datapath.c 56.8 KB
Newer Older
1
/*
2
 * Copyright (c) 2007-2013 Nicira, Inc.
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of version 2 of the GNU General Public
 * License as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
 * 02110-1301, USA
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/init.h>
#include <linux/module.h>
#include <linux/if_arp.h>
#include <linux/if_vlan.h>
#include <linux/in.h>
#include <linux/ip.h>
#include <linux/jhash.h>
#include <linux/delay.h>
#include <linux/time.h>
#include <linux/etherdevice.h>
#include <linux/genetlink.h>
#include <linux/kernel.h>
#include <linux/kthread.h>
#include <linux/mutex.h>
#include <linux/percpu.h>
#include <linux/rcupdate.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/ethtool.h>
#include <linux/wait.h>
#include <asm/div64.h>
#include <linux/highmem.h>
#include <linux/netfilter_bridge.h>
#include <linux/netfilter_ipv4.h>
#include <linux/inetdevice.h>
#include <linux/list.h>
47
#include <linux/lockdep.h>
48 49 50 51 52
#include <linux/openvswitch.h>
#include <linux/rculist.h>
#include <linux/dmi.h>
#include <linux/workqueue.h>
#include <net/genetlink.h>
53 54
#include <net/net_namespace.h>
#include <net/netns/generic.h>
55 56 57 58

#include "datapath.h"
#include "flow.h"
#include "vport-internal_dev.h"
59
#include "vport-netdev.h"
60

61 62 63

#define REHASH_FLOW_INTERVAL (10 * 60 * HZ)

64 65
int ovs_net_id __read_mostly;

66 67 68 69 70 71 72
static void ovs_notify(struct sk_buff *skb, struct genl_info *info,
		       struct genl_multicast_group *grp)
{
	genl_notify(skb, genl_info_net(info), info->snd_portid,
		    grp->id, info->nlhdr, GFP_KERNEL);
}

73 74 75
/**
 * DOC: Locking:
 *
76 77 78 79
 * All writes e.g. Writes to device state (add/remove datapath, port, set
 * operations on vports, etc.), Writes to other state (flow table
 * modifications, set miscellaneous datapath parameters, etc.) are protected
 * by ovs_lock.
80 81 82 83 84 85
 *
 * Reads are protected by RCU.
 *
 * There are a few special cases (mostly stats) that have their own
 * synchronization but they nest under all of above and don't interact with
 * each other.
86 87
 *
 * The RTNL lock nests inside ovs_mutex.
88 89
 */

90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111
static DEFINE_MUTEX(ovs_mutex);

void ovs_lock(void)
{
	mutex_lock(&ovs_mutex);
}

void ovs_unlock(void)
{
	mutex_unlock(&ovs_mutex);
}

#ifdef CONFIG_LOCKDEP
int lockdep_ovsl_is_held(void)
{
	if (debug_locks)
		return lockdep_is_held(&ovs_mutex);
	else
		return 1;
}
#endif

112
static struct vport *new_vport(const struct vport_parms *);
113
static int queue_gso_packets(struct net *, int dp_ifindex, struct sk_buff *,
114
			     const struct dp_upcall_info *);
115 116
static int queue_userspace_packet(struct net *, int dp_ifindex,
				  struct sk_buff *,
117 118
				  const struct dp_upcall_info *);

119
/* Must be called with rcu_read_lock or ovs_mutex. */
120
static struct datapath *get_dp(struct net *net, int dp_ifindex)
121 122 123 124 125
{
	struct datapath *dp = NULL;
	struct net_device *dev;

	rcu_read_lock();
126
	dev = dev_get_by_index_rcu(net, dp_ifindex);
127 128 129 130 131 132 133 134 135 136
	if (dev) {
		struct vport *vport = ovs_internal_dev_get_vport(dev);
		if (vport)
			dp = vport->dp;
	}
	rcu_read_unlock();

	return dp;
}

137
/* Must be called with rcu_read_lock or ovs_mutex. */
138 139
const char *ovs_dp_name(const struct datapath *dp)
{
140
	struct vport *vport = ovs_vport_ovsl_rcu(dp, OVSP_LOCAL);
141 142 143 144 145 146 147 148 149 150
	return vport->ops->get_name(vport);
}

static int get_dpifindex(struct datapath *dp)
{
	struct vport *local;
	int ifindex;

	rcu_read_lock();

151
	local = ovs_vport_rcu(dp, OVSP_LOCAL);
152
	if (local)
153
		ifindex = netdev_vport_priv(local)->dev->ifindex;
154 155 156 157 158 159 160 161 162 163 164 165
	else
		ifindex = 0;

	rcu_read_unlock();

	return ifindex;
}

static void destroy_dp_rcu(struct rcu_head *rcu)
{
	struct datapath *dp = container_of(rcu, struct datapath, rcu);

166
	ovs_flow_tbl_destroy((__force struct flow_table *)dp->table, false);
167
	free_percpu(dp->stats_percpu);
168
	release_net(ovs_dp_get_net(dp));
169
	kfree(dp->ports);
170 171 172
	kfree(dp);
}

173 174 175 176 177 178 179 180 181 182 183 184
static struct hlist_head *vport_hash_bucket(const struct datapath *dp,
					    u16 port_no)
{
	return &dp->ports[port_no & (DP_VPORT_HASH_BUCKETS - 1)];
}

struct vport *ovs_lookup_vport(const struct datapath *dp, u16 port_no)
{
	struct vport *vport;
	struct hlist_head *head;

	head = vport_hash_bucket(dp, port_no);
185
	hlist_for_each_entry_rcu(vport, head, dp_hash_node) {
186 187 188 189 190 191
		if (vport->port_no == port_no)
			return vport;
	}
	return NULL;
}

192
/* Called with ovs_mutex. */
193 194 195 196 197 198 199
static struct vport *new_vport(const struct vport_parms *parms)
{
	struct vport *vport;

	vport = ovs_vport_add(parms);
	if (!IS_ERR(vport)) {
		struct datapath *dp = parms->dp;
200
		struct hlist_head *head = vport_hash_bucket(dp, vport->port_no);
201

202
		hlist_add_head_rcu(&vport->dp_hash_node, head);
203 204 205 206 207 208
	}
	return vport;
}

void ovs_dp_detach_port(struct vport *p)
{
209
	ASSERT_OVSL();
210 211

	/* First drop references to device. */
212
	hlist_del_rcu(&p->dp_hash_node);
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227

	/* Then destroy it. */
	ovs_vport_del(p);
}

/* Must be called with rcu_read_lock. */
void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
{
	struct datapath *dp = p->dp;
	struct sw_flow *flow;
	struct dp_stats_percpu *stats;
	struct sw_flow_key key;
	u64 *stats_counter;
	int error;

228
	stats = this_cpu_ptr(dp->stats_percpu);
229 230

	/* Extract flow from 'skb' into 'key'. */
231
	error = ovs_flow_extract(skb, p->port_no, &key);
232 233 234 235 236 237
	if (unlikely(error)) {
		kfree_skb(skb);
		return;
	}

	/* Look up flow. */
238
	flow = ovs_flow_lookup(rcu_dereference(dp->table), &key);
239 240 241 242 243 244
	if (unlikely(!flow)) {
		struct dp_upcall_info upcall;

		upcall.cmd = OVS_PACKET_CMD_MISS;
		upcall.key = &key;
		upcall.userdata = NULL;
245
		upcall.portid = p->upcall_portid;
246 247 248 249 250 251 252
		ovs_dp_upcall(dp, skb, &upcall);
		consume_skb(skb);
		stats_counter = &stats->n_missed;
		goto out;
	}

	OVS_CB(skb)->flow = flow;
253
	OVS_CB(skb)->pkt_key = &key;
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270

	stats_counter = &stats->n_hit;
	ovs_flow_used(OVS_CB(skb)->flow, skb);
	ovs_execute_actions(dp, skb);

out:
	/* Update datapath statistics. */
	u64_stats_update_begin(&stats->sync);
	(*stats_counter)++;
	u64_stats_update_end(&stats->sync);
}

static struct genl_family dp_packet_genl_family = {
	.id = GENL_ID_GENERATE,
	.hdrsize = sizeof(struct ovs_header),
	.name = OVS_PACKET_FAMILY,
	.version = OVS_PACKET_VERSION,
271
	.maxattr = OVS_PACKET_ATTR_MAX,
272 273
	.netnsok = true,
	.parallel_ops = true,
274 275 276
};

int ovs_dp_upcall(struct datapath *dp, struct sk_buff *skb,
277
		  const struct dp_upcall_info *upcall_info)
278 279 280 281 282
{
	struct dp_stats_percpu *stats;
	int dp_ifindex;
	int err;

283
	if (upcall_info->portid == 0) {
284 285 286 287 288 289 290 291 292 293 294
		err = -ENOTCONN;
		goto err;
	}

	dp_ifindex = get_dpifindex(dp);
	if (!dp_ifindex) {
		err = -ENODEV;
		goto err;
	}

	if (!skb_is_gso(skb))
295
		err = queue_userspace_packet(ovs_dp_get_net(dp), dp_ifindex, skb, upcall_info);
296
	else
297
		err = queue_gso_packets(ovs_dp_get_net(dp), dp_ifindex, skb, upcall_info);
298 299 300 301 302 303
	if (err)
		goto err;

	return 0;

err:
304
	stats = this_cpu_ptr(dp->stats_percpu);
305 306 307 308 309 310 311 312

	u64_stats_update_begin(&stats->sync);
	stats->n_lost++;
	u64_stats_update_end(&stats->sync);

	return err;
}

313 314
static int queue_gso_packets(struct net *net, int dp_ifindex,
			     struct sk_buff *skb,
315 316
			     const struct dp_upcall_info *upcall_info)
{
317
	unsigned short gso_type = skb_shinfo(skb)->gso_type;
318 319 320 321 322
	struct dp_upcall_info later_info;
	struct sw_flow_key later_key;
	struct sk_buff *segs, *nskb;
	int err;

323
	segs = __skb_gso_segment(skb, NETIF_F_SG | NETIF_F_HW_CSUM, false);
324 325
	if (IS_ERR(segs))
		return PTR_ERR(segs);
326 327 328 329

	/* Queue all of the segments. */
	skb = segs;
	do {
330
		err = queue_userspace_packet(net, dp_ifindex, skb, upcall_info);
331 332 333
		if (err)
			break;

334
		if (skb == segs && gso_type & SKB_GSO_UDP) {
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
			/* The initial flow key extracted by ovs_flow_extract()
			 * in this case is for a first fragment, so we need to
			 * properly mark later fragments.
			 */
			later_key = *upcall_info->key;
			later_key.ip.frag = OVS_FRAG_TYPE_LATER;

			later_info = *upcall_info;
			later_info.key = &later_key;
			upcall_info = &later_info;
		}
	} while ((skb = skb->next));

	/* Free all of the segments. */
	skb = segs;
	do {
		nskb = skb->next;
		if (err)
			kfree_skb(skb);
		else
			consume_skb(skb);
	} while ((skb = nskb));
	return err;
}

360 361 362
static size_t key_attr_size(void)
{
	return    nla_total_size(4)   /* OVS_KEY_ATTR_PRIORITY */
363 364 365 366 367 368 369 370
		+ nla_total_size(0)   /* OVS_KEY_ATTR_TUNNEL */
		  + nla_total_size(8)   /* OVS_TUNNEL_KEY_ATTR_ID */
		  + nla_total_size(4)   /* OVS_TUNNEL_KEY_ATTR_IPV4_SRC */
		  + nla_total_size(4)   /* OVS_TUNNEL_KEY_ATTR_IPV4_DST */
		  + nla_total_size(1)   /* OVS_TUNNEL_KEY_ATTR_TOS */
		  + nla_total_size(1)   /* OVS_TUNNEL_KEY_ATTR_TTL */
		  + nla_total_size(0)   /* OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT */
		  + nla_total_size(0)   /* OVS_TUNNEL_KEY_ATTR_CSUM */
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
		+ nla_total_size(4)   /* OVS_KEY_ATTR_IN_PORT */
		+ nla_total_size(4)   /* OVS_KEY_ATTR_SKB_MARK */
		+ nla_total_size(12)  /* OVS_KEY_ATTR_ETHERNET */
		+ nla_total_size(2)   /* OVS_KEY_ATTR_ETHERTYPE */
		+ nla_total_size(4)   /* OVS_KEY_ATTR_8021Q */
		+ nla_total_size(0)   /* OVS_KEY_ATTR_ENCAP */
		+ nla_total_size(2)   /* OVS_KEY_ATTR_ETHERTYPE */
		+ nla_total_size(40)  /* OVS_KEY_ATTR_IPV6 */
		+ nla_total_size(2)   /* OVS_KEY_ATTR_ICMPV6 */
		+ nla_total_size(28); /* OVS_KEY_ATTR_ND */
}

static size_t upcall_msg_size(const struct sk_buff *skb,
			      const struct nlattr *userdata)
{
	size_t size = NLMSG_ALIGN(sizeof(struct ovs_header))
		+ nla_total_size(skb->len) /* OVS_PACKET_ATTR_PACKET */
		+ nla_total_size(key_attr_size()); /* OVS_PACKET_ATTR_KEY */

	/* OVS_PACKET_ATTR_USERDATA */
	if (userdata)
		size += NLA_ALIGN(userdata->nla_len);

	return size;
}

397 398
static int queue_userspace_packet(struct net *net, int dp_ifindex,
				  struct sk_buff *skb,
399 400 401 402 403 404 405 406 407 408 409 410 411
				  const struct dp_upcall_info *upcall_info)
{
	struct ovs_header *upcall;
	struct sk_buff *nskb = NULL;
	struct sk_buff *user_skb; /* to be queued to userspace */
	struct nlattr *nla;
	int err;

	if (vlan_tx_tag_present(skb)) {
		nskb = skb_clone(skb, GFP_ATOMIC);
		if (!nskb)
			return -ENOMEM;

412
		nskb = __vlan_put_tag(nskb, nskb->vlan_proto, vlan_tx_tag_get(nskb));
413
		if (!nskb)
414 415 416 417 418 419 420 421 422 423 424
			return -ENOMEM;

		nskb->vlan_tci = 0;
		skb = nskb;
	}

	if (nla_attr_size(skb->len) > USHRT_MAX) {
		err = -EFBIG;
		goto out;
	}

425
	user_skb = genlmsg_new(upcall_msg_size(skb, upcall_info->userdata), GFP_ATOMIC);
426 427 428 429 430 431 432 433 434 435
	if (!user_skb) {
		err = -ENOMEM;
		goto out;
	}

	upcall = genlmsg_put(user_skb, 0, 0, &dp_packet_genl_family,
			     0, upcall_info->cmd);
	upcall->dp_ifindex = dp_ifindex;

	nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_KEY);
436
	ovs_flow_to_nlattrs(upcall_info->key, upcall_info->key, user_skb);
437 438 439
	nla_nest_end(user_skb, nla);

	if (upcall_info->userdata)
440 441 442
		__nla_put(user_skb, OVS_PACKET_ATTR_USERDATA,
			  nla_len(upcall_info->userdata),
			  nla_data(upcall_info->userdata));
443 444 445 446 447

	nla = __nla_reserve(user_skb, OVS_PACKET_ATTR_PACKET, skb->len);

	skb_copy_and_csum_dev(skb, nla_data(nla));

448
	genlmsg_end(user_skb, upcall);
449
	err = genlmsg_unicast(net, user_skb, upcall_info->portid);
450 451 452 453 454 455

out:
	kfree_skb(nskb);
	return err;
}

456
/* Called with ovs_mutex. */
457
static int flush_flows(struct datapath *dp)
458 459 460 461
{
	struct flow_table *old_table;
	struct flow_table *new_table;

462
	old_table = ovsl_dereference(dp->table);
463 464 465 466 467 468
	new_table = ovs_flow_tbl_alloc(TBL_MIN_BUCKETS);
	if (!new_table)
		return -ENOMEM;

	rcu_assign_pointer(dp->table, new_table);

469
	ovs_flow_tbl_destroy(old_table, true);
470 471 472
	return 0;
}

473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546
static struct nlattr *reserve_sfa_size(struct sw_flow_actions **sfa, int attr_len)
{

	struct sw_flow_actions *acts;
	int new_acts_size;
	int req_size = NLA_ALIGN(attr_len);
	int next_offset = offsetof(struct sw_flow_actions, actions) +
					(*sfa)->actions_len;

	if (req_size <= (ksize(*sfa) - next_offset))
		goto out;

	new_acts_size = ksize(*sfa) * 2;

	if (new_acts_size > MAX_ACTIONS_BUFSIZE) {
		if ((MAX_ACTIONS_BUFSIZE - next_offset) < req_size)
			return ERR_PTR(-EMSGSIZE);
		new_acts_size = MAX_ACTIONS_BUFSIZE;
	}

	acts = ovs_flow_actions_alloc(new_acts_size);
	if (IS_ERR(acts))
		return (void *)acts;

	memcpy(acts->actions, (*sfa)->actions, (*sfa)->actions_len);
	acts->actions_len = (*sfa)->actions_len;
	kfree(*sfa);
	*sfa = acts;

out:
	(*sfa)->actions_len += req_size;
	return  (struct nlattr *) ((unsigned char *)(*sfa) + next_offset);
}

static int add_action(struct sw_flow_actions **sfa, int attrtype, void *data, int len)
{
	struct nlattr *a;

	a = reserve_sfa_size(sfa, nla_attr_size(len));
	if (IS_ERR(a))
		return PTR_ERR(a);

	a->nla_type = attrtype;
	a->nla_len = nla_attr_size(len);

	if (data)
		memcpy(nla_data(a), data, len);
	memset((unsigned char *) a + a->nla_len, 0, nla_padlen(len));

	return 0;
}

static inline int add_nested_action_start(struct sw_flow_actions **sfa, int attrtype)
{
	int used = (*sfa)->actions_len;
	int err;

	err = add_action(sfa, attrtype, NULL, 0);
	if (err)
		return err;

	return used;
}

static inline void add_nested_action_end(struct sw_flow_actions *sfa, int st_offset)
{
	struct nlattr *a = (struct nlattr *) ((unsigned char *)sfa->actions + st_offset);

	a->nla_len = sfa->actions_len - st_offset;
}

static int validate_and_copy_actions(const struct nlattr *attr,
				     const struct sw_flow_key *key, int depth,
				     struct sw_flow_actions **sfa);
547

548 549 550
static int validate_and_copy_sample(const struct nlattr *attr,
				    const struct sw_flow_key *key, int depth,
				    struct sw_flow_actions **sfa)
551 552 553 554
{
	const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
	const struct nlattr *probability, *actions;
	const struct nlattr *a;
555
	int rem, start, err, st_acts;
556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573

	memset(attrs, 0, sizeof(attrs));
	nla_for_each_nested(a, attr, rem) {
		int type = nla_type(a);
		if (!type || type > OVS_SAMPLE_ATTR_MAX || attrs[type])
			return -EINVAL;
		attrs[type] = a;
	}
	if (rem)
		return -EINVAL;

	probability = attrs[OVS_SAMPLE_ATTR_PROBABILITY];
	if (!probability || nla_len(probability) != sizeof(u32))
		return -EINVAL;

	actions = attrs[OVS_SAMPLE_ATTR_ACTIONS];
	if (!actions || (nla_len(actions) && nla_len(actions) < NLA_HDRLEN))
		return -EINVAL;
574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593

	/* validation done, copy sample action. */
	start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SAMPLE);
	if (start < 0)
		return start;
	err = add_action(sfa, OVS_SAMPLE_ATTR_PROBABILITY, nla_data(probability), sizeof(u32));
	if (err)
		return err;
	st_acts = add_nested_action_start(sfa, OVS_SAMPLE_ATTR_ACTIONS);
	if (st_acts < 0)
		return st_acts;

	err = validate_and_copy_actions(actions, key, depth + 1, sfa);
	if (err)
		return err;

	add_nested_action_end(*sfa, st_acts);
	add_nested_action_end(*sfa, start);

	return 0;
594 595
}

596 597 598
static int validate_tp_port(const struct sw_flow_key *flow_key)
{
	if (flow_key->eth.type == htons(ETH_P_IP)) {
599
		if (flow_key->ipv4.tp.src || flow_key->ipv4.tp.dst)
600 601
			return 0;
	} else if (flow_key->eth.type == htons(ETH_P_IPV6)) {
602
		if (flow_key->ipv6.tp.src || flow_key->ipv6.tp.dst)
603 604 605 606 607 608
			return 0;
	}

	return -EINVAL;
}

609 610 611
static int validate_and_copy_set_tun(const struct nlattr *attr,
				     struct sw_flow_actions **sfa)
{
612 613
	struct sw_flow_match match;
	struct sw_flow_key key;
614 615
	int err, start;

616 617
	ovs_match_init(&match, &key, NULL);
	err = ovs_ipv4_tun_from_nlattr(nla_data(attr), &match, false);
618 619 620 621 622 623 624
	if (err)
		return err;

	start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SET);
	if (start < 0)
		return start;

625 626
	err = add_action(sfa, OVS_KEY_ATTR_IPV4_TUNNEL, &match.key->tun_key,
			sizeof(match.key->tun_key));
627 628 629 630 631
	add_nested_action_end(*sfa, start);

	return err;
}

632
static int validate_set(const struct nlattr *a,
633 634 635
			const struct sw_flow_key *flow_key,
			struct sw_flow_actions **sfa,
			bool *set_tun)
636 637 638 639 640 641 642 643 644
{
	const struct nlattr *ovs_key = nla_data(a);
	int key_type = nla_type(ovs_key);

	/* There can be only one key in a action */
	if (nla_total_size(nla_len(ovs_key)) != nla_len(a))
		return -EINVAL;

	if (key_type > OVS_KEY_ATTR_MAX ||
645 646
	   (ovs_key_lens[key_type] != nla_len(ovs_key) &&
	    ovs_key_lens[key_type] != -1))
647 648 649 650
		return -EINVAL;

	switch (key_type) {
	const struct ovs_key_ipv4 *ipv4_key;
651
	const struct ovs_key_ipv6 *ipv6_key;
652
	int err;
653 654

	case OVS_KEY_ATTR_PRIORITY:
655
	case OVS_KEY_ATTR_SKB_MARK:
656 657 658
	case OVS_KEY_ATTR_ETHERNET:
		break;

659 660 661 662 663 664 665
	case OVS_KEY_ATTR_TUNNEL:
		*set_tun = true;
		err = validate_and_copy_set_tun(a, sfa);
		if (err)
			return err;
		break;

666 667 668 669
	case OVS_KEY_ATTR_IPV4:
		if (flow_key->eth.type != htons(ETH_P_IP))
			return -EINVAL;

670
		if (!flow_key->ip.proto)
671 672 673 674 675 676 677 678 679 680 681
			return -EINVAL;

		ipv4_key = nla_data(ovs_key);
		if (ipv4_key->ipv4_proto != flow_key->ip.proto)
			return -EINVAL;

		if (ipv4_key->ipv4_frag != flow_key->ip.frag)
			return -EINVAL;

		break;

682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700
	case OVS_KEY_ATTR_IPV6:
		if (flow_key->eth.type != htons(ETH_P_IPV6))
			return -EINVAL;

		if (!flow_key->ip.proto)
			return -EINVAL;

		ipv6_key = nla_data(ovs_key);
		if (ipv6_key->ipv6_proto != flow_key->ip.proto)
			return -EINVAL;

		if (ipv6_key->ipv6_frag != flow_key->ip.frag)
			return -EINVAL;

		if (ntohl(ipv6_key->ipv6_label) & 0xFFF00000)
			return -EINVAL;

		break;

701 702 703 704
	case OVS_KEY_ATTR_TCP:
		if (flow_key->ip.proto != IPPROTO_TCP)
			return -EINVAL;

705
		return validate_tp_port(flow_key);
706 707 708 709 710

	case OVS_KEY_ATTR_UDP:
		if (flow_key->ip.proto != IPPROTO_UDP)
			return -EINVAL;

711
		return validate_tp_port(flow_key);
712

Joe Stringer's avatar
Joe Stringer committed
713 714 715 716 717 718
	case OVS_KEY_ATTR_SCTP:
		if (flow_key->ip.proto != IPPROTO_SCTP)
			return -EINVAL;

		return validate_tp_port(flow_key);

719 720 721 722 723 724 725 726 727 728 729
	default:
		return -EINVAL;
	}

	return 0;
}

static int validate_userspace(const struct nlattr *attr)
{
	static const struct nla_policy userspace_policy[OVS_USERSPACE_ATTR_MAX + 1] =	{
		[OVS_USERSPACE_ATTR_PID] = {.type = NLA_U32 },
730
		[OVS_USERSPACE_ATTR_USERDATA] = {.type = NLA_UNSPEC },
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746
	};
	struct nlattr *a[OVS_USERSPACE_ATTR_MAX + 1];
	int error;

	error = nla_parse_nested(a, OVS_USERSPACE_ATTR_MAX,
				 attr, userspace_policy);
	if (error)
		return error;

	if (!a[OVS_USERSPACE_ATTR_PID] ||
	    !nla_get_u32(a[OVS_USERSPACE_ATTR_PID]))
		return -EINVAL;

	return 0;
}

747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764
static int copy_action(const struct nlattr *from,
		       struct sw_flow_actions **sfa)
{
	int totlen = NLA_ALIGN(from->nla_len);
	struct nlattr *to;

	to = reserve_sfa_size(sfa, from->nla_len);
	if (IS_ERR(to))
		return PTR_ERR(to);

	memcpy(to, from, totlen);
	return 0;
}

static int validate_and_copy_actions(const struct nlattr *attr,
				     const struct sw_flow_key *key,
				     int depth,
				     struct sw_flow_actions **sfa)
765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783
{
	const struct nlattr *a;
	int rem, err;

	if (depth >= SAMPLE_ACTION_DEPTH)
		return -EOVERFLOW;

	nla_for_each_nested(a, attr, rem) {
		/* Expected argument lengths, (u32)-1 for variable length. */
		static const u32 action_lens[OVS_ACTION_ATTR_MAX + 1] = {
			[OVS_ACTION_ATTR_OUTPUT] = sizeof(u32),
			[OVS_ACTION_ATTR_USERSPACE] = (u32)-1,
			[OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct ovs_action_push_vlan),
			[OVS_ACTION_ATTR_POP_VLAN] = 0,
			[OVS_ACTION_ATTR_SET] = (u32)-1,
			[OVS_ACTION_ATTR_SAMPLE] = (u32)-1
		};
		const struct ovs_action_push_vlan *vlan;
		int type = nla_type(a);
784
		bool skip_copy;
785 786 787 788 789 790

		if (type > OVS_ACTION_ATTR_MAX ||
		    (action_lens[type] != nla_len(a) &&
		     action_lens[type] != (u32)-1))
			return -EINVAL;

791
		skip_copy = false;
792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819
		switch (type) {
		case OVS_ACTION_ATTR_UNSPEC:
			return -EINVAL;

		case OVS_ACTION_ATTR_USERSPACE:
			err = validate_userspace(a);
			if (err)
				return err;
			break;

		case OVS_ACTION_ATTR_OUTPUT:
			if (nla_get_u32(a) >= DP_MAX_PORTS)
				return -EINVAL;
			break;


		case OVS_ACTION_ATTR_POP_VLAN:
			break;

		case OVS_ACTION_ATTR_PUSH_VLAN:
			vlan = nla_data(a);
			if (vlan->vlan_tpid != htons(ETH_P_8021Q))
				return -EINVAL;
			if (!(vlan->vlan_tci & htons(VLAN_TAG_PRESENT)))
				return -EINVAL;
			break;

		case OVS_ACTION_ATTR_SET:
820
			err = validate_set(a, key, sfa, &skip_copy);
821 822 823 824 825
			if (err)
				return err;
			break;

		case OVS_ACTION_ATTR_SAMPLE:
826
			err = validate_and_copy_sample(a, key, depth, sfa);
827 828
			if (err)
				return err;
829
			skip_copy = true;
830 831 832 833 834
			break;

		default:
			return -EINVAL;
		}
835 836 837 838 839
		if (!skip_copy) {
			err = copy_action(a, sfa);
			if (err)
				return err;
		}
840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869
	}

	if (rem > 0)
		return -EINVAL;

	return 0;
}

static void clear_stats(struct sw_flow *flow)
{
	flow->used = 0;
	flow->tcp_flags = 0;
	flow->packet_count = 0;
	flow->byte_count = 0;
}

static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
{
	struct ovs_header *ovs_header = info->userhdr;
	struct nlattr **a = info->attrs;
	struct sw_flow_actions *acts;
	struct sk_buff *packet;
	struct sw_flow *flow;
	struct datapath *dp;
	struct ethhdr *eth;
	int len;
	int err;

	err = -EINVAL;
	if (!a[OVS_PACKET_ATTR_PACKET] || !a[OVS_PACKET_ATTR_KEY] ||
870
	    !a[OVS_PACKET_ATTR_ACTIONS])
871 872 873 874 875 876 877 878 879
		goto err;

	len = nla_len(a[OVS_PACKET_ATTR_PACKET]);
	packet = __dev_alloc_skb(NET_IP_ALIGN + len, GFP_KERNEL);
	err = -ENOMEM;
	if (!packet)
		goto err;
	skb_reserve(packet, NET_IP_ALIGN);

880
	nla_memcpy(__skb_put(packet, len), a[OVS_PACKET_ATTR_PACKET], len);
881 882 883 884 885 886 887

	skb_reset_mac_header(packet);
	eth = eth_hdr(packet);

	/* Normally, setting the skb 'protocol' field would be handled by a
	 * call to eth_type_trans(), but it assumes there's a sending
	 * device, which we may not have. */
Simon Horman's avatar
Simon Horman committed
888
	if (ntohs(eth->h_proto) >= ETH_P_802_3_MIN)
889 890 891 892 893 894 895 896 897 898
		packet->protocol = eth->h_proto;
	else
		packet->protocol = htons(ETH_P_802_2);

	/* Build an sw_flow for sending this packet. */
	flow = ovs_flow_alloc();
	err = PTR_ERR(flow);
	if (IS_ERR(flow))
		goto err_kfree_skb;

899
	err = ovs_flow_extract(packet, -1, &flow->key);
900 901 902
	if (err)
		goto err_flow_free;

903
	err = ovs_flow_metadata_from_nlattrs(flow, a[OVS_PACKET_ATTR_KEY]);
904 905
	if (err)
		goto err_flow_free;
906
	acts = ovs_flow_actions_alloc(nla_len(a[OVS_PACKET_ATTR_ACTIONS]));
907 908 909
	err = PTR_ERR(acts);
	if (IS_ERR(acts))
		goto err_flow_free;
910 911

	err = validate_and_copy_actions(a[OVS_PACKET_ATTR_ACTIONS], &flow->key, 0, &acts);
912
	rcu_assign_pointer(flow->sf_acts, acts);
913 914
	if (err)
		goto err_flow_free;
915 916

	OVS_CB(packet)->flow = flow;
917
	OVS_CB(packet)->pkt_key = &flow->key;
918
	packet->priority = flow->key.phy.priority;
919
	packet->mark = flow->key.phy.skb_mark;
920 921

	rcu_read_lock();
922
	dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
923 924 925 926 927 928 929 930 931
	err = -ENODEV;
	if (!dp)
		goto err_unlock;

	local_bh_disable();
	err = ovs_execute_actions(dp, packet);
	local_bh_enable();
	rcu_read_unlock();

932
	ovs_flow_free(flow, false);
933 934 935 936 937
	return err;

err_unlock:
	rcu_read_unlock();
err_flow_free:
938
	ovs_flow_free(flow, false);
939 940 941 942 943 944 945
err_kfree_skb:
	kfree_skb(packet);
err:
	return err;
}

static const struct nla_policy packet_policy[OVS_PACKET_ATTR_MAX + 1] = {
946
	[OVS_PACKET_ATTR_PACKET] = { .len = ETH_HLEN },
947 948 949 950 951 952 953 954 955 956 957 958 959 960
	[OVS_PACKET_ATTR_KEY] = { .type = NLA_NESTED },
	[OVS_PACKET_ATTR_ACTIONS] = { .type = NLA_NESTED },
};

static struct genl_ops dp_packet_genl_ops[] = {
	{ .cmd = OVS_PACKET_CMD_EXECUTE,
	  .flags = GENL_ADMIN_PERM, /* Requires CAP_NET_ADMIN privilege. */
	  .policy = packet_policy,
	  .doit = ovs_packet_cmd_execute
	}
};

static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats)
{
961
	struct flow_table *table;
962 963
	int i;

964
	table = rcu_dereference_check(dp->table, lockdep_ovsl_is_held());
965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996
	stats->n_flows = ovs_flow_tbl_count(table);

	stats->n_hit = stats->n_missed = stats->n_lost = 0;
	for_each_possible_cpu(i) {
		const struct dp_stats_percpu *percpu_stats;
		struct dp_stats_percpu local_stats;
		unsigned int start;

		percpu_stats = per_cpu_ptr(dp->stats_percpu, i);

		do {
			start = u64_stats_fetch_begin_bh(&percpu_stats->sync);
			local_stats = *percpu_stats;
		} while (u64_stats_fetch_retry_bh(&percpu_stats->sync, start));

		stats->n_hit += local_stats.n_hit;
		stats->n_missed += local_stats.n_missed;
		stats->n_lost += local_stats.n_lost;
	}
}

static const struct nla_policy flow_policy[OVS_FLOW_ATTR_MAX + 1] = {
	[OVS_FLOW_ATTR_KEY] = { .type = NLA_NESTED },
	[OVS_FLOW_ATTR_ACTIONS] = { .type = NLA_NESTED },
	[OVS_FLOW_ATTR_CLEAR] = { .type = NLA_FLAG },
};

static struct genl_family dp_flow_genl_family = {
	.id = GENL_ID_GENERATE,
	.hdrsize = sizeof(struct ovs_header),
	.name = OVS_FLOW_FAMILY,
	.version = OVS_FLOW_VERSION,
997
	.maxattr = OVS_FLOW_ATTR_MAX,
998 999
	.netnsok = true,
	.parallel_ops = true,
1000 1001 1002 1003 1004 1005
};

static struct genl_multicast_group ovs_dp_flow_multicast_group = {
	.name = OVS_FLOW_MCGROUP
};

1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041
static int actions_to_attr(const struct nlattr *attr, int len, struct sk_buff *skb);
static int sample_action_to_attr(const struct nlattr *attr, struct sk_buff *skb)
{
	const struct nlattr *a;
	struct nlattr *start;
	int err = 0, rem;

	start = nla_nest_start(skb, OVS_ACTION_ATTR_SAMPLE);
	if (!start)
		return -EMSGSIZE;

	nla_for_each_nested(a, attr, rem) {
		int type = nla_type(a);
		struct nlattr *st_sample;

		switch (type) {
		case OVS_SAMPLE_ATTR_PROBABILITY:
			if (nla_put(skb, OVS_SAMPLE_ATTR_PROBABILITY, sizeof(u32), nla_data(a)))
				return -EMSGSIZE;
			break;
		case OVS_SAMPLE_ATTR_ACTIONS:
			st_sample = nla_nest_start(skb, OVS_SAMPLE_ATTR_ACTIONS);
			if (!st_sample)
				return -EMSGSIZE;
			err = actions_to_attr(nla_data(a), nla_len(a), skb);
			if (err)
				return err;
			nla_nest_end(skb, st_sample);
			break;
		}
	}

	nla_nest_end(skb, start);
	return err;
}

1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
static int set_action_to_attr(const struct nlattr *a, struct sk_buff *skb)
{
	const struct nlattr *ovs_key = nla_data(a);
	int key_type = nla_type(ovs_key);
	struct nlattr *start;
	int err;

	switch (key_type) {
	case OVS_KEY_ATTR_IPV4_TUNNEL:
		start = nla_nest_start(skb, OVS_ACTION_ATTR_SET);
		if (!start)
			return -EMSGSIZE;

1055 1056
		err = ovs_ipv4_tun_to_nlattr(skb, nla_data(ovs_key),
					     nla_data(ovs_key));
1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069
		if (err)
			return err;
		nla_nest_end(skb, start);
		break;
	default:
		if (nla_put(skb, OVS_ACTION_ATTR_SET, nla_len(a), ovs_key))
			return -EMSGSIZE;
		break;
	}

	return 0;
}

1070 1071 1072 1073 1074 1075 1076 1077 1078
static int actions_to_attr(const struct nlattr *attr, int len, struct sk_buff *skb)
{
	const struct nlattr *a;
	int rem, err;

	nla_for_each_attr(a, attr, len, rem) {
		int type = nla_type(a);

		switch (type) {
1079 1080 1081 1082 1083 1084
		case OVS_ACTION_ATTR_SET:
			err = set_action_to_attr(a, skb);
			if (err)
				return err;
			break;

1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099
		case OVS_ACTION_ATTR_SAMPLE:
			err = sample_action_to_attr(a, skb);
			if (err)
				return err;
			break;
		default:
			if (nla_put(skb, type, nla_len(a), nla_data(a)))
				return -EMSGSIZE;
			break;
		}
	}

	return 0;
}

1100 1101 1102 1103
static size_t ovs_flow_cmd_msg_size(const struct sw_flow_actions *acts)
{
	return NLMSG_ALIGN(sizeof(struct ovs_header))
		+ nla_total_size(key_attr_size()) /* OVS_FLOW_ATTR_KEY */
1104
		+ nla_total_size(key_attr_size()) /* OVS_FLOW_ATTR_MASK */
1105 1106 1107 1108 1109 1110
		+ nla_total_size(sizeof(struct ovs_flow_stats)) /* OVS_FLOW_ATTR_STATS */
		+ nla_total_size(1) /* OVS_FLOW_ATTR_TCP_FLAGS */
		+ nla_total_size(8) /* OVS_FLOW_ATTR_USED */
		+ nla_total_size(acts->actions_len); /* OVS_FLOW_ATTR_ACTIONS */
}

1111
/* Called with ovs_mutex. */
1112
static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
1113
				  struct sk_buff *skb, u32 portid,
1114 1115 1116
				  u32 seq, u32 flags, u8 cmd)
{
	const int skb_orig_len = skb->len;
1117
	struct nlattr *start;
1118 1119 1120 1121 1122 1123 1124
	struct ovs_flow_stats stats;
	struct ovs_header *ovs_header;
	struct nlattr *nla;
	unsigned long used;
	u8 tcp_flags;
	int err;

1125
	ovs_header = genlmsg_put(skb, portid, seq, &dp_flow_genl_family, flags, cmd);
1126 1127 1128 1129 1130
	if (!ovs_header)
		return -EMSGSIZE;

	ovs_header->dp_ifindex = get_dpifindex(dp);

1131
	/* Fill flow key. */
1132 1133 1134
	nla = nla_nest_start(skb, OVS_FLOW_ATTR_KEY);
	if (!nla)
		goto nla_put_failure;
1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146

	err = ovs_flow_to_nlattrs(&flow->unmasked_key,
			&flow->unmasked_key, skb);
	if (err)
		goto error;
	nla_nest_end(skb, nla);

	nla = nla_nest_start(skb, OVS_FLOW_ATTR_MASK);
	if (!nla)
		goto nla_put_failure;

	err = ovs_flow_to_nlattrs(&flow->key, &flow->mask->key, skb);
1147 1148
	if (err)
		goto error;
1149

1150 1151 1152 1153 1154 1155 1156 1157 1158
	nla_nest_end(skb, nla);

	spin_lock_bh(&flow->lock);
	used = flow->used;
	stats.n_packets = flow->packet_count;
	stats.n_bytes = flow->byte_count;
	tcp_flags = flow->tcp_flags;
	spin_unlock_bh(&flow->lock);

1159 1160 1161
	if (used &&
	    nla_put_u64(skb, OVS_FLOW_ATTR_USED, ovs_flow_used_time(used)))
		goto nla_put_failure;
1162

1163 1164 1165 1166
	if (stats.n_packets &&
	    nla_put(skb, OVS_FLOW_ATTR_STATS,
		    sizeof(struct ovs_flow_stats), &stats))
		goto nla_put_failure;
1167

1168 1169 1170
	if (tcp_flags &&
	    nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS, tcp_flags))
		goto nla_put_failure;
1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181

	/* If OVS_FLOW_ATTR_ACTIONS doesn't fit, skip dumping the actions if
	 * this is the first flow to be dumped into 'skb'.  This is unusual for
	 * Netlink but individual action lists can be longer than
	 * NLMSG_GOODSIZE and thus entirely undumpable if we didn't do this.
	 * The userspace caller can always fetch the actions separately if it
	 * really wants them.  (Most userspace callers in fact don't care.)
	 *
	 * This can only fail for dump operations because the skb is always
	 * properly sized for single flows.
	 */
1182 1183
	start = nla_nest_start(skb, OVS_FLOW_ATTR_ACTIONS);
	if (start) {
1184 1185 1186 1187 1188
		const struct sw_flow_actions *sf_acts;

		sf_acts = rcu_dereference_check(flow->sf_acts,
						lockdep_ovsl_is_held());

1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199
		err = actions_to_attr(sf_acts->actions, sf_acts->actions_len, skb);
		if (!err)
			nla_nest_end(skb, start);
		else {
			if (skb_orig_len)
				goto error;

			nla_nest_cancel(skb, start);
		}
	} else if (skb_orig_len)
		goto nla_put_failure;
1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213

	return genlmsg_end(skb, ovs_header);

nla_put_failure:
	err = -EMSGSIZE;
error:
	genlmsg_cancel(skb, ovs_header);
	return err;
}

static struct sk_buff *ovs_flow_cmd_alloc_info(struct sw_flow *flow)
{
	const struct sw_flow_actions *sf_acts;

1214
	sf_acts = ovsl_dereference(flow->sf_acts);
1215

1216
	return genlmsg_new(ovs_flow_cmd_msg_size(sf_acts), GFP_KERNEL);
1217 1218 1219 1220
}

static struct sk_buff *ovs_flow_cmd_build_info(struct sw_flow *flow,
					       struct datapath *dp,
1221
					       u32 portid, u32 seq, u8 cmd)
1222 1223 1224 1225 1226 1227 1228 1229
{
	struct sk_buff *skb;
	int retval;

	skb = ovs_flow_cmd_alloc_info(flow);
	if (!skb)
		return ERR_PTR(-ENOMEM);

1230
	retval = ovs_flow_cmd_fill_info(flow, dp, skb, portid, seq, 0, cmd);
1231 1232 1233 1234 1235 1236 1237 1238
	BUG_ON(retval < 0);
	return skb;
}

static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
{
	struct nlattr **a = info->attrs;
	struct ovs_header *ovs_header = info->userhdr;
1239 1240 1241
	struct sw_flow_key key, masked_key;
	struct sw_flow *flow = NULL;
	struct sw_flow_mask mask;
1242 1243 1244
	struct sk_buff *reply;
	struct datapath *dp;
	struct flow_table *table;
1245
	struct sw_flow_actions *acts = NULL;
1246
	struct sw_flow_match match;
1247 1248 1249 1250 1251 1252
	int error;

	/* Extract key. */
	error = -EINVAL;
	if (!a[OVS_FLOW_ATTR_KEY])
		goto error;
1253 1254 1255 1256

	ovs_match_init(&match, &key, &mask);
	error = ovs_match_from_nlattrs(&match,
			a[OVS_FLOW_ATTR_KEY], a[OVS_FLOW_ATTR_MASK]);
1257 1258 1259 1260 1261
	if (error)
		goto error;

	/* Validate actions. */
	if (a[OVS_FLOW_ATTR_ACTIONS]) {
1262 1263 1264
		acts = ovs_flow_actions_alloc(nla_len(a[OVS_FLOW_ATTR_ACTIONS]));
		error = PTR_ERR(acts);
		if (IS_ERR(acts))
1265
			goto error;
1266

1267 1268 1269 1270 1271
		ovs_flow_key_mask(&masked_key, &key, &mask);
		error = validate_and_copy_actions(a[OVS_FLOW_ATTR_ACTIONS],
						  &masked_key, 0, &acts);
		if (error) {
			OVS_NLERR("Flow actions may not be safe on all matching packets.\n");
1272
			goto err_kfree;
1273
		}
1274 1275 1276 1277 1278
	} else if (info->genlhdr->cmd == OVS_FLOW_CMD_NEW) {
		error = -EINVAL;
		goto error;
	}

1279
	ovs_lock();
1280
	dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
1281 1282
	error = -ENODEV;
	if (!dp)
1283
		goto err_unlock_ovs;
1284

1285
	table = ovsl_dereference(dp->table);
1286 1287 1288

	/* Check if this is a duplicate flow */
	flow = ovs_flow_lookup(table, &key);
1289
	if (!flow) {
1290
		struct flow_table *new_table = NULL;
1291
		struct sw_flow_mask *mask_p;
1292

1293 1294 1295
		/* Bail out if we're not allowed to create a new flow. */
		error = -ENOENT;
		if (info->genlhdr->cmd == OVS_FLOW_CMD_SET)
1296
			goto err_unlock_ovs;
1297 1298

		/* Expand table, if necessary, to make room. */
1299
		if (ovs_flow_tbl_need_to_expand(table))
1300
			new_table = ovs_flow_tbl_expand(table);
1301 1302 1303 1304 1305 1306 1307 1308
		else if (time_after(jiffies, dp->last_rehash + REHASH_FLOW_INTERVAL))
			new_table = ovs_flow_tbl_rehash(table);

		if (new_table && !IS_ERR(new_table)) {
			rcu_assign_pointer(dp->table, new_table);
			ovs_flow_tbl_destroy(table, true);
			table = ovsl_dereference(dp->table);
			dp->last_rehash = jiffies;
1309 1310 1311 1312 1313 1314
		}

		/* Allocate flow. */
		flow = ovs_flow_alloc();
		if (IS_ERR(flow)) {
			error = PTR_ERR(flow);
1315
			goto err_unlock_ovs;
1316 1317 1318
		}
		clear_stats(flow);

1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335
		flow->key = masked_key;
		flow->unmasked_key = key;

		/* Make sure mask is unique in the system */
		mask_p = ovs_sw_flow_mask_find(table, &mask);
		if (!mask_p) {
			/* Allocate a new mask if none exsits. */
			mask_p = ovs_sw_flow_mask_alloc();
			if (!mask_p)
				goto err_flow_free;
			mask_p->key = mask.key;
			mask_p->range = mask.range;
			ovs_sw_flow_mask_insert(table, mask_p);
		}

		ovs_sw_flow_mask_add_ref(mask_p);
		flow->mask = mask_p;