ebtables.c 39.4 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
/*
 *  ebtables
 *
 *  Author:
 *  Bart De Schuymer		<bdschuym@pandora.be>
 *
 *  ebtables.c,v 2.0, July, 2002
 *
 *  This code is stongly inspired on the iptables code which is
 *  Copyright (C) 1999 Paul `Rusty' Russell & Michael J. Neuling
 *
 *  This program is free software; you can redistribute it and/or
 *  modify it under the terms of the GNU General Public License
 *  as published by the Free Software Foundation; either version
 *  2 of the License, or (at your option) any later version.
 */

/* used for print_string */
#include <linux/sched.h>
#include <linux/tty.h>

#include <linux/kmod.h>
#include <linux/module.h>
#include <linux/vmalloc.h>
#include <linux/netfilter_bridge/ebtables.h>
#include <linux/spinlock.h>
P
Patrick McHardy 已提交
27
#include <linux/mutex.h>
L
Linus Torvalds 已提交
28 29
#include <asm/uaccess.h>
#include <linux/smp.h>
30
#include <linux/cpumask.h>
L
Linus Torvalds 已提交
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58
#include <net/sock.h>
/* needed for logical [in,out]-dev filtering */
#include "../br_private.h"

#define BUGPRINT(format, args...) printk("kernel msg: ebtables bug: please "\
                                         "report to author: "format, ## args)
/* #define BUGPRINT(format, args...) */
#define MEMPRINT(format, args...) printk("kernel msg: ebtables "\
                                         ": out of memory: "format, ## args)
/* #define MEMPRINT(format, args...) */



/*
 * Each cpu has its own set of counters, so there is no need for write_lock in
 * the softirq
 * For reading or updating the counters, the user context needs to
 * get a write_lock
 */

/* The size of each set of counters is altered to get cache alignment */
#define SMP_ALIGN(x) (((x) + SMP_CACHE_BYTES-1) & ~(SMP_CACHE_BYTES-1))
#define COUNTER_OFFSET(n) (SMP_ALIGN(n * sizeof(struct ebt_counter)))
#define COUNTER_BASE(c, n, cpu) ((struct ebt_counter *)(((char *)c) + \
   COUNTER_OFFSET(n) * cpu))



I
Ingo Molnar 已提交
59
static DEFINE_MUTEX(ebt_mutex);
L
Linus Torvalds 已提交
60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
static LIST_HEAD(ebt_tables);
static LIST_HEAD(ebt_targets);
static LIST_HEAD(ebt_matches);
static LIST_HEAD(ebt_watchers);

static struct ebt_target ebt_standard_target =
{ {NULL, NULL}, EBT_STANDARD_TARGET, NULL, NULL, NULL, NULL};

static inline int ebt_do_watcher (struct ebt_entry_watcher *w,
   const struct sk_buff *skb, unsigned int hooknr, const struct net_device *in,
   const struct net_device *out)
{
	w->u.watcher->watcher(skb, hooknr, in, out, w->data,
	   w->watcher_size);
	/* watchers don't give a verdict */
	return 0;
}

static inline int ebt_do_match (struct ebt_entry_match *m,
   const struct sk_buff *skb, const struct net_device *in,
   const struct net_device *out)
{
	return m->u.match->match(skb, in, out, m->data,
	   m->match_size);
}

static inline int ebt_dev_check(char *entry, const struct net_device *device)
{
	int i = 0;
89
	const char *devname = device->name;
L
Linus Torvalds 已提交
90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273

	if (*entry == '\0')
		return 0;
	if (!device)
		return 1;
	/* 1 is the wildcard token */
	while (entry[i] != '\0' && entry[i] != 1 && entry[i] == devname[i])
		i++;
	return (devname[i] != entry[i] && entry[i] != 1);
}

#define FWINV2(bool,invflg) ((bool) ^ !!(e->invflags & invflg))
/* process standard matches */
static inline int ebt_basic_match(struct ebt_entry *e, struct ethhdr *h,
   const struct net_device *in, const struct net_device *out)
{
	int verdict, i;

	if (e->bitmask & EBT_802_3) {
		if (FWINV2(ntohs(h->h_proto) >= 1536, EBT_IPROTO))
			return 1;
	} else if (!(e->bitmask & EBT_NOPROTO) &&
	   FWINV2(e->ethproto != h->h_proto, EBT_IPROTO))
		return 1;

	if (FWINV2(ebt_dev_check(e->in, in), EBT_IIN))
		return 1;
	if (FWINV2(ebt_dev_check(e->out, out), EBT_IOUT))
		return 1;
	if ((!in || !in->br_port) ? 0 : FWINV2(ebt_dev_check(
	   e->logical_in, in->br_port->br->dev), EBT_ILOGICALIN))
		return 1;
	if ((!out || !out->br_port) ? 0 : FWINV2(ebt_dev_check(
	   e->logical_out, out->br_port->br->dev), EBT_ILOGICALOUT))
		return 1;

	if (e->bitmask & EBT_SOURCEMAC) {
		verdict = 0;
		for (i = 0; i < 6; i++)
			verdict |= (h->h_source[i] ^ e->sourcemac[i]) &
			   e->sourcemsk[i];
		if (FWINV2(verdict != 0, EBT_ISOURCE) )
			return 1;
	}
	if (e->bitmask & EBT_DESTMAC) {
		verdict = 0;
		for (i = 0; i < 6; i++)
			verdict |= (h->h_dest[i] ^ e->destmac[i]) &
			   e->destmsk[i];
		if (FWINV2(verdict != 0, EBT_IDEST) )
			return 1;
	}
	return 0;
}

/* Do some firewalling */
unsigned int ebt_do_table (unsigned int hook, struct sk_buff **pskb,
   const struct net_device *in, const struct net_device *out,
   struct ebt_table *table)
{
	int i, nentries;
	struct ebt_entry *point;
	struct ebt_counter *counter_base, *cb_base;
	struct ebt_entry_target *t;
	int verdict, sp = 0;
	struct ebt_chainstack *cs;
	struct ebt_entries *chaininfo;
	char *base;
	struct ebt_table_info *private;

	read_lock_bh(&table->lock);
	private = table->private;
	cb_base = COUNTER_BASE(private->counters, private->nentries,
	   smp_processor_id());
	if (private->chainstack)
		cs = private->chainstack[smp_processor_id()];
	else
		cs = NULL;
	chaininfo = private->hook_entry[hook];
	nentries = private->hook_entry[hook]->nentries;
	point = (struct ebt_entry *)(private->hook_entry[hook]->data);
	counter_base = cb_base + private->hook_entry[hook]->counter_offset;
	/* base for chain jumps */
	base = private->entries;
	i = 0;
	while (i < nentries) {
		if (ebt_basic_match(point, eth_hdr(*pskb), in, out))
			goto letscontinue;

		if (EBT_MATCH_ITERATE(point, ebt_do_match, *pskb, in, out) != 0)
			goto letscontinue;

		/* increase counter */
		(*(counter_base + i)).pcnt++;
		(*(counter_base + i)).bcnt+=(**pskb).len;

		/* these should only watch: not modify, nor tell us
		   what to do with the packet */
		EBT_WATCHER_ITERATE(point, ebt_do_watcher, *pskb, hook, in,
		   out);

		t = (struct ebt_entry_target *)
		   (((char *)point) + point->target_offset);
		/* standard target */
		if (!t->u.target->target)
			verdict = ((struct ebt_standard_target *)t)->verdict;
		else
			verdict = t->u.target->target(pskb, hook,
			   in, out, t->data, t->target_size);
		if (verdict == EBT_ACCEPT) {
			read_unlock_bh(&table->lock);
			return NF_ACCEPT;
		}
		if (verdict == EBT_DROP) {
			read_unlock_bh(&table->lock);
			return NF_DROP;
		}
		if (verdict == EBT_RETURN) {
letsreturn:
#ifdef CONFIG_NETFILTER_DEBUG
			if (sp == 0) {
				BUGPRINT("RETURN on base chain");
				/* act like this is EBT_CONTINUE */
				goto letscontinue;
			}
#endif
			sp--;
			/* put all the local variables right */
			i = cs[sp].n;
			chaininfo = cs[sp].chaininfo;
			nentries = chaininfo->nentries;
			point = cs[sp].e;
			counter_base = cb_base +
			   chaininfo->counter_offset;
			continue;
		}
		if (verdict == EBT_CONTINUE)
			goto letscontinue;
#ifdef CONFIG_NETFILTER_DEBUG
		if (verdict < 0) {
			BUGPRINT("bogus standard verdict\n");
			read_unlock_bh(&table->lock);
			return NF_DROP;
		}
#endif
		/* jump to a udc */
		cs[sp].n = i + 1;
		cs[sp].chaininfo = chaininfo;
		cs[sp].e = (struct ebt_entry *)
		   (((char *)point) + point->next_offset);
		i = 0;
		chaininfo = (struct ebt_entries *) (base + verdict);
#ifdef CONFIG_NETFILTER_DEBUG
		if (chaininfo->distinguisher) {
			BUGPRINT("jump to non-chain\n");
			read_unlock_bh(&table->lock);
			return NF_DROP;
		}
#endif
		nentries = chaininfo->nentries;
		point = (struct ebt_entry *)chaininfo->data;
		counter_base = cb_base + chaininfo->counter_offset;
		sp++;
		continue;
letscontinue:
		point = (struct ebt_entry *)
		   (((char *)point) + point->next_offset);
		i++;
	}

	/* I actually like this :) */
	if (chaininfo->policy == EBT_RETURN)
		goto letsreturn;
	if (chaininfo->policy == EBT_ACCEPT) {
		read_unlock_bh(&table->lock);
		return NF_ACCEPT;
	}
	read_unlock_bh(&table->lock);
	return NF_DROP;
}

/* If it succeeds, returns element and locks mutex */
static inline void *
find_inlist_lock_noload(struct list_head *head, const char *name, int *error,
I
Ingo Molnar 已提交
274
   struct mutex *mutex)
L
Linus Torvalds 已提交
275
{
P
Patrick McHardy 已提交
276 277 278 279
	struct {
		struct list_head list;
		char name[EBT_FUNCTION_MAXNAMELEN];
	} *e;
L
Linus Torvalds 已提交
280

I
Ingo Molnar 已提交
281
	*error = mutex_lock_interruptible(mutex);
L
Linus Torvalds 已提交
282 283 284
	if (*error != 0)
		return NULL;

P
Patrick McHardy 已提交
285 286 287
	list_for_each_entry(e, head, list) {
		if (strcmp(e->name, name) == 0)
			return e;
L
Linus Torvalds 已提交
288
	}
P
Patrick McHardy 已提交
289 290 291
	*error = -ENOENT;
	mutex_unlock(mutex);
	return NULL;
L
Linus Torvalds 已提交
292 293 294 295 296 297 298
}

#ifndef CONFIG_KMOD
#define find_inlist_lock(h,n,p,e,m) find_inlist_lock_noload((h),(n),(e),(m))
#else
static void *
find_inlist_lock(struct list_head *head, const char *name, const char *prefix,
I
Ingo Molnar 已提交
299
   int *error, struct mutex *mutex)
L
Linus Torvalds 已提交
300 301 302 303 304 305 306 307 308 309 310 311 312
{
	void *ret;

	ret = find_inlist_lock_noload(head, name, error, mutex);
	if (!ret) {
		request_module("%s%s", prefix, name);
		ret = find_inlist_lock_noload(head, name, error, mutex);
	}
	return ret;
}
#endif

static inline struct ebt_table *
I
Ingo Molnar 已提交
313
find_table_lock(const char *name, int *error, struct mutex *mutex)
L
Linus Torvalds 已提交
314 315 316 317 318
{
	return find_inlist_lock(&ebt_tables, name, "ebtable_", error, mutex);
}

static inline struct ebt_match *
I
Ingo Molnar 已提交
319
find_match_lock(const char *name, int *error, struct mutex *mutex)
L
Linus Torvalds 已提交
320 321 322 323 324
{
	return find_inlist_lock(&ebt_matches, name, "ebt_", error, mutex);
}

static inline struct ebt_watcher *
I
Ingo Molnar 已提交
325
find_watcher_lock(const char *name, int *error, struct mutex *mutex)
L
Linus Torvalds 已提交
326 327 328 329 330
{
	return find_inlist_lock(&ebt_watchers, name, "ebt_", error, mutex);
}

static inline struct ebt_target *
I
Ingo Molnar 已提交
331
find_target_lock(const char *name, int *error, struct mutex *mutex)
L
Linus Torvalds 已提交
332 333 334 335 336 337 338 339 340
{
	return find_inlist_lock(&ebt_targets, name, "ebt_", error, mutex);
}

static inline int
ebt_check_match(struct ebt_entry_match *m, struct ebt_entry *e,
   const char *name, unsigned int hookmask, unsigned int *cnt)
{
	struct ebt_match *match;
341
	size_t left = ((char *)e + e->watchers_offset) - (char *)m;
L
Linus Torvalds 已提交
342 343
	int ret;

344 345
	if (left < sizeof(struct ebt_entry_match) ||
	    left - sizeof(struct ebt_entry_match) < m->match_size)
L
Linus Torvalds 已提交
346 347 348 349 350 351
		return -EINVAL;
	match = find_match_lock(m->u.name, &ret, &ebt_mutex);
	if (!match)
		return ret;
	m->u.match = match;
	if (!try_module_get(match->me)) {
I
Ingo Molnar 已提交
352
		mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
353 354
		return -ENOENT;
	}
I
Ingo Molnar 已提交
355
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
	if (match->check &&
	   match->check(name, hookmask, e, m->data, m->match_size) != 0) {
		BUGPRINT("match->check failed\n");
		module_put(match->me);
		return -EINVAL;
	}
	(*cnt)++;
	return 0;
}

static inline int
ebt_check_watcher(struct ebt_entry_watcher *w, struct ebt_entry *e,
   const char *name, unsigned int hookmask, unsigned int *cnt)
{
	struct ebt_watcher *watcher;
371
	size_t left = ((char *)e + e->target_offset) - (char *)w;
L
Linus Torvalds 已提交
372 373
	int ret;

374 375
	if (left < sizeof(struct ebt_entry_watcher) ||
	   left - sizeof(struct ebt_entry_watcher) < w->watcher_size)
L
Linus Torvalds 已提交
376 377 378 379 380 381
		return -EINVAL;
	watcher = find_watcher_lock(w->u.name, &ret, &ebt_mutex);
	if (!watcher)
		return ret;
	w->u.watcher = watcher;
	if (!try_module_get(watcher->me)) {
I
Ingo Molnar 已提交
382
		mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
383 384
		return -ENOENT;
	}
I
Ingo Molnar 已提交
385
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
386 387 388 389 390 391 392 393 394 395
	if (watcher->check &&
	   watcher->check(name, hookmask, e, w->data, w->watcher_size) != 0) {
		BUGPRINT("watcher->check failed\n");
		module_put(watcher->me);
		return -EINVAL;
	}
	(*cnt)++;
	return 0;
}

396 397
static int ebt_verify_pointers(struct ebt_replace *repl,
			       struct ebt_table_info *newinfo)
L
Linus Torvalds 已提交
398
{
399 400 401
	unsigned int limit = repl->entries_size;
	unsigned int valid_hooks = repl->valid_hooks;
	unsigned int offset = 0;
L
Linus Torvalds 已提交
402 403
	int i;

404 405 406 407 408 409
	for (i = 0; i < NF_BR_NUMHOOKS; i++)
		newinfo->hook_entry[i] = NULL;

	newinfo->entries_size = repl->entries_size;
	newinfo->nentries = repl->nentries;

410 411 412
	while (offset < limit) {
		size_t left = limit - offset;
		struct ebt_entry *e = (void *)newinfo->entries + offset;
413

414
		if (left < sizeof(unsigned int))
L
Linus Torvalds 已提交
415
			break;
416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442

		for (i = 0; i < NF_BR_NUMHOOKS; i++) {
			if ((valid_hooks & (1 << i)) == 0)
				continue;
			if ((char *)repl->hook_entry[i] == repl->entries + offset)
				break;
		}

		if (i != NF_BR_NUMHOOKS || !(e->bitmask & EBT_ENTRY_OR_ENTRIES)) {
			if (e->bitmask != 0) {
				/* we make userspace set this right,
				   so there is no misunderstanding */
				BUGPRINT("EBT_ENTRY_OR_ENTRIES shouldn't be set "
					 "in distinguisher\n");
				return -EINVAL;
			}
			if (i != NF_BR_NUMHOOKS)
				newinfo->hook_entry[i] = (struct ebt_entries *)e;
			if (left < sizeof(struct ebt_entries))
				break;
			offset += sizeof(struct ebt_entries);
		} else {
			if (left < sizeof(struct ebt_entry))
				break;
			if (left < e->next_offset)
				break;
			offset += e->next_offset;
L
Linus Torvalds 已提交
443
		}
444
	}
445 446 447 448
	if (offset != limit) {
		BUGPRINT("entries_size too small\n");
		return -EINVAL;
	}
449 450 451 452 453 454 455 456 457

	/* check if all valid hooks have a chain */
	for (i = 0; i < NF_BR_NUMHOOKS; i++) {
		if (!newinfo->hook_entry[i] &&
		   (valid_hooks & (1 << i))) {
			BUGPRINT("Valid hook without chain\n");
			return -EINVAL;
		}
	}
458 459 460 461 462 463 464 465 466
	return 0;
}

/*
 * this one is very careful, as it is the first function
 * to parse the userspace data
 */
static inline int
ebt_check_entry_size_and_hooks(struct ebt_entry *e,
467 468 469
   struct ebt_table_info *newinfo,
   unsigned int *n, unsigned int *cnt,
   unsigned int *totalcnt, unsigned int *udc_cnt)
470 471 472 473
{
	int i;

	for (i = 0; i < NF_BR_NUMHOOKS; i++) {
474
		if ((void *)e == (void *)newinfo->hook_entry[i])
475 476 477 478 479
			break;
	}
	/* beginning of a new chain
	   if i == NF_BR_NUMHOOKS it must be a user defined chain */
	if (i != NF_BR_NUMHOOKS || !e->bitmask) {
L
Linus Torvalds 已提交
480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541
		/* this checks if the previous chain has as many entries
		   as it said it has */
		if (*n != *cnt) {
			BUGPRINT("nentries does not equal the nr of entries "
		                 "in the chain\n");
			return -EINVAL;
		}
		if (((struct ebt_entries *)e)->policy != EBT_DROP &&
		   ((struct ebt_entries *)e)->policy != EBT_ACCEPT) {
			/* only RETURN from udc */
			if (i != NF_BR_NUMHOOKS ||
			   ((struct ebt_entries *)e)->policy != EBT_RETURN) {
				BUGPRINT("bad policy\n");
				return -EINVAL;
			}
		}
		if (i == NF_BR_NUMHOOKS) /* it's a user defined chain */
			(*udc_cnt)++;
		if (((struct ebt_entries *)e)->counter_offset != *totalcnt) {
			BUGPRINT("counter_offset != totalcnt");
			return -EINVAL;
		}
		*n = ((struct ebt_entries *)e)->nentries;
		*cnt = 0;
		return 0;
	}
	/* a plain old entry, heh */
	if (sizeof(struct ebt_entry) > e->watchers_offset ||
	   e->watchers_offset > e->target_offset ||
	   e->target_offset >= e->next_offset) {
		BUGPRINT("entry offsets not in right order\n");
		return -EINVAL;
	}
	/* this is not checked anywhere else */
	if (e->next_offset - e->target_offset < sizeof(struct ebt_entry_target)) {
		BUGPRINT("target size too small\n");
		return -EINVAL;
	}
	(*cnt)++;
	(*totalcnt)++;
	return 0;
}

struct ebt_cl_stack
{
	struct ebt_chainstack cs;
	int from;
	unsigned int hookmask;
};

/*
 * we need these positions to check that the jumps to a different part of the
 * entries is a jump to the beginning of a new chain.
 */
static inline int
ebt_get_udc_positions(struct ebt_entry *e, struct ebt_table_info *newinfo,
   struct ebt_entries **hook_entries, unsigned int *n, unsigned int valid_hooks,
   struct ebt_cl_stack *udc)
{
	int i;

	/* we're only interested in chain starts */
542
	if (e->bitmask)
L
Linus Torvalds 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591
		return 0;
	for (i = 0; i < NF_BR_NUMHOOKS; i++) {
		if ((valid_hooks & (1 << i)) == 0)
			continue;
		if (newinfo->hook_entry[i] == (struct ebt_entries *)e)
			break;
	}
	/* only care about udc */
	if (i != NF_BR_NUMHOOKS)
		return 0;

	udc[*n].cs.chaininfo = (struct ebt_entries *)e;
	/* these initialisations are depended on later in check_chainloops() */
	udc[*n].cs.n = 0;
	udc[*n].hookmask = 0;

	(*n)++;
	return 0;
}

static inline int
ebt_cleanup_match(struct ebt_entry_match *m, unsigned int *i)
{
	if (i && (*i)-- == 0)
		return 1;
	if (m->u.match->destroy)
		m->u.match->destroy(m->data, m->match_size);
	module_put(m->u.match->me);

	return 0;
}

static inline int
ebt_cleanup_watcher(struct ebt_entry_watcher *w, unsigned int *i)
{
	if (i && (*i)-- == 0)
		return 1;
	if (w->u.watcher->destroy)
		w->u.watcher->destroy(w->data, w->watcher_size);
	module_put(w->u.watcher->me);

	return 0;
}

static inline int
ebt_cleanup_entry(struct ebt_entry *e, unsigned int *cnt)
{
	struct ebt_entry_target *t;

592
	if (e->bitmask == 0)
L
Linus Torvalds 已提交
593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614
		return 0;
	/* we're done */
	if (cnt && (*cnt)-- == 0)
		return 1;
	EBT_WATCHER_ITERATE(e, ebt_cleanup_watcher, NULL);
	EBT_MATCH_ITERATE(e, ebt_cleanup_match, NULL);
	t = (struct ebt_entry_target *)(((char *)e) + e->target_offset);
	if (t->u.target->destroy)
		t->u.target->destroy(t->data, t->target_size);
	module_put(t->u.target->me);

	return 0;
}

static inline int
ebt_check_entry(struct ebt_entry *e, struct ebt_table_info *newinfo,
   const char *name, unsigned int *cnt, unsigned int valid_hooks,
   struct ebt_cl_stack *cl_s, unsigned int udc_cnt)
{
	struct ebt_entry_target *t;
	struct ebt_target *target;
	unsigned int i, j, hook = 0, hookmask = 0;
615
	size_t gap = e->next_offset - e->target_offset;
L
Linus Torvalds 已提交
616 617 618
	int ret;

	/* don't mess with the struct ebt_entries */
619
	if (e->bitmask == 0)
L
Linus Torvalds 已提交
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668
		return 0;

	if (e->bitmask & ~EBT_F_MASK) {
		BUGPRINT("Unknown flag for bitmask\n");
		return -EINVAL;
	}
	if (e->invflags & ~EBT_INV_MASK) {
		BUGPRINT("Unknown flag for inv bitmask\n");
		return -EINVAL;
	}
	if ( (e->bitmask & EBT_NOPROTO) && (e->bitmask & EBT_802_3) ) {
		BUGPRINT("NOPROTO & 802_3 not allowed\n");
		return -EINVAL;
	}
	/* what hook do we belong to? */
	for (i = 0; i < NF_BR_NUMHOOKS; i++) {
		if ((valid_hooks & (1 << i)) == 0)
			continue;
		if ((char *)newinfo->hook_entry[i] < (char *)e)
			hook = i;
		else
			break;
	}
	/* (1 << NF_BR_NUMHOOKS) tells the check functions the rule is on
	   a base chain */
	if (i < NF_BR_NUMHOOKS)
		hookmask = (1 << hook) | (1 << NF_BR_NUMHOOKS);
	else {
		for (i = 0; i < udc_cnt; i++)
			if ((char *)(cl_s[i].cs.chaininfo) > (char *)e)
				break;
		if (i == 0)
			hookmask = (1 << hook) | (1 << NF_BR_NUMHOOKS);
		else
			hookmask = cl_s[i - 1].hookmask;
	}
	i = 0;
	ret = EBT_MATCH_ITERATE(e, ebt_check_match, e, name, hookmask, &i);
	if (ret != 0)
		goto cleanup_matches;
	j = 0;
	ret = EBT_WATCHER_ITERATE(e, ebt_check_watcher, e, name, hookmask, &j);
	if (ret != 0)
		goto cleanup_watchers;
	t = (struct ebt_entry_target *)(((char *)e) + e->target_offset);
	target = find_target_lock(t->u.name, &ret, &ebt_mutex);
	if (!target)
		goto cleanup_watchers;
	if (!try_module_get(target->me)) {
I
Ingo Molnar 已提交
669
		mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
670 671 672
		ret = -ENOENT;
		goto cleanup_watchers;
	}
I
Ingo Molnar 已提交
673
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
674 675 676

	t->u.target = target;
	if (t->u.target == &ebt_standard_target) {
677
		if (gap < sizeof(struct ebt_standard_target)) {
L
Linus Torvalds 已提交
678 679 680 681 682 683 684 685 686 687
			BUGPRINT("Standard target size too big\n");
			ret = -EFAULT;
			goto cleanup_watchers;
		}
		if (((struct ebt_standard_target *)t)->verdict <
		   -NUM_STANDARD_TARGETS) {
			BUGPRINT("Invalid standard target\n");
			ret = -EFAULT;
			goto cleanup_watchers;
		}
688
	} else if (t->target_size > gap - sizeof(struct ebt_entry_target) ||
L
Linus Torvalds 已提交
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757
	   (t->u.target->check &&
	   t->u.target->check(name, hookmask, e, t->data, t->target_size) != 0)){
		module_put(t->u.target->me);
		ret = -EFAULT;
		goto cleanup_watchers;
	}
	(*cnt)++;
	return 0;
cleanup_watchers:
	EBT_WATCHER_ITERATE(e, ebt_cleanup_watcher, &j);
cleanup_matches:
	EBT_MATCH_ITERATE(e, ebt_cleanup_match, &i);
	return ret;
}

/*
 * checks for loops and sets the hook mask for udc
 * the hook mask for udc tells us from which base chains the udc can be
 * accessed. This mask is a parameter to the check() functions of the extensions
 */
static int check_chainloops(struct ebt_entries *chain, struct ebt_cl_stack *cl_s,
   unsigned int udc_cnt, unsigned int hooknr, char *base)
{
	int i, chain_nr = -1, pos = 0, nentries = chain->nentries, verdict;
	struct ebt_entry *e = (struct ebt_entry *)chain->data;
	struct ebt_entry_target *t;

	while (pos < nentries || chain_nr != -1) {
		/* end of udc, go back one 'recursion' step */
		if (pos == nentries) {
			/* put back values of the time when this chain was called */
			e = cl_s[chain_nr].cs.e;
			if (cl_s[chain_nr].from != -1)
				nentries =
				cl_s[cl_s[chain_nr].from].cs.chaininfo->nentries;
			else
				nentries = chain->nentries;
			pos = cl_s[chain_nr].cs.n;
			/* make sure we won't see a loop that isn't one */
			cl_s[chain_nr].cs.n = 0;
			chain_nr = cl_s[chain_nr].from;
			if (pos == nentries)
				continue;
		}
		t = (struct ebt_entry_target *)
		   (((char *)e) + e->target_offset);
		if (strcmp(t->u.name, EBT_STANDARD_TARGET))
			goto letscontinue;
		if (e->target_offset + sizeof(struct ebt_standard_target) >
		   e->next_offset) {
			BUGPRINT("Standard target size too big\n");
			return -1;
		}
		verdict = ((struct ebt_standard_target *)t)->verdict;
		if (verdict >= 0) { /* jump to another chain */
			struct ebt_entries *hlp2 =
			   (struct ebt_entries *)(base + verdict);
			for (i = 0; i < udc_cnt; i++)
				if (hlp2 == cl_s[i].cs.chaininfo)
					break;
			/* bad destination or loop */
			if (i == udc_cnt) {
				BUGPRINT("bad destination\n");
				return -1;
			}
			if (cl_s[i].cs.n) {
				BUGPRINT("loop\n");
				return -1;
			}
758 759 760
			if (cl_s[i].hookmask & (1 << hooknr))
				goto letscontinue;
			/* this can't be 0, so the loop test is correct */
L
Linus Torvalds 已提交
761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786
			cl_s[i].cs.n = pos + 1;
			pos = 0;
			cl_s[i].cs.e = ((void *)e + e->next_offset);
			e = (struct ebt_entry *)(hlp2->data);
			nentries = hlp2->nentries;
			cl_s[i].from = chain_nr;
			chain_nr = i;
			/* this udc is accessible from the base chain for hooknr */
			cl_s[i].hookmask |= (1 << hooknr);
			continue;
		}
letscontinue:
		e = (void *)e + e->next_offset;
		pos++;
	}
	return 0;
}

/* do the parsing of the table/chains/entries/matches/watchers/targets, heh */
static int translate_table(struct ebt_replace *repl,
   struct ebt_table_info *newinfo)
{
	unsigned int i, j, k, udc_cnt;
	int ret;
	struct ebt_cl_stack *cl_s = NULL; /* used in the checking for chain loops */

787 788 789 790
	ret = ebt_verify_pointers(repl, newinfo);
	if (ret != 0)
		return ret;

L
Linus Torvalds 已提交
791
	i = 0;
792
	while (i < NF_BR_NUMHOOKS && !newinfo->hook_entry[i])
L
Linus Torvalds 已提交
793 794 795 796 797
		i++;
	if (i == NF_BR_NUMHOOKS) {
		BUGPRINT("No valid hooks specified\n");
		return -EINVAL;
	}
798
	if (newinfo->hook_entry[i] != (struct ebt_entries *)newinfo->entries) {
L
Linus Torvalds 已提交
799 800 801 802 803 804
		BUGPRINT("Chains don't start at beginning\n");
		return -EINVAL;
	}
	/* make sure chains are ordered after each other in same order
	   as their corresponding hooks */
	for (j = i + 1; j < NF_BR_NUMHOOKS; j++) {
805
		if (!newinfo->hook_entry[j])
L
Linus Torvalds 已提交
806
			continue;
807
		if (newinfo->hook_entry[j] <= newinfo->hook_entry[i]) {
L
Linus Torvalds 已提交
808 809 810 811 812 813 814 815 816 817 818 819 820
			BUGPRINT("Hook order must be followed\n");
			return -EINVAL;
		}
		i = j;
	}

	/* do some early checkings and initialize some things */
	i = 0; /* holds the expected nr. of entries for the chain */
	j = 0; /* holds the up to now counted entries for the chain */
	k = 0; /* holds the total nr. of entries, should equal
	          newinfo->nentries afterwards */
	udc_cnt = 0; /* will hold the nr. of user defined chains (udc) */
	ret = EBT_ENTRY_ITERATE(newinfo->entries, newinfo->entries_size,
821 822
	   ebt_check_entry_size_and_hooks, newinfo,
	   &i, &j, &k, &udc_cnt);
L
Linus Torvalds 已提交
823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841

	if (ret != 0)
		return ret;

	if (i != j) {
		BUGPRINT("nentries does not equal the nr of entries in the "
		         "(last) chain\n");
		return -EINVAL;
	}
	if (k != newinfo->nentries) {
		BUGPRINT("Total nentries is wrong\n");
		return -EINVAL;
	}

	/* get the location of the udc, put them in an array
	   while we're at it, allocate the chainstack */
	if (udc_cnt) {
		/* this will get free'd in do_replace()/ebt_register_table()
		   if an error occurs */
842 843 844
		newinfo->chainstack =
			vmalloc((highest_possible_processor_id()+1)
				   	* sizeof(*(newinfo->chainstack)));
L
Linus Torvalds 已提交
845 846
		if (!newinfo->chainstack)
			return -ENOMEM;
847
		for_each_possible_cpu(i) {
L
Linus Torvalds 已提交
848
			newinfo->chainstack[i] =
849
			  vmalloc(udc_cnt * sizeof(*(newinfo->chainstack[0])));
L
Linus Torvalds 已提交
850 851 852 853 854 855 856 857 858
			if (!newinfo->chainstack[i]) {
				while (i)
					vfree(newinfo->chainstack[--i]);
				vfree(newinfo->chainstack);
				newinfo->chainstack = NULL;
				return -ENOMEM;
			}
		}

859
		cl_s = vmalloc(udc_cnt * sizeof(*cl_s));
L
Linus Torvalds 已提交
860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875
		if (!cl_s)
			return -ENOMEM;
		i = 0; /* the i'th udc */
		EBT_ENTRY_ITERATE(newinfo->entries, newinfo->entries_size,
		   ebt_get_udc_positions, newinfo, repl->hook_entry, &i,
		   repl->valid_hooks, cl_s);
		/* sanity check */
		if (i != udc_cnt) {
			BUGPRINT("i != udc_cnt\n");
			vfree(cl_s);
			return -EFAULT;
		}
	}

	/* Check for loops */
	for (i = 0; i < NF_BR_NUMHOOKS; i++)
876
		if (newinfo->hook_entry[i])
L
Linus Torvalds 已提交
877 878
			if (check_chainloops(newinfo->hook_entry[i],
			   cl_s, udc_cnt, i, newinfo->entries)) {
879
				vfree(cl_s);
L
Linus Torvalds 已提交
880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901
				return -EINVAL;
			}

	/* we now know the following (along with E=mc):
	   - the nr of entries in each chain is right
	   - the size of the allocated space is right
	   - all valid hooks have a corresponding chain
	   - there are no loops
	   - wrong data can still be on the level of a single entry
	   - could be there are jumps to places that are not the
	     beginning of a chain. This can only occur in chains that
	     are not accessible from any base chains, so we don't care. */

	/* used to know what we need to clean up if something goes wrong */
	i = 0;
	ret = EBT_ENTRY_ITERATE(newinfo->entries, newinfo->entries_size,
	   ebt_check_entry, newinfo, repl->name, &i, repl->valid_hooks,
	   cl_s, udc_cnt);
	if (ret != 0) {
		EBT_ENTRY_ITERATE(newinfo->entries, newinfo->entries_size,
		   ebt_cleanup_entry, &i);
	}
902
	vfree(cl_s);
L
Linus Torvalds 已提交
903 904 905 906 907 908 909 910 911 912 913 914
	return ret;
}

/* called under write_lock */
static void get_counters(struct ebt_counter *oldcounters,
   struct ebt_counter *counters, unsigned int nentries)
{
	int i, cpu;
	struct ebt_counter *counter_base;

	/* counters of cpu 0 */
	memcpy(counters, oldcounters,
915 916
	       sizeof(struct ebt_counter) * nentries);

L
Linus Torvalds 已提交
917
	/* add other counters to those of cpu 0 */
918
	for_each_possible_cpu(cpu) {
919 920
		if (cpu == 0)
			continue;
L
Linus Torvalds 已提交
921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951
		counter_base = COUNTER_BASE(oldcounters, nentries, cpu);
		for (i = 0; i < nentries; i++) {
			counters[i].pcnt += counter_base[i].pcnt;
			counters[i].bcnt += counter_base[i].bcnt;
		}
	}
}

/* replace the table */
static int do_replace(void __user *user, unsigned int len)
{
	int ret, i, countersize;
	struct ebt_table_info *newinfo;
	struct ebt_replace tmp;
	struct ebt_table *t;
	struct ebt_counter *counterstmp = NULL;
	/* used to be able to unlock earlier */
	struct ebt_table_info *table;

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

	if (len != sizeof(tmp) + tmp.entries_size) {
		BUGPRINT("Wrong len argument\n");
		return -EINVAL;
	}

	if (tmp.entries_size == 0) {
		BUGPRINT("Entries_size never zero\n");
		return -EINVAL;
	}
952 953 954 955 956 957 958
	/* overflow check */
	if (tmp.nentries >= ((INT_MAX - sizeof(struct ebt_table_info)) / NR_CPUS -
			SMP_CACHE_BYTES) / sizeof(struct ebt_counter))
		return -ENOMEM;
	if (tmp.num_counters >= INT_MAX / sizeof(struct ebt_counter))
		return -ENOMEM;

959 960
	countersize = COUNTER_OFFSET(tmp.nentries) * 
					(highest_possible_processor_id()+1);
961
	newinfo = vmalloc(sizeof(*newinfo) + countersize);
L
Linus Torvalds 已提交
962 963 964 965 966 967
	if (!newinfo)
		return -ENOMEM;

	if (countersize)
		memset(newinfo->counters, 0, countersize);

968
	newinfo->entries = vmalloc(tmp.entries_size);
L
Linus Torvalds 已提交
969 970 971 972 973 974 975 976 977 978 979 980 981 982
	if (!newinfo->entries) {
		ret = -ENOMEM;
		goto free_newinfo;
	}
	if (copy_from_user(
	   newinfo->entries, tmp.entries, tmp.entries_size) != 0) {
		BUGPRINT("Couldn't copy entries from userspace\n");
		ret = -EFAULT;
		goto free_entries;
	}

	/* the user wants counters back
	   the check on the size is done later, when we have the lock */
	if (tmp.num_counters) {
983
		counterstmp = vmalloc(tmp.num_counters * sizeof(*counterstmp));
L
Linus Torvalds 已提交
984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030
		if (!counterstmp) {
			ret = -ENOMEM;
			goto free_entries;
		}
	}
	else
		counterstmp = NULL;

	/* this can get initialized by translate_table() */
	newinfo->chainstack = NULL;
	ret = translate_table(&tmp, newinfo);

	if (ret != 0)
		goto free_counterstmp;

	t = find_table_lock(tmp.name, &ret, &ebt_mutex);
	if (!t) {
		ret = -ENOENT;
		goto free_iterate;
	}

	/* the table doesn't like it */
	if (t->check && (ret = t->check(newinfo, tmp.valid_hooks)))
		goto free_unlock;

	if (tmp.num_counters && tmp.num_counters != t->private->nentries) {
		BUGPRINT("Wrong nr. of counters requested\n");
		ret = -EINVAL;
		goto free_unlock;
	}

	/* we have the mutex lock, so no danger in reading this pointer */
	table = t->private;
	/* make sure the table can only be rmmod'ed if it contains no rules */
	if (!table->nentries && newinfo->nentries && !try_module_get(t->me)) {
		ret = -ENOENT;
		goto free_unlock;
	} else if (table->nentries && !newinfo->nentries)
		module_put(t->me);
	/* we need an atomic snapshot of the counters */
	write_lock_bh(&t->lock);
	if (tmp.num_counters)
		get_counters(t->private->counters, counterstmp,
		   t->private->nentries);

	t->private = newinfo;
	write_unlock_bh(&t->lock);
I
Ingo Molnar 已提交
1031
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
	/* so, a user can change the chains while having messed up her counter
	   allocation. Only reason why this is done is because this way the lock
	   is held only once, while this doesn't bring the kernel into a
	   dangerous state. */
	if (tmp.num_counters &&
	   copy_to_user(tmp.counters, counterstmp,
	   tmp.num_counters * sizeof(struct ebt_counter))) {
		BUGPRINT("Couldn't copy counters to userspace\n");
		ret = -EFAULT;
	}
	else
		ret = 0;

	/* decrease module count and free resources */
	EBT_ENTRY_ITERATE(table->entries, table->entries_size,
	   ebt_cleanup_entry, NULL);

	vfree(table->entries);
	if (table->chainstack) {
1051
		for_each_possible_cpu(i)
L
Linus Torvalds 已提交
1052 1053 1054 1055 1056
			vfree(table->chainstack[i]);
		vfree(table->chainstack);
	}
	vfree(table);

1057
	vfree(counterstmp);
L
Linus Torvalds 已提交
1058 1059 1060
	return ret;

free_unlock:
I
Ingo Molnar 已提交
1061
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1062 1063 1064 1065
free_iterate:
	EBT_ENTRY_ITERATE(newinfo->entries, newinfo->entries_size,
	   ebt_cleanup_entry, NULL);
free_counterstmp:
1066
	vfree(counterstmp);
L
Linus Torvalds 已提交
1067 1068
	/* can be initialized in translate_table() */
	if (newinfo->chainstack) {
1069
		for_each_possible_cpu(i)
L
Linus Torvalds 已提交
1070 1071 1072 1073
			vfree(newinfo->chainstack[i]);
		vfree(newinfo->chainstack);
	}
free_entries:
1074
	vfree(newinfo->entries);
L
Linus Torvalds 已提交
1075
free_newinfo:
1076
	vfree(newinfo);
L
Linus Torvalds 已提交
1077 1078 1079 1080 1081
	return ret;
}

int ebt_register_target(struct ebt_target *target)
{
P
Patrick McHardy 已提交
1082
	struct ebt_target *t;
L
Linus Torvalds 已提交
1083 1084
	int ret;

I
Ingo Molnar 已提交
1085
	ret = mutex_lock_interruptible(&ebt_mutex);
L
Linus Torvalds 已提交
1086 1087
	if (ret != 0)
		return ret;
P
Patrick McHardy 已提交
1088 1089 1090 1091 1092
	list_for_each_entry(t, &ebt_targets, list) {
		if (strcmp(t->name, target->name) == 0) {
			mutex_unlock(&ebt_mutex);
			return -EEXIST;
		}
L
Linus Torvalds 已提交
1093
	}
P
Patrick McHardy 已提交
1094
	list_add(&target->list, &ebt_targets);
I
Ingo Molnar 已提交
1095
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1096 1097 1098 1099 1100 1101

	return 0;
}

void ebt_unregister_target(struct ebt_target *target)
{
I
Ingo Molnar 已提交
1102
	mutex_lock(&ebt_mutex);
P
Patrick McHardy 已提交
1103
	list_del(&target->list);
I
Ingo Molnar 已提交
1104
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1105 1106 1107 1108
}

int ebt_register_match(struct ebt_match *match)
{
P
Patrick McHardy 已提交
1109
	struct ebt_match *m;
L
Linus Torvalds 已提交
1110 1111
	int ret;

I
Ingo Molnar 已提交
1112
	ret = mutex_lock_interruptible(&ebt_mutex);
L
Linus Torvalds 已提交
1113 1114
	if (ret != 0)
		return ret;
P
Patrick McHardy 已提交
1115 1116 1117 1118 1119
	list_for_each_entry(m, &ebt_matches, list) {
		if (strcmp(m->name, match->name) == 0) {
			mutex_unlock(&ebt_mutex);
			return -EEXIST;
		}
L
Linus Torvalds 已提交
1120
	}
P
Patrick McHardy 已提交
1121
	list_add(&match->list, &ebt_matches);
I
Ingo Molnar 已提交
1122
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1123 1124 1125 1126 1127 1128

	return 0;
}

void ebt_unregister_match(struct ebt_match *match)
{
I
Ingo Molnar 已提交
1129
	mutex_lock(&ebt_mutex);
P
Patrick McHardy 已提交
1130
	list_del(&match->list);
I
Ingo Molnar 已提交
1131
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1132 1133 1134 1135
}

int ebt_register_watcher(struct ebt_watcher *watcher)
{
P
Patrick McHardy 已提交
1136
	struct ebt_watcher *w;
L
Linus Torvalds 已提交
1137 1138
	int ret;

I
Ingo Molnar 已提交
1139
	ret = mutex_lock_interruptible(&ebt_mutex);
L
Linus Torvalds 已提交
1140 1141
	if (ret != 0)
		return ret;
P
Patrick McHardy 已提交
1142 1143 1144 1145 1146
	list_for_each_entry(w, &ebt_watchers, list) {
		if (strcmp(w->name, watcher->name) == 0) {
			mutex_unlock(&ebt_mutex);
			return -EEXIST;
		}
L
Linus Torvalds 已提交
1147
	}
P
Patrick McHardy 已提交
1148
	list_add(&watcher->list, &ebt_watchers);
I
Ingo Molnar 已提交
1149
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1150 1151 1152 1153 1154 1155

	return 0;
}

void ebt_unregister_watcher(struct ebt_watcher *watcher)
{
I
Ingo Molnar 已提交
1156
	mutex_lock(&ebt_mutex);
P
Patrick McHardy 已提交
1157
	list_del(&watcher->list);
I
Ingo Molnar 已提交
1158
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1159 1160 1161 1162 1163
}

int ebt_register_table(struct ebt_table *table)
{
	struct ebt_table_info *newinfo;
P
Patrick McHardy 已提交
1164
	struct ebt_table *t;
L
Linus Torvalds 已提交
1165 1166 1167 1168 1169 1170 1171 1172 1173
	int ret, i, countersize;

	if (!table || !table->table ||!table->table->entries ||
	    table->table->entries_size == 0 ||
	    table->table->counters || table->private) {
		BUGPRINT("Bad table data for ebt_register_table!!!\n");
		return -EINVAL;
	}

1174 1175
	countersize = COUNTER_OFFSET(table->table->nentries) *
					(highest_possible_processor_id()+1);
1176
	newinfo = vmalloc(sizeof(*newinfo) + countersize);
L
Linus Torvalds 已提交
1177 1178 1179 1180
	ret = -ENOMEM;
	if (!newinfo)
		return -ENOMEM;

1181
	newinfo->entries = vmalloc(table->table->entries_size);
L
Linus Torvalds 已提交
1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205
	if (!(newinfo->entries))
		goto free_newinfo;

	memcpy(newinfo->entries, table->table->entries,
	   table->table->entries_size);

	if (countersize)
		memset(newinfo->counters, 0, countersize);

	/* fill in newinfo and parse the entries */
	newinfo->chainstack = NULL;
	ret = translate_table(table->table, newinfo);
	if (ret != 0) {
		BUGPRINT("Translate_table failed\n");
		goto free_chainstack;
	}

	if (table->check && table->check(newinfo, table->valid_hooks)) {
		BUGPRINT("The table doesn't like its own initial data, lol\n");
		return -EINVAL;
	}

	table->private = newinfo;
	rwlock_init(&table->lock);
I
Ingo Molnar 已提交
1206
	ret = mutex_lock_interruptible(&ebt_mutex);
L
Linus Torvalds 已提交
1207 1208 1209
	if (ret != 0)
		goto free_chainstack;

P
Patrick McHardy 已提交
1210 1211 1212 1213 1214 1215
	list_for_each_entry(t, &ebt_tables, list) {
		if (strcmp(t->name, table->name) == 0) {
			ret = -EEXIST;
			BUGPRINT("Table name already exists\n");
			goto free_unlock;
		}
L
Linus Torvalds 已提交
1216 1217 1218 1219 1220 1221 1222
	}

	/* Hold a reference count if the chains aren't empty */
	if (newinfo->nentries && !try_module_get(table->me)) {
		ret = -ENOENT;
		goto free_unlock;
	}
P
Patrick McHardy 已提交
1223
	list_add(&table->list, &ebt_tables);
I
Ingo Molnar 已提交
1224
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1225 1226
	return 0;
free_unlock:
I
Ingo Molnar 已提交
1227
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1228 1229
free_chainstack:
	if (newinfo->chainstack) {
1230
		for_each_possible_cpu(i)
L
Linus Torvalds 已提交
1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247
			vfree(newinfo->chainstack[i]);
		vfree(newinfo->chainstack);
	}
	vfree(newinfo->entries);
free_newinfo:
	vfree(newinfo);
	return ret;
}

void ebt_unregister_table(struct ebt_table *table)
{
	int i;

	if (!table) {
		BUGPRINT("Request to unregister NULL table!!!\n");
		return;
	}
I
Ingo Molnar 已提交
1248
	mutex_lock(&ebt_mutex);
P
Patrick McHardy 已提交
1249
	list_del(&table->list);
I
Ingo Molnar 已提交
1250
	mutex_unlock(&ebt_mutex);
1251
	vfree(table->private->entries);
L
Linus Torvalds 已提交
1252
	if (table->private->chainstack) {
1253
		for_each_possible_cpu(i)
L
Linus Torvalds 已提交
1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275
			vfree(table->private->chainstack[i]);
		vfree(table->private->chainstack);
	}
	vfree(table->private);
}

/* userspace just supplied us with counters */
static int update_counters(void __user *user, unsigned int len)
{
	int i, ret;
	struct ebt_counter *tmp;
	struct ebt_replace hlp;
	struct ebt_table *t;

	if (copy_from_user(&hlp, user, sizeof(hlp)))
		return -EFAULT;

	if (len != sizeof(hlp) + hlp.num_counters * sizeof(struct ebt_counter))
		return -EINVAL;
	if (hlp.num_counters == 0)
		return -EINVAL;

1276
	if (!(tmp = vmalloc(hlp.num_counters * sizeof(*tmp)))) {
L
Linus Torvalds 已提交
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
		MEMPRINT("Update_counters && nomemory\n");
		return -ENOMEM;
	}

	t = find_table_lock(hlp.name, &ret, &ebt_mutex);
	if (!t)
		goto free_tmp;

	if (hlp.num_counters != t->private->nentries) {
		BUGPRINT("Wrong nr of counters\n");
		ret = -EINVAL;
		goto unlock_mutex;
	}

	if ( copy_from_user(tmp, hlp.counters,
	   hlp.num_counters * sizeof(struct ebt_counter)) ) {
		BUGPRINT("Updata_counters && !cfu\n");
		ret = -EFAULT;
		goto unlock_mutex;
	}

	/* we want an atomic add of the counters */
	write_lock_bh(&t->lock);

	/* we add to the counters of the first cpu */
	for (i = 0; i < hlp.num_counters; i++) {
		t->private->counters[i].pcnt += tmp[i].pcnt;
		t->private->counters[i].bcnt += tmp[i].bcnt;
	}

	write_unlock_bh(&t->lock);
	ret = 0;
unlock_mutex:
I
Ingo Molnar 已提交
1310
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339
free_tmp:
	vfree(tmp);
	return ret;
}

static inline int ebt_make_matchname(struct ebt_entry_match *m,
   char *base, char *ubase)
{
	char *hlp = ubase - base + (char *)m;
	if (copy_to_user(hlp, m->u.match->name, EBT_FUNCTION_MAXNAMELEN))
		return -EFAULT;
	return 0;
}

static inline int ebt_make_watchername(struct ebt_entry_watcher *w,
   char *base, char *ubase)
{
	char *hlp = ubase - base + (char *)w;
	if (copy_to_user(hlp , w->u.watcher->name, EBT_FUNCTION_MAXNAMELEN))
		return -EFAULT;
	return 0;
}

static inline int ebt_make_names(struct ebt_entry *e, char *base, char *ubase)
{
	int ret;
	char *hlp;
	struct ebt_entry_target *t;

1340
	if (e->bitmask == 0)
L
Linus Torvalds 已提交
1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356
		return 0;

	hlp = ubase - base + (char *)e + e->target_offset;
	t = (struct ebt_entry_target *)(((char *)e) + e->target_offset);
	
	ret = EBT_MATCH_ITERATE(e, ebt_make_matchname, base, ubase);
	if (ret != 0)
		return ret;
	ret = EBT_WATCHER_ITERATE(e, ebt_make_watchername, base, ubase);
	if (ret != 0)
		return ret;
	if (copy_to_user(hlp, t->u.target->name, EBT_FUNCTION_MAXNAMELEN))
		return -EFAULT;
	return 0;
}

I
Ingo Molnar 已提交
1357
/* called with ebt_mutex locked */
L
Linus Torvalds 已提交
1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404
static int copy_everything_to_user(struct ebt_table *t, void __user *user,
   int *len, int cmd)
{
	struct ebt_replace tmp;
	struct ebt_counter *counterstmp, *oldcounters;
	unsigned int entries_size, nentries;
	char *entries;

	if (cmd == EBT_SO_GET_ENTRIES) {
		entries_size = t->private->entries_size;
		nentries = t->private->nentries;
		entries = t->private->entries;
		oldcounters = t->private->counters;
	} else {
		entries_size = t->table->entries_size;
		nentries = t->table->nentries;
		entries = t->table->entries;
		oldcounters = t->table->counters;
	}

	if (copy_from_user(&tmp, user, sizeof(tmp))) {
		BUGPRINT("Cfu didn't work\n");
		return -EFAULT;
	}

	if (*len != sizeof(struct ebt_replace) + entries_size +
	   (tmp.num_counters? nentries * sizeof(struct ebt_counter): 0)) {
		BUGPRINT("Wrong size\n");
		return -EINVAL;
	}

	if (tmp.nentries != nentries) {
		BUGPRINT("Nentries wrong\n");
		return -EINVAL;
	}

	if (tmp.entries_size != entries_size) {
		BUGPRINT("Wrong size\n");
		return -EINVAL;
	}

	/* userspace might not need the counters */
	if (tmp.num_counters) {
		if (tmp.num_counters != nentries) {
			BUGPRINT("Num_counters wrong\n");
			return -EINVAL;
		}
1405
		counterstmp = vmalloc(nentries * sizeof(*counterstmp));
L
Linus Torvalds 已提交
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467
		if (!counterstmp) {
			MEMPRINT("Couldn't copy counters, out of memory\n");
			return -ENOMEM;
		}
		write_lock_bh(&t->lock);
		get_counters(oldcounters, counterstmp, nentries);
		write_unlock_bh(&t->lock);

		if (copy_to_user(tmp.counters, counterstmp,
		   nentries * sizeof(struct ebt_counter))) {
			BUGPRINT("Couldn't copy counters to userspace\n");
			vfree(counterstmp);
			return -EFAULT;
		}
		vfree(counterstmp);
	}

	if (copy_to_user(tmp.entries, entries, entries_size)) {
		BUGPRINT("Couldn't copy entries to userspace\n");
		return -EFAULT;
	}
	/* set the match/watcher/target names right */
	return EBT_ENTRY_ITERATE(entries, entries_size,
	   ebt_make_names, entries, tmp.entries);
}

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

	switch(cmd) {
	case EBT_SO_SET_ENTRIES:
		ret = do_replace(user, len);
		break;
	case EBT_SO_SET_COUNTERS:
		ret = update_counters(user, len);
		break;
	default:
		ret = -EINVAL;
  }
	return ret;
}

static int do_ebt_get_ctl(struct sock *sk, int cmd, void __user *user, int *len)
{
	int ret;
	struct ebt_replace tmp;
	struct ebt_table *t;

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

	t = find_table_lock(tmp.name, &ret, &ebt_mutex);
	if (!t)
		return ret;

	switch(cmd) {
	case EBT_SO_GET_INFO:
	case EBT_SO_GET_INIT_INFO:
		if (*len != sizeof(struct ebt_replace)){
			ret = -EINVAL;
I
Ingo Molnar 已提交
1468
			mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479
			break;
		}
		if (cmd == EBT_SO_GET_INFO) {
			tmp.nentries = t->private->nentries;
			tmp.entries_size = t->private->entries_size;
			tmp.valid_hooks = t->valid_hooks;
		} else {
			tmp.nentries = t->table->nentries;
			tmp.entries_size = t->table->entries_size;
			tmp.valid_hooks = t->table->valid_hooks;
		}
I
Ingo Molnar 已提交
1480
		mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491
		if (copy_to_user(user, &tmp, *len) != 0){
			BUGPRINT("c2u Didn't work\n");
			ret = -EFAULT;
			break;
		}
		ret = 0;
		break;

	case EBT_SO_GET_ENTRIES:
	case EBT_SO_GET_INIT_ENTRIES:
		ret = copy_everything_to_user(t, user, len, cmd);
I
Ingo Molnar 已提交
1492
		mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1493 1494 1495
		break;

	default:
I
Ingo Molnar 已提交
1496
		mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1497 1498 1499 1500 1501 1502 1503
		ret = -EINVAL;
	}

	return ret;
}

static struct nf_sockopt_ops ebt_sockopts =
A
Andrew Morton 已提交
1504 1505 1506 1507 1508 1509 1510 1511
{
	.pf		= PF_INET,
	.set_optmin	= EBT_BASE_CTL,
	.set_optmax	= EBT_SO_SET_MAX + 1,
	.set		= do_ebt_set_ctl,
	.get_optmin	= EBT_BASE_CTL,
	.get_optmax	= EBT_SO_GET_MAX + 1,
	.get		= do_ebt_get_ctl,
L
Linus Torvalds 已提交
1512 1513
};

1514
static int __init ebtables_init(void)
L
Linus Torvalds 已提交
1515 1516 1517
{
	int ret;

I
Ingo Molnar 已提交
1518
	mutex_lock(&ebt_mutex);
P
Patrick McHardy 已提交
1519
	list_add(&ebt_standard_target.list, &ebt_targets);
I
Ingo Molnar 已提交
1520
	mutex_unlock(&ebt_mutex);
L
Linus Torvalds 已提交
1521 1522 1523 1524 1525 1526 1527
	if ((ret = nf_register_sockopt(&ebt_sockopts)) < 0)
		return ret;

	printk(KERN_NOTICE "Ebtables v2.0 registered\n");
	return 0;
}

1528
static void __exit ebtables_fini(void)
L
Linus Torvalds 已提交
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542
{
	nf_unregister_sockopt(&ebt_sockopts);
	printk(KERN_NOTICE "Ebtables v2.0 unregistered\n");
}

EXPORT_SYMBOL(ebt_register_table);
EXPORT_SYMBOL(ebt_unregister_table);
EXPORT_SYMBOL(ebt_register_match);
EXPORT_SYMBOL(ebt_unregister_match);
EXPORT_SYMBOL(ebt_register_watcher);
EXPORT_SYMBOL(ebt_unregister_watcher);
EXPORT_SYMBOL(ebt_register_target);
EXPORT_SYMBOL(ebt_unregister_target);
EXPORT_SYMBOL(ebt_do_table);
1543 1544
module_init(ebtables_init);
module_exit(ebtables_fini);
L
Linus Torvalds 已提交
1545
MODULE_LICENSE("GPL");