xfrm_policy.c 38.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12
/* 
 * xfrm_policy.c
 *
 * Changes:
 *	Mitsuru KANDA @USAGI
 * 	Kazunori MIYAZAWA @USAGI
 * 	Kunihiro Ishiguro <kunihiro@ipinfusion.com>
 * 		IPv6 support
 * 	Kazunori MIYAZAWA @USAGI
 * 	YOSHIFUJI Hideaki
 * 		Split up af-specific portion
 *	Derek Atkins <derek@ihtfp.com>		Add the post_input processor
13
 *
L
Linus Torvalds 已提交
14 15 16 17 18 19 20 21 22
 */

#include <linux/slab.h>
#include <linux/kmod.h>
#include <linux/list.h>
#include <linux/spinlock.h>
#include <linux/workqueue.h>
#include <linux/notifier.h>
#include <linux/netdevice.h>
23
#include <linux/netfilter.h>
L
Linus Torvalds 已提交
24 25 26 27
#include <linux/module.h>
#include <net/xfrm.h>
#include <net/ip.h>

A
Arjan van de Ven 已提交
28 29
DEFINE_MUTEX(xfrm_cfg_mutex);
EXPORT_SYMBOL(xfrm_cfg_mutex);
L
Linus Torvalds 已提交
30 31 32 33 34

static DEFINE_RWLOCK(xfrm_policy_lock);

struct xfrm_policy *xfrm_policy_list[XFRM_POLICY_MAX*2];
EXPORT_SYMBOL(xfrm_policy_list);
35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52
#ifdef CONFIG_XFRM_SUB_POLICY
struct xfrm_policy *xfrm_policy_list_sub[XFRM_POLICY_MAX*2];
EXPORT_SYMBOL(xfrm_policy_list_sub);

#define XFRM_POLICY_LISTS(type) \
	((type == XFRM_POLICY_TYPE_SUB) ? xfrm_policy_list_sub : \
	 xfrm_policy_list)
#define XFRM_POLICY_LISTHEAD(type, dir) \
	((type == XFRM_POLICY_TYPE_SUB) ? xfrm_policy_list_sub[dir] : \
	 xfrm_policy_list[dir])
#define XFRM_POLICY_LISTHEADP(type, dir) \
	((type == XFRM_POLICY_TYPE_SUB) ? &xfrm_policy_list_sub[dir] : \
	 &xfrm_policy_list[dir])
#else
#define XFRM_POLICY_LISTS(type)              xfrm_policy_list
#define XFRM_POLICY_LISTHEAD(type, dif)      xfrm_policy_list[dir]
#define XFRM_POLICY_LISTHEADP(type, dif)     &xfrm_policy_list[dir]
#endif
L
Linus Torvalds 已提交
53 54 55 56

static DEFINE_RWLOCK(xfrm_policy_afinfo_lock);
static struct xfrm_policy_afinfo *xfrm_policy_afinfo[NPROTO];

57
static kmem_cache_t *xfrm_dst_cache __read_mostly;
L
Linus Torvalds 已提交
58 59 60 61 62 63 64 65

static struct work_struct xfrm_policy_gc_work;
static struct list_head xfrm_policy_gc_list =
	LIST_HEAD_INIT(xfrm_policy_gc_list);
static DEFINE_SPINLOCK(xfrm_policy_gc_lock);

static struct xfrm_policy_afinfo *xfrm_policy_get_afinfo(unsigned short family);
static void xfrm_policy_put_afinfo(struct xfrm_policy_afinfo *afinfo);
66 67
static struct xfrm_policy_afinfo *xfrm_policy_lock_afinfo(unsigned int family);
static void xfrm_policy_unlock_afinfo(struct xfrm_policy_afinfo *afinfo);
L
Linus Torvalds 已提交
68 69 70

int xfrm_register_type(struct xfrm_type *type, unsigned short family)
{
71 72
	struct xfrm_policy_afinfo *afinfo = xfrm_policy_lock_afinfo(family);
	struct xfrm_type **typemap;
L
Linus Torvalds 已提交
73 74 75 76 77 78
	int err = 0;

	if (unlikely(afinfo == NULL))
		return -EAFNOSUPPORT;
	typemap = afinfo->type_map;

79 80
	if (likely(typemap[type->proto] == NULL))
		typemap[type->proto] = type;
L
Linus Torvalds 已提交
81 82
	else
		err = -EEXIST;
83
	xfrm_policy_unlock_afinfo(afinfo);
L
Linus Torvalds 已提交
84 85 86 87 88 89
	return err;
}
EXPORT_SYMBOL(xfrm_register_type);

int xfrm_unregister_type(struct xfrm_type *type, unsigned short family)
{
90 91
	struct xfrm_policy_afinfo *afinfo = xfrm_policy_lock_afinfo(family);
	struct xfrm_type **typemap;
L
Linus Torvalds 已提交
92 93 94 95 96 97
	int err = 0;

	if (unlikely(afinfo == NULL))
		return -EAFNOSUPPORT;
	typemap = afinfo->type_map;

98
	if (unlikely(typemap[type->proto] != type))
L
Linus Torvalds 已提交
99 100
		err = -ENOENT;
	else
101 102
		typemap[type->proto] = NULL;
	xfrm_policy_unlock_afinfo(afinfo);
L
Linus Torvalds 已提交
103 104 105 106 107 108 109
	return err;
}
EXPORT_SYMBOL(xfrm_unregister_type);

struct xfrm_type *xfrm_get_type(u8 proto, unsigned short family)
{
	struct xfrm_policy_afinfo *afinfo;
110
	struct xfrm_type **typemap;
L
Linus Torvalds 已提交
111 112 113 114 115 116 117 118 119
	struct xfrm_type *type;
	int modload_attempted = 0;

retry:
	afinfo = xfrm_policy_get_afinfo(family);
	if (unlikely(afinfo == NULL))
		return NULL;
	typemap = afinfo->type_map;

120
	type = typemap[proto];
L
Linus Torvalds 已提交
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
	if (unlikely(type && !try_module_get(type->owner)))
		type = NULL;
	if (!type && !modload_attempted) {
		xfrm_policy_put_afinfo(afinfo);
		request_module("xfrm-type-%d-%d",
			       (int) family, (int) proto);
		modload_attempted = 1;
		goto retry;
	}

	xfrm_policy_put_afinfo(afinfo);
	return type;
}

int xfrm_dst_lookup(struct xfrm_dst **dst, struct flowi *fl, 
		    unsigned short family)
{
	struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
	int err = 0;

	if (unlikely(afinfo == NULL))
		return -EAFNOSUPPORT;

	if (likely(afinfo->dst_lookup != NULL))
		err = afinfo->dst_lookup(dst, fl);
	else
		err = -EINVAL;
	xfrm_policy_put_afinfo(afinfo);
	return err;
}
EXPORT_SYMBOL(xfrm_dst_lookup);

void xfrm_put_type(struct xfrm_type *type)
{
	module_put(type->owner);
}

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
int xfrm_register_mode(struct xfrm_mode *mode, int family)
{
	struct xfrm_policy_afinfo *afinfo;
	struct xfrm_mode **modemap;
	int err;

	if (unlikely(mode->encap >= XFRM_MODE_MAX))
		return -EINVAL;

	afinfo = xfrm_policy_lock_afinfo(family);
	if (unlikely(afinfo == NULL))
		return -EAFNOSUPPORT;

	err = -EEXIST;
	modemap = afinfo->mode_map;
	if (likely(modemap[mode->encap] == NULL)) {
		modemap[mode->encap] = mode;
		err = 0;
	}

	xfrm_policy_unlock_afinfo(afinfo);
	return err;
}
EXPORT_SYMBOL(xfrm_register_mode);

int xfrm_unregister_mode(struct xfrm_mode *mode, int family)
{
	struct xfrm_policy_afinfo *afinfo;
	struct xfrm_mode **modemap;
	int err;

	if (unlikely(mode->encap >= XFRM_MODE_MAX))
		return -EINVAL;

	afinfo = xfrm_policy_lock_afinfo(family);
	if (unlikely(afinfo == NULL))
		return -EAFNOSUPPORT;

	err = -ENOENT;
	modemap = afinfo->mode_map;
	if (likely(modemap[mode->encap] == mode)) {
		modemap[mode->encap] = NULL;
		err = 0;
	}

	xfrm_policy_unlock_afinfo(afinfo);
	return err;
}
EXPORT_SYMBOL(xfrm_unregister_mode);

struct xfrm_mode *xfrm_get_mode(unsigned int encap, int family)
{
	struct xfrm_policy_afinfo *afinfo;
	struct xfrm_mode *mode;
	int modload_attempted = 0;

	if (unlikely(encap >= XFRM_MODE_MAX))
		return NULL;

retry:
	afinfo = xfrm_policy_get_afinfo(family);
	if (unlikely(afinfo == NULL))
		return NULL;

	mode = afinfo->mode_map[encap];
	if (unlikely(mode && !try_module_get(mode->owner)))
		mode = NULL;
	if (!mode && !modload_attempted) {
		xfrm_policy_put_afinfo(afinfo);
		request_module("xfrm-mode-%d-%d", family, encap);
		modload_attempted = 1;
		goto retry;
	}

	xfrm_policy_put_afinfo(afinfo);
	return mode;
}

void xfrm_put_mode(struct xfrm_mode *mode)
{
	module_put(mode->owner);
}

L
Linus Torvalds 已提交
241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
static inline unsigned long make_jiffies(long secs)
{
	if (secs >= (MAX_SCHEDULE_TIMEOUT-1)/HZ)
		return MAX_SCHEDULE_TIMEOUT-1;
	else
	        return secs*HZ;
}

static void xfrm_policy_timer(unsigned long data)
{
	struct xfrm_policy *xp = (struct xfrm_policy*)data;
	unsigned long now = (unsigned long)xtime.tv_sec;
	long next = LONG_MAX;
	int warn = 0;
	int dir;

	read_lock(&xp->lock);

	if (xp->dead)
		goto out;

262
	dir = xfrm_policy_id2dir(xp->index);
L
Linus Torvalds 已提交
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301

	if (xp->lft.hard_add_expires_seconds) {
		long tmo = xp->lft.hard_add_expires_seconds +
			xp->curlft.add_time - now;
		if (tmo <= 0)
			goto expired;
		if (tmo < next)
			next = tmo;
	}
	if (xp->lft.hard_use_expires_seconds) {
		long tmo = xp->lft.hard_use_expires_seconds +
			(xp->curlft.use_time ? : xp->curlft.add_time) - now;
		if (tmo <= 0)
			goto expired;
		if (tmo < next)
			next = tmo;
	}
	if (xp->lft.soft_add_expires_seconds) {
		long tmo = xp->lft.soft_add_expires_seconds +
			xp->curlft.add_time - now;
		if (tmo <= 0) {
			warn = 1;
			tmo = XFRM_KM_TIMEOUT;
		}
		if (tmo < next)
			next = tmo;
	}
	if (xp->lft.soft_use_expires_seconds) {
		long tmo = xp->lft.soft_use_expires_seconds +
			(xp->curlft.use_time ? : xp->curlft.add_time) - now;
		if (tmo <= 0) {
			warn = 1;
			tmo = XFRM_KM_TIMEOUT;
		}
		if (tmo < next)
			next = tmo;
	}

	if (warn)
302
		km_policy_expired(xp, dir, 0, 0);
L
Linus Torvalds 已提交
303 304 305 306 307 308 309 310 311 312 313
	if (next != LONG_MAX &&
	    !mod_timer(&xp->timer, jiffies + make_jiffies(next)))
		xfrm_pol_hold(xp);

out:
	read_unlock(&xp->lock);
	xfrm_pol_put(xp);
	return;

expired:
	read_unlock(&xp->lock);
314
	if (!xfrm_policy_delete(xp, dir))
315
		km_policy_expired(xp, dir, 1, 0);
L
Linus Torvalds 已提交
316 317 318 319 320 321 322 323
	xfrm_pol_put(xp);
}


/* Allocate xfrm_policy. Not used here, it is supposed to be used by pfkeyv2
 * SPD calls.
 */

A
Al Viro 已提交
324
struct xfrm_policy *xfrm_policy_alloc(gfp_t gfp)
L
Linus Torvalds 已提交
325 326 327
{
	struct xfrm_policy *policy;

328
	policy = kzalloc(sizeof(struct xfrm_policy), gfp);
L
Linus Torvalds 已提交
329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344

	if (policy) {
		atomic_set(&policy->refcnt, 1);
		rwlock_init(&policy->lock);
		init_timer(&policy->timer);
		policy->timer.data = (unsigned long)policy;
		policy->timer.function = xfrm_policy_timer;
	}
	return policy;
}
EXPORT_SYMBOL(xfrm_policy_alloc);

/* Destroy xfrm_policy: descendant resources must be released to this moment. */

void __xfrm_policy_destroy(struct xfrm_policy *policy)
{
345
	BUG_ON(!policy->dead);
L
Linus Torvalds 已提交
346

347
	BUG_ON(policy->bundles);
L
Linus Torvalds 已提交
348 349 350 351

	if (del_timer(&policy->timer))
		BUG();

352
	security_xfrm_policy_free(policy);
L
Linus Torvalds 已提交
353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417
	kfree(policy);
}
EXPORT_SYMBOL(__xfrm_policy_destroy);

static void xfrm_policy_gc_kill(struct xfrm_policy *policy)
{
	struct dst_entry *dst;

	while ((dst = policy->bundles) != NULL) {
		policy->bundles = dst->next;
		dst_free(dst);
	}

	if (del_timer(&policy->timer))
		atomic_dec(&policy->refcnt);

	if (atomic_read(&policy->refcnt) > 1)
		flow_cache_flush();

	xfrm_pol_put(policy);
}

static void xfrm_policy_gc_task(void *data)
{
	struct xfrm_policy *policy;
	struct list_head *entry, *tmp;
	struct list_head gc_list = LIST_HEAD_INIT(gc_list);

	spin_lock_bh(&xfrm_policy_gc_lock);
	list_splice_init(&xfrm_policy_gc_list, &gc_list);
	spin_unlock_bh(&xfrm_policy_gc_lock);

	list_for_each_safe(entry, tmp, &gc_list) {
		policy = list_entry(entry, struct xfrm_policy, list);
		xfrm_policy_gc_kill(policy);
	}
}

/* Rule must be locked. Release descentant resources, announce
 * entry dead. The rule must be unlinked from lists to the moment.
 */

static void xfrm_policy_kill(struct xfrm_policy *policy)
{
	int dead;

	write_lock_bh(&policy->lock);
	dead = policy->dead;
	policy->dead = 1;
	write_unlock_bh(&policy->lock);

	if (unlikely(dead)) {
		WARN_ON(1);
		return;
	}

	spin_lock(&xfrm_policy_gc_lock);
	list_add(&policy->list, &xfrm_policy_gc_list);
	spin_unlock(&xfrm_policy_gc_lock);

	schedule_work(&xfrm_policy_gc_work);
}

/* Generate new index... KAME seems to generate them ordered by cost
 * of an absolute inpredictability of ordering of rules. This will not pass. */
418
static u32 xfrm_gen_index(u8 type, int dir)
L
Linus Torvalds 已提交
419 420 421 422 423 424 425 426 427 428
{
	u32 idx;
	struct xfrm_policy *p;
	static u32 idx_generator;

	for (;;) {
		idx = (idx_generator | dir);
		idx_generator += 8;
		if (idx == 0)
			idx = 8;
429
		for (p = XFRM_POLICY_LISTHEAD(type, dir); p; p = p->next) {
L
Linus Torvalds 已提交
430 431 432 433 434 435 436 437 438 439 440 441 442
			if (p->index == idx)
				break;
		}
		if (!p)
			return idx;
	}
}

int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl)
{
	struct xfrm_policy *pol, **p;
	struct xfrm_policy *delpol = NULL;
	struct xfrm_policy **newpos = NULL;
443
	struct dst_entry *gc_list;
L
Linus Torvalds 已提交
444 445

	write_lock_bh(&xfrm_policy_lock);
446
	for (p = XFRM_POLICY_LISTHEADP(policy->type, dir); (pol=*p)!=NULL;) {
447 448
		if (!delpol && memcmp(&policy->selector, &pol->selector, sizeof(pol->selector)) == 0 &&
		    xfrm_sec_ctx_match(pol->security, policy->security)) {
L
Linus Torvalds 已提交
449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472
			if (excl) {
				write_unlock_bh(&xfrm_policy_lock);
				return -EEXIST;
			}
			*p = pol->next;
			delpol = pol;
			if (policy->priority > pol->priority)
				continue;
		} else if (policy->priority >= pol->priority) {
			p = &pol->next;
			continue;
		}
		if (!newpos)
			newpos = p;
		if (delpol)
			break;
		p = &pol->next;
	}
	if (newpos)
		p = newpos;
	xfrm_pol_hold(policy);
	policy->next = *p;
	*p = policy;
	atomic_inc(&flow_cache_genid);
473
	policy->index = delpol ? delpol->index : xfrm_gen_index(policy->type, dir);
L
Linus Torvalds 已提交
474 475 476 477 478 479
	policy->curlft.add_time = (unsigned long)xtime.tv_sec;
	policy->curlft.use_time = 0;
	if (!mod_timer(&policy->timer, jiffies + HZ))
		xfrm_pol_hold(policy);
	write_unlock_bh(&xfrm_policy_lock);

480
	if (delpol)
L
Linus Torvalds 已提交
481
		xfrm_policy_kill(delpol);
482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499

	read_lock_bh(&xfrm_policy_lock);
	gc_list = NULL;
	for (policy = policy->next; policy; policy = policy->next) {
		struct dst_entry *dst;

		write_lock(&policy->lock);
		dst = policy->bundles;
		if (dst) {
			struct dst_entry *tail = dst;
			while (tail->next)
				tail = tail->next;
			tail->next = gc_list;
			gc_list = dst;

			policy->bundles = NULL;
		}
		write_unlock(&policy->lock);
L
Linus Torvalds 已提交
500
	}
501 502 503 504 505 506 507 508 509
	read_unlock_bh(&xfrm_policy_lock);

	while (gc_list) {
		struct dst_entry *dst = gc_list;

		gc_list = dst->next;
		dst_free(dst);
	}

L
Linus Torvalds 已提交
510 511 512 513
	return 0;
}
EXPORT_SYMBOL(xfrm_policy_insert);

514 515
struct xfrm_policy *xfrm_policy_bysel_ctx(u8 type, int dir,
					  struct xfrm_selector *sel,
516
					  struct xfrm_sec_ctx *ctx, int delete)
L
Linus Torvalds 已提交
517 518 519 520
{
	struct xfrm_policy *pol, **p;

	write_lock_bh(&xfrm_policy_lock);
521
	for (p = XFRM_POLICY_LISTHEADP(type, dir); (pol=*p)!=NULL; p = &pol->next) {
522 523
		if ((memcmp(sel, &pol->selector, sizeof(*sel)) == 0) &&
		    (xfrm_sec_ctx_match(ctx, pol->security))) {
L
Linus Torvalds 已提交
524 525 526 527 528 529 530 531 532 533 534 535 536 537
			xfrm_pol_hold(pol);
			if (delete)
				*p = pol->next;
			break;
		}
	}
	write_unlock_bh(&xfrm_policy_lock);

	if (pol && delete) {
		atomic_inc(&flow_cache_genid);
		xfrm_policy_kill(pol);
	}
	return pol;
}
538
EXPORT_SYMBOL(xfrm_policy_bysel_ctx);
L
Linus Torvalds 已提交
539

540
struct xfrm_policy *xfrm_policy_byid(u8 type, int dir, u32 id, int delete)
L
Linus Torvalds 已提交
541 542 543 544
{
	struct xfrm_policy *pol, **p;

	write_lock_bh(&xfrm_policy_lock);
545
	for (p = XFRM_POLICY_LISTHEADP(type, dir); (pol=*p)!=NULL; p = &pol->next) {
L
Linus Torvalds 已提交
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562
		if (pol->index == id) {
			xfrm_pol_hold(pol);
			if (delete)
				*p = pol->next;
			break;
		}
	}
	write_unlock_bh(&xfrm_policy_lock);

	if (pol && delete) {
		atomic_inc(&flow_cache_genid);
		xfrm_policy_kill(pol);
	}
	return pol;
}
EXPORT_SYMBOL(xfrm_policy_byid);

563
void xfrm_policy_flush(u8 type)
L
Linus Torvalds 已提交
564 565
{
	struct xfrm_policy *xp;
566
	struct xfrm_policy **p_list = XFRM_POLICY_LISTS(type);
L
Linus Torvalds 已提交
567 568 569 570
	int dir;

	write_lock_bh(&xfrm_policy_lock);
	for (dir = 0; dir < XFRM_POLICY_MAX; dir++) {
571 572
		while ((xp = p_list[dir]) != NULL) {
			p_list[dir] = xp->next;
L
Linus Torvalds 已提交
573 574 575 576 577 578 579 580 581 582 583 584
			write_unlock_bh(&xfrm_policy_lock);

			xfrm_policy_kill(xp);

			write_lock_bh(&xfrm_policy_lock);
		}
	}
	atomic_inc(&flow_cache_genid);
	write_unlock_bh(&xfrm_policy_lock);
}
EXPORT_SYMBOL(xfrm_policy_flush);

585
int xfrm_policy_walk(u8 type, int (*func)(struct xfrm_policy *, int, int, void*),
L
Linus Torvalds 已提交
586 587 588 589 590 591 592 593 594
		     void *data)
{
	struct xfrm_policy *xp;
	int dir;
	int count = 0;
	int error = 0;

	read_lock_bh(&xfrm_policy_lock);
	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
595
		for (xp = XFRM_POLICY_LISTHEAD(type, dir); xp; xp = xp->next)
L
Linus Torvalds 已提交
596 597 598 599 600 601 602 603 604
			count++;
	}

	if (count == 0) {
		error = -ENOENT;
		goto out;
	}

	for (dir = 0; dir < 2*XFRM_POLICY_MAX; dir++) {
605
		for (xp = XFRM_POLICY_LISTHEAD(type, dir); xp; xp = xp->next) {
L
Linus Torvalds 已提交
606 607 608 609 610 611 612 613 614 615 616 617 618 619
			error = func(xp, dir%XFRM_POLICY_MAX, --count, data);
			if (error)
				goto out;
		}
	}

out:
	read_unlock_bh(&xfrm_policy_lock);
	return error;
}
EXPORT_SYMBOL(xfrm_policy_walk);

/* Find policy to apply to this flow. */

620 621
static struct xfrm_policy *xfrm_policy_lookup_bytype(u8 type, struct flowi *fl,
						     u16 family, u8 dir)
L
Linus Torvalds 已提交
622 623 624 625
{
	struct xfrm_policy *pol;

	read_lock_bh(&xfrm_policy_lock);
626
	for (pol = XFRM_POLICY_LISTHEAD(type, dir); pol; pol = pol->next) {
L
Linus Torvalds 已提交
627 628 629 630 631 632 633
		struct xfrm_selector *sel = &pol->selector;
		int match;

		if (pol->family != family)
			continue;

		match = xfrm_selector_match(sel, fl, family);
634

L
Linus Torvalds 已提交
635
		if (match) {
636
 			if (!security_xfrm_policy_lookup(pol, fl->secid, dir)) {
637 638 639
				xfrm_pol_hold(pol);
				break;
			}
L
Linus Torvalds 已提交
640 641 642
		}
	}
	read_unlock_bh(&xfrm_policy_lock);
643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661

	return pol;
}

static void xfrm_policy_lookup(struct flowi *fl, u16 family, u8 dir,
			       void **objp, atomic_t **obj_refp)
{
	struct xfrm_policy *pol;

#ifdef CONFIG_XFRM_SUB_POLICY
	pol = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_SUB, fl, family, dir);
	if (pol)
		goto end;
#endif
	pol = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN, fl, family, dir);

#ifdef CONFIG_XFRM_SUB_POLICY
 end:
#endif
L
Linus Torvalds 已提交
662 663 664 665
	if ((*objp = (void *) pol) != NULL)
		*obj_refp = &pol->refcnt;
}

666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
static inline int policy_to_flow_dir(int dir)
{
	if (XFRM_POLICY_IN == FLOW_DIR_IN &&
 	    XFRM_POLICY_OUT == FLOW_DIR_OUT &&
 	    XFRM_POLICY_FWD == FLOW_DIR_FWD)
 		return dir;
 	switch (dir) {
 	default:
 	case XFRM_POLICY_IN:
 		return FLOW_DIR_IN;
 	case XFRM_POLICY_OUT:
 		return FLOW_DIR_OUT;
 	case XFRM_POLICY_FWD:
 		return FLOW_DIR_FWD;
	};
}

683
static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, struct flowi *fl)
L
Linus Torvalds 已提交
684 685 686 687 688
{
	struct xfrm_policy *pol;

	read_lock_bh(&xfrm_policy_lock);
	if ((pol = sk->sk_policy[dir]) != NULL) {
689
 		int match = xfrm_selector_match(&pol->selector, fl,
L
Linus Torvalds 已提交
690
						sk->sk_family);
691 692
 		int err = 0;

L
Linus Torvalds 已提交
693
		if (match)
694
		  err = security_xfrm_policy_lookup(pol, fl->secid, policy_to_flow_dir(dir));
695 696

 		if (match && !err)
L
Linus Torvalds 已提交
697 698 699 700 701 702 703 704 705 706
			xfrm_pol_hold(pol);
		else
			pol = NULL;
	}
	read_unlock_bh(&xfrm_policy_lock);
	return pol;
}

static void __xfrm_policy_link(struct xfrm_policy *pol, int dir)
{
707 708 709 710
	struct xfrm_policy **p_list = XFRM_POLICY_LISTS(pol->type);

	pol->next = p_list[dir];
	p_list[dir] = pol;
L
Linus Torvalds 已提交
711 712 713 714 715 716 717 718
	xfrm_pol_hold(pol);
}

static struct xfrm_policy *__xfrm_policy_unlink(struct xfrm_policy *pol,
						int dir)
{
	struct xfrm_policy **polp;

719
	for (polp = XFRM_POLICY_LISTHEADP(pol->type, dir);
L
Linus Torvalds 已提交
720 721 722 723 724 725 726 727 728
	     *polp != NULL; polp = &(*polp)->next) {
		if (*polp == pol) {
			*polp = pol->next;
			return pol;
		}
	}
	return NULL;
}

729
int xfrm_policy_delete(struct xfrm_policy *pol, int dir)
L
Linus Torvalds 已提交
730 731 732 733 734 735 736 737
{
	write_lock_bh(&xfrm_policy_lock);
	pol = __xfrm_policy_unlink(pol, dir);
	write_unlock_bh(&xfrm_policy_lock);
	if (pol) {
		if (dir < XFRM_POLICY_MAX)
			atomic_inc(&flow_cache_genid);
		xfrm_policy_kill(pol);
738
		return 0;
L
Linus Torvalds 已提交
739
	}
740
	return -ENOENT;
L
Linus Torvalds 已提交
741
}
742
EXPORT_SYMBOL(xfrm_policy_delete);
L
Linus Torvalds 已提交
743 744 745 746 747

int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol)
{
	struct xfrm_policy *old_pol;

748 749 750 751 752
#ifdef CONFIG_XFRM_SUB_POLICY
	if (pol && pol->type != XFRM_POLICY_TYPE_MAIN)
		return -EINVAL;
#endif

L
Linus Torvalds 已提交
753 754 755 756 757
	write_lock_bh(&xfrm_policy_lock);
	old_pol = sk->sk_policy[dir];
	sk->sk_policy[dir] = pol;
	if (pol) {
		pol->curlft.add_time = (unsigned long)xtime.tv_sec;
758
		pol->index = xfrm_gen_index(pol->type, XFRM_POLICY_MAX+dir);
L
Linus Torvalds 已提交
759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776
		__xfrm_policy_link(pol, XFRM_POLICY_MAX+dir);
	}
	if (old_pol)
		__xfrm_policy_unlink(old_pol, XFRM_POLICY_MAX+dir);
	write_unlock_bh(&xfrm_policy_lock);

	if (old_pol) {
		xfrm_policy_kill(old_pol);
	}
	return 0;
}

static struct xfrm_policy *clone_policy(struct xfrm_policy *old, int dir)
{
	struct xfrm_policy *newp = xfrm_policy_alloc(GFP_ATOMIC);

	if (newp) {
		newp->selector = old->selector;
777 778 779 780
		if (security_xfrm_policy_clone(old, newp)) {
			kfree(newp);
			return NULL;  /* ENOMEM */
		}
L
Linus Torvalds 已提交
781 782 783 784 785 786
		newp->lft = old->lft;
		newp->curlft = old->curlft;
		newp->action = old->action;
		newp->flags = old->flags;
		newp->xfrm_nr = old->xfrm_nr;
		newp->index = old->index;
787
		newp->type = old->type;
L
Linus Torvalds 已提交
788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813
		memcpy(newp->xfrm_vec, old->xfrm_vec,
		       newp->xfrm_nr*sizeof(struct xfrm_tmpl));
		write_lock_bh(&xfrm_policy_lock);
		__xfrm_policy_link(newp, XFRM_POLICY_MAX+dir);
		write_unlock_bh(&xfrm_policy_lock);
		xfrm_pol_put(newp);
	}
	return newp;
}

int __xfrm_sk_clone_policy(struct sock *sk)
{
	struct xfrm_policy *p0 = sk->sk_policy[0],
			   *p1 = sk->sk_policy[1];

	sk->sk_policy[0] = sk->sk_policy[1] = NULL;
	if (p0 && (sk->sk_policy[0] = clone_policy(p0, 0)) == NULL)
		return -ENOMEM;
	if (p1 && (sk->sk_policy[1] = clone_policy(p1, 1)) == NULL)
		return -ENOMEM;
	return 0;
}

/* Resolve list of templates for the flow, given policy. */

static int
814 815 816
xfrm_tmpl_resolve_one(struct xfrm_policy *policy, struct flowi *fl,
		      struct xfrm_state **xfrm,
		      unsigned short family)
L
Linus Torvalds 已提交
817 818 819 820 821 822 823 824 825 826 827 828
{
	int nx;
	int i, error;
	xfrm_address_t *daddr = xfrm_flowi_daddr(fl, family);
	xfrm_address_t *saddr = xfrm_flowi_saddr(fl, family);

	for (nx=0, i = 0; i < policy->xfrm_nr; i++) {
		struct xfrm_state *x;
		xfrm_address_t *remote = daddr;
		xfrm_address_t *local  = saddr;
		struct xfrm_tmpl *tmpl = &policy->xfrm_vec[i];

829
		if (tmpl->mode == XFRM_MODE_TUNNEL) {
L
Linus Torvalds 已提交
830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858
			remote = &tmpl->id.daddr;
			local = &tmpl->saddr;
		}

		x = xfrm_state_find(remote, local, fl, tmpl, policy, &error, family);

		if (x && x->km.state == XFRM_STATE_VALID) {
			xfrm[nx++] = x;
			daddr = remote;
			saddr = local;
			continue;
		}
		if (x) {
			error = (x->km.state == XFRM_STATE_ERROR ?
				 -EINVAL : -EAGAIN);
			xfrm_state_put(x);
		}

		if (!tmpl->optional)
			goto fail;
	}
	return nx;

fail:
	for (nx--; nx>=0; nx--)
		xfrm_state_put(xfrm[nx]);
	return error;
}

859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890
static int
xfrm_tmpl_resolve(struct xfrm_policy **pols, int npols, struct flowi *fl,
		  struct xfrm_state **xfrm,
		  unsigned short family)
{
	int cnx = 0;
	int error;
	int ret;
	int i;

	for (i = 0; i < npols; i++) {
		if (cnx + pols[i]->xfrm_nr >= XFRM_MAX_DEPTH) {
			error = -ENOBUFS;
			goto fail;
		}
		ret = xfrm_tmpl_resolve_one(pols[i], fl, &xfrm[cnx], family);
		if (ret < 0) {
			error = ret;
			goto fail;
		} else
			cnx += ret;
	}

	return cnx;

 fail:
	for (cnx--; cnx>=0; cnx--)
		xfrm_state_put(xfrm[cnx]);
	return error;

}

L
Linus Torvalds 已提交
891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936
/* Check that the bundle accepts the flow and its components are
 * still valid.
 */

static struct dst_entry *
xfrm_find_bundle(struct flowi *fl, struct xfrm_policy *policy, unsigned short family)
{
	struct dst_entry *x;
	struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
	if (unlikely(afinfo == NULL))
		return ERR_PTR(-EINVAL);
	x = afinfo->find_bundle(fl, policy);
	xfrm_policy_put_afinfo(afinfo);
	return x;
}

/* Allocate chain of dst_entry's, attach known xfrm's, calculate
 * all the metrics... Shortly, bundle a bundle.
 */

static int
xfrm_bundle_create(struct xfrm_policy *policy, struct xfrm_state **xfrm, int nx,
		   struct flowi *fl, struct dst_entry **dst_p,
		   unsigned short family)
{
	int err;
	struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
	if (unlikely(afinfo == NULL))
		return -EINVAL;
	err = afinfo->bundle_create(policy, xfrm, nx, fl, dst_p);
	xfrm_policy_put_afinfo(afinfo);
	return err;
}


static int stale_bundle(struct dst_entry *dst);

/* Main function: finds/creates a bundle for given flow.
 *
 * At the moment we eat a raw IP route. Mostly to speed up lookups
 * on interfaces with disabled IPsec.
 */
int xfrm_lookup(struct dst_entry **dst_p, struct flowi *fl,
		struct sock *sk, int flags)
{
	struct xfrm_policy *policy;
937 938 939 940 941
	struct xfrm_policy *pols[XFRM_POLICY_TYPE_MAX];
	int npols;
	int pol_dead;
	int xfrm_nr;
	int pi;
L
Linus Torvalds 已提交
942 943 944 945 946
	struct xfrm_state *xfrm[XFRM_MAX_DEPTH];
	struct dst_entry *dst, *dst_orig = *dst_p;
	int nx = 0;
	int err;
	u32 genid;
947
	u16 family;
948
	u8 dir = policy_to_flow_dir(XFRM_POLICY_OUT);
949

L
Linus Torvalds 已提交
950 951 952
restart:
	genid = atomic_read(&flow_cache_genid);
	policy = NULL;
953 954 955 956 957 958
	for (pi = 0; pi < ARRAY_SIZE(pols); pi++)
		pols[pi] = NULL;
	npols = 0;
	pol_dead = 0;
	xfrm_nr = 0;

L
Linus Torvalds 已提交
959
	if (sk && sk->sk_policy[1])
960
		policy = xfrm_sk_policy_lookup(sk, XFRM_POLICY_OUT, fl);
L
Linus Torvalds 已提交
961 962 963

	if (!policy) {
		/* To accelerate a bit...  */
964
		if ((dst_orig->flags & DST_NOXFRM) || xfrm_policy_lists_empty(XFRM_POLICY_OUT))
L
Linus Torvalds 已提交
965 966
			return 0;

967
		policy = flow_cache_lookup(fl, dst_orig->ops->family,
968
					   dir, xfrm_policy_lookup);
L
Linus Torvalds 已提交
969 970 971 972 973
	}

	if (!policy)
		return 0;

974
	family = dst_orig->ops->family;
L
Linus Torvalds 已提交
975
	policy->curlft.use_time = (unsigned long)xtime.tv_sec;
976 977 978
	pols[0] = policy;
	npols ++;
	xfrm_nr += pols[0]->xfrm_nr;
L
Linus Torvalds 已提交
979 980 981 982

	switch (policy->action) {
	case XFRM_POLICY_BLOCK:
		/* Prohibit the flow */
983 984
		err = -EPERM;
		goto error;
L
Linus Torvalds 已提交
985 986

	case XFRM_POLICY_ALLOW:
987
#ifndef CONFIG_XFRM_SUB_POLICY
L
Linus Torvalds 已提交
988 989 990 991 992
		if (policy->xfrm_nr == 0) {
			/* Flow passes not transformed. */
			xfrm_pol_put(policy);
			return 0;
		}
993
#endif
L
Linus Torvalds 已提交
994 995 996 997 998 999 1000 1001

		/* Try to find matching bundle.
		 *
		 * LATER: help from flow cache. It is optional, this
		 * is required only for output policy.
		 */
		dst = xfrm_find_bundle(fl, policy, family);
		if (IS_ERR(dst)) {
1002 1003
			err = PTR_ERR(dst);
			goto error;
L
Linus Torvalds 已提交
1004 1005 1006 1007 1008
		}

		if (dst)
			break;

1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038
#ifdef CONFIG_XFRM_SUB_POLICY
		if (pols[0]->type != XFRM_POLICY_TYPE_MAIN) {
			pols[1] = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN,
							    fl, family,
							    XFRM_POLICY_OUT);
			if (pols[1]) {
				if (pols[1]->action == XFRM_POLICY_BLOCK) {
					err = -EPERM;
					goto error;
				}
				npols ++;
				xfrm_nr += pols[1]->xfrm_nr;
			}
		}

		/*
		 * Because neither flowi nor bundle information knows about
		 * transformation template size. On more than one policy usage
		 * we can realize whether all of them is bypass or not after
		 * they are searched. See above not-transformed bypass
		 * is surrounded by non-sub policy configuration, too.
		 */
		if (xfrm_nr == 0) {
			/* Flow passes not transformed. */
			xfrm_pols_put(pols, npols);
			return 0;
		}

#endif
		nx = xfrm_tmpl_resolve(pols, npols, fl, xfrm, family);
L
Linus Torvalds 已提交
1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050

		if (unlikely(nx<0)) {
			err = nx;
			if (err == -EAGAIN && flags) {
				DECLARE_WAITQUEUE(wait, current);

				add_wait_queue(&km_waitq, &wait);
				set_current_state(TASK_INTERRUPTIBLE);
				schedule();
				set_current_state(TASK_RUNNING);
				remove_wait_queue(&km_waitq, &wait);

1051
				nx = xfrm_tmpl_resolve(pols, npols, fl, xfrm, family);
L
Linus Torvalds 已提交
1052 1053 1054 1055 1056 1057 1058

				if (nx == -EAGAIN && signal_pending(current)) {
					err = -ERESTART;
					goto error;
				}
				if (nx == -EAGAIN ||
				    genid != atomic_read(&flow_cache_genid)) {
1059
					xfrm_pols_put(pols, npols);
L
Linus Torvalds 已提交
1060 1061 1062 1063 1064 1065 1066 1067 1068
					goto restart;
				}
				err = nx;
			}
			if (err < 0)
				goto error;
		}
		if (nx == 0) {
			/* Flow passes not transformed. */
1069
			xfrm_pols_put(pols, npols);
L
Linus Torvalds 已提交
1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082
			return 0;
		}

		dst = dst_orig;
		err = xfrm_bundle_create(policy, xfrm, nx, fl, &dst, family);

		if (unlikely(err)) {
			int i;
			for (i=0; i<nx; i++)
				xfrm_state_put(xfrm[i]);
			goto error;
		}

1083 1084 1085 1086 1087 1088
		for (pi = 0; pi < npols; pi++) {
			read_lock_bh(&pols[pi]->lock);
			pol_dead |= pols[pi]->dead;
			read_unlock_bh(&pols[pi]->lock);
		}

L
Linus Torvalds 已提交
1089
		write_lock_bh(&policy->lock);
1090
		if (unlikely(pol_dead || stale_bundle(dst))) {
L
Linus Torvalds 已提交
1091 1092 1093 1094 1095 1096 1097 1098
			/* Wow! While we worked on resolving, this
			 * policy has gone. Retry. It is not paranoia,
			 * we just cannot enlist new bundle to dead object.
			 * We can't enlist stable bundles either.
			 */
			write_unlock_bh(&policy->lock);
			if (dst)
				dst_free(dst);
H
Herbert Xu 已提交
1099 1100 1101

			err = -EHOSTUNREACH;
			goto error;
L
Linus Torvalds 已提交
1102 1103 1104 1105 1106 1107 1108 1109
		}
		dst->next = policy->bundles;
		policy->bundles = dst;
		dst_hold(dst);
		write_unlock_bh(&policy->lock);
	}
	*dst_p = dst;
	dst_release(dst_orig);
1110
 	xfrm_pols_put(pols, npols);
L
Linus Torvalds 已提交
1111 1112 1113 1114
	return 0;

error:
	dst_release(dst_orig);
1115
	xfrm_pols_put(pols, npols);
L
Linus Torvalds 已提交
1116 1117 1118 1119 1120
	*dst_p = NULL;
	return err;
}
EXPORT_SYMBOL(xfrm_lookup);

1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137
static inline int
xfrm_secpath_reject(int idx, struct sk_buff *skb, struct flowi *fl)
{
	struct xfrm_state *x;
	int err;

	if (!skb->sp || idx < 0 || idx >= skb->sp->len)
		return 0;
	x = skb->sp->xvec[idx];
	if (!x->type->reject)
		return 0;
	xfrm_state_hold(x);
	err = x->type->reject(x, skb, fl);
	xfrm_state_put(x);
	return err;
}

L
Linus Torvalds 已提交
1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153
/* When skb is transformed back to its "native" form, we have to
 * check policy restrictions. At the moment we make this in maximally
 * stupid way. Shame on me. :-) Of course, connected sockets must
 * have policy cached at them.
 */

static inline int
xfrm_state_ok(struct xfrm_tmpl *tmpl, struct xfrm_state *x, 
	      unsigned short family)
{
	if (xfrm_state_kern(x))
		return tmpl->optional && !xfrm_state_addr_cmp(tmpl, x, family);
	return	x->id.proto == tmpl->id.proto &&
		(x->id.spi == tmpl->id.spi || !tmpl->id.spi) &&
		(x->props.reqid == tmpl->reqid || !tmpl->reqid) &&
		x->props.mode == tmpl->mode &&
1154 1155
		((tmpl->aalgos & (1<<x->props.aalgo)) ||
		 !(xfrm_id_proto_match(tmpl->id.proto, IPSEC_PROTO_ANY))) &&
1156 1157
		!(x->props.mode != XFRM_MODE_TRANSPORT &&
		  xfrm_state_addr_cmp(tmpl, x, family));
L
Linus Torvalds 已提交
1158 1159
}

1160 1161 1162 1163 1164 1165 1166
/*
 * 0 or more than 0 is returned when validation is succeeded (either bypass
 * because of optional transport mode, or next index of the mathced secpath
 * state with the template.
 * -1 is returned when no matching template is found.
 * Otherwise "-2 - errored_index" is returned.
 */
L
Linus Torvalds 已提交
1167 1168 1169 1170 1171 1172 1173
static inline int
xfrm_policy_ok(struct xfrm_tmpl *tmpl, struct sec_path *sp, int start,
	       unsigned short family)
{
	int idx = start;

	if (tmpl->optional) {
1174
		if (tmpl->mode == XFRM_MODE_TRANSPORT)
L
Linus Torvalds 已提交
1175 1176 1177 1178
			return start;
	} else
		start = -1;
	for (; idx < sp->len; idx++) {
1179
		if (xfrm_state_ok(tmpl, sp->xvec[idx], family))
L
Linus Torvalds 已提交
1180
			return ++idx;
1181 1182 1183
		if (sp->xvec[idx]->props.mode != XFRM_MODE_TRANSPORT) {
			if (start == -1)
				start = -2-idx;
L
Linus Torvalds 已提交
1184
			break;
1185
		}
L
Linus Torvalds 已提交
1186 1187 1188 1189
	}
	return start;
}

1190 1191
int
xfrm_decode_session(struct sk_buff *skb, struct flowi *fl, unsigned short family)
L
Linus Torvalds 已提交
1192 1193
{
	struct xfrm_policy_afinfo *afinfo = xfrm_policy_get_afinfo(family);
1194
	int err;
L
Linus Torvalds 已提交
1195 1196 1197 1198 1199

	if (unlikely(afinfo == NULL))
		return -EAFNOSUPPORT;

	afinfo->decode_session(skb, fl);
V
Venkat Yekkirala 已提交
1200
	err = security_xfrm_decode_session(skb, &fl->secid);
L
Linus Torvalds 已提交
1201
	xfrm_policy_put_afinfo(afinfo);
1202
	return err;
L
Linus Torvalds 已提交
1203
}
1204
EXPORT_SYMBOL(xfrm_decode_session);
L
Linus Torvalds 已提交
1205

1206
static inline int secpath_has_nontransport(struct sec_path *sp, int k, int *idxp)
L
Linus Torvalds 已提交
1207 1208
{
	for (; k < sp->len; k++) {
1209 1210 1211
		if (sp->xvec[k]->props.mode != XFRM_MODE_TRANSPORT) {
			if (idxp)
				*idxp = k;
L
Linus Torvalds 已提交
1212
			return 1;
1213
		}
L
Linus Torvalds 已提交
1214 1215 1216 1217 1218 1219 1220 1221 1222
	}

	return 0;
}

int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb, 
			unsigned short family)
{
	struct xfrm_policy *pol;
1223 1224 1225 1226
	struct xfrm_policy *pols[XFRM_POLICY_TYPE_MAX];
	int npols = 0;
	int xfrm_nr;
	int pi;
L
Linus Torvalds 已提交
1227
	struct flowi fl;
1228
	u8 fl_dir = policy_to_flow_dir(dir);
1229 1230
	int xerr_idx = -1;
	int *xerr_idxp = &xerr_idx;
L
Linus Torvalds 已提交
1231

1232
	if (xfrm_decode_session(skb, &fl, family) < 0)
L
Linus Torvalds 已提交
1233
		return 0;
1234
	nf_nat_decode_session(skb, &fl, family);
L
Linus Torvalds 已提交
1235 1236 1237 1238 1239 1240

	/* First, check used SA against their selectors. */
	if (skb->sp) {
		int i;

		for (i=skb->sp->len-1; i>=0; i--) {
1241 1242
			struct xfrm_state *x = skb->sp->xvec[i];
			if (!xfrm_selector_match(&x->sel, &fl, family))
L
Linus Torvalds 已提交
1243 1244 1245 1246 1247 1248
				return 0;
		}
	}

	pol = NULL;
	if (sk && sk->sk_policy[dir])
1249
		pol = xfrm_sk_policy_lookup(sk, dir, &fl);
L
Linus Torvalds 已提交
1250 1251

	if (!pol)
1252
		pol = flow_cache_lookup(&fl, family, fl_dir,
L
Linus Torvalds 已提交
1253 1254
					xfrm_policy_lookup);

1255 1256 1257 1258 1259 1260 1261
	if (!pol) {
		if (skb->sp && secpath_has_nontransport(skb->sp, 0, xerr_idxp)) {
			xfrm_secpath_reject(xerr_idx, skb, &fl);
			return 0;
		}
		return 1;
	}
L
Linus Torvalds 已提交
1262 1263 1264

	pol->curlft.use_time = (unsigned long)xtime.tv_sec;

1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278
	pols[0] = pol;
	npols ++;
#ifdef CONFIG_XFRM_SUB_POLICY
	if (pols[0]->type != XFRM_POLICY_TYPE_MAIN) {
		pols[1] = xfrm_policy_lookup_bytype(XFRM_POLICY_TYPE_MAIN,
						    &fl, family,
						    XFRM_POLICY_IN);
		if (pols[1]) {
			pols[1]->curlft.use_time = (unsigned long)xtime.tv_sec;
			npols ++;
		}
	}
#endif

L
Linus Torvalds 已提交
1279 1280 1281
	if (pol->action == XFRM_POLICY_ALLOW) {
		struct sec_path *sp;
		static struct sec_path dummy;
1282 1283 1284
		struct xfrm_tmpl *tp[XFRM_MAX_DEPTH];
		struct xfrm_tmpl **tpp = tp;
		int ti = 0;
L
Linus Torvalds 已提交
1285 1286 1287 1288 1289
		int i, k;

		if ((sp = skb->sp) == NULL)
			sp = &dummy;

1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300
		for (pi = 0; pi < npols; pi++) {
			if (pols[pi] != pol &&
			    pols[pi]->action != XFRM_POLICY_ALLOW)
				goto reject;
			if (ti + pols[pi]->xfrm_nr >= XFRM_MAX_DEPTH)
				goto reject_error;
			for (i = 0; i < pols[pi]->xfrm_nr; i++)
				tpp[ti++] = &pols[pi]->xfrm_vec[i];
		}
		xfrm_nr = ti;

L
Linus Torvalds 已提交
1301 1302 1303 1304 1305 1306
		/* For each tunnel xfrm, find the first matching tmpl.
		 * For each tmpl before that, find corresponding xfrm.
		 * Order is _important_. Later we will implement
		 * some barriers, but at the moment barriers
		 * are implied between each two transformations.
		 */
1307 1308
		for (i = xfrm_nr-1, k = 0; i >= 0; i--) {
			k = xfrm_policy_ok(tpp[i], sp, k, family);
1309 1310 1311
			if (k < 0) {
				if (k < -1 && xerr_idxp)
					*xerr_idxp = -(2+k);
L
Linus Torvalds 已提交
1312
				goto reject;
1313
			}
L
Linus Torvalds 已提交
1314 1315
		}

1316
		if (secpath_has_nontransport(sp, k, xerr_idxp))
L
Linus Torvalds 已提交
1317 1318
			goto reject;

1319
		xfrm_pols_put(pols, npols);
L
Linus Torvalds 已提交
1320 1321 1322 1323
		return 1;
	}

reject:
1324
	xfrm_secpath_reject(xerr_idx, skb, &fl);
1325 1326
reject_error:
	xfrm_pols_put(pols, npols);
L
Linus Torvalds 已提交
1327 1328 1329 1330 1331 1332 1333 1334
	return 0;
}
EXPORT_SYMBOL(__xfrm_policy_check);

int __xfrm_route_forward(struct sk_buff *skb, unsigned short family)
{
	struct flowi fl;

1335
	if (xfrm_decode_session(skb, &fl, family) < 0)
L
Linus Torvalds 已提交
1336 1337 1338 1339 1340 1341
		return 0;

	return xfrm_lookup(&skb->dst, &fl, NULL, 0) == 0;
}
EXPORT_SYMBOL(__xfrm_route_forward);

1342 1343
/* Optimize later using cookies and generation ids. */

L
Linus Torvalds 已提交
1344 1345
static struct dst_entry *xfrm_dst_check(struct dst_entry *dst, u32 cookie)
{
1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364
	/* Code (such as __xfrm4_bundle_create()) sets dst->obsolete
	 * to "-1" to force all XFRM destinations to get validated by
	 * dst_ops->check on every use.  We do this because when a
	 * normal route referenced by an XFRM dst is obsoleted we do
	 * not go looking around for all parent referencing XFRM dsts
	 * so that we can invalidate them.  It is just too much work.
	 * Instead we make the checks here on every use.  For example:
	 *
	 *	XFRM dst A --> IPv4 dst X
	 *
	 * X is the "xdst->route" of A (X is also the "dst->path" of A
	 * in this example).  If X is marked obsolete, "A" will not
	 * notice.  That's what we are validating here via the
	 * stale_bundle() check.
	 *
	 * When a policy's bundle is pruned, we dst_free() the XFRM
	 * dst which causes it's ->obsolete field to be set to a
	 * positive non-zero integer.  If an XFRM dst has been pruned
	 * like this, we want to force a new route lookup.
1365
	 */
1366 1367 1368
	if (dst->obsolete < 0 && !stale_bundle(dst))
		return dst;

L
Linus Torvalds 已提交
1369 1370 1371 1372 1373
	return NULL;
}

static int stale_bundle(struct dst_entry *dst)
{
1374
	return !xfrm_bundle_ok((struct xfrm_dst *)dst, NULL, AF_UNSPEC, 0);
L
Linus Torvalds 已提交
1375 1376
}

H
Herbert Xu 已提交
1377
void xfrm_dst_ifdown(struct dst_entry *dst, struct net_device *dev)
L
Linus Torvalds 已提交
1378 1379 1380 1381 1382 1383 1384
{
	while ((dst = dst->child) && dst->xfrm && dst->dev == dev) {
		dst->dev = &loopback_dev;
		dev_hold(&loopback_dev);
		dev_put(dev);
	}
}
H
Herbert Xu 已提交
1385
EXPORT_SYMBOL(xfrm_dst_ifdown);
L
Linus Torvalds 已提交
1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411

static void xfrm_link_failure(struct sk_buff *skb)
{
	/* Impossible. Such dst must be popped before reaches point of failure. */
	return;
}

static struct dst_entry *xfrm_negative_advice(struct dst_entry *dst)
{
	if (dst) {
		if (dst->obsolete) {
			dst_release(dst);
			dst = NULL;
		}
	}
	return dst;
}

static void xfrm_prune_bundles(int (*func)(struct dst_entry *))
{
	int i;
	struct xfrm_policy *pol;
	struct dst_entry *dst, **dstp, *gc_list = NULL;

	read_lock_bh(&xfrm_policy_lock);
	for (i=0; i<2*XFRM_POLICY_MAX; i++) {
1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428
#ifdef CONFIG_XFRM_SUB_POLICY
		for (pol = xfrm_policy_list_sub[i]; pol; pol = pol->next) {
			write_lock(&pol->lock);
			dstp = &pol->bundles;
			while ((dst=*dstp) != NULL) {
				if (func(dst)) {
					*dstp = dst->next;
					dst->next = gc_list;
					gc_list = dst;
				} else {
					dstp = &dst->next;
				}
			}
			write_unlock(&pol->lock);
		}

#endif
L
Linus Torvalds 已提交
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 1468
		for (pol = xfrm_policy_list[i]; pol; pol = pol->next) {
			write_lock(&pol->lock);
			dstp = &pol->bundles;
			while ((dst=*dstp) != NULL) {
				if (func(dst)) {
					*dstp = dst->next;
					dst->next = gc_list;
					gc_list = dst;
				} else {
					dstp = &dst->next;
				}
			}
			write_unlock(&pol->lock);
		}
	}
	read_unlock_bh(&xfrm_policy_lock);

	while (gc_list) {
		dst = gc_list;
		gc_list = dst->next;
		dst_free(dst);
	}
}

static int unused_bundle(struct dst_entry *dst)
{
	return !atomic_read(&dst->__refcnt);
}

static void __xfrm_garbage_collect(void)
{
	xfrm_prune_bundles(unused_bundle);
}

int xfrm_flush_bundles(void)
{
	xfrm_prune_bundles(stale_bundle);
	return 0;
}

1469 1470 1471 1472 1473 1474 1475 1476 1477 1478
static int always_true(struct dst_entry *dst)
{
	return 1;
}

void xfrm_flush_all_bundles(void)
{
	xfrm_prune_bundles(always_true);
}

L
Linus Torvalds 已提交
1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505
void xfrm_init_pmtu(struct dst_entry *dst)
{
	do {
		struct xfrm_dst *xdst = (struct xfrm_dst *)dst;
		u32 pmtu, route_mtu_cached;

		pmtu = dst_mtu(dst->child);
		xdst->child_mtu_cached = pmtu;

		pmtu = xfrm_state_mtu(dst->xfrm, pmtu);

		route_mtu_cached = dst_mtu(xdst->route);
		xdst->route_mtu_cached = route_mtu_cached;

		if (pmtu > route_mtu_cached)
			pmtu = route_mtu_cached;

		dst->metrics[RTAX_MTU-1] = pmtu;
	} while ((dst = dst->next));
}

EXPORT_SYMBOL(xfrm_init_pmtu);

/* Check that the bundle accepts the flow and its components are
 * still valid.
 */

1506
int xfrm_bundle_ok(struct xfrm_dst *first, struct flowi *fl, int family, int strict)
L
Linus Torvalds 已提交
1507 1508 1509 1510 1511
{
	struct dst_entry *dst = &first->u.dst;
	struct xfrm_dst *last;
	u32 mtu;

1512
	if (!dst_check(dst->path, ((struct xfrm_dst *)dst)->path_cookie) ||
L
Linus Torvalds 已提交
1513 1514 1515 1516 1517 1518 1519 1520 1521 1522
	    (dst->dev && !netif_running(dst->dev)))
		return 0;

	last = NULL;

	do {
		struct xfrm_dst *xdst = (struct xfrm_dst *)dst;

		if (fl && !xfrm_selector_match(&dst->xfrm->sel, fl, family))
			return 0;
1523 1524
		if (fl && !security_xfrm_flow_state_match(fl, dst->xfrm))
			return 0;
L
Linus Torvalds 已提交
1525 1526
		if (dst->xfrm->km.state != XFRM_STATE_VALID)
			return 0;
1527 1528 1529 1530

		if (strict && fl && dst->xfrm->props.mode != XFRM_MODE_TUNNEL &&
		    !xfrm_state_addr_flow_check(dst->xfrm, fl, family))
			return 0;
L
Linus Torvalds 已提交
1531 1532 1533 1534 1535 1536 1537

		mtu = dst_mtu(dst->child);
		if (xdst->child_mtu_cached != mtu) {
			last = xdst;
			xdst->child_mtu_cached = mtu;
		}

1538
		if (!dst_check(xdst->route, xdst->route_cookie))
L
Linus Torvalds 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579
			return 0;
		mtu = dst_mtu(xdst->route);
		if (xdst->route_mtu_cached != mtu) {
			last = xdst;
			xdst->route_mtu_cached = mtu;
		}

		dst = dst->child;
	} while (dst->xfrm);

	if (likely(!last))
		return 1;

	mtu = last->child_mtu_cached;
	for (;;) {
		dst = &last->u.dst;

		mtu = xfrm_state_mtu(dst->xfrm, mtu);
		if (mtu > last->route_mtu_cached)
			mtu = last->route_mtu_cached;
		dst->metrics[RTAX_MTU-1] = mtu;

		if (last == first)
			break;

		last = last->u.next;
		last->child_mtu_cached = mtu;
	}

	return 1;
}

EXPORT_SYMBOL(xfrm_bundle_ok);

int xfrm_policy_register_afinfo(struct xfrm_policy_afinfo *afinfo)
{
	int err = 0;
	if (unlikely(afinfo == NULL))
		return -EINVAL;
	if (unlikely(afinfo->family >= NPROTO))
		return -EAFNOSUPPORT;
1580
	write_lock_bh(&xfrm_policy_afinfo_lock);
L
Linus Torvalds 已提交
1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596
	if (unlikely(xfrm_policy_afinfo[afinfo->family] != NULL))
		err = -ENOBUFS;
	else {
		struct dst_ops *dst_ops = afinfo->dst_ops;
		if (likely(dst_ops->kmem_cachep == NULL))
			dst_ops->kmem_cachep = xfrm_dst_cache;
		if (likely(dst_ops->check == NULL))
			dst_ops->check = xfrm_dst_check;
		if (likely(dst_ops->negative_advice == NULL))
			dst_ops->negative_advice = xfrm_negative_advice;
		if (likely(dst_ops->link_failure == NULL))
			dst_ops->link_failure = xfrm_link_failure;
		if (likely(afinfo->garbage_collect == NULL))
			afinfo->garbage_collect = __xfrm_garbage_collect;
		xfrm_policy_afinfo[afinfo->family] = afinfo;
	}
1597
	write_unlock_bh(&xfrm_policy_afinfo_lock);
L
Linus Torvalds 已提交
1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608
	return err;
}
EXPORT_SYMBOL(xfrm_policy_register_afinfo);

int xfrm_policy_unregister_afinfo(struct xfrm_policy_afinfo *afinfo)
{
	int err = 0;
	if (unlikely(afinfo == NULL))
		return -EINVAL;
	if (unlikely(afinfo->family >= NPROTO))
		return -EAFNOSUPPORT;
1609
	write_lock_bh(&xfrm_policy_afinfo_lock);
L
Linus Torvalds 已提交
1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622
	if (likely(xfrm_policy_afinfo[afinfo->family] != NULL)) {
		if (unlikely(xfrm_policy_afinfo[afinfo->family] != afinfo))
			err = -EINVAL;
		else {
			struct dst_ops *dst_ops = afinfo->dst_ops;
			xfrm_policy_afinfo[afinfo->family] = NULL;
			dst_ops->kmem_cachep = NULL;
			dst_ops->check = NULL;
			dst_ops->negative_advice = NULL;
			dst_ops->link_failure = NULL;
			afinfo->garbage_collect = NULL;
		}
	}
1623
	write_unlock_bh(&xfrm_policy_afinfo_lock);
L
Linus Torvalds 已提交
1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
	return err;
}
EXPORT_SYMBOL(xfrm_policy_unregister_afinfo);

static struct xfrm_policy_afinfo *xfrm_policy_get_afinfo(unsigned short family)
{
	struct xfrm_policy_afinfo *afinfo;
	if (unlikely(family >= NPROTO))
		return NULL;
	read_lock(&xfrm_policy_afinfo_lock);
	afinfo = xfrm_policy_afinfo[family];
1635 1636
	if (unlikely(!afinfo))
		read_unlock(&xfrm_policy_afinfo_lock);
L
Linus Torvalds 已提交
1637 1638 1639 1640 1641
	return afinfo;
}

static void xfrm_policy_put_afinfo(struct xfrm_policy_afinfo *afinfo)
{
1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
	read_unlock(&xfrm_policy_afinfo_lock);
}

static struct xfrm_policy_afinfo *xfrm_policy_lock_afinfo(unsigned int family)
{
	struct xfrm_policy_afinfo *afinfo;
	if (unlikely(family >= NPROTO))
		return NULL;
	write_lock_bh(&xfrm_policy_afinfo_lock);
	afinfo = xfrm_policy_afinfo[family];
	if (unlikely(!afinfo))
		write_unlock_bh(&xfrm_policy_afinfo_lock);
	return afinfo;
}

static void xfrm_policy_unlock_afinfo(struct xfrm_policy_afinfo *afinfo)
{
	write_unlock_bh(&xfrm_policy_afinfo_lock);
L
Linus Torvalds 已提交
1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696
}

static int xfrm_dev_event(struct notifier_block *this, unsigned long event, void *ptr)
{
	switch (event) {
	case NETDEV_DOWN:
		xfrm_flush_bundles();
	}
	return NOTIFY_DONE;
}

static struct notifier_block xfrm_dev_notifier = {
	xfrm_dev_event,
	NULL,
	0
};

static void __init xfrm_policy_init(void)
{
	xfrm_dst_cache = kmem_cache_create("xfrm_dst_cache",
					   sizeof(struct xfrm_dst),
					   0, SLAB_HWCACHE_ALIGN,
					   NULL, NULL);
	if (!xfrm_dst_cache)
		panic("XFRM: failed to allocate xfrm_dst_cache\n");

	INIT_WORK(&xfrm_policy_gc_work, xfrm_policy_gc_task, NULL);
	register_netdevice_notifier(&xfrm_dev_notifier);
}

void __init xfrm_init(void)
{
	xfrm_state_init();
	xfrm_policy_init();
	xfrm_input_init();
}