/* Check for an extension */
87
int
L
Linus Torvalds 已提交
88 89
ip6t_ext_hdr(u8 nexthdr)
{
90 91 92 93 94 95 96
	return ( (nexthdr == IPPROTO_HOPOPTS)   ||
		 (nexthdr == IPPROTO_ROUTING)   ||
		 (nexthdr == IPPROTO_FRAGMENT)  ||
		 (nexthdr == IPPROTO_ESP)       ||
		 (nexthdr == IPPROTO_AH)        ||
		 (nexthdr == IPPROTO_NONE)      ||
		 (nexthdr == IPPROTO_DSTOPTS) );
L
Linus Torvalds 已提交
97 98 99
}

/* Returns whether matches rule or not. */
100
/* Performance critical - called for every packet */
101
static inline bool
L
Linus Torvalds 已提交
102 103 104 105 106
ip6_packet_match(const struct sk_buff *skb,
		 const char *indev,
		 const char *outdev,
		 const struct ip6t_ip6 *ip6info,
		 unsigned int *protoff,
107
		 int *fragoff, bool *hotdrop)
L
Linus Torvalds 已提交
108 109
{
	unsigned long ret;
110
	const struct ipv6hdr *ipv6 = ipv6_hdr(skb);
L
Linus Torvalds 已提交
111

112
#define FWINV(bool, invflg) ((bool) ^ !!(ip6info->invflags & (invflg)))
L
Linus Torvalds 已提交
113

114
	if (FWINV(ipv6_masked_addr_cmp(&ipv6->saddr, &ip6info->smsk,
115 116 117
				       &ip6info->src), IP6T_INV_SRCIP) ||
	    FWINV(ipv6_masked_addr_cmp(&ipv6->daddr, &ip6info->dmsk,
				       &ip6info->dst), IP6T_INV_DSTIP)) {
L
Linus Torvalds 已提交
118 119 120 121 122 123 124 125
		dprintf("Source or dest mismatch.\n");
/*
		dprintf("SRC: %u. Mask: %u. Target: %u.%s\n", ip->saddr,
			ipinfo->smsk.s_addr, ipinfo->src.s_addr,
			ipinfo->invflags & IP6T_INV_SRCIP ? " (INV)" : "");
		dprintf("DST: %u. Mask: %u. Target: %u.%s\n", ip->daddr,
			ipinfo->dmsk.s_addr, ipinfo->dst.s_addr,
			ipinfo->invflags & IP6T_INV_DSTIP ? " (INV)" : "");*/
126
		return false;
L
Linus Torvalds 已提交
127 128
	}

129
	ret = ifname_compare_aligned(indev, ip6info->iniface, ip6info->iniface_mask);
L
Linus Torvalds 已提交
130 131 132 133 134

	if (FWINV(ret != 0, IP6T_INV_VIA_IN)) {
		dprintf("VIA in mismatch (%s vs %s).%s\n",
			indev, ip6info->iniface,
			ip6info->invflags&IP6T_INV_VIA_IN ?" (INV)":"");
135
		return false;
L
Linus Torvalds 已提交
136 137
	}

138
	ret = ifname_compare_aligned(outdev, ip6info->outiface, ip6info->outiface_mask);
L
Linus Torvalds 已提交
139 140 141 142 143

	if (FWINV(ret != 0, IP6T_INV_VIA_OUT)) {
		dprintf("VIA out mismatch (%s vs %s).%s\n",
			outdev, ip6info->outiface,
			ip6info->invflags&IP6T_INV_VIA_OUT ?" (INV)":"");
144
		return false;
L
Linus Torvalds 已提交
145 146 147 148 149 150
	}

/* ... might want to do something with class and flowlabel here ... */

	/* look for the desired protocol header */
	if((ip6info->flags & IP6T_F_PROTO)) {
151 152
		int protohdr;
		unsigned short _frag_off;
L
Linus Torvalds 已提交
153

154
		protohdr = ipv6_find_hdr(skb, protoff, -1, &_frag_off);
155 156
		if (protohdr < 0) {
			if (_frag_off == 0)
157
				*hotdrop = true;
158
			return false;
159
		}
160
		*fragoff = _frag_off;
L
Linus Torvalds 已提交
161 162

		dprintf("Packet protocol %hi ?= %s%hi.\n",
163
				protohdr,
L
Linus Torvalds 已提交
164 165 166
				ip6info->invflags & IP6T_INV_PROTO ? "!":"",
				ip6info->proto);

167
		if (ip6info->proto == protohdr) {
L
Linus Torvalds 已提交
168
			if(ip6info->invflags & IP6T_INV_PROTO) {
169
				return false;
L
Linus Torvalds 已提交
170
			}
171
			return true;
L
Linus Torvalds 已提交
172 173 174 175 176
		}

		/* We need match for the '-p all', too! */
		if ((ip6info->proto != 0) &&
			!(ip6info->invflags & IP6T_INV_PROTO))
177
			return false;
L
Linus Torvalds 已提交
178
	}
179
	return true;
L
Linus Torvalds 已提交
180 181 182
}

/* should be ip6 safe */
183
static bool
L
Linus Torvalds 已提交
184 185 186 187 188
ip6_checkentry(const struct ip6t_ip6 *ipv6)
{
	if (ipv6->flags & ~IP6T_F_MASK) {
		duprintf("Unknown flag bits set: %08X\n",
			 ipv6->flags & ~IP6T_F_MASK);
189
		return false;
L
Linus Torvalds 已提交
190 191 192 193
	}
	if (ipv6->invflags & ~IP6T_INV_MASK) {
		duprintf("Unknown invflag bits set: %08X\n",
			 ipv6->invflags & ~IP6T_INV_MASK);
194
		return false;
L
Linus Torvalds 已提交
195
	}
196
	return true;
L
Linus Torvalds 已提交
197 198 199
}

static unsigned int
200
ip6t_error(struct sk_buff *skb, const struct xt_target_param *par)
L
Linus Torvalds 已提交
201 202
{
	if (net_ratelimit())
203
		pr_info("error: `%s'\n", (const char *)par->targinfo);
L
Linus Torvalds 已提交
204 205 206 207

	return NF_DROP;
}

208 209
/* Performance critical - called for every packet */
static inline bool
210
do_match(const struct ip6t_entry_match *m, const struct sk_buff *skb,
211
	 struct xt_match_param *par)
L
Linus Torvalds 已提交
212
{
213 214 215
	par->match     = m->u.kernel.match;
	par->matchinfo = m->data;

L
Linus Torvalds 已提交
216
	/* Stop iteration if it doesn't match */
217
	if (!m->u.kernel.match->match(skb, par))
218
		return true;
L
Linus Torvalds 已提交
219
	else
220
		return false;
L
Linus Torvalds 已提交
221 222 223
}

static inline struct ip6t_entry *
224
get_entry(const void *base, unsigned int offset)
L
Linus Torvalds 已提交
225 226 227 228
{
	return (struct ip6t_entry *)(base + offset);
}

229
/* All zeroes == unconditional rule. */
230
/* Mildly perf critical (only if packet tracing is on) */
231
static inline bool unconditional(const struct ip6t_ip6 *ipv6)
232
{
233
	static const struct ip6t_ip6 uncond;
234

235
	return memcmp(ipv6, &uncond, sizeof(uncond)) == 0;
236 237
}

238 239 240 241 242 243
static inline const struct ip6t_entry_target *
ip6t_get_target_c(const struct ip6t_entry *e)
{
	return ip6t_get_target((struct ip6t_entry *)e);
}

244 245 246
#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
/* This cries for unification! */
247
static const char *const hooknames[] = {
248 249 250 251 252
	[NF_INET_PRE_ROUTING]		= "PREROUTING",
	[NF_INET_LOCAL_IN]		= "INPUT",
	[NF_INET_FORWARD]		= "FORWARD",
	[NF_INET_LOCAL_OUT]		= "OUTPUT",
	[NF_INET_POST_ROUTING]		= "POSTROUTING",
253 254 255 256 257 258 259 260
};

enum nf_ip_trace_comments {
	NF_IP6_TRACE_COMMENT_RULE,
	NF_IP6_TRACE_COMMENT_RETURN,
	NF_IP6_TRACE_COMMENT_POLICY,
};

261
static const char *const comments[] = {
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276
	[NF_IP6_TRACE_COMMENT_RULE]	= "rule",
	[NF_IP6_TRACE_COMMENT_RETURN]	= "return",
	[NF_IP6_TRACE_COMMENT_POLICY]	= "policy",
};

static struct nf_loginfo trace_loginfo = {
	.type = NF_LOG_TYPE_LOG,
	.u = {
		.log = {
			.level = 4,
			.logflags = NF_LOG_MASK,
		},
	},
};

277
/* Mildly perf critical (only if packet tracing is on) */
278
static inline int
279
get_chainname_rulenum(const struct ip6t_entry *s, const struct ip6t_entry *e,
280 281
		      const char *hookname, const char **chainname,
		      const char **comment, unsigned int *rulenum)
282
{
283
	const struct ip6t_standard_target *t = (void *)ip6t_get_target_c(s);
284 285 286 287 288 289 290 291

	if (strcmp(t->target.u.kernel.target->name, IP6T_ERROR_TARGET) == 0) {
		/* Head of user chain: ERROR target with chainname */
		*chainname = t->target.data;
		(*rulenum) = 0;
	} else if (s == e) {
		(*rulenum)++;

292 293 294 295 296
		if (s->target_offset == sizeof(struct ip6t_entry) &&
		    strcmp(t->target.u.kernel.target->name,
			   IP6T_STANDARD_TARGET) == 0 &&
		    t->verdict < 0 &&
		    unconditional(&s->ipv6)) {
297 298
			/* Tail of chains: STANDARD target (return/policy) */
			*comment = *chainname == hookname
299 300
				? comments[NF_IP6_TRACE_COMMENT_POLICY]
				: comments[NF_IP6_TRACE_COMMENT_RETURN];
301 302 303 304 305 306 307 308
		}
		return 1;
	} else
		(*rulenum)++;

	return 0;
}

309
static void trace_packet(const struct sk_buff *skb,
310 311 312
			 unsigned int hook,
			 const struct net_device *in,
			 const struct net_device *out,
313
			 const char *tablename,
314 315
			 const struct xt_table_info *private,
			 const struct ip6t_entry *e)
316
{
317
	const void *table_base;
318
	const struct ip6t_entry *root;
319
	const char *hookname, *chainname, *comment;
320
	const struct ip6t_entry *iter;
321 322
	unsigned int rulenum = 0;

323
	table_base = private->entries[smp_processor_id()];
324 325
	root = get_entry(table_base, private->hook_entry[hook]);

326 327
	hookname = chainname = hooknames[hook];
	comment = comments[NF_IP6_TRACE_COMMENT_RULE];
328

329 330 331 332
	xt_entry_foreach(iter, root, private->size - private->hook_entry[hook])
		if (get_chainname_rulenum(iter, e, hookname,
		    &chainname, &comment, &rulenum) != 0)
			break;
333 334 335 336 337 338 339

	nf_log_packet(AF_INET6, hook, skb, in, out, &trace_loginfo,
		      "TRACE: %s:%s:%s:%u ",
		      tablename, chainname, comment, rulenum);
}
#endif

340 341 342 343 344 345
static inline __pure struct ip6t_entry *
ip6t_next_entry(const struct ip6t_entry *entry)
{
	return (void *)entry + entry->next_offset;
}

L
Linus Torvalds 已提交
346 347
/* Returns one of the generic firewall policies, like NF_ACCEPT. */
unsigned int
348
ip6t_do_table(struct sk_buff *skb,
L
Linus Torvalds 已提交
349 350 351
	      unsigned int hook,
	      const struct net_device *in,
	      const struct net_device *out,
352
	      struct xt_table *table)
L
Linus Torvalds 已提交
353
{
354 355
#define tb_comefrom ((struct ip6t_entry *)table_base)->comefrom

356
	static const char nulldevname[IFNAMSIZ] __attribute__((aligned(sizeof(long))));
357
	bool hotdrop = false;
L
Linus Torvalds 已提交
358 359 360
	/* Initializing verdict to NF_DROP keeps gcc happy. */
	unsigned int verdict = NF_DROP;
	const char *indev, *outdev;
361
	const void *table_base;
L
Linus Torvalds 已提交
362
	struct ip6t_entry *e, *back;
363
	const struct xt_table_info *private;
364
	struct xt_match_param mtpar;
365
	struct xt_target_param tgpar;
L
Linus Torvalds 已提交
366 367 368 369 370 371 372 373 374 375

	/* Initialization */
	indev = in ? in->name : nulldevname;
	outdev = out ? out->name : nulldevname;
	/* We handle fragments by dealing with the first fragment as
	 * if it was a normal packet.  All other fragments are treated
	 * normally, except that they will NEVER match rules that ask
	 * things we don't know, ie. tcp syn flag or ports).  If the
	 * rule is also a fragment-specific rule, non-fragments won't
	 * match it. */
376
	mtpar.hotdrop = &hotdrop;
377 378
	mtpar.in      = tgpar.in  = in;
	mtpar.out     = tgpar.out = out;
379
	mtpar.family  = tgpar.family = NFPROTO_IPV6;
380
	mtpar.hooknum = tgpar.hooknum = hook;
L
Linus Torvalds 已提交
381 382

	IP_NF_ASSERT(table->valid_hooks & (1 << hook));
383

384 385 386
	xt_info_rdlock_bh();
	private = table->private;
	table_base = private->entries[smp_processor_id()];
387

388
	e = get_entry(table_base, private->hook_entry[hook]);
L
Linus Torvalds 已提交
389 390

	/* For return from builtin chain */
391
	back = get_entry(table_base, private->underflow[hook]);
L
Linus Torvalds 已提交
392 393

	do {
394
		const struct ip6t_entry_target *t;
395
		const struct xt_entry_match *ematch;
396

L
Linus Torvalds 已提交
397 398
		IP_NF_ASSERT(e);
		IP_NF_ASSERT(back);
399
		if (!ip6_packet_match(skb, indev, outdev, &e->ipv6,
400 401
		    &mtpar.thoff, &mtpar.fragoff, &hotdrop)) {
 no_match:
402 403 404
			e = ip6t_next_entry(e);
			continue;
		}
L
Linus Torvalds 已提交
405

406 407 408 409
		xt_ematch_foreach(ematch, e)
			if (do_match(ematch, skb, &mtpar) != 0)
				goto no_match;

410 411 412
		ADD_COUNTER(e->counters,
			    ntohs(ipv6_hdr(skb)->payload_len) +
			    sizeof(struct ipv6hdr), 1);
L
Linus Torvalds 已提交
413

414
		t = ip6t_get_target_c(e);
415
		IP_NF_ASSERT(t->u.kernel.target);
416 417 418

#if defined(CONFIG_NETFILTER_XT_TARGET_TRACE) || \
    defined(CONFIG_NETFILTER_XT_TARGET_TRACE_MODULE)
419 420 421 422
		/* The packet is traced: log it */
		if (unlikely(skb->nf_trace))
			trace_packet(skb, hook, in, out,
				     table->name, private, e);
423
#endif
424 425 426 427 428 429 430 431 432 433
		/* Standard target? */
		if (!t->u.kernel.target->target) {
			int v;

			v = ((struct ip6t_standard_target *)t)->verdict;
			if (v < 0) {
				/* Pop from stack? */
				if (v != IP6T_RETURN) {
					verdict = (unsigned)(-v) - 1;
					break;
L
Linus Torvalds 已提交
434
				}
435 436 437 438
				e = back;
				back = get_entry(table_base, back->comefrom);
				continue;
			}
439 440
			if (table_base + v != ip6t_next_entry(e) &&
			    !(e->ipv6.flags & IP6T_F_GOTO)) {
441 442 443 444 445 446
				/* Save old back ptr in next entry */
				struct ip6t_entry *next = ip6t_next_entry(e);
				next->comefrom = (void *)back - table_base;
				/* set back pointer to next entry */
				back = next;
			}
L
Linus Torvalds 已提交
447

448
			e = get_entry(table_base, v);
449 450 451 452 453 454 455
			continue;
		}

		/* Targets which reenter must return
		   abs. verdicts */
		tgpar.target   = t->u.kernel.target;
		tgpar.targinfo = t->data;
456

L
Linus Torvalds 已提交
457
#ifdef CONFIG_NETFILTER_DEBUG
458
		tb_comefrom = 0xeeeeeeec;
L
Linus Torvalds 已提交
459
#endif
460
		verdict = t->u.kernel.target->target(skb, &tgpar);
L
Linus Torvalds 已提交
461 462

#ifdef CONFIG_NETFILTER_DEBUG
463
		if (tb_comefrom != 0xeeeeeeec && verdict == IP6T_CONTINUE) {
464 465 466
			printk("Target %s reentered!\n",
			       t->u.kernel.target->name);
			verdict = NF_DROP;
L
Linus Torvalds 已提交
467
		}
468
		tb_comefrom = 0x57acc001;
469 470 471 472 473 474
#endif
		if (verdict == IP6T_CONTINUE)
			e = ip6t_next_entry(e);
		else
			/* Verdict */
			break;
L
Linus Torvalds 已提交
475 476 477
	} while (!hotdrop);

#ifdef CONFIG_NETFILTER_DEBUG
478
	tb_comefrom = NETFILTER_LINK_POISON;
L
Linus Torvalds 已提交
479
#endif
480
	xt_info_rdunlock_bh();
L
Linus Torvalds 已提交
481 482 483 484 485 486 487 488

#ifdef DEBUG_ALLOW_ALL
	return NF_ACCEPT;
#else
	if (hotdrop)
		return NF_DROP;
	else return verdict;
#endif
489 490

#undef tb_comefrom
L
Linus Torvalds 已提交
491 492 493 494 495
}

/* Figures out from what hook each rule can be called: returns 0 if
   there are loops.  Puts hook bitmask in comefrom. */
static int
496
mark_source_chains(const struct xt_table_info *newinfo,
497
		   unsigned int valid_hooks, void *entry0)
L
Linus Torvalds 已提交
498 499 500 501 502
{
	unsigned int hook;

	/* No recursion; use packet counter to save back ptrs (reset
	   to 0 as we leave), and comefrom to save source hook bitmask */
503
	for (hook = 0; hook < NF_INET_NUMHOOKS; hook++) {
L
Linus Torvalds 已提交
504
		unsigned int pos = newinfo->hook_entry[hook];
505
		struct ip6t_entry *e = (struct ip6t_entry *)(entry0 + pos);
L
Linus Torvalds 已提交
506 507 508 509 510 511 512 513

		if (!(valid_hooks & (1 << hook)))
			continue;

		/* Set initial back pointer. */
		e->counters.pcnt = pos;

		for (;;) {
514 515
			const struct ip6t_standard_target *t
				= (void *)ip6t_get_target_c(e);
516
			int visited = e->comefrom & (1 << hook);
L
Linus Torvalds 已提交
517

518
			if (e->comefrom & (1 << NF_INET_NUMHOOKS)) {
L
Linus Torvalds 已提交
519 520 521 522
				printk("iptables: loop hook %u pos %u %08X.\n",
				       hook, pos, e->comefrom);
				return 0;
			}
523
			e->comefrom |= ((1 << hook) | (1 << NF_INET_NUMHOOKS));
L
Linus Torvalds 已提交
524 525

			/* Unconditional return/END. */
526 527 528 529 530
			if ((e->target_offset == sizeof(struct ip6t_entry) &&
			     (strcmp(t->target.u.user.name,
				     IP6T_STANDARD_TARGET) == 0) &&
			     t->verdict < 0 &&
			     unconditional(&e->ipv6)) || visited) {
L
Linus Torvalds 已提交
531 532
				unsigned int oldpos, size;

533 534 535
				if ((strcmp(t->target.u.user.name,
					    IP6T_STANDARD_TARGET) == 0) &&
				    t->verdict < -NF_MAX_VERDICT - 1) {
536 537 538 539 540 541
					duprintf("mark_source_chains: bad "
						"negative verdict (%i)\n",
								t->verdict);
					return 0;
				}

L
Linus Torvalds 已提交
542 543 544
				/* Return: backtrack through the last
				   big jump. */
				do {
545
					e->comefrom ^= (1<<NF_INET_NUMHOOKS);
L
Linus Torvalds 已提交
546 547
#ifdef DEBUG_IP_FIREWALL_USER
					if (e->comefrom
548
					    & (1 << NF_INET_NUMHOOKS)) {
L
Linus Torvalds 已提交
549 550 551 552 553 554 555 556 557 558 559 560 561 562 563
						duprintf("Back unset "
							 "on hook %u "
							 "rule %u\n",
							 hook, pos);
					}
#endif
					oldpos = pos;
					pos = e->counters.pcnt;
					e->counters.pcnt = 0;

					/* We're at the start. */
					if (pos == oldpos)
						goto next;

					e = (struct ip6t_entry *)
564
						(entry0 + pos);
L
Linus Torvalds 已提交
565 566 567 568 569
				} while (oldpos == pos + e->next_offset);

				/* Move along one */
				size = e->next_offset;
				e = (struct ip6t_entry *)
570
					(entry0 + pos + size);
L
Linus Torvalds 已提交
571 572 573 574 575 576
				e->counters.pcnt = pos;
				pos += size;
			} else {
				int newpos = t->verdict;

				if (strcmp(t->target.u.user.name,
577 578
					   IP6T_STANDARD_TARGET) == 0 &&
				    newpos >= 0) {
579 580 581 582 583 584 585
					if (newpos > newinfo->size -
						sizeof(struct ip6t_entry)) {
						duprintf("mark_source_chains: "
							"bad verdict (%i)\n",
								newpos);
						return 0;
					}
L
Linus Torvalds 已提交
586 587 588 589 590 591 592 593
					/* This a jump; chase it. */
					duprintf("Jump rule %u -> %u\n",
						 pos, newpos);
				} else {
					/* ... this is a fallthru */
					newpos = pos + e->next_offset;
				}
				e = (struct ip6t_entry *)
594
					(entry0 + newpos);
L
Linus Torvalds 已提交
595 596 597 598 599 600 601 602 603 604
				e->counters.pcnt = pos;
				pos = newpos;
			}
		}
		next:
		duprintf("Finished chain %u\n", hook);
	}
	return 1;
}

605
static void cleanup_match(struct ip6t_entry_match *m, struct net *net)
L
Linus Torvalds 已提交
606
{
607 608
	struct xt_mtdtor_param par;

609
	par.net       = net;
610 611
	par.match     = m->u.kernel.match;
	par.matchinfo = m->data;
612
	par.family    = NFPROTO_IPV6;
613 614 615
	if (par.match->destroy != NULL)
		par.match->destroy(&par);
	module_put(par.match->me);
L
Linus Torvalds 已提交
616 617
}

618
static int
619
check_entry(const struct ip6t_entry *e, const char *name)
620
{
621
	const struct ip6t_entry_target *t;
622 623 624 625 626 627 628 629 630 631

	if (!ip6_checkentry(&e->ipv6)) {
		duprintf("ip_tables: ip check failed %p %s.\n", e, name);
		return -EINVAL;
	}

	if (e->target_offset + sizeof(struct ip6t_entry_target) >
	    e->next_offset)
		return -EINVAL;

632
	t = ip6t_get_target_c(e);
633 634 635 636 637 638
	if (e->target_offset + t->u.target_size > e->next_offset)
		return -EINVAL;

	return 0;
}

639
static int check_match(struct ip6t_entry_match *m, struct xt_mtchk_param *par)
640
{
641
	const struct ip6t_ip6 *ipv6 = par->entryinfo;
642 643
	int ret;

644 645 646
	par->match     = m->u.kernel.match;
	par->matchinfo = m->data;

647
	ret = xt_check_match(par, m->u.match_size - sizeof(*m),
648
			     ipv6->proto, ipv6->invflags & IP6T_INV_PROTO);
649
	if (ret < 0) {
650
		duprintf("ip_tables: check failed for `%s'.\n",
651
			 par.match->name);
652
		return ret;
653
	}
654
	return 0;
655 656
}

657
static int
658
find_check_match(struct ip6t_entry_match *m, struct xt_mtchk_param *par)
L
Linus Torvalds 已提交
659
{
660
	struct xt_match *match;
661
	int ret;
L
Linus Torvalds 已提交
662

663 664 665
	match = xt_request_find_match(NFPROTO_IPV6, m->u.user.name,
				      m->u.user.revision);
	if (IS_ERR(match)) {
666
		duprintf("find_check_match: `%s' not found\n", m->u.user.name);
667
		return PTR_ERR(match);
L
Linus Torvalds 已提交
668 669 670
	}
	m->u.kernel.match = match;

671
	ret = check_match(m, par);
672 673 674
	if (ret)
		goto err;

L
Linus Torvalds 已提交
675
	return 0;
676 677 678
err:
	module_put(m->u.kernel.match->me);
	return ret;
L
Linus Torvalds 已提交
679 680
}

681
static int check_target(struct ip6t_entry *e, struct net *net, const char *name)
L
Linus Torvalds 已提交
682
{
683 684
	struct ip6t_entry_target *t = ip6t_get_target(e);
	struct xt_tgchk_param par = {
685
		.net       = net,
686 687 688 689 690
		.table     = name,
		.entryinfo = e,
		.target    = t->u.kernel.target,
		.targinfo  = t->data,
		.hook_mask = e->comefrom,
691
		.family    = NFPROTO_IPV6,
692
	};
L
Linus Torvalds 已提交
693 694
	int ret;

695
	t = ip6t_get_target(e);
696
	ret = xt_check_target(&par, t->u.target_size - sizeof(*t),
697
	      e->ipv6.proto, e->ipv6.invflags & IP6T_INV_PROTO);
698
	if (ret < 0) {
699 700
		duprintf("ip_tables: check failed for `%s'.\n",
			 t->u.kernel.target->name);
701
		return ret;
L
Linus Torvalds 已提交
702
	}
703
	return 0;
704
}
L
Linus Torvalds 已提交
705

706
static int
707
find_check_entry(struct ip6t_entry *e, struct net *net, const char *name,
708
		 unsigned int size)
709 710 711 712 713
{
	struct ip6t_entry_target *t;
	struct xt_target *target;
	int ret;
	unsigned int j;
714
	struct xt_mtchk_param mtpar;
715
	struct xt_entry_match *ematch;
716 717 718 719

	ret = check_entry(e, name);
	if (ret)
		return ret;
720

L
Linus Torvalds 已提交
721
	j = 0;
722
	mtpar.net	= net;
723 724 725
	mtpar.table     = name;
	mtpar.entryinfo = &e->ipv6;
	mtpar.hook_mask = e->comefrom;
726
	mtpar.family    = NFPROTO_IPV6;
727
	xt_ematch_foreach(ematch, e) {
728
		ret = find_check_match(ematch, &mtpar);
729
		if (ret != 0)
730 731
			goto cleanup_matches;
		++j;
732
	}
L
Linus Torvalds 已提交
733 734

	t = ip6t_get_target(e);
735 736 737
	target = xt_request_find_target(NFPROTO_IPV6, t->u.user.name,
					t->u.user.revision);
	if (IS_ERR(target)) {
738
		duprintf("find_check_entry: `%s' not found\n", t->u.user.name);
739
		ret = PTR_ERR(target);
L
Linus Torvalds 已提交
740 741 742
		goto cleanup_matches;
	}
	t->u.kernel.target = target;
743

744
	ret = check_target(e, net, name);
745 746
	if (ret)
		goto err;
L
Linus Torvalds 已提交
747
	return 0;
748 749
 err:
	module_put(t->u.kernel.target->me);
L
Linus Torvalds 已提交
750
 cleanup_matches:
751 752
	xt_ematch_foreach(ematch, e) {
		if (j-- == 0)
753
			break;
754 755
		cleanup_match(ematch, net);
	}
L
Linus Torvalds 已提交
756 757 758
	return ret;
}

759
static bool check_underflow(const struct ip6t_entry *e)
760 761 762 763 764 765
{
	const struct ip6t_entry_target *t;
	unsigned int verdict;

	if (!unconditional(&e->ipv6))
		return false;
766
	t = ip6t_get_target_c(e);
767 768 769 770 771 772 773
	if (strcmp(t->u.user.name, XT_STANDARD_TARGET) != 0)
		return false;
	verdict = ((struct ip6t_standard_target *)t)->verdict;
	verdict = -verdict - 1;
	return verdict == NF_DROP || verdict == NF_ACCEPT;
}

774
static int
L
Linus Torvalds 已提交
775
check_entry_size_and_hooks(struct ip6t_entry *e,
776
			   struct xt_table_info *newinfo,
777 778
			   const unsigned char *base,
			   const unsigned char *limit,
L
Linus Torvalds 已提交
779 780
			   const unsigned int *hook_entries,
			   const unsigned int *underflows,
781
			   unsigned int valid_hooks)
L
Linus Torvalds 已提交
782 783 784
{
	unsigned int h;

785 786
	if ((unsigned long)e % __alignof__(struct ip6t_entry) != 0 ||
	    (unsigned char *)e + sizeof(struct ip6t_entry) >= limit) {
L
Linus Torvalds 已提交
787 788 789 790 791 792 793 794 795 796 797 798
		duprintf("Bad offset %p\n", e);
		return -EINVAL;
	}

	if (e->next_offset
	    < sizeof(struct ip6t_entry) + sizeof(struct ip6t_entry_target)) {
		duprintf("checking: element %p size %u\n",
			 e, e->next_offset);
		return -EINVAL;
	}

	/* Check hooks & underflows */
799
	for (h = 0; h < NF_INET_NUMHOOKS; h++) {
800 801
		if (!(valid_hooks & (1 << h)))
			continue;
L
Linus Torvalds 已提交
802 803
		if ((unsigned char *)e - base == hook_entries[h])
			newinfo->hook_entry[h] = hook_entries[h];
804
		if ((unsigned char *)e - base == underflows[h]) {
805 806 807 808
			if (!check_underflow(e)) {
				pr_err("Underflows must be unconditional and "
				       "use the STANDARD target with "
				       "ACCEPT/DROP\n");
809 810
				return -EINVAL;
			}
L
Linus Torvalds 已提交
811
			newinfo->underflow[h] = underflows[h];
812
		}
L
Linus Torvalds 已提交
813 814 815
	}

	/* Clear counters and comefrom */
816
	e->counters = ((struct xt_counters) { 0, 0 });
L
Linus Torvalds 已提交
817 818 819 820
	e->comefrom = 0;
	return 0;
}

821
static void cleanup_entry(struct ip6t_entry *e, struct net *net)
L
Linus Torvalds 已提交
822
{
823
	struct xt_tgdtor_param par;
L
Linus Torvalds 已提交
824
	struct ip6t_entry_target *t;
825
	struct xt_entry_match *ematch;
L
Linus Torvalds 已提交
826 827

	/* Cleanup all matches */
828
	xt_ematch_foreach(ematch, e)
829
		cleanup_match(ematch, net);
L
Linus Torvalds 已提交
830
	t = ip6t_get_target(e);
831

832
	par.net      = net;
833 834
	par.target   = t->u.kernel.target;
	par.targinfo = t->data;
835
	par.family   = NFPROTO_IPV6;
836 837 838
	if (par.target->destroy != NULL)
		par.target->destroy(&par);
	module_put(par.target->me);
L
Linus Torvalds 已提交
839 840 841 842 843
}

/* Checks and translates the user-supplied table segment (held in
   newinfo) */
static int
844 845
translate_table(struct net *net, struct xt_table_info *newinfo, void *entry0,
                const struct ip6t_replace *repl)
L
Linus Torvalds 已提交
846
{
847
	struct ip6t_entry *iter;
L
Linus Torvalds 已提交
848
	unsigned int i;
849
	int ret = 0;
L
Linus Torvalds 已提交
850

851 852
	newinfo->size = repl->size;
	newinfo->number = repl->num_entries;
L
Linus Torvalds 已提交
853 854

	/* Init all hooks to impossible value. */
855
	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
L
Linus Torvalds 已提交
856 857 858 859 860 861 862
		newinfo->hook_entry[i] = 0xFFFFFFFF;
		newinfo->underflow[i] = 0xFFFFFFFF;
	}

	duprintf("translate_table: size %u\n", newinfo->size);
	i = 0;
	/* Walk through entries, checking offsets. */
863 864
	xt_entry_foreach(iter, entry0, newinfo->size) {
		ret = check_entry_size_and_hooks(iter, newinfo, entry0,
865 866 867 868
						 entry0 + repl->size,
						 repl->hook_entry,
						 repl->underflow,
						 repl->valid_hooks);
869
		if (ret != 0)
870 871
			return ret;
		++i;
872
	}
L
Linus Torvalds 已提交
873

874
	if (i != repl->num_entries) {
L
Linus Torvalds 已提交
875
		duprintf("translate_table: %u not %u entries\n",
876
			 i, repl->num_entries);
L
Linus Torvalds 已提交
877 878 879 880
		return -EINVAL;
	}

	/* Check hooks all assigned */
881
	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
L
Linus Torvalds 已提交
882
		/* Only hooks which are valid */
883
		if (!(repl->valid_hooks & (1 << i)))
L
Linus Torvalds 已提交
884 885 886
			continue;
		if (newinfo->hook_entry[i] == 0xFFFFFFFF) {
			duprintf("Invalid hook entry %u %u\n",
887
				 i, repl->hook_entry[i]);
L
Linus Torvalds 已提交
888 889 890 891
			return -EINVAL;
		}
		if (newinfo->underflow[i] == 0xFFFFFFFF) {
			duprintf("Invalid underflow %u %u\n",
892
				 i, repl->underflow[i]);
L
Linus Torvalds 已提交
893 894 895 896
			return -EINVAL;
		}
	}

897
	if (!mark_source_chains(newinfo, repl->valid_hooks, entry0))
898 899
		return -ELOOP;

L
Linus Torvalds 已提交
900 901
	/* Finally, each sanity check must pass */
	i = 0;
902
	xt_entry_foreach(iter, entry0, newinfo->size) {
903
		ret = find_check_entry(iter, net, repl->name, repl->size);
904 905
		if (ret != 0)
			break;
906
		++i;
907
	}
L
Linus Torvalds 已提交
908

909
	if (ret != 0) {
910 911
		xt_entry_foreach(iter, entry0, newinfo->size) {
			if (i-- == 0)
912
				break;
913 914
			cleanup_entry(iter, net);
		}
915 916
		return ret;
	}
L
Linus Torvalds 已提交
917 918

	/* And one copy for every other CPU */
919
	for_each_possible_cpu(i) {
920 921
		if (newinfo->entries[i] && newinfo->entries[i] != entry0)
			memcpy(newinfo->entries[i], entry0, newinfo->size);
L
Linus Torvalds 已提交
922 923
	}

924
	return ret;
L
Linus Torvalds 已提交
925 926 927
}

static void
928 929
get_counters(const struct xt_table_info *t,
	     struct xt_counters counters[])
L
Linus Torvalds 已提交
930
{
931
	struct ip6t_entry *iter;
L
Linus Torvalds 已提交
932 933
	unsigned int cpu;
	unsigned int i;
934 935 936 937 938
	unsigned int curcpu;

	/* Instead of clearing (by a previous call to memset())
	 * the counters and using adds, we set the counters
	 * with data used by 'current' CPU
939 940 941
	 *
	 * Bottom half has to be disabled to prevent deadlock
	 * if new softirq were to run and call ipt_do_table
942
	 */
943 944
	local_bh_disable();
	curcpu = smp_processor_id();
945 946

	i = 0;
947 948
	xt_entry_foreach(iter, t->entries[curcpu], t->size) {
		SET_COUNTER(counters[i], iter->counters.bcnt,
949
			    iter->counters.pcnt);
950 951
		++i;
	}
L
Linus Torvalds 已提交
952

953
	for_each_possible_cpu(cpu) {
954 955
		if (cpu == curcpu)
			continue;
L
Linus Torvalds 已提交
956
		i = 0;
957
		xt_info_wrlock(cpu);
958 959
		xt_entry_foreach(iter, t->entries[cpu], t->size) {
			ADD_COUNTER(counters[i], iter->counters.bcnt,
960
				    iter->counters.pcnt);
961 962
			++i;
		}
963
		xt_info_wrunlock(cpu);
L
Linus Torvalds 已提交
964
	}
965 966 967
	local_bh_enable();
}

968
static struct xt_counters *alloc_counters(const struct xt_table *table)
L
Linus Torvalds 已提交
969
{
970
	unsigned int countersize;
971
	struct xt_counters *counters;
972
	const struct xt_table_info *private = table->private;
L
Linus Torvalds 已提交
973 974 975 976

	/* We need atomic snapshot of counters: rest doesn't change
	   (other than comefrom, which userspace doesn't care
	   about). */
977
	countersize = sizeof(struct xt_counters) * private->number;
978
	counters = vmalloc_node(countersize, numa_node_id());
L
Linus Torvalds 已提交
979 980

	if (counters == NULL)
981
		return ERR_PTR(-ENOMEM);
982

983
	get_counters(private, counters);
984

985
	return counters;
986 987 988 989
}

static int
copy_entries_to_user(unsigned int total_size,
990
		     const struct xt_table *table,
991 992 993
		     void __user *userptr)
{
	unsigned int off, num;
994
	const struct ip6t_entry *e;
995
	struct xt_counters *counters;
996
	const struct xt_table_info *private = table->private;
997
	int ret = 0;
998
	const void *loc_cpu_entry;
999 1000 1001 1002 1003

	counters = alloc_counters(table);
	if (IS_ERR(counters))
		return PTR_ERR(counters);

1004 1005 1006 1007
	/* choose the copy that is on our node/cpu, ...
	 * This choice is lazy (because current thread is
	 * allowed to migrate to another cpu)
	 */
1008
	loc_cpu_entry = private->entries[raw_smp_processor_id()];
1009
	if (copy_to_user(userptr, loc_cpu_entry, total_size) != 0) {
L
Linus Torvalds 已提交
1010 1011 1012 1013 1014 1015 1016 1017
		ret = -EFAULT;
		goto free_counters;
	}

	/* FIXME: use iterator macros --RR */
	/* ... then go back and fix counters and names */
	for (off = 0, num = 0; off < total_size; off += e->next_offset, num++){
		unsigned int i;
1018 1019
		const struct ip6t_entry_match *m;
		const struct ip6t_entry_target *t;
L
Linus Torvalds 已提交
1020

1021
		e = (struct ip6t_entry *)(loc_cpu_entry + off);
L
Linus Torvalds 已提交
1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045
		if (copy_to_user(userptr + off
				 + offsetof(struct ip6t_entry, counters),
				 &counters[num],
				 sizeof(counters[num])) != 0) {
			ret = -EFAULT;
			goto free_counters;
		}

		for (i = sizeof(struct ip6t_entry);
		     i < e->target_offset;
		     i += m->u.match_size) {
			m = (void *)e + i;

			if (copy_to_user(userptr + off + i
					 + offsetof(struct ip6t_entry_match,
						    u.user.name),
					 m->u.kernel.match->name,
					 strlen(m->u.kernel.match->name)+1)
			    != 0) {
				ret = -EFAULT;
				goto free_counters;
			}
		}

1046
		t = ip6t_get_target_c(e);
L
Linus Torvalds 已提交
1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
		if (copy_to_user(userptr + off + e->target_offset
				 + offsetof(struct ip6t_entry_target,
					    u.user.name),
				 t->u.kernel.target->name,
				 strlen(t->u.kernel.target->name)+1) != 0) {
			ret = -EFAULT;
			goto free_counters;
		}
	}

 free_counters:
	vfree(counters);
	return ret;
}

1062
#ifdef CONFIG_COMPAT
1063
static void compat_standard_from_user(void *dst, const void *src)
1064 1065 1066 1067 1068 1069 1070 1071
{
	int v = *(compat_int_t *)src;

	if (v > 0)
		v += xt_compat_calc_jump(AF_INET6, v);
	memcpy(dst, &v, sizeof(v));
}

1072
static int compat_standard_to_user(void __user *dst, const void *src)
1073 1074 1075 1076 1077 1078 1079 1080
{
	compat_int_t cv = *(int *)src;

	if (cv > 0)
		cv -= xt_compat_calc_jump(AF_INET6, cv);
	return copy_to_user(dst, &cv, sizeof(cv)) ? -EFAULT : 0;
}

1081
static int compat_calc_entry(const struct ip6t_entry *e,
1082
			     const struct xt_table_info *info,
1083
			     const void *base, struct xt_table_info *newinfo)
1084
{
1085
	const struct xt_entry_match *ematch;
1086
	const struct ip6t_entry_target *t;
1087 1088 1089 1090 1091
	unsigned int entry_offset;
	int off, i, ret;

	off = sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);
	entry_offset = (void *)e - base;
1092
	xt_ematch_foreach(ematch, e)
1093
		off += xt_compat_match_offset(ematch->u.kernel.match);
1094
	t = ip6t_get_target_c(e);
1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114
	off += xt_compat_target_offset(t->u.kernel.target);
	newinfo->size -= off;
	ret = xt_compat_add_offset(AF_INET6, entry_offset, off);
	if (ret)
		return ret;

	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
		if (info->hook_entry[i] &&
		    (e < (struct ip6t_entry *)(base + info->hook_entry[i])))
			newinfo->hook_entry[i] -= off;
		if (info->underflow[i] &&
		    (e < (struct ip6t_entry *)(base + info->underflow[i])))
			newinfo->underflow[i] -= off;
	}
	return 0;
}

static int compat_table_info(const struct xt_table_info *info,
			     struct xt_table_info *newinfo)
{
1115
	struct ip6t_entry *iter;
1116
	void *loc_cpu_entry;
1117
	int ret;
1118 1119 1120 1121 1122 1123 1124 1125

	if (!newinfo || !info)
		return -EINVAL;

	/* we dont care about newinfo->entries[] */
	memcpy(newinfo, info, offsetof(struct xt_table_info, entries));
	newinfo->initial_entries = 0;
	loc_cpu_entry = info->entries[raw_smp_processor_id()];
1126 1127 1128
	xt_entry_foreach(iter, loc_cpu_entry, info->size) {
		ret = compat_calc_entry(iter, info, loc_cpu_entry, newinfo);
		if (ret != 0)
1129
			return ret;
1130
	}
1131
	return 0;
1132 1133 1134
}
#endif

1135 1136
static int get_info(struct net *net, void __user *user,
                    const int *len, int compat)
1137 1138 1139 1140 1141 1142
{
	char name[IP6T_TABLE_MAXNAMELEN];
	struct xt_table *t;
	int ret;

	if (*len != sizeof(struct ip6t_getinfo)) {
1143
		duprintf("length %u != %zu\n", *len,
1144 1145 1146 1147 1148 1149 1150 1151
			 sizeof(struct ip6t_getinfo));
		return -EINVAL;
	}

	if (copy_from_user(name, user, sizeof(name)) != 0)
		return -EFAULT;

	name[IP6T_TABLE_MAXNAMELEN-1] = '\0';
1152 1153 1154 1155
#ifdef CONFIG_COMPAT
	if (compat)
		xt_compat_lock(AF_INET6);
#endif
1156
	t = try_then_request_module(xt_find_table_lock(net, AF_INET6, name),
1157 1158 1159
				    "ip6table_%s", name);
	if (t && !IS_ERR(t)) {
		struct ip6t_getinfo info;
1160
		const struct xt_table_info *private = t->private;
1161
#ifdef CONFIG_COMPAT
1162 1163
		struct xt_table_info tmp;

1164 1165 1166 1167 1168 1169
		if (compat) {
			ret = compat_table_info(private, &tmp);
			xt_compat_flush_offsets(AF_INET6);
			private = &tmp;
		}
#endif
1170 1171 1172 1173 1174 1175 1176
		info.valid_hooks = t->valid_hooks;
		memcpy(info.hook_entry, private->hook_entry,
		       sizeof(info.hook_entry));
		memcpy(info.underflow, private->underflow,
		       sizeof(info.underflow));
		info.num_entries = private->number;
		info.size = private->size;
1177
		strcpy(info.name, name);
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187

		if (copy_to_user(user, &info, *len) != 0)
			ret = -EFAULT;
		else
			ret = 0;

		xt_table_unlock(t);
		module_put(t->me);
	} else
		ret = t ? PTR_ERR(t) : -ENOENT;
1188 1189 1190 1191
#ifdef CONFIG_COMPAT
	if (compat)
		xt_compat_unlock(AF_INET6);
#endif
1192 1193 1194
	return ret;
}

L
Linus Torvalds 已提交
1195
static int
1196 1197
get_entries(struct net *net, struct ip6t_get_entries __user *uptr,
            const int *len)
L
Linus Torvalds 已提交
1198 1199
{
	int ret;
1200
	struct ip6t_get_entries get;
1201
	struct xt_table *t;
L
Linus Torvalds 已提交
1202

1203
	if (*len < sizeof(get)) {
1204
		duprintf("get_entries: %u < %zu\n", *len, sizeof(get));
1205 1206 1207 1208 1209
		return -EINVAL;
	}
	if (copy_from_user(&get, uptr, sizeof(get)) != 0)
		return -EFAULT;
	if (*len != sizeof(struct ip6t_get_entries) + get.size) {
1210 1211
		duprintf("get_entries: %u != %zu\n",
			 *len, sizeof(get) + get.size);
1212 1213 1214
		return -EINVAL;
	}

1215
	t = xt_find_table_lock(net, AF_INET6, get.name);
1216
	if (t && !IS_ERR(t)) {
1217 1218
		struct xt_table_info *private = t->private;
		duprintf("t->private->number = %u\n", private->number);
1219
		if (get.size == private->size)
1220
			ret = copy_entries_to_user(private->size,
L
Linus Torvalds 已提交
1221 1222 1223
						   t, uptr->entrytable);
		else {
			duprintf("get_entries: I've got %u not %u!\n",
1224
				 private->size, get.size);
1225
			ret = -EAGAIN;
L
Linus Torvalds 已提交
1226
		}
1227
		module_put(t->me);
1228
		xt_table_unlock(t);
L
Linus Torvalds 已提交
1229
	} else
1230
		ret = t ? PTR_ERR(t) : -ENOENT;
L
Linus Torvalds 已提交
1231 1232 1233 1234 1235

	return ret;
}

static int
1236
__do_replace(struct net *net, const char *name, unsigned int valid_hooks,
1237 1238
	     struct xt_table_info *newinfo, unsigned int num_counters,
	     void __user *counters_ptr)
L
Linus Torvalds 已提交
1239 1240
{
	int ret;
1241
	struct xt_table *t;
1242
	struct xt_table_info *oldinfo;
1243
	struct xt_counters *counters;
1244
	const void *loc_cpu_old_entry;
1245
	struct ip6t_entry *iter;
L
Linus Torvalds 已提交
1246

1247 1248
	ret = 0;
	counters = vmalloc_node(num_counters * sizeof(struct xt_counters),
1249
				numa_node_id());
L
Linus Torvalds 已提交
1250 1251
	if (!counters) {
		ret = -ENOMEM;
1252
		goto out;
L
Linus Torvalds 已提交
1253 1254
	}

1255
	t = try_then_request_module(xt_find_table_lock(net, AF_INET6, name),
1256
				    "ip6table_%s", name);
1257 1258
	if (!t || IS_ERR(t)) {
		ret = t ? PTR_ERR(t) : -ENOENT;
L
Linus Torvalds 已提交
1259
		goto free_newinfo_counters_untrans;
1260
	}
L
Linus Torvalds 已提交
1261 1262

	/* You lied! */
1263
	if (valid_hooks != t->valid_hooks) {
L
Linus Torvalds 已提交
1264
		duprintf("Valid hook crap: %08X vs %08X\n",
1265
			 valid_hooks, t->valid_hooks);
L
Linus Torvalds 已提交
1266
		ret = -EINVAL;
1267
		goto put_module;
L
Linus Torvalds 已提交
1268 1269
	}

1270
	oldinfo = xt_replace_table(t, num_counters, newinfo, &ret);
L
Linus Torvalds 已提交
1271 1272 1273 1274 1275 1276
	if (!oldinfo)
		goto put_module;

	/* Update module usage count based on number of rules */
	duprintf("do_replace: oldnum=%u, initnum=%u, newnum=%u\n",
		oldinfo->number, oldinfo->initial_entries, newinfo->number);
1277 1278
	if ((oldinfo->number > oldinfo->initial_entries) ||
	    (newinfo->number <= oldinfo->initial_entries))
L
Linus Torvalds 已提交
1279 1280 1281 1282 1283
		module_put(t->me);
	if ((oldinfo->number > oldinfo->initial_entries) &&
	    (newinfo->number <= oldinfo->initial_entries))
		module_put(t->me);

1284
	/* Get the old counters, and synchronize with replace */
L
Linus Torvalds 已提交
1285
	get_counters(oldinfo, counters);
1286

L
Linus Torvalds 已提交
1287
	/* Decrease module usage counts and free resource */
1288
	loc_cpu_old_entry = oldinfo->entries[raw_smp_processor_id()];
1289
	xt_entry_foreach(iter, loc_cpu_old_entry, oldinfo->size)
1290
		cleanup_entry(iter, net);
1291

1292
	xt_free_table_info(oldinfo);
1293 1294
	if (copy_to_user(counters_ptr, counters,
			 sizeof(struct xt_counters) * num_counters) != 0)
L
Linus Torvalds 已提交
1295 1296
		ret = -EFAULT;
	vfree(counters);
1297
	xt_table_unlock(t);
L
Linus Torvalds 已提交
1298 1299 1300 1301
	return ret;

 put_module:
	module_put(t->me);
1302
	xt_table_unlock(t);
L
Linus Torvalds 已提交
1303 1304
 free_newinfo_counters_untrans:
	vfree(counters);
1305 1306 1307 1308 1309
 out:
	return ret;
}

static int
1310
do_replace(struct net *net, const void __user *user, unsigned int len)
1311 1312 1313 1314 1315
{
	int ret;
	struct ip6t_replace tmp;
	struct xt_table_info *newinfo;
	void *loc_cpu_entry;
1316
	struct ip6t_entry *iter;
1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336

	if (copy_from_user(&tmp, user, sizeof(tmp)) != 0)
		return -EFAULT;

	/* overflow check */
	if (tmp.num_counters >= INT_MAX / sizeof(struct xt_counters))
		return -ENOMEM;

	newinfo = xt_alloc_table_info(tmp.size);
	if (!newinfo)
		return -ENOMEM;

	/* choose the copy that is on our node/cpu */
	loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
	if (copy_from_user(loc_cpu_entry, user + sizeof(tmp),
			   tmp.size) != 0) {
		ret = -EFAULT;
		goto free_newinfo;
	}

1337
	ret = translate_table(net, newinfo, loc_cpu_entry, &tmp);
1338 1339 1340 1341 1342
	if (ret != 0)
		goto free_newinfo;

	duprintf("ip_tables: Translated table\n");

1343
	ret = __do_replace(net, tmp.name, tmp.valid_hooks, newinfo,
1344 1345 1346 1347 1348 1349
			   tmp.num_counters, tmp.counters);
	if (ret)
		goto free_newinfo_untrans;
	return 0;

 free_newinfo_untrans:
1350
	xt_entry_foreach(iter, loc_cpu_entry, newinfo->size)
1351
		cleanup_entry(iter, net);
L
Linus Torvalds 已提交
1352
 free_newinfo:
1353
	xt_free_table_info(newinfo);
L
Linus Torvalds 已提交
1354 1355 1356 1357
	return ret;
}

static int
1358
do_add_counters(struct net *net, const void __user *user, unsigned int len,
1359
		int compat)
L
Linus Torvalds 已提交
1360
{
1361
	unsigned int i, curcpu;
1362 1363 1364 1365 1366 1367
	struct xt_counters_info tmp;
	struct xt_counters *paddc;
	unsigned int num_counters;
	char *name;
	int size;
	void *ptmp;
1368
	struct xt_table *t;
1369
	const struct xt_table_info *private;
1370
	int ret = 0;
1371
	const void *loc_cpu_entry;
1372
	struct ip6t_entry *iter;
1373 1374
#ifdef CONFIG_COMPAT
	struct compat_xt_counters_info compat_tmp;
L
Linus Torvalds 已提交
1375

1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386
	if (compat) {
		ptmp = &compat_tmp;
		size = sizeof(struct compat_xt_counters_info);
	} else
#endif
	{
		ptmp = &tmp;
		size = sizeof(struct xt_counters_info);
	}

	if (copy_from_user(ptmp, user, size) != 0)
L
Linus Torvalds 已提交
1387 1388
		return -EFAULT;

1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400
#ifdef CONFIG_COMPAT
	if (compat) {
		num_counters = compat_tmp.num_counters;
		name = compat_tmp.name;
	} else
#endif
	{
		num_counters = tmp.num_counters;
		name = tmp.name;
	}

	if (len != size + num_counters * sizeof(struct xt_counters))
L
Linus Torvalds 已提交
1401 1402
		return -EINVAL;

1403
	paddc = vmalloc_node(len - size, numa_node_id());
L
Linus Torvalds 已提交
1404 1405 1406
	if (!paddc)
		return -ENOMEM;

1407
	if (copy_from_user(paddc, user + size, len - size) != 0) {
L
Linus Torvalds 已提交
1408 1409 1410 1411
		ret = -EFAULT;
		goto free;
	}

1412
	t = xt_find_table_lock(net, AF_INET6, name);
1413 1414
	if (!t || IS_ERR(t)) {
		ret = t ? PTR_ERR(t) : -ENOENT;
L
Linus Torvalds 已提交
1415
		goto free;
1416
	}
L
Linus Torvalds 已提交
1417

1418 1419

	local_bh_disable();
1420
	private = t->private;
1421
	if (private->number != num_counters) {
L
Linus Torvalds 已提交
1422 1423 1424 1425 1426
		ret = -EINVAL;
		goto unlock_up_free;
	}

	i = 0;
1427
	/* Choose the copy that is on our node */
1428 1429 1430
	curcpu = smp_processor_id();
	xt_info_wrlock(curcpu);
	loc_cpu_entry = private->entries[curcpu];
1431 1432 1433 1434
	xt_entry_foreach(iter, loc_cpu_entry, private->size) {
		ADD_COUNTER(iter->counters, paddc[i].bcnt, paddc[i].pcnt);
		++i;
	}
1435 1436
	xt_info_wrunlock(curcpu);

L
Linus Torvalds 已提交
1437
 unlock_up_free:
1438
	local_bh_enable();
1439
	xt_table_unlock(t);
1440
	module_put(t->me);
L
Linus Torvalds 已提交
1441 1442 1443 1444 1445 1446
 free:
	vfree(paddc);

	return ret;
}

1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461
#ifdef CONFIG_COMPAT
struct compat_ip6t_replace {
	char			name[IP6T_TABLE_MAXNAMELEN];
	u32			valid_hooks;
	u32			num_entries;
	u32			size;
	u32			hook_entry[NF_INET_NUMHOOKS];
	u32			underflow[NF_INET_NUMHOOKS];
	u32			num_counters;
	compat_uptr_t		counters;	/* struct ip6t_counters * */
	struct compat_ip6t_entry entries[0];
};

static int
compat_copy_entry_to_user(struct ip6t_entry *e, void __user **dstptr,
1462
			  unsigned int *size, struct xt_counters *counters,
1463
			  unsigned int i)
1464 1465 1466 1467 1468
{
	struct ip6t_entry_target *t;
	struct compat_ip6t_entry __user *ce;
	u_int16_t target_offset, next_offset;
	compat_uint_t origsize;
1469 1470
	const struct xt_entry_match *ematch;
	int ret = 0;
1471 1472 1473

	origsize = *size;
	ce = (struct compat_ip6t_entry __user *)*dstptr;
1474 1475 1476 1477
	if (copy_to_user(ce, e, sizeof(struct ip6t_entry)) != 0 ||
	    copy_to_user(&ce->counters, &counters[i],
	    sizeof(counters[i])) != 0)
		return -EFAULT;
1478 1479 1480 1481

	*dstptr += sizeof(struct compat_ip6t_entry);
	*size -= sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);

1482 1483 1484
	xt_ematch_foreach(ematch, e) {
		ret = xt_compat_match_to_user(ematch, dstptr, size);
		if (ret != 0)
1485
			return ret;
1486
	}
1487 1488 1489 1490
	target_offset = e->target_offset - (origsize - *size);
	t = ip6t_get_target(e);
	ret = xt_compat_target_to_user(t, dstptr, size);
	if (ret)
1491
		return ret;
1492
	next_offset = e->next_offset - (origsize - *size);
1493 1494 1495
	if (put_user(target_offset, &ce->target_offset) != 0 ||
	    put_user(next_offset, &ce->next_offset) != 0)
		return -EFAULT;
1496 1497 1498
	return 0;
}

1499
static int
1500 1501 1502 1503
compat_find_calc_match(struct ip6t_entry_match *m,
		       const char *name,
		       const struct ip6t_ip6 *ipv6,
		       unsigned int hookmask,
1504
		       int *size)
1505 1506 1507
{
	struct xt_match *match;

1508 1509 1510
	match = xt_request_find_match(NFPROTO_IPV6, m->u.user.name,
				      m->u.user.revision);
	if (IS_ERR(match)) {
1511 1512
		duprintf("compat_check_calc_match: `%s' not found\n",
			 m->u.user.name);
1513
		return PTR_ERR(match);
1514 1515 1516 1517 1518 1519
	}
	m->u.kernel.match = match;
	*size += xt_compat_match_offset(match);
	return 0;
}

1520
static void compat_release_entry(struct compat_ip6t_entry *e)
1521 1522
{
	struct ip6t_entry_target *t;
1523
	struct xt_entry_match *ematch;
1524 1525

	/* Cleanup all matches */
1526
	xt_ematch_foreach(ematch, e)
1527
		module_put(ematch->u.kernel.match->me);
1528 1529 1530 1531
	t = compat_ip6t_get_target(e);
	module_put(t->u.kernel.target->me);
}

1532
static int
1533 1534 1535
check_compat_entry_size_and_hooks(struct compat_ip6t_entry *e,
				  struct xt_table_info *newinfo,
				  unsigned int *size,
1536 1537 1538 1539
				  const unsigned char *base,
				  const unsigned char *limit,
				  const unsigned int *hook_entries,
				  const unsigned int *underflows,
1540 1541
				  const char *name)
{
1542
	struct xt_entry_match *ematch;
1543 1544 1545
	struct ip6t_entry_target *t;
	struct xt_target *target;
	unsigned int entry_offset;
1546 1547
	unsigned int j;
	int ret, off, h;
1548 1549

	duprintf("check_compat_entry_size_and_hooks %p\n", e);
1550 1551
	if ((unsigned long)e % __alignof__(struct compat_ip6t_entry) != 0 ||
	    (unsigned char *)e + sizeof(struct compat_ip6t_entry) >= limit) {
1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570
		duprintf("Bad offset %p, limit = %p\n", e, limit);
		return -EINVAL;
	}

	if (e->next_offset < sizeof(struct compat_ip6t_entry) +
			     sizeof(struct compat_xt_entry_target)) {
		duprintf("checking: element %p size %u\n",
			 e, e->next_offset);
		return -EINVAL;
	}

	/* For purposes of check_entry casting the compat entry is fine */
	ret = check_entry((struct ip6t_entry *)e, name);
	if (ret)
		return ret;

	off = sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);
	entry_offset = (void *)e - (void *)base;
	j = 0;
1571 1572
	xt_ematch_foreach(ematch, e) {
		ret = compat_find_calc_match(ematch, name,
1573
					     &e->ipv6, e->comefrom, &off);
1574
		if (ret != 0)
1575 1576
			goto release_matches;
		++j;
1577
	}
1578 1579

	t = compat_ip6t_get_target(e);
1580 1581 1582
	target = xt_request_find_target(NFPROTO_IPV6, t->u.user.name,
					t->u.user.revision);
	if (IS_ERR(target)) {
1583 1584
		duprintf("check_compat_entry_size_and_hooks: `%s' not found\n",
			 t->u.user.name);
1585
		ret = PTR_ERR(target);
1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611
		goto release_matches;
	}
	t->u.kernel.target = target;

	off += xt_compat_target_offset(target);
	*size += off;
	ret = xt_compat_add_offset(AF_INET6, entry_offset, off);
	if (ret)
		goto out;

	/* Check hooks & underflows */
	for (h = 0; h < NF_INET_NUMHOOKS; h++) {
		if ((unsigned char *)e - base == hook_entries[h])
			newinfo->hook_entry[h] = hook_entries[h];
		if ((unsigned char *)e - base == underflows[h])
			newinfo->underflow[h] = underflows[h];
	}

	/* Clear counters and comefrom */
	memset(&e->counters, 0, sizeof(e->counters));
	e->comefrom = 0;
	return 0;

out:
	module_put(t->u.kernel.target->me);
release_matches:
1612 1613
	xt_ematch_foreach(ematch, e) {
		if (j-- == 0)
1614
			break;
1615 1616
		module_put(ematch->u.kernel.match->me);
	}
1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629
	return ret;
}

static int
compat_copy_entry_from_user(struct compat_ip6t_entry *e, void **dstptr,
			    unsigned int *size, const char *name,
			    struct xt_table_info *newinfo, unsigned char *base)
{
	struct ip6t_entry_target *t;
	struct xt_target *target;
	struct ip6t_entry *de;
	unsigned int origsize;
	int ret, h;
1630
	struct xt_entry_match *ematch;
1631 1632 1633 1634 1635 1636 1637 1638 1639 1640

	ret = 0;
	origsize = *size;
	de = (struct ip6t_entry *)*dstptr;
	memcpy(de, e, sizeof(struct ip6t_entry));
	memcpy(&de->counters, &e->counters, sizeof(e->counters));

	*dstptr += sizeof(struct ip6t_entry);
	*size += sizeof(struct ip6t_entry) - sizeof(struct compat_ip6t_entry);

1641 1642 1643
	xt_ematch_foreach(ematch, e) {
		ret = xt_compat_match_from_user(ematch, dstptr, size);
		if (ret != 0)
1644
			return ret;
1645
	}
1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660
	de->target_offset = e->target_offset - (origsize - *size);
	t = compat_ip6t_get_target(e);
	target = t->u.kernel.target;
	xt_compat_target_from_user(t, dstptr, size);

	de->next_offset = e->next_offset - (origsize - *size);
	for (h = 0; h < NF_INET_NUMHOOKS; h++) {
		if ((unsigned char *)de - base < newinfo->hook_entry[h])
			newinfo->hook_entry[h] -= origsize - *size;
		if ((unsigned char *)de - base < newinfo->underflow[h])
			newinfo->underflow[h] -= origsize - *size;
	}
	return ret;
}

1661
static int compat_check_entry(struct ip6t_entry *e, struct net *net,
1662
			      const char *name)
1663
{
1664
	unsigned int j;
1665
	int ret = 0;
1666
	struct xt_mtchk_param mtpar;
1667
	struct xt_entry_match *ematch;
1668 1669

	j = 0;
1670
	mtpar.net	= net;
1671 1672 1673
	mtpar.table     = name;
	mtpar.entryinfo = &e->ipv6;
	mtpar.hook_mask = e->comefrom;
1674
	mtpar.family    = NFPROTO_IPV6;
1675
	xt_ematch_foreach(ematch, e) {
1676
		ret = check_match(ematch, &mtpar);
1677
		if (ret != 0)
1678 1679
			goto cleanup_matches;
		++j;
1680
	}
1681

1682
	ret = check_target(e, net, name);
1683 1684 1685 1686 1687
	if (ret)
		goto cleanup_matches;
	return 0;

 cleanup_matches:
1688 1689
	xt_ematch_foreach(ematch, e) {
		if (j-- == 0)
1690
			break;
1691 1692
		cleanup_match(ematch, net);
	}
1693 1694 1695 1696
	return ret;
}

static int
1697 1698
translate_compat_table(struct net *net,
		       const char *name,
1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
		       unsigned int valid_hooks,
		       struct xt_table_info **pinfo,
		       void **pentry0,
		       unsigned int total_size,
		       unsigned int number,
		       unsigned int *hook_entries,
		       unsigned int *underflows)
{
	unsigned int i, j;
	struct xt_table_info *newinfo, *info;
	void *pos, *entry0, *entry1;
1710 1711
	struct compat_ip6t_entry *iter0;
	struct ip6t_entry *iter1;
1712
	unsigned int size;
1713
	int ret = 0;
1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729

	info = *pinfo;
	entry0 = *pentry0;
	size = total_size;
	info->number = number;

	/* Init all hooks to impossible value. */
	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
		info->hook_entry[i] = 0xFFFFFFFF;
		info->underflow[i] = 0xFFFFFFFF;
	}

	duprintf("translate_compat_table: size %u\n", info->size);
	j = 0;
	xt_compat_lock(AF_INET6);
	/* Walk through entries, checking offsets. */
1730 1731
	xt_entry_foreach(iter0, entry0, total_size) {
		ret = check_compat_entry_size_and_hooks(iter0, info, &size,
1732 1733 1734 1735 1736
							entry0,
							entry0 + total_size,
							hook_entries,
							underflows,
							name);
1737
		if (ret != 0)
1738 1739
			goto out_unlock;
		++j;
1740
	}
1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778

	ret = -EINVAL;
	if (j != number) {
		duprintf("translate_compat_table: %u not %u entries\n",
			 j, number);
		goto out_unlock;
	}

	/* Check hooks all assigned */
	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
		/* Only hooks which are valid */
		if (!(valid_hooks & (1 << i)))
			continue;
		if (info->hook_entry[i] == 0xFFFFFFFF) {
			duprintf("Invalid hook entry %u %u\n",
				 i, hook_entries[i]);
			goto out_unlock;
		}
		if (info->underflow[i] == 0xFFFFFFFF) {
			duprintf("Invalid underflow %u %u\n",
				 i, underflows[i]);
			goto out_unlock;
		}
	}

	ret = -ENOMEM;
	newinfo = xt_alloc_table_info(size);
	if (!newinfo)
		goto out_unlock;

	newinfo->number = number;
	for (i = 0; i < NF_INET_NUMHOOKS; i++) {
		newinfo->hook_entry[i] = info->hook_entry[i];
		newinfo->underflow[i] = info->underflow[i];
	}
	entry1 = newinfo->entries[raw_smp_processor_id()];
	pos = entry1;
	size = total_size;
1779
	xt_entry_foreach(iter0, entry0, total_size) {
1780 1781
		ret = compat_copy_entry_from_user(iter0, &pos, &size,
						  name, newinfo, entry1);
1782 1783 1784
		if (ret != 0)
			break;
	}
1785 1786 1787 1788 1789 1790 1791 1792 1793 1794
	xt_compat_flush_offsets(AF_INET6);
	xt_compat_unlock(AF_INET6);
	if (ret)
		goto free_newinfo;

	ret = -ELOOP;
	if (!mark_source_chains(newinfo, valid_hooks, entry1))
		goto free_newinfo;

	i = 0;
1795
	xt_entry_foreach(iter1, entry1, newinfo->size) {
1796
		ret = compat_check_entry(iter1, net, name);
1797 1798
		if (ret != 0)
			break;
1799
		++i;
1800
	}
1801
	if (ret) {
1802 1803 1804 1805 1806 1807
		/*
		 * The first i matches need cleanup_entry (calls ->destroy)
		 * because they had called ->check already. The other j-i
		 * entries need only release.
		 */
		int skip = i;
1808
		j -= i;
1809 1810 1811
		xt_entry_foreach(iter0, entry0, newinfo->size) {
			if (skip-- > 0)
				continue;
1812
			if (j-- == 0)
1813
				break;
1814
			compat_release_entry(iter0);
1815
		}
1816 1817
		xt_entry_foreach(iter1, entry1, newinfo->size) {
			if (i-- == 0)
1818
				break;
1819 1820
			cleanup_entry(iter1, net);
		}
1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837
		xt_free_table_info(newinfo);
		return ret;
	}

	/* And one copy for every other CPU */
	for_each_possible_cpu(i)
		if (newinfo->entries[i] && newinfo->entries[i] != entry1)
			memcpy(newinfo->entries[i], entry1, newinfo->size);

	*pinfo = newinfo;
	*pentry0 = entry1;
	xt_free_table_info(info);
	return 0;

free_newinfo:
	xt_free_table_info(newinfo);
out:
1838 1839
	xt_entry_foreach(iter0, entry0, total_size) {
		if (j-- == 0)
1840
			break;
1841 1842
		compat_release_entry(iter0);
	}
1843 1844 1845 1846 1847 1848 1849 1850
	return ret;
out_unlock:
	xt_compat_flush_offsets(AF_INET6);
	xt_compat_unlock(AF_INET6);
	goto out;
}

static int
1851
compat_do_replace(struct net *net, void __user *user, unsigned int len)
1852 1853 1854 1855 1856
{
	int ret;
	struct compat_ip6t_replace tmp;
	struct xt_table_info *newinfo;
	void *loc_cpu_entry;
1857
	struct ip6t_entry *iter;
1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871

	if (copy_from_user(&tmp, user, sizeof(tmp)) != 0)
		return -EFAULT;

	/* overflow check */
	if (tmp.size >= INT_MAX / num_possible_cpus())
		return -ENOMEM;
	if (tmp.num_counters >= INT_MAX / sizeof(struct xt_counters))
		return -ENOMEM;

	newinfo = xt_alloc_table_info(tmp.size);
	if (!newinfo)
		return -ENOMEM;

1872
	/* choose the copy that is on our node/cpu */
1873 1874 1875 1876 1877 1878 1879
	loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
	if (copy_from_user(loc_cpu_entry, user + sizeof(tmp),
			   tmp.size) != 0) {
		ret = -EFAULT;
		goto free_newinfo;
	}

1880
	ret = translate_compat_table(net, tmp.name, tmp.valid_hooks,
1881 1882 1883 1884 1885 1886 1887 1888
				     &newinfo, &loc_cpu_entry, tmp.size,
				     tmp.num_entries, tmp.hook_entry,
				     tmp.underflow);
	if (ret != 0)
		goto free_newinfo;

	duprintf("compat_do_replace: Translated table\n");

1889
	ret = __do_replace(net, tmp.name, tmp.valid_hooks, newinfo,
1890 1891 1892 1893 1894 1895
			   tmp.num_counters, compat_ptr(tmp.counters));
	if (ret)
		goto free_newinfo_untrans;
	return 0;

 free_newinfo_untrans:
1896
	xt_entry_foreach(iter, loc_cpu_entry, newinfo->size)
1897
		cleanup_entry(iter, net);
1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913
 free_newinfo:
	xt_free_table_info(newinfo);
	return ret;
}

static int
compat_do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user,
		       unsigned int len)
{
	int ret;

	if (!capable(CAP_NET_ADMIN))
		return -EPERM;

	switch (cmd) {
	case IP6T_SO_SET_REPLACE:
1914
		ret = compat_do_replace(sock_net(sk), user, len);
1915 1916 1917
		break;

	case IP6T_SO_SET_ADD_COUNTERS:
1918
		ret = do_add_counters(sock_net(sk), user, len, 1);
1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939
		break;

	default:
		duprintf("do_ip6t_set_ctl:  unknown request %i\n", cmd);
		ret = -EINVAL;
	}

	return ret;
}

struct compat_ip6t_get_entries {
	char name[IP6T_TABLE_MAXNAMELEN];
	compat_uint_t size;
	struct compat_ip6t_entry entrytable[0];
};

static int
compat_copy_entries_to_user(unsigned int total_size, struct xt_table *table,
			    void __user *userptr)
{
	struct xt_counters *counters;
1940
	const struct xt_table_info *private = table->private;
1941 1942 1943
	void __user *pos;
	unsigned int size;
	int ret = 0;
1944
	const void *loc_cpu_entry;
1945
	unsigned int i = 0;
1946
	struct ip6t_entry *iter;
1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958

	counters = alloc_counters(table);
	if (IS_ERR(counters))
		return PTR_ERR(counters);

	/* choose the copy that is on our node/cpu, ...
	 * This choice is lazy (because current thread is
	 * allowed to migrate to another cpu)
	 */
	loc_cpu_entry = private->entries[raw_smp_processor_id()];
	pos = userptr;
	size = total_size;
1959 1960
	xt_entry_foreach(iter, loc_cpu_entry, total_size) {
		ret = compat_copy_entry_to_user(iter, &pos,
1961
						&size, counters, i++);
1962 1963 1964
		if (ret != 0)
			break;
	}
1965 1966 1967 1968 1969 1970

	vfree(counters);
	return ret;
}

static int
1971 1972
compat_get_entries(struct net *net, struct compat_ip6t_get_entries __user *uptr,
		   int *len)
1973 1974 1975 1976 1977 1978
{
	int ret;
	struct compat_ip6t_get_entries get;
	struct xt_table *t;

	if (*len < sizeof(get)) {
1979
		duprintf("compat_get_entries: %u < %zu\n", *len, sizeof(get));
1980 1981 1982 1983 1984 1985 1986
		return -EINVAL;
	}

	if (copy_from_user(&get, uptr, sizeof(get)) != 0)
		return -EFAULT;

	if (*len != sizeof(struct compat_ip6t_get_entries) + get.size) {
1987 1988
		duprintf("compat_get_entries: %u != %zu\n",
			 *len, sizeof(get) + get.size);
1989 1990 1991 1992
		return -EINVAL;
	}

	xt_compat_lock(AF_INET6);
1993
	t = xt_find_table_lock(net, AF_INET6, get.name);
1994
	if (t && !IS_ERR(t)) {
1995
		const struct xt_table_info *private = t->private;
1996
		struct xt_table_info info;
1997
		duprintf("t->private->number = %u\n", private->number);
1998 1999 2000 2001 2002 2003
		ret = compat_table_info(private, &info);
		if (!ret && get.size == info.size) {
			ret = compat_copy_entries_to_user(private->size,
							  t, uptr->entrytable);
		} else if (!ret) {
			duprintf("compat_get_entries: I've got %u not %u!\n",
2004
				 private->size, get.size);
2005
			ret = -EAGAIN;
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028
		}
		xt_compat_flush_offsets(AF_INET6);
		module_put(t->me);
		xt_table_unlock(t);
	} else
		ret = t ? PTR_ERR(t) : -ENOENT;

	xt_compat_unlock(AF_INET6);
	return ret;
}

static int do_ip6t_get_ctl(struct sock *, int, void __user *, int *);

static int
compat_do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
{
	int ret;

	if (!capable(CAP_NET_ADMIN))
		return -EPERM;

	switch (cmd) {
	case IP6T_SO_GET_INFO:
2029
		ret = get_info(sock_net(sk), user, len, 1);
2030 2031
		break;
	case IP6T_SO_GET_ENTRIES:
2032
		ret = compat_get_entries(sock_net(sk), user, len);
2033 2034 2035 2036 2037 2038 2039 2040
		break;
	default:
		ret = do_ip6t_get_ctl(sk, cmd, user, len);
	}
	return ret;
}
#endif

L
Linus Torvalds 已提交
2041 2042 2043 2044 2045 2046 2047 2048 2049 2050
static int
do_ip6t_set_ctl(struct sock *sk, int cmd, void __user *user, unsigned int len)
{
	int ret;

	if (!capable(CAP_NET_ADMIN))
		return -EPERM;

	switch (cmd) {
	case IP6T_SO_SET_REPLACE:
2051
		ret = do_replace(sock_net(sk), user, len);
L
Linus Torvalds 已提交
2052 2053 2054
		break;

	case IP6T_SO_SET_ADD_COUNTERS:
2055
		ret = do_add_counters(sock_net(sk), user, len, 0);
L
Linus Torvalds 已提交
2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074
		break;

	default:
		duprintf("do_ip6t_set_ctl:  unknown request %i\n", cmd);
		ret = -EINVAL;
	}

	return ret;
}

static int
do_ip6t_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
{
	int ret;

	if (!capable(CAP_NET_ADMIN))
		return -EPERM;

	switch (cmd) {
2075
	case IP6T_SO_GET_INFO:
2076
		ret = get_info(sock_net(sk), user, len, 0);
2077
		break;
L
Linus Torvalds 已提交
2078

2079
	case IP6T_SO_GET_ENTRIES:
2080
		ret = get_entries(sock_net(sk), user, len);
L
Linus Torvalds 已提交
2081 2082
		break;

2083 2084 2085
	case IP6T_SO_GET_REVISION_MATCH:
	case IP6T_SO_GET_REVISION_TARGET: {
		struct ip6t_get_revision rev;
2086
		int target;
2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097

		if (*len != sizeof(rev)) {
			ret = -EINVAL;
			break;
		}
		if (copy_from_user(&rev, user, sizeof(rev)) != 0) {
			ret = -EFAULT;
			break;
		}

		if (cmd == IP6T_SO_GET_REVISION_TARGET)
2098
			target = 1;
2099
		else
2100
			target = 0;
2101

2102 2103 2104
		try_then_request_module(xt_find_revision(AF_INET6, rev.name,
							 rev.revision,
							 target, &ret),
2105 2106 2107 2108
					"ip6t_%s", rev.name);
		break;
	}

L
Linus Torvalds 已提交
2109 2110 2111 2112 2113 2114 2115 2116
	default:
		duprintf("do_ip6t_get_ctl: unknown request %i\n", cmd);
		ret = -EINVAL;
	}

	return ret;
}

2117 2118
struct xt_table *ip6t_register_table(struct net *net,
				     const struct xt_table *table,
2119
				     const struct ip6t_replace *repl)
L
Linus Torvalds 已提交
2120 2121
{
	int ret;
2122
	struct xt_table_info *newinfo;
2123
	struct xt_table_info bootstrap
L
Linus Torvalds 已提交
2124
		= { 0, 0, 0, { 0 }, { 0 }, { } };
2125
	void *loc_cpu_entry;
2126
	struct xt_table *new_table;
L
Linus Torvalds 已提交
2127

2128
	newinfo = xt_alloc_table_info(repl->size);
2129 2130 2131 2132
	if (!newinfo) {
		ret = -ENOMEM;
		goto out;
	}
L
Linus Torvalds 已提交
2133

2134
	/* choose the copy on our node/cpu, but dont care about preemption */
2135 2136
	loc_cpu_entry = newinfo->entries[raw_smp_processor_id()];
	memcpy(loc_cpu_entry, repl->entries, repl->size);
L
Linus Torvalds 已提交
2137

2138
	ret = translate_table(net, newinfo, loc_cpu_entry, repl);
2139 2140
	if (ret != 0)
		goto out_free;
L
Linus Torvalds 已提交
2141

2142
	new_table = xt_register_table(net, table, &bootstrap, newinfo);
2143
	if (IS_ERR(new_table)) {
2144 2145
		ret = PTR_ERR(new_table);
		goto out_free;
L
Linus Torvalds 已提交
2146
	}
2147
	return new_table;
L
Linus Torvalds 已提交
2148

2149 2150 2151 2152
out_free:
	xt_free_table_info(newinfo);
out:
	return ERR_PTR(ret);
L
Linus Torvalds 已提交
2153 2154
}

2155
void ip6t_unregister_table(struct net *net, struct xt_table *table)
L
Linus Torvalds 已提交
2156
{
2157
	struct xt_table_info *private;
2158
	void *loc_cpu_entry;
2159
	struct module *table_owner = table->me;
2160
	struct ip6t_entry *iter;
2161

2162
	private = xt_unregister_table(table);
L
Linus Torvalds 已提交
2163 2164

	/* Decrease module usage counts and free resources */
2165
	loc_cpu_entry = private->entries[raw_smp_processor_id()];
2166
	xt_entry_foreach(iter, loc_cpu_entry, private->size)
2167
		cleanup_entry(iter, net);
2168 2169
	if (private->number > private->initial_entries)
		module_put(table_owner);
2170
	xt_free_table_info(private);
L
Linus Torvalds 已提交
2171 2172 2173
}

/* Returns 1 if the type and code is matched by the range, 0 otherwise */
2174
static inline bool
L
Linus Torvalds 已提交
2175 2176
icmp6_type_code_match(u_int8_t test_type, u_int8_t min_code, u_int8_t max_code,
		     u_int8_t type, u_int8_t code,
2177
		     bool invert)
L
Linus Torvalds 已提交
2178 2179 2180 2181 2182
{
	return (type == test_type && code >= min_code && code <= max_code)
		^ invert;
}

2183
static bool
2184
icmp6_match(const struct sk_buff *skb, const struct xt_match_param *par)
L
Linus Torvalds 已提交
2185
{
2186 2187
	const struct icmp6hdr *ic;
	struct icmp6hdr _icmph;
2188
	const struct ip6t_icmp *icmpinfo = par->matchinfo;
L
Linus Torvalds 已提交
2189 2190

	/* Must not be a fragment. */
2191
	if (par->fragoff != 0)
2192
		return false;
L
Linus Torvalds 已提交
2193

2194
	ic = skb_header_pointer(skb, par->thoff, sizeof(_icmph), &_icmph);
L
Linus Torvalds 已提交
2195 2196
	if (ic == NULL) {
		/* We've been asked to examine this packet, and we
2197 2198
		 * can't.  Hence, no choice but to drop.
		 */
L
Linus Torvalds 已提交
2199
		duprintf("Dropping evil ICMP tinygram.\n");
2200
		*par->hotdrop = true;
2201
		return false;
L
Linus Torvalds 已提交
2202 2203 2204 2205 2206 2207 2208 2209 2210 2211
	}

	return icmp6_type_code_match(icmpinfo->type,
				     icmpinfo->code[0],
				     icmpinfo->code[1],
				     ic->icmp6_type, ic->icmp6_code,
				     !!(icmpinfo->invflags&IP6T_ICMP_INV));
}

/* Called when user tries to insert an entry of this type. */
2212
static int icmp6_checkentry(const struct xt_mtchk_param *par)
L
Linus Torvalds 已提交
2213
{
2214
	const struct ip6t_icmp *icmpinfo = par->matchinfo;
L
Linus Torvalds 已提交
2215

2216 2217
	/* Must specify no unknown invflags */
	return !(icmpinfo->invflags & ~IP6T_ICMP_INV);
L
Linus Torvalds 已提交
2218 2219 2220
}

/* The built-in targets: standard (NULL) and error. */
2221
static struct xt_target ip6t_standard_target __read_mostly = {
L
Linus Torvalds 已提交
2222
	.name		= IP6T_STANDARD_TARGET,
2223
	.targetsize	= sizeof(int),
2224
	.family		= NFPROTO_IPV6,
2225 2226 2227 2228 2229
#ifdef CONFIG_COMPAT
	.compatsize	= sizeof(compat_int_t),
	.compat_from_user = compat_standard_from_user,
	.compat_to_user	= compat_standard_to_user,
#endif
L
Linus Torvalds 已提交
2230 2231
};

2232
static struct xt_target ip6t_error_target __read_mostly = {
L
Linus Torvalds 已提交
2233 2234
	.name		= IP6T_ERROR_TARGET,
	.target		= ip6t_error,
2235
	.targetsize	= IP6T_FUNCTION_MAXNAMELEN,
2236
	.family		= NFPROTO_IPV6,
L
Linus Torvalds 已提交
2237 2238 2239 2240 2241 2242 2243
};

static struct nf_sockopt_ops ip6t_sockopts = {
	.pf		= PF_INET6,
	.set_optmin	= IP6T_BASE_CTL,
	.set_optmax	= IP6T_SO_SET_MAX+1,
	.set		= do_ip6t_set_ctl,
2244 2245 2246
#ifdef CONFIG_COMPAT
	.compat_set	= compat_do_ip6t_set_ctl,
#endif
L
Linus Torvalds 已提交
2247 2248 2249
	.get_optmin	= IP6T_BASE_CTL,
	.get_optmax	= IP6T_SO_GET_MAX+1,
	.get		= do_ip6t_get_ctl,
2250 2251 2252
#ifdef CONFIG_COMPAT
	.compat_get	= compat_do_ip6t_get_ctl,
#endif
2253
	.owner		= THIS_MODULE,
L
Linus Torvalds 已提交
2254 2255
};

2256
static struct xt_match icmp6_matchstruct __read_mostly = {
L
Linus Torvalds 已提交
2257
	.name		= "icmp6",
2258
	.match		= icmp6_match,
2259 2260 2261
	.matchsize	= sizeof(struct ip6t_icmp),
	.checkentry	= icmp6_checkentry,
	.proto		= IPPROTO_ICMPV6,
2262
	.family		= NFPROTO_IPV6,
L
Linus Torvalds 已提交
2263 2264
};

2265 2266
static int __net_init ip6_tables_net_init(struct net *net)
{
2267
	return xt_proto_init(net, NFPROTO_IPV6);
2268 2269 2270 2271
}

static void __net_exit ip6_tables_net_exit(struct net *net)
{
2272
	xt_proto_fini(net, NFPROTO_IPV6);
2273 2274 2275 2276 2277 2278 2279
}

static struct pernet_operations ip6_tables_net_ops = {
	.init = ip6_tables_net_init,
	.exit = ip6_tables_net_exit,
};

2280
static int __init ip6_tables_init(void)
L
Linus Torvalds 已提交
2281 2282 2283
{
	int ret;

2284
	ret = register_pernet_subsys(&ip6_tables_net_ops);
2285 2286
	if (ret < 0)
		goto err1;
2287

L
Linus Torvalds 已提交
2288
	/* Noone else will be downing sem now, so we won't sleep */
2289 2290 2291 2292 2293 2294 2295 2296 2297
	ret = xt_register_target(&ip6t_standard_target);
	if (ret < 0)
		goto err2;
	ret = xt_register_target(&ip6t_error_target);
	if (ret < 0)
		goto err3;
	ret = xt_register_match(&icmp6_matchstruct);
	if (ret < 0)
		goto err4;
L
Linus Torvalds 已提交
2298 2299 2300

	/* Register setsockopt */
	ret = nf_register_sockopt(&ip6t_sockopts);
2301 2302
	if (ret < 0)
		goto err5;
L
Linus Torvalds 已提交
2303

2304
	pr_info("(C) 2000-2006 Netfilter Core Team\n");
L
Linus Torvalds 已提交
2305
	return 0;
2306 2307 2308 2309 2310 2311 2312 2313

err5:
	xt_unregister_match(&icmp6_matchstruct);
err4:
	xt_unregister_target(&ip6t_error_target);
err3:
	xt_unregister_target(&ip6t_standard_target);
err2:
2314
	unregister_pernet_subsys(&ip6_tables_net_ops);
2315 2316
err1:
	return ret;
L
Linus Torvalds 已提交
2317 2318
}

2319
static void __exit ip6_tables_fini(void)
L
Linus Torvalds 已提交
2320 2321
{
	nf_unregister_sockopt(&ip6t_sockopts);
2322

2323 2324 2325
	xt_unregister_match(&icmp6_matchstruct);
	xt_unregister_target(&ip6t_error_target);
	xt_unregister_target(&ip6t_standard_target);
2326 2327

	unregister_pernet_subsys(&ip6_tables_net_ops);
L
Linus Torvalds 已提交
2328 2329
}

2330
/*
2331 2332 2333 2334 2335 2336 2337
 * find the offset to specified header or the protocol number of last header
 * if target < 0. "last header" is transport protocol header, ESP, or
 * "No next header".
 *
 * If target header is found, its offset is set in *offset and return protocol
 * number. Otherwise, return -1.
 *
2338 2339 2340
 * If the first fragment doesn't contain the final protocol header or
 * NEXTHDR_NONE it is considered invalid.
 *
2341 2342 2343 2344
 * Note that non-1st fragment is special case that "the protocol number
 * of last header" is "next header" field in Fragment header. In this case,
 * *offset is meaningless and fragment offset is stored in *fragoff if fragoff
 * isn't NULL.
2345 2346
 *
 */
2347 2348
int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset,
		  int target, unsigned short *fragoff)
2349
{
2350
	unsigned int start = skb_network_offset(skb) + sizeof(struct ipv6hdr);
2351
	u8 nexthdr = ipv6_hdr(skb)->nexthdr;
2352 2353
	unsigned int len = skb->len - start;

2354 2355 2356
	if (fragoff)
		*fragoff = 0;

2357 2358 2359 2360
	while (nexthdr != target) {
		struct ipv6_opt_hdr _hdr, *hp;
		unsigned int hdrlen;

2361 2362 2363
		if ((!ipv6_ext_hdr(nexthdr)) || nexthdr == NEXTHDR_NONE) {
			if (target < 0)
				break;
2364
			return -ENOENT;
2365 2366
		}

2367 2368
		hp = skb_header_pointer(skb, start, sizeof(_hdr), &_hdr);
		if (hp == NULL)
2369
			return -EBADMSG;
2370
		if (nexthdr == NEXTHDR_FRAGMENT) {
A
Al Viro 已提交
2371 2372
			unsigned short _frag_off;
			__be16 *fp;
2373 2374 2375 2376 2377 2378
			fp = skb_header_pointer(skb,
						start+offsetof(struct frag_hdr,
							       frag_off),
						sizeof(_frag_off),
						&_frag_off);
			if (fp == NULL)
2379
				return -EBADMSG;
2380

2381 2382 2383 2384
			_frag_off = ntohs(*fp) & ~0x7;
			if (_frag_off) {
				if (target < 0 &&
				    ((!ipv6_ext_hdr(hp->nexthdr)) ||
2385
				     hp->nexthdr == NEXTHDR_NONE)) {
2386 2387 2388 2389
					if (fragoff)
						*fragoff = _frag_off;
					return hp->nexthdr;
				}
2390
				return -ENOENT;
2391
			}
2392 2393
			hdrlen = 8;
		} else if (nexthdr == NEXTHDR_AUTH)
2394
			hdrlen = (hp->hdrlen + 2) << 2;
2395
		else
2396
			hdrlen = ipv6_optlen(hp);
2397 2398 2399 2400 2401 2402 2403

		nexthdr = hp->nexthdr;
		len -= hdrlen;
		start += hdrlen;
	}

	*offset = start;
2404
	return nexthdr;
2405 2406
}

L
Linus Torvalds 已提交
2407 2408 2409 2410
EXPORT_SYMBOL(ip6t_register_table);
EXPORT_SYMBOL(ip6t_unregister_table);
EXPORT_SYMBOL(ip6t_do_table);
EXPORT_SYMBOL(ip6t_ext_hdr);
2411
EXPORT_SYMBOL(ipv6_find_hdr);
L
Linus Torvalds 已提交
2412

2413 2414
module_init(ip6_tables_init);
module_exit(ip6_tables_fini);