cls_api.c 39.1 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
/*
 * net/sched/cls_api.c	Packet classifier API.
 *
 *		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.
 *
 * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *
 * Changes:
 *
 * Eduardo J. Blanco <ejbs@netlabs.com.uy> :990222: kmod support
 *
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/errno.h>
22
#include <linux/err.h>
L
Linus Torvalds 已提交
23 24 25
#include <linux/skbuff.h>
#include <linux/init.h>
#include <linux/kmod.h>
26
#include <linux/slab.h>
27
#include <linux/idr.h>
28 29
#include <net/net_namespace.h>
#include <net/sock.h>
30
#include <net/netlink.h>
L
Linus Torvalds 已提交
31 32 33 34
#include <net/pkt_sched.h>
#include <net/pkt_cls.h>

/* The list of all installed classifier types */
35
static LIST_HEAD(tcf_proto_base);
L
Linus Torvalds 已提交
36 37 38 39 40 41

/* Protects list of registered TC modules. It is pure SMP lock. */
static DEFINE_RWLOCK(cls_mod_lock);

/* Find classifier type by string name */

42
static const struct tcf_proto_ops *tcf_proto_lookup_ops(const char *kind)
L
Linus Torvalds 已提交
43
{
44
	const struct tcf_proto_ops *t, *res = NULL;
L
Linus Torvalds 已提交
45 46 47

	if (kind) {
		read_lock(&cls_mod_lock);
48
		list_for_each_entry(t, &tcf_proto_base, head) {
49
			if (strcmp(kind, t->kind) == 0) {
50 51
				if (try_module_get(t->owner))
					res = t;
L
Linus Torvalds 已提交
52 53 54 55 56
				break;
			}
		}
		read_unlock(&cls_mod_lock);
	}
57
	return res;
L
Linus Torvalds 已提交
58 59 60 61 62 63
}

/* Register(unregister) new classifier type */

int register_tcf_proto_ops(struct tcf_proto_ops *ops)
{
64
	struct tcf_proto_ops *t;
L
Linus Torvalds 已提交
65 66 67
	int rc = -EEXIST;

	write_lock(&cls_mod_lock);
68
	list_for_each_entry(t, &tcf_proto_base, head)
L
Linus Torvalds 已提交
69 70 71
		if (!strcmp(ops->kind, t->kind))
			goto out;

72
	list_add_tail(&ops->head, &tcf_proto_base);
L
Linus Torvalds 已提交
73 74 75 76 77
	rc = 0;
out:
	write_unlock(&cls_mod_lock);
	return rc;
}
78
EXPORT_SYMBOL(register_tcf_proto_ops);
L
Linus Torvalds 已提交
79

80 81
static struct workqueue_struct *tc_filter_wq;

L
Linus Torvalds 已提交
82 83
int unregister_tcf_proto_ops(struct tcf_proto_ops *ops)
{
84
	struct tcf_proto_ops *t;
L
Linus Torvalds 已提交
85 86
	int rc = -ENOENT;

87 88 89 90
	/* Wait for outstanding call_rcu()s, if any, from a
	 * tcf_proto_ops's destroy() handler.
	 */
	rcu_barrier();
91
	flush_workqueue(tc_filter_wq);
92

L
Linus Torvalds 已提交
93
	write_lock(&cls_mod_lock);
94 95 96 97
	list_for_each_entry(t, &tcf_proto_base, head) {
		if (t == ops) {
			list_del(&t->head);
			rc = 0;
L
Linus Torvalds 已提交
98
			break;
99 100
		}
	}
L
Linus Torvalds 已提交
101 102 103
	write_unlock(&cls_mod_lock);
	return rc;
}
104
EXPORT_SYMBOL(unregister_tcf_proto_ops);
L
Linus Torvalds 已提交
105

106 107 108 109 110 111
bool tcf_queue_work(struct work_struct *work)
{
	return queue_work(tc_filter_wq, work);
}
EXPORT_SYMBOL(tcf_queue_work);

L
Linus Torvalds 已提交
112 113
/* Select new prio value from the range, managed by kernel. */

114
static inline u32 tcf_auto_prio(struct tcf_proto *tp)
L
Linus Torvalds 已提交
115
{
116
	u32 first = TC_H_MAKE(0xC0000000U, 0U);
L
Linus Torvalds 已提交
117 118

	if (tp)
E
Eric Dumazet 已提交
119
		first = tp->prio - 1;
L
Linus Torvalds 已提交
120

121
	return TC_H_MAJ(first);
L
Linus Torvalds 已提交
122 123
}

124
static struct tcf_proto *tcf_proto_create(const char *kind, u32 protocol,
125 126
					  u32 prio, struct tcf_chain *chain,
					  struct netlink_ext_ack *extack)
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
{
	struct tcf_proto *tp;
	int err;

	tp = kzalloc(sizeof(*tp), GFP_KERNEL);
	if (!tp)
		return ERR_PTR(-ENOBUFS);

	err = -ENOENT;
	tp->ops = tcf_proto_lookup_ops(kind);
	if (!tp->ops) {
#ifdef CONFIG_MODULES
		rtnl_unlock();
		request_module("cls_%s", kind);
		rtnl_lock();
		tp->ops = tcf_proto_lookup_ops(kind);
		/* We dropped the RTNL semaphore in order to perform
		 * the module load. So, even if we succeeded in loading
		 * the module we have to replay the request. We indicate
		 * this using -EAGAIN.
		 */
		if (tp->ops) {
			module_put(tp->ops->owner);
			err = -EAGAIN;
		} else {
152
			NL_SET_ERR_MSG(extack, "TC classifier not found");
153 154 155 156 157 158 159 160
			err = -ENOENT;
		}
		goto errout;
#endif
	}
	tp->classify = tp->ops->classify;
	tp->protocol = protocol;
	tp->prio = prio;
161
	tp->chain = chain;
162 163 164 165 166 167 168 169 170 171 172 173 174

	err = tp->ops->init(tp);
	if (err) {
		module_put(tp->ops->owner);
		goto errout;
	}
	return tp;

errout:
	kfree(tp);
	return ERR_PTR(err);
}

175
static void tcf_proto_destroy(struct tcf_proto *tp)
176
{
177 178 179
	tp->ops->destroy(tp);
	module_put(tp->ops->owner);
	kfree_rcu(tp, rcu);
180 181
}

182 183 184 185 186 187
struct tcf_filter_chain_list_item {
	struct list_head list;
	tcf_chain_head_change_t *chain_head_change;
	void *chain_head_change_priv;
};

188 189
static struct tcf_chain *tcf_chain_create(struct tcf_block *block,
					  u32 chain_index)
190
{
191 192 193 194 195
	struct tcf_chain *chain;

	chain = kzalloc(sizeof(*chain), GFP_KERNEL);
	if (!chain)
		return NULL;
196
	INIT_LIST_HEAD(&chain->filter_chain_list);
197 198 199
	list_add_tail(&chain->list, &block->chain_list);
	chain->block = block;
	chain->index = chain_index;
200
	chain->refcnt = 1;
201
	return chain;
202 203
}

204 205 206 207 208 209
static void tcf_chain_head_change_item(struct tcf_filter_chain_list_item *item,
				       struct tcf_proto *tp_head)
{
	if (item->chain_head_change)
		item->chain_head_change(tp_head, item->chain_head_change_priv);
}
210 211 212
static void tcf_chain_head_change(struct tcf_chain *chain,
				  struct tcf_proto *tp_head)
{
213 214 215 216
	struct tcf_filter_chain_list_item *item;

	list_for_each_entry(item, &chain->filter_chain_list, list)
		tcf_chain_head_change_item(item, tp_head);
217 218
}

J
Jiri Pirko 已提交
219
static void tcf_chain_flush(struct tcf_chain *chain)
220
{
221
	struct tcf_proto *tp = rtnl_dereference(chain->filter_chain);
222

223
	tcf_chain_head_change(chain, NULL);
224
	while (tp) {
225
		RCU_INIT_POINTER(chain->filter_chain, tp->next);
226
		tcf_proto_destroy(tp);
227 228
		tp = rtnl_dereference(chain->filter_chain);
		tcf_chain_put(chain);
229
	}
J
Jiri Pirko 已提交
230 231 232 233
}

static void tcf_chain_destroy(struct tcf_chain *chain)
{
234 235
	struct tcf_block *block = chain->block;

236 237
	list_del(&chain->list);
	kfree(chain);
238 239
	if (list_empty(&block->chain_list))
		kfree(block);
240
}
241

242 243 244
static void tcf_chain_hold(struct tcf_chain *chain)
{
	++chain->refcnt;
245 246
}

247 248
struct tcf_chain *tcf_chain_get(struct tcf_block *block, u32 chain_index,
				bool create)
249 250 251 252
{
	struct tcf_chain *chain;

	list_for_each_entry(chain, &block->chain_list, list) {
253 254 255 256
		if (chain->index == chain_index) {
			tcf_chain_hold(chain);
			return chain;
		}
257
	}
258

259
	return create ? tcf_chain_create(block, chain_index) : NULL;
260 261 262 263 264
}
EXPORT_SYMBOL(tcf_chain_get);

void tcf_chain_put(struct tcf_chain *chain)
{
265
	if (--chain->refcnt == 0)
266 267 268 269
		tcf_chain_destroy(chain);
}
EXPORT_SYMBOL(tcf_chain_put);

270 271 272 273 274 275 276 277 278
static bool tcf_block_offload_in_use(struct tcf_block *block)
{
	return block->offloadcnt;
}

static int tcf_block_offload_cmd(struct tcf_block *block,
				 struct net_device *dev,
				 struct tcf_block_ext_info *ei,
				 enum tc_block_command command)
279 280 281 282 283 284
{
	struct tc_block_offload bo = {};

	bo.command = command;
	bo.binder_type = ei->binder_type;
	bo.block = block;
285
	return dev->netdev_ops->ndo_setup_tc(dev, TC_SETUP_BLOCK, &bo);
286 287
}

288 289
static int tcf_block_offload_bind(struct tcf_block *block, struct Qdisc *q,
				  struct tcf_block_ext_info *ei)
290
{
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312
	struct net_device *dev = q->dev_queue->dev;
	int err;

	if (!dev->netdev_ops->ndo_setup_tc)
		goto no_offload_dev_inc;

	/* If tc offload feature is disabled and the block we try to bind
	 * to already has some offloaded filters, forbid to bind.
	 */
	if (!tc_can_offload(dev) && tcf_block_offload_in_use(block))
		return -EOPNOTSUPP;

	err = tcf_block_offload_cmd(block, dev, ei, TC_BLOCK_BIND);
	if (err == -EOPNOTSUPP)
		goto no_offload_dev_inc;
	return err;

no_offload_dev_inc:
	if (tcf_block_offload_in_use(block))
		return -EOPNOTSUPP;
	block->nooffloaddevcnt++;
	return 0;
313 314 315 316 317
}

static void tcf_block_offload_unbind(struct tcf_block *block, struct Qdisc *q,
				     struct tcf_block_ext_info *ei)
{
318 319 320 321 322 323 324 325 326 327 328 329
	struct net_device *dev = q->dev_queue->dev;
	int err;

	if (!dev->netdev_ops->ndo_setup_tc)
		goto no_offload_dev_dec;
	err = tcf_block_offload_cmd(block, dev, ei, TC_BLOCK_UNBIND);
	if (err == -EOPNOTSUPP)
		goto no_offload_dev_dec;
	return;

no_offload_dev_dec:
	WARN_ON(block->nooffloaddevcnt-- == 0);
330 331
}

332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
static int
tcf_chain_head_change_cb_add(struct tcf_chain *chain,
			     struct tcf_block_ext_info *ei,
			     struct netlink_ext_ack *extack)
{
	struct tcf_filter_chain_list_item *item;

	item = kmalloc(sizeof(*item), GFP_KERNEL);
	if (!item) {
		NL_SET_ERR_MSG(extack, "Memory allocation for head change callback item failed");
		return -ENOMEM;
	}
	item->chain_head_change = ei->chain_head_change;
	item->chain_head_change_priv = ei->chain_head_change_priv;
	if (chain->filter_chain)
		tcf_chain_head_change_item(item, chain->filter_chain);
	list_add(&item->list, &chain->filter_chain_list);
	return 0;
}

static void
tcf_chain_head_change_cb_del(struct tcf_chain *chain,
			     struct tcf_block_ext_info *ei)
{
	struct tcf_filter_chain_list_item *item;

	list_for_each_entry(item, &chain->filter_chain_list, list) {
		if ((!ei->chain_head_change && !ei->chain_head_change_priv) ||
		    (item->chain_head_change == ei->chain_head_change &&
		     item->chain_head_change_priv == ei->chain_head_change_priv)) {
			tcf_chain_head_change_item(item, NULL);
			list_del(&item->list);
			kfree(item);
			return;
		}
	}
	WARN_ON(1);
}

371 372 373 374 375 376 377 378
struct tcf_net {
	struct idr idr;
};

static unsigned int tcf_net_id;

static int tcf_block_insert(struct tcf_block *block, struct net *net,
			    u32 block_index, struct netlink_ext_ack *extack)
379
{
380 381 382 383 384 385 386 387 388
	struct tcf_net *tn = net_generic(net, tcf_net_id);
	int err;

	err = idr_alloc_ext(&tn->idr, block, NULL, block_index,
			    block_index + 1, GFP_KERNEL);
	if (err)
		return err;
	block->index = block_index;
	return 0;
389 390
}

391 392 393 394 395 396 397 398 399
static void tcf_block_remove(struct tcf_block *block, struct net *net)
{
	struct tcf_net *tn = net_generic(net, tcf_net_id);

	idr_remove_ext(&tn->idr, block->index);
}

static struct tcf_block *tcf_block_create(struct net *net, struct Qdisc *q,
					  struct netlink_ext_ack *extack)
400
{
401
	struct tcf_block *block;
402
	struct tcf_chain *chain;
403
	int err;
404

405
	block = kzalloc(sizeof(*block), GFP_KERNEL);
406 407
	if (!block) {
		NL_SET_ERR_MSG(extack, "Memory allocation for block failed");
408
		return ERR_PTR(-ENOMEM);
409
	}
410
	INIT_LIST_HEAD(&block->chain_list);
411
	INIT_LIST_HEAD(&block->cb_list);
412
	INIT_LIST_HEAD(&block->owner_list);
413

414 415 416
	/* Create chain 0 by default, it has to be always present. */
	chain = tcf_chain_create(block, 0);
	if (!chain) {
417
		NL_SET_ERR_MSG(extack, "Failed to create new tcf chain");
418 419 420
		err = -ENOMEM;
		goto err_chain_create;
	}
421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443
	block->net = qdisc_net(q);
	block->refcnt = 1;
	block->net = net;
	block->q = q;
	return block;

err_chain_create:
	kfree(block);
	return ERR_PTR(err);
}

static struct tcf_block *tcf_block_lookup(struct net *net, u32 block_index)
{
	struct tcf_net *tn = net_generic(net, tcf_net_id);

	return idr_find_ext(&tn->idr, block_index);
}

static struct tcf_chain *tcf_block_chain_zero(struct tcf_block *block)
{
	return list_first_entry(&block->chain_list, struct tcf_chain, list);
}

444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502
struct tcf_block_owner_item {
	struct list_head list;
	struct Qdisc *q;
	enum tcf_block_binder_type binder_type;
};

static void
tcf_block_owner_netif_keep_dst(struct tcf_block *block,
			       struct Qdisc *q,
			       enum tcf_block_binder_type binder_type)
{
	if (block->keep_dst &&
	    binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS &&
	    binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS)
		netif_keep_dst(qdisc_dev(q));
}

void tcf_block_netif_keep_dst(struct tcf_block *block)
{
	struct tcf_block_owner_item *item;

	block->keep_dst = true;
	list_for_each_entry(item, &block->owner_list, list)
		tcf_block_owner_netif_keep_dst(block, item->q,
					       item->binder_type);
}
EXPORT_SYMBOL(tcf_block_netif_keep_dst);

static int tcf_block_owner_add(struct tcf_block *block,
			       struct Qdisc *q,
			       enum tcf_block_binder_type binder_type)
{
	struct tcf_block_owner_item *item;

	item = kmalloc(sizeof(*item), GFP_KERNEL);
	if (!item)
		return -ENOMEM;
	item->q = q;
	item->binder_type = binder_type;
	list_add(&item->list, &block->owner_list);
	return 0;
}

static void tcf_block_owner_del(struct tcf_block *block,
				struct Qdisc *q,
				enum tcf_block_binder_type binder_type)
{
	struct tcf_block_owner_item *item;

	list_for_each_entry(item, &block->owner_list, list) {
		if (item->q == q && item->binder_type == binder_type) {
			list_del(&item->list);
			kfree(item);
			return;
		}
	}
	WARN_ON(1);
}

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
int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q,
		      struct tcf_block_ext_info *ei,
		      struct netlink_ext_ack *extack)
{
	struct net *net = qdisc_net(q);
	struct tcf_block *block = NULL;
	bool created = false;
	int err;

	if (ei->block_index) {
		/* block_index not 0 means the shared block is requested */
		block = tcf_block_lookup(net, ei->block_index);
		if (block)
			block->refcnt++;
	}

	if (!block) {
		block = tcf_block_create(net, q, extack);
		if (IS_ERR(block))
			return PTR_ERR(block);
		created = true;
		if (ei->block_index) {
			err = tcf_block_insert(block, net,
					       ei->block_index, extack);
			if (err)
				goto err_block_insert;
		}
	}

532 533 534 535 536 537
	err = tcf_block_owner_add(block, q, ei->binder_type);
	if (err)
		goto err_block_owner_add;

	tcf_block_owner_netif_keep_dst(block, q, ei->binder_type);

538 539 540 541
	err = tcf_chain_head_change_cb_add(tcf_block_chain_zero(block),
					   ei, extack);
	if (err)
		goto err_chain_head_change_cb_add;
542 543 544 545 546

	err = tcf_block_offload_bind(block, q, ei);
	if (err)
		goto err_block_offload_bind;

547 548
	*p_block = block;
	return 0;
549

550 551
err_block_offload_bind:
	tcf_chain_head_change_cb_del(tcf_block_chain_zero(block), ei);
552
err_chain_head_change_cb_add:
553 554
	tcf_block_owner_del(block, q, ei->binder_type);
err_block_owner_add:
555 556 557 558 559 560 561 562 563
	if (created) {
		if (tcf_block_shared(block))
			tcf_block_remove(block, net);
err_block_insert:
		kfree(tcf_block_chain_zero(block));
		kfree(block);
	} else {
		block->refcnt--;
	}
564
	return err;
565
}
566 567
EXPORT_SYMBOL(tcf_block_get_ext);

568 569 570 571 572 573 574
static void tcf_chain_head_change_dflt(struct tcf_proto *tp_head, void *priv)
{
	struct tcf_proto __rcu **p_filter_chain = priv;

	rcu_assign_pointer(*p_filter_chain, tp_head);
}

575
int tcf_block_get(struct tcf_block **p_block,
576 577
		  struct tcf_proto __rcu **p_filter_chain, struct Qdisc *q,
		  struct netlink_ext_ack *extack)
578
{
579 580 581 582
	struct tcf_block_ext_info ei = {
		.chain_head_change = tcf_chain_head_change_dflt,
		.chain_head_change_priv = p_filter_chain,
	};
583

584
	WARN_ON(!p_filter_chain);
585
	return tcf_block_get_ext(p_block, q, &ei, extack);
586
}
587 588
EXPORT_SYMBOL(tcf_block_get);

589
/* XXX: Standalone actions are not allowed to jump to any chain, and bound
590
 * actions should be all removed after flushing.
591
 */
592
void tcf_block_put_ext(struct tcf_block *block, struct Qdisc *q,
593
		       struct tcf_block_ext_info *ei)
594
{
595
	struct tcf_chain *chain, *tmp;
596

597 598
	if (!block)
		return;
599
	tcf_chain_head_change_cb_del(tcf_block_chain_zero(block), ei);
600
	tcf_block_owner_del(block, q, ei->binder_type);
601

602 603 604 605 606 607 608 609 610 611 612 613 614
	if (--block->refcnt == 0) {
		if (tcf_block_shared(block))
			tcf_block_remove(block, block->net);

		/* Hold a refcnt for all chains, so that they don't disappear
		 * while we are iterating.
		 */
		list_for_each_entry(chain, &block->chain_list, list)
			tcf_chain_hold(chain);

		list_for_each_entry(chain, &block->chain_list, list)
			tcf_chain_flush(chain);
	}
615

616 617
	tcf_block_offload_unbind(block, q, ei);

618 619 620 621
	if (block->refcnt == 0) {
		/* At this point, all the chains should have refcnt >= 1. */
		list_for_each_entry_safe(chain, tmp, &block->chain_list, list)
			tcf_chain_put(chain);
622

623 624 625
		/* Finally, put chain 0 and allow block to be freed. */
		tcf_chain_put(tcf_block_chain_zero(block));
	}
626
}
627 628 629 630 631 632
EXPORT_SYMBOL(tcf_block_put_ext);

void tcf_block_put(struct tcf_block *block)
{
	struct tcf_block_ext_info ei = {0, };

633 634
	if (!block)
		return;
635
	tcf_block_put_ext(block, block->q, &ei);
636
}
637

638
EXPORT_SYMBOL(tcf_block_put);
639

640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682
struct tcf_block_cb {
	struct list_head list;
	tc_setup_cb_t *cb;
	void *cb_ident;
	void *cb_priv;
	unsigned int refcnt;
};

void *tcf_block_cb_priv(struct tcf_block_cb *block_cb)
{
	return block_cb->cb_priv;
}
EXPORT_SYMBOL(tcf_block_cb_priv);

struct tcf_block_cb *tcf_block_cb_lookup(struct tcf_block *block,
					 tc_setup_cb_t *cb, void *cb_ident)
{	struct tcf_block_cb *block_cb;

	list_for_each_entry(block_cb, &block->cb_list, list)
		if (block_cb->cb == cb && block_cb->cb_ident == cb_ident)
			return block_cb;
	return NULL;
}
EXPORT_SYMBOL(tcf_block_cb_lookup);

void tcf_block_cb_incref(struct tcf_block_cb *block_cb)
{
	block_cb->refcnt++;
}
EXPORT_SYMBOL(tcf_block_cb_incref);

unsigned int tcf_block_cb_decref(struct tcf_block_cb *block_cb)
{
	return --block_cb->refcnt;
}
EXPORT_SYMBOL(tcf_block_cb_decref);

struct tcf_block_cb *__tcf_block_cb_register(struct tcf_block *block,
					     tc_setup_cb_t *cb, void *cb_ident,
					     void *cb_priv)
{
	struct tcf_block_cb *block_cb;

683 684 685 686 687 688 689
	/* At this point, playback of previous block cb calls is not supported,
	 * so forbid to register to block which already has some offloaded
	 * filters present.
	 */
	if (tcf_block_offload_in_use(block))
		return ERR_PTR(-EOPNOTSUPP);

690 691
	block_cb = kzalloc(sizeof(*block_cb), GFP_KERNEL);
	if (!block_cb)
692
		return ERR_PTR(-ENOMEM);
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707
	block_cb->cb = cb;
	block_cb->cb_ident = cb_ident;
	block_cb->cb_priv = cb_priv;
	list_add(&block_cb->list, &block->cb_list);
	return block_cb;
}
EXPORT_SYMBOL(__tcf_block_cb_register);

int tcf_block_cb_register(struct tcf_block *block,
			  tc_setup_cb_t *cb, void *cb_ident,
			  void *cb_priv)
{
	struct tcf_block_cb *block_cb;

	block_cb = __tcf_block_cb_register(block, cb, cb_ident, cb_priv);
708
	return IS_ERR(block_cb) ? PTR_ERR(block_cb) : 0;
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
}
EXPORT_SYMBOL(tcf_block_cb_register);

void __tcf_block_cb_unregister(struct tcf_block_cb *block_cb)
{
	list_del(&block_cb->list);
	kfree(block_cb);
}
EXPORT_SYMBOL(__tcf_block_cb_unregister);

void tcf_block_cb_unregister(struct tcf_block *block,
			     tc_setup_cb_t *cb, void *cb_ident)
{
	struct tcf_block_cb *block_cb;

	block_cb = tcf_block_cb_lookup(block, cb, cb_ident);
	if (!block_cb)
		return;
	__tcf_block_cb_unregister(block_cb);
}
EXPORT_SYMBOL(tcf_block_cb_unregister);

static int tcf_block_cb_call(struct tcf_block *block, enum tc_setup_type type,
			     void *type_data, bool err_stop)
{
	struct tcf_block_cb *block_cb;
	int ok_count = 0;
	int err;

738 739 740 741
	/* Make sure all netdevs sharing this block are offload-capable. */
	if (block->nooffloaddevcnt && err_stop)
		return -EOPNOTSUPP;

742 743 744 745 746 747 748 749 750 751 752 753
	list_for_each_entry(block_cb, &block->cb_list, list) {
		err = block_cb->cb(type, type_data, block_cb->cb_priv);
		if (err) {
			if (err_stop)
				return err;
		} else {
			ok_count++;
		}
	}
	return ok_count;
}

754 755 756 757 758 759 760 761 762 763
/* Main classifier routine: scans classifier chain attached
 * to this qdisc, (optionally) tests for protocol and asks
 * specific classifiers.
 */
int tcf_classify(struct sk_buff *skb, const struct tcf_proto *tp,
		 struct tcf_result *res, bool compat_mode)
{
	__be16 protocol = tc_skb_protocol(skb);
#ifdef CONFIG_NET_CLS_ACT
	const int max_reclassify_loop = 4;
764 765
	const struct tcf_proto *orig_tp = tp;
	const struct tcf_proto *first_tp;
766 767 768 769 770 771 772 773 774 775 776 777 778
	int limit = 0;

reclassify:
#endif
	for (; tp; tp = rcu_dereference_bh(tp->next)) {
		int err;

		if (tp->protocol != protocol &&
		    tp->protocol != htons(ETH_P_ALL))
			continue;

		err = tp->classify(skb, tp, res);
#ifdef CONFIG_NET_CLS_ACT
779
		if (unlikely(err == TC_ACT_RECLASSIFY && !compat_mode)) {
780
			first_tp = orig_tp;
781
			goto reset;
782
		} else if (unlikely(TC_ACT_EXT_CMP(err, TC_ACT_GOTO_CHAIN))) {
783
			first_tp = res->goto_tp;
784 785
			goto reset;
		}
786 787 788 789 790 791 792 793 794
#endif
		if (err >= 0)
			return err;
	}

	return TC_ACT_UNSPEC; /* signal: continue lookup */
#ifdef CONFIG_NET_CLS_ACT
reset:
	if (unlikely(limit++ >= max_reclassify_loop)) {
795 796 797
		net_notice_ratelimited("%u: reclassify loop, rule prio %u, protocol %02x\n",
				       tp->chain->block->index,
				       tp->prio & 0xffff,
798 799 800 801
				       ntohs(tp->protocol));
		return TC_ACT_SHOT;
	}

802
	tp = first_tp;
803 804 805 806 807 808
	protocol = tc_skb_protocol(skb);
	goto reclassify;
#endif
}
EXPORT_SYMBOL(tcf_classify);

809 810 811 812 813 814 815 816 817 818 819 820 821 822
struct tcf_chain_info {
	struct tcf_proto __rcu **pprev;
	struct tcf_proto __rcu *next;
};

static struct tcf_proto *tcf_chain_tp_prev(struct tcf_chain_info *chain_info)
{
	return rtnl_dereference(*chain_info->pprev);
}

static void tcf_chain_tp_insert(struct tcf_chain *chain,
				struct tcf_chain_info *chain_info,
				struct tcf_proto *tp)
{
823 824
	if (*chain_info->pprev == chain->filter_chain)
		tcf_chain_head_change(chain, tp);
825 826
	RCU_INIT_POINTER(tp->next, tcf_chain_tp_prev(chain_info));
	rcu_assign_pointer(*chain_info->pprev, tp);
827
	tcf_chain_hold(chain);
828 829 830 831 832 833 834 835
}

static void tcf_chain_tp_remove(struct tcf_chain *chain,
				struct tcf_chain_info *chain_info,
				struct tcf_proto *tp)
{
	struct tcf_proto *next = rtnl_dereference(chain_info->next);

836 837
	if (tp == chain->filter_chain)
		tcf_chain_head_change(chain, next);
838
	RCU_INIT_POINTER(*chain_info->pprev, next);
839
	tcf_chain_put(chain);
840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868
}

static struct tcf_proto *tcf_chain_tp_find(struct tcf_chain *chain,
					   struct tcf_chain_info *chain_info,
					   u32 protocol, u32 prio,
					   bool prio_allocate)
{
	struct tcf_proto **pprev;
	struct tcf_proto *tp;

	/* Check the chain for existence of proto-tcf with this priority */
	for (pprev = &chain->filter_chain;
	     (tp = rtnl_dereference(*pprev)); pprev = &tp->next) {
		if (tp->prio >= prio) {
			if (tp->prio == prio) {
				if (prio_allocate ||
				    (tp->protocol != protocol && protocol))
					return ERR_PTR(-EINVAL);
			} else {
				tp = NULL;
			}
			break;
		}
	}
	chain_info->pprev = pprev;
	chain_info->next = tp ? tp->next : NULL;
	return tp;
}

869
static int tcf_fill_node(struct net *net, struct sk_buff *skb,
870 871 872
			 struct tcf_proto *tp, struct tcf_block *block,
			 struct Qdisc *q, u32 parent, void *fh,
			 u32 portid, u32 seq, u16 flags, int event)
873 874 875 876 877 878 879 880 881 882 883 884
{
	struct tcmsg *tcm;
	struct nlmsghdr  *nlh;
	unsigned char *b = skb_tail_pointer(skb);

	nlh = nlmsg_put(skb, portid, seq, event, sizeof(*tcm), flags);
	if (!nlh)
		goto out_nlmsg_trim;
	tcm = nlmsg_data(nlh);
	tcm->tcm_family = AF_UNSPEC;
	tcm->tcm__pad1 = 0;
	tcm->tcm__pad2 = 0;
885 886 887 888 889 890 891
	if (q) {
		tcm->tcm_ifindex = qdisc_dev(q)->ifindex;
		tcm->tcm_parent = parent;
	} else {
		tcm->tcm_ifindex = TCM_IFINDEX_MAGIC_BLOCK;
		tcm->tcm_block_index = block->index;
	}
892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913
	tcm->tcm_info = TC_H_MAKE(tp->prio, tp->protocol);
	if (nla_put_string(skb, TCA_KIND, tp->ops->kind))
		goto nla_put_failure;
	if (nla_put_u32(skb, TCA_CHAIN, tp->chain->index))
		goto nla_put_failure;
	if (!fh) {
		tcm->tcm_handle = 0;
	} else {
		if (tp->ops->dump && tp->ops->dump(net, tp, fh, skb, tcm) < 0)
			goto nla_put_failure;
	}
	nlh->nlmsg_len = skb_tail_pointer(skb) - b;
	return skb->len;

out_nlmsg_trim:
nla_put_failure:
	nlmsg_trim(skb, b);
	return -1;
}

static int tfilter_notify(struct net *net, struct sk_buff *oskb,
			  struct nlmsghdr *n, struct tcf_proto *tp,
914 915
			  struct tcf_block *block, struct Qdisc *q,
			  u32 parent, void *fh, int event, bool unicast)
916 917 918 919 920 921 922 923
{
	struct sk_buff *skb;
	u32 portid = oskb ? NETLINK_CB(oskb).portid : 0;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!skb)
		return -ENOBUFS;

924 925
	if (tcf_fill_node(net, skb, tp, block, q, parent, fh, portid,
			  n->nlmsg_seq, n->nlmsg_flags, event) <= 0) {
926 927 928 929 930 931 932 933 934 935 936 937 938
		kfree_skb(skb);
		return -EINVAL;
	}

	if (unicast)
		return netlink_unicast(net->rtnl, skb, portid, MSG_DONTWAIT);

	return rtnetlink_send(skb, net, portid, RTNLGRP_TC,
			      n->nlmsg_flags & NLM_F_ECHO);
}

static int tfilter_del_notify(struct net *net, struct sk_buff *oskb,
			      struct nlmsghdr *n, struct tcf_proto *tp,
939
			      struct tcf_block *block, struct Qdisc *q,
940 941
			      u32 parent, void *fh, bool unicast, bool *last,
			      struct netlink_ext_ack *extack)
942 943 944 945 946 947 948 949 950
{
	struct sk_buff *skb;
	u32 portid = oskb ? NETLINK_CB(oskb).portid : 0;
	int err;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!skb)
		return -ENOBUFS;

951 952
	if (tcf_fill_node(net, skb, tp, block, q, parent, fh, portid,
			  n->nlmsg_seq, n->nlmsg_flags, RTM_DELTFILTER) <= 0) {
953
		NL_SET_ERR_MSG(extack, "Failed to build del event notification");
954 955 956 957 958 959 960 961 962 963 964 965 966
		kfree_skb(skb);
		return -EINVAL;
	}

	err = tp->ops->delete(tp, fh, last);
	if (err) {
		kfree_skb(skb);
		return err;
	}

	if (unicast)
		return netlink_unicast(net->rtnl, skb, portid, MSG_DONTWAIT);

967 968 969 970 971
	err = rtnetlink_send(skb, net, portid, RTNLGRP_TC,
			     n->nlmsg_flags & NLM_F_ECHO);
	if (err < 0)
		NL_SET_ERR_MSG(extack, "Failed to send filter delete notification");
	return err;
972 973 974
}

static void tfilter_notify_chain(struct net *net, struct sk_buff *oskb,
975 976
				 struct tcf_block *block, struct Qdisc *q,
				 u32 parent, struct nlmsghdr *n,
977 978 979 980 981 982
				 struct tcf_chain *chain, int event)
{
	struct tcf_proto *tp;

	for (tp = rtnl_dereference(chain->filter_chain);
	     tp; tp = rtnl_dereference(tp->next))
983 984
		tfilter_notify(net, oskb, n, tp, block,
			       q, parent, 0, event, false);
985 986
}

L
Linus Torvalds 已提交
987 988
/* Add/change/delete/get a filter node */

989 990
static int tc_ctl_tfilter(struct sk_buff *skb, struct nlmsghdr *n,
			  struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
991
{
992
	struct net *net = sock_net(skb->sk);
993
	struct nlattr *tca[TCA_MAX + 1];
L
Linus Torvalds 已提交
994 995 996
	struct tcmsg *t;
	u32 protocol;
	u32 prio;
997
	bool prio_allocate;
L
Linus Torvalds 已提交
998
	u32 parent;
999
	u32 chain_index;
1000
	struct Qdisc *q = NULL;
1001
	struct tcf_chain_info chain_info;
1002
	struct tcf_chain *chain = NULL;
1003
	struct tcf_block *block;
L
Linus Torvalds 已提交
1004 1005
	struct tcf_proto *tp;
	unsigned long cl;
1006
	void *fh;
L
Linus Torvalds 已提交
1007
	int err;
1008
	int tp_created;
L
Linus Torvalds 已提交
1009

1010
	if ((n->nlmsg_type != RTM_GETTFILTER) &&
1011
	    !netlink_ns_capable(skb, net->user_ns, CAP_NET_ADMIN))
1012
		return -EPERM;
1013

L
Linus Torvalds 已提交
1014
replay:
1015 1016
	tp_created = 0;

1017
	err = nlmsg_parse(n, sizeof(*t), tca, TCA_MAX, NULL, extack);
1018 1019 1020
	if (err < 0)
		return err;

1021
	t = nlmsg_data(n);
L
Linus Torvalds 已提交
1022 1023
	protocol = TC_H_MIN(t->tcm_info);
	prio = TC_H_MAJ(t->tcm_info);
1024
	prio_allocate = false;
L
Linus Torvalds 已提交
1025 1026 1027 1028
	parent = t->tcm_parent;
	cl = 0;

	if (prio == 0) {
1029 1030
		switch (n->nlmsg_type) {
		case RTM_DELTFILTER:
1031 1032
			if (protocol || t->tcm_handle || tca[TCA_KIND]) {
				NL_SET_ERR_MSG(extack, "Cannot flush filters with protocol, handle or kind set");
1033
				return -ENOENT;
1034
			}
1035 1036 1037 1038 1039 1040 1041
			break;
		case RTM_NEWTFILTER:
			/* If no priority is provided by the user,
			 * we allocate one.
			 */
			if (n->nlmsg_flags & NLM_F_CREATE) {
				prio = TC_H_MAKE(0x80000000U, 0U);
1042
				prio_allocate = true;
1043 1044 1045 1046
				break;
			}
			/* fall-through */
		default:
1047
			NL_SET_ERR_MSG(extack, "Invalid filter command with priority of zero");
L
Linus Torvalds 已提交
1048
			return -ENOENT;
1049
		}
L
Linus Torvalds 已提交
1050 1051 1052 1053
	}

	/* Find head of filter chain. */

1054 1055 1056 1057 1058 1059 1060
	if (t->tcm_ifindex == TCM_IFINDEX_MAGIC_BLOCK) {
		block = tcf_block_lookup(net, t->tcm_block_index);
		if (!block) {
			NL_SET_ERR_MSG(extack, "Block of given index was not found");
			err = -EINVAL;
			goto errout;
		}
1061
	} else {
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
		const struct Qdisc_class_ops *cops;
		struct net_device *dev;

		/* Find link */
		dev = __dev_get_by_index(net, t->tcm_ifindex);
		if (!dev)
			return -ENODEV;

		/* Find qdisc */
		if (!parent) {
			q = dev->qdisc;
			parent = q->handle;
		} else {
			q = qdisc_lookup(dev, TC_H_MAJ(t->tcm_parent));
1076 1077
			if (!q) {
				NL_SET_ERR_MSG(extack, "Parent Qdisc doesn't exists");
1078
				return -EINVAL;
1079
			}
1080
		}
L
Linus Torvalds 已提交
1081

1082 1083
		/* Is it classful? */
		cops = q->ops->cl_ops;
1084 1085
		if (!cops) {
			NL_SET_ERR_MSG(extack, "Qdisc not classful");
1086
			return -EINVAL;
1087
		}
L
Linus Torvalds 已提交
1088

1089 1090
		if (!cops->tcf_block) {
			NL_SET_ERR_MSG(extack, "Class doesn't support blocks");
1091
			return -EOPNOTSUPP;
1092
		}
1093

1094 1095 1096
		/* Do we search for filter, attached to class? */
		if (TC_H_MIN(parent)) {
			cl = cops->find(q, parent);
1097 1098
			if (cl == 0) {
				NL_SET_ERR_MSG(extack, "Specified class doesn't exist");
1099
				return -ENOENT;
1100
			}
1101
		}
L
Linus Torvalds 已提交
1102

1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
		/* And the last stroke */
		block = cops->tcf_block(q, cl, extack);
		if (!block) {
			err = -EINVAL;
			goto errout;
		}
		if (tcf_block_shared(block)) {
			NL_SET_ERR_MSG(extack, "This filter block is shared. Please use the block index to manipulate the filters");
			err = -EOPNOTSUPP;
			goto errout;
		}
1114
	}
1115 1116 1117

	chain_index = tca[TCA_CHAIN] ? nla_get_u32(tca[TCA_CHAIN]) : 0;
	if (chain_index > TC_ACT_EXT_VAL_MASK) {
1118
		NL_SET_ERR_MSG(extack, "Specified chain index exceeds upper limit");
1119 1120 1121
		err = -EINVAL;
		goto errout;
	}
1122 1123
	chain = tcf_chain_get(block, chain_index,
			      n->nlmsg_type == RTM_NEWTFILTER);
1124
	if (!chain) {
1125
		NL_SET_ERR_MSG(extack, "Cannot find specified filter chain");
1126
		err = n->nlmsg_type == RTM_NEWTFILTER ? -ENOMEM : -EINVAL;
1127 1128
		goto errout;
	}
1129

1130
	if (n->nlmsg_type == RTM_DELTFILTER && prio == 0) {
1131
		tfilter_notify_chain(net, skb, block, q, parent, n,
1132
				     chain, RTM_DELTFILTER);
J
Jiri Pirko 已提交
1133
		tcf_chain_flush(chain);
1134 1135 1136
		err = 0;
		goto errout;
	}
L
Linus Torvalds 已提交
1137

1138 1139 1140
	tp = tcf_chain_tp_find(chain, &chain_info, protocol,
			       prio, prio_allocate);
	if (IS_ERR(tp)) {
1141
		NL_SET_ERR_MSG(extack, "Filter with specified priority/protocol not found");
1142 1143
		err = PTR_ERR(tp);
		goto errout;
L
Linus Torvalds 已提交
1144 1145 1146 1147 1148
	}

	if (tp == NULL) {
		/* Proto-tcf does not exist, create new one */

1149
		if (tca[TCA_KIND] == NULL || !protocol) {
1150
			NL_SET_ERR_MSG(extack, "Filter kind and protocol must be specified");
1151
			err = -EINVAL;
L
Linus Torvalds 已提交
1152
			goto errout;
1153
		}
L
Linus Torvalds 已提交
1154

E
Eric Dumazet 已提交
1155
		if (n->nlmsg_type != RTM_NEWTFILTER ||
1156
		    !(n->nlmsg_flags & NLM_F_CREATE)) {
1157
			NL_SET_ERR_MSG(extack, "Need both RTM_NEWTFILTER and NLM_F_CREATE to create a new filter");
1158
			err = -ENOENT;
L
Linus Torvalds 已提交
1159
			goto errout;
1160
		}
L
Linus Torvalds 已提交
1161

1162
		if (prio_allocate)
1163
			prio = tcf_auto_prio(tcf_chain_tp_prev(&chain_info));
L
Linus Torvalds 已提交
1164

1165
		tp = tcf_proto_create(nla_data(tca[TCA_KIND]),
1166
				      protocol, prio, chain, extack);
1167 1168
		if (IS_ERR(tp)) {
			err = PTR_ERR(tp);
L
Linus Torvalds 已提交
1169 1170
			goto errout;
		}
1171
		tp_created = 1;
1172
	} else if (tca[TCA_KIND] && nla_strcmp(tca[TCA_KIND], tp->ops->kind)) {
1173
		NL_SET_ERR_MSG(extack, "Specified filter kind does not match existing one");
1174
		err = -EINVAL;
L
Linus Torvalds 已提交
1175
		goto errout;
1176
	}
L
Linus Torvalds 已提交
1177 1178 1179

	fh = tp->ops->get(tp, t->tcm_handle);

1180
	if (!fh) {
L
Linus Torvalds 已提交
1181
		if (n->nlmsg_type == RTM_DELTFILTER && t->tcm_handle == 0) {
1182
			tcf_chain_tp_remove(chain, &chain_info, tp);
1183
			tfilter_notify(net, skb, n, tp, block, q, parent, fh,
1184
				       RTM_DELTFILTER, false);
1185
			tcf_proto_destroy(tp);
L
Linus Torvalds 已提交
1186 1187 1188 1189
			err = 0;
			goto errout;
		}

1190
		if (n->nlmsg_type != RTM_NEWTFILTER ||
1191
		    !(n->nlmsg_flags & NLM_F_CREATE)) {
1192
			NL_SET_ERR_MSG(extack, "Need both RTM_NEWTFILTER and NLM_F_CREATE to create a new filter");
1193
			err = -ENOENT;
L
Linus Torvalds 已提交
1194
			goto errout;
1195
		}
L
Linus Torvalds 已提交
1196
	} else {
1197 1198
		bool last;

L
Linus Torvalds 已提交
1199
		switch (n->nlmsg_type) {
1200
		case RTM_NEWTFILTER:
1201 1202
			if (n->nlmsg_flags & NLM_F_EXCL) {
				if (tp_created)
1203
					tcf_proto_destroy(tp);
1204
				NL_SET_ERR_MSG(extack, "Filter already exists");
1205
				err = -EEXIST;
L
Linus Torvalds 已提交
1206
				goto errout;
1207
			}
L
Linus Torvalds 已提交
1208 1209
			break;
		case RTM_DELTFILTER:
1210
			err = tfilter_del_notify(net, skb, n, tp, block,
1211 1212
						 q, parent, fh, false, &last,
						 extack);
1213 1214
			if (err)
				goto errout;
1215
			if (last) {
1216
				tcf_chain_tp_remove(chain, &chain_info, tp);
1217 1218
				tcf_proto_destroy(tp);
			}
1219
			goto errout;
L
Linus Torvalds 已提交
1220
		case RTM_GETTFILTER:
1221 1222
			err = tfilter_notify(net, skb, n, tp, block, q, parent,
					     fh, RTM_NEWTFILTER, true);
1223 1224
			if (err < 0)
				NL_SET_ERR_MSG(extack, "Failed to send filter notify message");
L
Linus Torvalds 已提交
1225 1226
			goto errout;
		default:
1227
			NL_SET_ERR_MSG(extack, "Invalid netlink message type");
L
Linus Torvalds 已提交
1228 1229 1230 1231 1232
			err = -EINVAL;
			goto errout;
		}
	}

1233
	err = tp->ops->change(net, skb, tp, cl, t->tcm_handle, tca, &fh,
1234 1235
			      n->nlmsg_flags & NLM_F_CREATE ? TCA_ACT_NOREPLACE : TCA_ACT_REPLACE,
			      extack);
1236
	if (err == 0) {
1237 1238
		if (tp_created)
			tcf_chain_tp_insert(chain, &chain_info, tp);
1239
		tfilter_notify(net, skb, n, tp, block, q, parent, fh,
1240
			       RTM_NEWTFILTER, false);
1241 1242
	} else {
		if (tp_created)
1243
			tcf_proto_destroy(tp);
1244
	}
L
Linus Torvalds 已提交
1245 1246

errout:
1247 1248
	if (chain)
		tcf_chain_put(chain);
L
Linus Torvalds 已提交
1249 1250 1251 1252 1253 1254
	if (err == -EAGAIN)
		/* Replay the request. */
		goto replay;
	return err;
}

1255
struct tcf_dump_args {
L
Linus Torvalds 已提交
1256 1257 1258
	struct tcf_walker w;
	struct sk_buff *skb;
	struct netlink_callback *cb;
1259
	struct tcf_block *block;
1260 1261
	struct Qdisc *q;
	u32 parent;
L
Linus Torvalds 已提交
1262 1263
};

1264
static int tcf_node_dump(struct tcf_proto *tp, void *n, struct tcf_walker *arg)
L
Linus Torvalds 已提交
1265
{
1266
	struct tcf_dump_args *a = (void *)arg;
1267
	struct net *net = sock_net(a->skb->sk);
L
Linus Torvalds 已提交
1268

1269
	return tcf_fill_node(net, a->skb, tp, a->block, a->q, a->parent,
1270
			     n, NETLINK_CB(a->cb->skb).portid,
J
Jamal Hadi Salim 已提交
1271 1272
			     a->cb->nlh->nlmsg_seq, NLM_F_MULTI,
			     RTM_NEWTFILTER);
L
Linus Torvalds 已提交
1273 1274
}

1275 1276
static bool tcf_chain_dump(struct tcf_chain *chain, struct Qdisc *q, u32 parent,
			   struct sk_buff *skb, struct netlink_callback *cb,
1277 1278 1279
			   long index_start, long *p_index)
{
	struct net *net = sock_net(skb->sk);
1280
	struct tcf_block *block = chain->block;
1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298
	struct tcmsg *tcm = nlmsg_data(cb->nlh);
	struct tcf_dump_args arg;
	struct tcf_proto *tp;

	for (tp = rtnl_dereference(chain->filter_chain);
	     tp; tp = rtnl_dereference(tp->next), (*p_index)++) {
		if (*p_index < index_start)
			continue;
		if (TC_H_MAJ(tcm->tcm_info) &&
		    TC_H_MAJ(tcm->tcm_info) != tp->prio)
			continue;
		if (TC_H_MIN(tcm->tcm_info) &&
		    TC_H_MIN(tcm->tcm_info) != tp->protocol)
			continue;
		if (*p_index > index_start)
			memset(&cb->args[1], 0,
			       sizeof(cb->args) - sizeof(cb->args[0]));
		if (cb->args[1] == 0) {
1299
			if (tcf_fill_node(net, skb, tp, block, q, parent, 0,
1300 1301 1302
					  NETLINK_CB(cb->skb).portid,
					  cb->nlh->nlmsg_seq, NLM_F_MULTI,
					  RTM_NEWTFILTER) <= 0)
1303
				return false;
1304 1305 1306 1307 1308 1309 1310 1311

			cb->args[1] = 1;
		}
		if (!tp->ops->walk)
			continue;
		arg.w.fn = tcf_node_dump;
		arg.skb = skb;
		arg.cb = cb;
1312
		arg.block = block;
1313 1314
		arg.q = q;
		arg.parent = parent;
1315 1316 1317 1318 1319 1320
		arg.w.stop = 0;
		arg.w.skip = cb->args[1] - 1;
		arg.w.count = 0;
		tp->ops->walk(tp, &arg.w);
		cb->args[1] = arg.w.count + 1;
		if (arg.w.stop)
1321
			return false;
1322
	}
1323
	return true;
1324 1325
}

E
Eric Dumazet 已提交
1326
/* called with RTNL */
L
Linus Torvalds 已提交
1327 1328
static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb)
{
1329
	struct net *net = sock_net(skb->sk);
1330
	struct nlattr *tca[TCA_MAX + 1];
1331
	struct Qdisc *q = NULL;
1332
	struct tcf_block *block;
1333
	struct tcf_chain *chain;
1334
	struct tcmsg *tcm = nlmsg_data(cb->nlh);
1335 1336
	long index_start;
	long index;
1337
	u32 parent;
1338
	int err;
L
Linus Torvalds 已提交
1339

1340
	if (nlmsg_len(cb->nlh) < sizeof(*tcm))
L
Linus Torvalds 已提交
1341
		return skb->len;
1342 1343 1344 1345 1346

	err = nlmsg_parse(cb->nlh, sizeof(*tcm), tca, TCA_MAX, NULL, NULL);
	if (err)
		return err;

1347 1348 1349 1350
	if (tcm->tcm_ifindex == TCM_IFINDEX_MAGIC_BLOCK) {
		block = tcf_block_lookup(net, tcm->tcm_block_index);
		if (!block)
			goto out;
1351 1352 1353 1354 1355 1356 1357
		/* If we work with block index, q is NULL and parent value
		 * will never be used in the following code. The check
		 * in tcf_fill_node prevents it. However, compiler does not
		 * see that far, so set parent to zero to silence the warning
		 * about parent being uninitialized.
		 */
		parent = 0;
1358
	} else {
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377
		const struct Qdisc_class_ops *cops;
		struct net_device *dev;
		unsigned long cl = 0;

		dev = __dev_get_by_index(net, tcm->tcm_ifindex);
		if (!dev)
			return skb->len;

		parent = tcm->tcm_parent;
		if (!parent) {
			q = dev->qdisc;
			parent = q->handle;
		} else {
			q = qdisc_lookup(dev, TC_H_MAJ(tcm->tcm_parent));
		}
		if (!q)
			goto out;
		cops = q->ops->cl_ops;
		if (!cops)
1378
			goto out;
1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390
		if (!cops->tcf_block)
			goto out;
		if (TC_H_MIN(tcm->tcm_parent)) {
			cl = cops->find(q, tcm->tcm_parent);
			if (cl == 0)
				goto out;
		}
		block = cops->tcf_block(q, cl, NULL);
		if (!block)
			goto out;
		if (tcf_block_shared(block))
			q = NULL;
L
Linus Torvalds 已提交
1391 1392
	}

1393 1394
	index_start = cb->args[0];
	index = 0;
1395 1396 1397 1398 1399

	list_for_each_entry(chain, &block->chain_list, list) {
		if (tca[TCA_CHAIN] &&
		    nla_get_u32(tca[TCA_CHAIN]) != chain->index)
			continue;
1400 1401
		if (!tcf_chain_dump(chain, q, parent, skb, cb,
				    index_start, &index))
1402 1403 1404
			break;
	}

1405
	cb->args[0] = index;
L
Linus Torvalds 已提交
1406 1407 1408 1409 1410

out:
	return skb->len;
}

1411
void tcf_exts_destroy(struct tcf_exts *exts)
L
Linus Torvalds 已提交
1412 1413
{
#ifdef CONFIG_NET_CLS_ACT
1414 1415
	LIST_HEAD(actions);

1416
	ASSERT_RTNL();
1417 1418 1419 1420
	tcf_exts_to_list(exts, &actions);
	tcf_action_destroy(&actions, TCA_ACT_UNBIND);
	kfree(exts->actions);
	exts->nr_actions = 0;
L
Linus Torvalds 已提交
1421 1422
#endif
}
1423
EXPORT_SYMBOL(tcf_exts_destroy);
L
Linus Torvalds 已提交
1424

1425
int tcf_exts_validate(struct net *net, struct tcf_proto *tp, struct nlattr **tb,
1426 1427
		      struct nlattr *rate_tlv, struct tcf_exts *exts, bool ovr,
		      struct netlink_ext_ack *extack)
L
Linus Torvalds 已提交
1428 1429 1430 1431 1432
{
#ifdef CONFIG_NET_CLS_ACT
	{
		struct tc_action *act;

1433
		if (exts->police && tb[exts->police]) {
1434 1435 1436
			act = tcf_action_init_1(net, tp, tb[exts->police],
						rate_tlv, "police", ovr,
						TCA_ACT_BIND);
1437 1438
			if (IS_ERR(act))
				return PTR_ERR(act);
L
Linus Torvalds 已提交
1439

1440
			act->type = exts->type = TCA_OLD_COMPAT;
1441 1442
			exts->actions[0] = act;
			exts->nr_actions = 1;
1443
		} else if (exts->action && tb[exts->action]) {
1444 1445 1446
			LIST_HEAD(actions);
			int err, i = 0;

1447 1448
			err = tcf_action_init(net, tp, tb[exts->action],
					      rate_tlv, NULL, ovr, TCA_ACT_BIND,
J
Jamal Hadi Salim 已提交
1449
					      &actions);
1450 1451
			if (err)
				return err;
1452 1453 1454
			list_for_each_entry(act, &actions, list)
				exts->actions[i++] = act;
			exts->nr_actions = i;
L
Linus Torvalds 已提交
1455
		}
1456
		exts->net = net;
L
Linus Torvalds 已提交
1457 1458
	}
#else
1459
	if ((exts->action && tb[exts->action]) ||
1460 1461
	    (exts->police && tb[exts->police])) {
		NL_SET_ERR_MSG(extack, "Classifier actions are not supported per compile options (CONFIG_NET_CLS_ACT)");
L
Linus Torvalds 已提交
1462
		return -EOPNOTSUPP;
1463
	}
L
Linus Torvalds 已提交
1464 1465 1466 1467
#endif

	return 0;
}
1468
EXPORT_SYMBOL(tcf_exts_validate);
L
Linus Torvalds 已提交
1469

1470
void tcf_exts_change(struct tcf_exts *dst, struct tcf_exts *src)
L
Linus Torvalds 已提交
1471 1472
{
#ifdef CONFIG_NET_CLS_ACT
1473 1474
	struct tcf_exts old = *dst;

1475
	*dst = *src;
1476
	tcf_exts_destroy(&old);
L
Linus Torvalds 已提交
1477 1478
#endif
}
1479
EXPORT_SYMBOL(tcf_exts_change);
L
Linus Torvalds 已提交
1480

1481 1482 1483 1484 1485 1486 1487 1488 1489
#ifdef CONFIG_NET_CLS_ACT
static struct tc_action *tcf_exts_first_act(struct tcf_exts *exts)
{
	if (exts->nr_actions == 0)
		return NULL;
	else
		return exts->actions[0];
}
#endif
1490

1491
int tcf_exts_dump(struct sk_buff *skb, struct tcf_exts *exts)
L
Linus Torvalds 已提交
1492 1493
{
#ifdef CONFIG_NET_CLS_ACT
1494 1495
	struct nlattr *nest;

1496
	if (exts->action && tcf_exts_has_actions(exts)) {
L
Linus Torvalds 已提交
1497 1498 1499 1500 1501
		/*
		 * again for backward compatible mode - we want
		 * to work with both old and new modes of entering
		 * tc data even if iproute2  was newer - jhs
		 */
1502
		if (exts->type != TCA_OLD_COMPAT) {
1503 1504
			LIST_HEAD(actions);

1505
			nest = nla_nest_start(skb, exts->action);
1506 1507
			if (nest == NULL)
				goto nla_put_failure;
1508 1509 1510

			tcf_exts_to_list(exts, &actions);
			if (tcf_action_dump(skb, &actions, 0, 0) < 0)
1511
				goto nla_put_failure;
1512
			nla_nest_end(skb, nest);
1513
		} else if (exts->police) {
1514
			struct tc_action *act = tcf_exts_first_act(exts);
1515
			nest = nla_nest_start(skb, exts->police);
1516
			if (nest == NULL || !act)
1517
				goto nla_put_failure;
1518
			if (tcf_action_dump_old(skb, act, 0, 0) < 0)
1519
				goto nla_put_failure;
1520
			nla_nest_end(skb, nest);
L
Linus Torvalds 已提交
1521 1522 1523
		}
	}
	return 0;
1524 1525 1526

nla_put_failure:
	nla_nest_cancel(skb, nest);
L
Linus Torvalds 已提交
1527
	return -1;
1528 1529 1530
#else
	return 0;
#endif
L
Linus Torvalds 已提交
1531
}
1532
EXPORT_SYMBOL(tcf_exts_dump);
L
Linus Torvalds 已提交
1533

1534

1535
int tcf_exts_dump_stats(struct sk_buff *skb, struct tcf_exts *exts)
L
Linus Torvalds 已提交
1536 1537
{
#ifdef CONFIG_NET_CLS_ACT
1538
	struct tc_action *a = tcf_exts_first_act(exts);
1539
	if (a != NULL && tcf_action_copy_stats(skb, a, 1) < 0)
1540
		return -1;
L
Linus Torvalds 已提交
1541 1542 1543
#endif
	return 0;
}
1544
EXPORT_SYMBOL(tcf_exts_dump_stats);
L
Linus Torvalds 已提交
1545

1546 1547 1548
static int tc_exts_setup_cb_egdev_call(struct tcf_exts *exts,
				       enum tc_setup_type type,
				       void *type_data, bool err_stop)
1549 1550 1551 1552 1553
{
	int ok_count = 0;
#ifdef CONFIG_NET_CLS_ACT
	const struct tc_action *a;
	struct net_device *dev;
1554
	int i, ret;
1555 1556 1557 1558

	if (!tcf_exts_has_actions(exts))
		return 0;

1559 1560
	for (i = 0; i < exts->nr_actions; i++) {
		a = exts->actions[i];
1561 1562 1563
		if (!a->ops->get_dev)
			continue;
		dev = a->ops->get_dev(a);
1564
		if (!dev)
1565 1566 1567 1568 1569 1570 1571 1572 1573
			continue;
		ret = tc_setup_cb_egdev_call(dev, type, type_data, err_stop);
		if (ret < 0)
			return ret;
		ok_count += ret;
	}
#endif
	return ok_count;
}
1574

1575 1576
int tc_setup_cb_call(struct tcf_block *block, struct tcf_exts *exts,
		     enum tc_setup_type type, void *type_data, bool err_stop)
1577
{
1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
	int ok_count;
	int ret;

	ret = tcf_block_cb_call(block, type, type_data, err_stop);
	if (ret < 0)
		return ret;
	ok_count = ret;

	if (!exts)
		return ok_count;
	ret = tc_exts_setup_cb_egdev_call(exts, type, type_data, err_stop);
	if (ret < 0)
		return ret;
	ok_count += ret;

	return ok_count;
1594 1595
}
EXPORT_SYMBOL(tc_setup_cb_call);
1596

1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618
static __net_init int tcf_net_init(struct net *net)
{
	struct tcf_net *tn = net_generic(net, tcf_net_id);

	idr_init(&tn->idr);
	return 0;
}

static void __net_exit tcf_net_exit(struct net *net)
{
	struct tcf_net *tn = net_generic(net, tcf_net_id);

	idr_destroy(&tn->idr);
}

static struct pernet_operations tcf_net_ops = {
	.init = tcf_net_init,
	.exit = tcf_net_exit,
	.id   = &tcf_net_id,
	.size = sizeof(struct tcf_net),
};

L
Linus Torvalds 已提交
1619 1620
static int __init tc_filter_init(void)
{
1621 1622
	int err;

1623 1624 1625 1626
	tc_filter_wq = alloc_ordered_workqueue("tc_filter_workqueue", 0);
	if (!tc_filter_wq)
		return -ENOMEM;

1627 1628 1629 1630
	err = register_pernet_subsys(&tcf_net_ops);
	if (err)
		goto err_register_pernet_subsys;

1631 1632
	rtnl_register(PF_UNSPEC, RTM_NEWTFILTER, tc_ctl_tfilter, NULL, 0);
	rtnl_register(PF_UNSPEC, RTM_DELTFILTER, tc_ctl_tfilter, NULL, 0);
1633
	rtnl_register(PF_UNSPEC, RTM_GETTFILTER, tc_ctl_tfilter,
1634
		      tc_dump_tfilter, 0);
L
Linus Torvalds 已提交
1635 1636

	return 0;
1637 1638 1639 1640

err_register_pernet_subsys:
	destroy_workqueue(tc_filter_wq);
	return err;
L
Linus Torvalds 已提交
1641 1642 1643
}

subsys_initcall(tc_filter_init);