team.c 72.6 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0-or-later
2
/*
3
 * drivers/net/team/team.c - Network team device driver
4 5 6 7 8 9 10 11 12 13 14 15 16
 * Copyright (c) 2011 Jiri Pirko <jpirko@redhat.com>
 */

#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/rcupdate.h>
#include <linux/errno.h>
#include <linux/ctype.h>
#include <linux/notifier.h>
#include <linux/netdevice.h>
J
Jiri Pirko 已提交
17
#include <linux/netpoll.h>
18
#include <linux/if_vlan.h>
19 20 21 22 23 24 25
#include <linux/if_arp.h>
#include <linux/socket.h>
#include <linux/etherdevice.h>
#include <linux/rtnetlink.h>
#include <net/rtnetlink.h>
#include <net/genetlink.h>
#include <net/netlink.h>
J
Jiri Pirko 已提交
26
#include <net/sch_generic.h>
F
Flavio Leitner 已提交
27
#include <generated/utsrelease.h>
28 29 30 31 32 33 34 35 36 37 38 39 40
#include <linux/if_team.h>

#define DRV_NAME "team"


/**********
 * Helpers
 **********/

static struct team_port *team_port_get_rtnl(const struct net_device *dev)
{
	struct team_port *port = rtnl_dereference(dev->rx_handler_data);

J
Julian Wiedmann 已提交
41
	return netif_is_team_port(dev) ? port : NULL;
42 43 44
}

/*
45
 * Since the ability to change device address for open port device is tested in
46 47
 * team_port_add, this function can be called without control of return value
 */
48 49
static int __set_port_dev_addr(struct net_device *port_dev,
			       const unsigned char *dev_addr)
50
{
51
	struct sockaddr_storage addr;
52

53 54
	memcpy(addr.__data, dev_addr, port_dev->addr_len);
	addr.ss_family = port_dev->type;
55
	return dev_set_mac_address(port_dev, (struct sockaddr *)&addr, NULL);
56 57
}

58
static int team_port_set_orig_dev_addr(struct team_port *port)
59
{
60
	return __set_port_dev_addr(port->dev, port->orig.dev_addr);
61 62
}

63 64
static int team_port_set_team_dev_addr(struct team *team,
				       struct team_port *port)
65
{
66
	return __set_port_dev_addr(port->dev, team->dev->dev_addr);
67
}
68 69 70 71 72 73 74 75 76 77 78 79 80

int team_modeop_port_enter(struct team *team, struct team_port *port)
{
	return team_port_set_team_dev_addr(team, port);
}
EXPORT_SYMBOL(team_modeop_port_enter);

void team_modeop_port_change_dev_addr(struct team *team,
				      struct team_port *port)
{
	team_port_set_team_dev_addr(team, port);
}
EXPORT_SYMBOL(team_modeop_port_change_dev_addr);
81

82 83 84 85 86 87 88 89 90
static void team_lower_state_changed(struct team_port *port)
{
	struct netdev_lag_lower_state_info info;

	info.link_up = port->linkup;
	info.tx_enabled = team_port_enabled(port);
	netdev_lower_state_changed(port->dev, &info);
}

91 92
static void team_refresh_port_linkup(struct team_port *port)
{
93 94 95 96 97 98 99
	bool new_linkup = port->user.linkup_enabled ? port->user.linkup :
						      port->state.linkup;

	if (port->linkup != new_linkup) {
		port->linkup = new_linkup;
		team_lower_state_changed(port);
	}
100
}
101

J
Jiri Pirko 已提交
102

103 104 105 106
/*******************
 * Options handling
 *******************/

107 108
struct team_option_inst { /* One for each option instance */
	struct list_head list;
109
	struct list_head tmp_list;
110
	struct team_option *option;
111
	struct team_option_inst_info info;
112 113 114 115 116 117
	bool changed;
	bool removed;
};

static struct team_option *__team_find_option(struct team *team,
					      const char *opt_name)
J
Jiri Pirko 已提交
118 119 120 121 122 123 124 125 126 127
{
	struct team_option *option;

	list_for_each_entry(option, &team->option_list, list) {
		if (strcmp(option->name, opt_name) == 0)
			return option;
	}
	return NULL;
}

128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144
static void __team_option_inst_del(struct team_option_inst *opt_inst)
{
	list_del(&opt_inst->list);
	kfree(opt_inst);
}

static void __team_option_inst_del_option(struct team *team,
					  struct team_option *option)
{
	struct team_option_inst *opt_inst, *tmp;

	list_for_each_entry_safe(opt_inst, tmp, &team->option_inst_list, list) {
		if (opt_inst->option == option)
			__team_option_inst_del(opt_inst);
	}
}

J
Jiri Pirko 已提交
145 146 147 148 149 150
static int __team_option_inst_add(struct team *team, struct team_option *option,
				  struct team_port *port)
{
	struct team_option_inst *opt_inst;
	unsigned int array_size;
	unsigned int i;
151
	int err;
J
Jiri Pirko 已提交
152 153 154 155 156 157 158 159 160 161

	array_size = option->array_size;
	if (!array_size)
		array_size = 1; /* No array but still need one instance */

	for (i = 0; i < array_size; i++) {
		opt_inst = kmalloc(sizeof(*opt_inst), GFP_KERNEL);
		if (!opt_inst)
			return -ENOMEM;
		opt_inst->option = option;
162 163
		opt_inst->info.port = port;
		opt_inst->info.array_index = i;
J
Jiri Pirko 已提交
164 165 166
		opt_inst->changed = true;
		opt_inst->removed = false;
		list_add_tail(&opt_inst->list, &team->option_inst_list);
167 168 169 170 171 172
		if (option->init) {
			err = option->init(team, &opt_inst->info);
			if (err)
				return err;
		}

J
Jiri Pirko 已提交
173 174 175 176
	}
	return 0;
}

177 178 179 180 181
static int __team_option_inst_add_option(struct team *team,
					 struct team_option *option)
{
	int err;

J
Jiri Pirko 已提交
182
	if (!option->per_port) {
183
		err = __team_option_inst_add(team, option, NULL);
J
Jiri Pirko 已提交
184 185 186
		if (err)
			goto inst_del_option;
	}
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
	return 0;

inst_del_option:
	__team_option_inst_del_option(team, option);
	return err;
}

static void __team_option_inst_mark_removed_option(struct team *team,
						   struct team_option *option)
{
	struct team_option_inst *opt_inst;

	list_for_each_entry(opt_inst, &team->option_inst_list, list) {
		if (opt_inst->option == option) {
			opt_inst->changed = true;
			opt_inst->removed = true;
		}
	}
}

static void __team_option_inst_del_port(struct team *team,
					struct team_port *port)
{
	struct team_option_inst *opt_inst, *tmp;

	list_for_each_entry_safe(opt_inst, tmp, &team->option_inst_list, list) {
		if (opt_inst->option->per_port &&
214
		    opt_inst->info.port == port)
215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244
			__team_option_inst_del(opt_inst);
	}
}

static int __team_option_inst_add_port(struct team *team,
				       struct team_port *port)
{
	struct team_option *option;
	int err;

	list_for_each_entry(option, &team->option_list, list) {
		if (!option->per_port)
			continue;
		err = __team_option_inst_add(team, option, port);
		if (err)
			goto inst_del_port;
	}
	return 0;

inst_del_port:
	__team_option_inst_del_port(team, port);
	return err;
}

static void __team_option_inst_mark_removed_port(struct team *team,
						 struct team_port *port)
{
	struct team_option_inst *opt_inst;

	list_for_each_entry(opt_inst, &team->option_inst_list, list) {
245
		if (opt_inst->info.port == port) {
246 247 248 249 250 251 252 253 254
			opt_inst->changed = true;
			opt_inst->removed = true;
		}
	}
}

static int __team_options_register(struct team *team,
				   const struct team_option *option,
				   size_t option_count)
255 256
{
	int i;
257
	struct team_option **dst_opts;
J
Jiri Pirko 已提交
258
	int err;
259

K
Kees Cook 已提交
260
	dst_opts = kcalloc(option_count, sizeof(struct team_option *),
261 262 263
			   GFP_KERNEL);
	if (!dst_opts)
		return -ENOMEM;
J
Jiri Pirko 已提交
264 265 266
	for (i = 0; i < option_count; i++, option++) {
		if (__team_find_option(team, option->name)) {
			err = -EEXIST;
267
			goto alloc_rollback;
J
Jiri Pirko 已提交
268
		}
269 270
		dst_opts[i] = kmemdup(option, sizeof(*option), GFP_KERNEL);
		if (!dst_opts[i]) {
J
Jiri Pirko 已提交
271
			err = -ENOMEM;
272
			goto alloc_rollback;
J
Jiri Pirko 已提交
273 274 275
		}
	}

276
	for (i = 0; i < option_count; i++) {
277 278 279
		err = __team_option_inst_add_option(team, dst_opts[i]);
		if (err)
			goto inst_rollback;
J
Jiri Pirko 已提交
280
		list_add_tail(&dst_opts[i]->list, &team->option_list);
281
	}
J
Jiri Pirko 已提交
282

283
	kfree(dst_opts);
J
Jiri Pirko 已提交
284 285
	return 0;

286 287 288 289 290 291 292
inst_rollback:
	for (i--; i >= 0; i--)
		__team_option_inst_del_option(team, dst_opts[i]);

	i = option_count - 1;
alloc_rollback:
	for (i--; i >= 0; i--)
J
Jiri Pirko 已提交
293 294
		kfree(dst_opts[i]);

295
	kfree(dst_opts);
J
Jiri Pirko 已提交
296
	return err;
297
}
J
Jiri Pirko 已提交
298

299 300 301 302 303 304 305 306
static void __team_options_mark_removed(struct team *team,
					const struct team_option *option,
					size_t option_count)
{
	int i;

	for (i = 0; i < option_count; i++, option++) {
		struct team_option *del_opt;
307

308
		del_opt = __team_find_option(team, option->name);
309 310
		if (del_opt)
			__team_option_inst_mark_removed_option(team, del_opt);
311 312
	}
}
313 314

static void __team_options_unregister(struct team *team,
J
Jiri Pirko 已提交
315
				      const struct team_option *option,
316 317 318 319
				      size_t option_count)
{
	int i;

J
Jiri Pirko 已提交
320 321 322 323 324
	for (i = 0; i < option_count; i++, option++) {
		struct team_option *del_opt;

		del_opt = __team_find_option(team, option->name);
		if (del_opt) {
325
			__team_option_inst_del_option(team, del_opt);
J
Jiri Pirko 已提交
326 327 328 329
			list_del(&del_opt->list);
			kfree(del_opt);
		}
	}
330 331
}

332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
static void __team_options_change_check(struct team *team);

int team_options_register(struct team *team,
			  const struct team_option *option,
			  size_t option_count)
{
	int err;

	err = __team_options_register(team, option, option_count);
	if (err)
		return err;
	__team_options_change_check(team);
	return 0;
}
EXPORT_SYMBOL(team_options_register);

J
Jiri Pirko 已提交
348 349
void team_options_unregister(struct team *team,
			     const struct team_option *option,
350 351
			     size_t option_count)
{
352 353
	__team_options_mark_removed(team, option, option_count);
	__team_options_change_check(team);
354 355 356 357
	__team_options_unregister(team, option, option_count);
}
EXPORT_SYMBOL(team_options_unregister);

358 359 360 361
static int team_option_get(struct team *team,
			   struct team_option_inst *opt_inst,
			   struct team_gsetter_ctx *ctx)
{
J
Jiri Pirko 已提交
362 363
	if (!opt_inst->option->getter)
		return -EOPNOTSUPP;
364 365 366 367 368 369
	return opt_inst->option->getter(team, ctx);
}

static int team_option_set(struct team *team,
			   struct team_option_inst *opt_inst,
			   struct team_gsetter_ctx *ctx)
370
{
J
Jiri Pirko 已提交
371 372
	if (!opt_inst->option->setter)
		return -EOPNOTSUPP;
373
	return opt_inst->option->setter(team, ctx);
374 375
}

J
Jiri Pirko 已提交
376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391
void team_option_inst_set_change(struct team_option_inst_info *opt_inst_info)
{
	struct team_option_inst *opt_inst;

	opt_inst = container_of(opt_inst_info, struct team_option_inst, info);
	opt_inst->changed = true;
}
EXPORT_SYMBOL(team_option_inst_set_change);

void team_options_change_check(struct team *team)
{
	__team_options_change_check(team);
}
EXPORT_SYMBOL(team_options_change_check);


392 393 394 395 396 397 398
/****************
 * Mode handling
 ****************/

static LIST_HEAD(mode_list);
static DEFINE_SPINLOCK(mode_list_lock);

J
Jiri Pirko 已提交
399 400 401 402 403 404
struct team_mode_item {
	struct list_head list;
	const struct team_mode *mode;
};

static struct team_mode_item *__find_mode(const char *kind)
405
{
J
Jiri Pirko 已提交
406
	struct team_mode_item *mitem;
407

J
Jiri Pirko 已提交
408 409 410
	list_for_each_entry(mitem, &mode_list, list) {
		if (strcmp(mitem->mode->kind, kind) == 0)
			return mitem;
411 412 413 414 415 416 417 418 419 420 421 422 423 424
	}
	return NULL;
}

static bool is_good_mode_name(const char *name)
{
	while (*name != '\0') {
		if (!isalpha(*name) && !isdigit(*name) && *name != '_')
			return false;
		name++;
	}
	return true;
}

J
Jiri Pirko 已提交
425
int team_mode_register(const struct team_mode *mode)
426 427
{
	int err = 0;
J
Jiri Pirko 已提交
428
	struct team_mode_item *mitem;
429 430 431 432

	if (!is_good_mode_name(mode->kind) ||
	    mode->priv_size > TEAM_MODE_PRIV_SIZE)
		return -EINVAL;
J
Jiri Pirko 已提交
433 434 435 436 437

	mitem = kmalloc(sizeof(*mitem), GFP_KERNEL);
	if (!mitem)
		return -ENOMEM;

438 439 440
	spin_lock(&mode_list_lock);
	if (__find_mode(mode->kind)) {
		err = -EEXIST;
J
Jiri Pirko 已提交
441
		kfree(mitem);
442 443
		goto unlock;
	}
J
Jiri Pirko 已提交
444 445
	mitem->mode = mode;
	list_add_tail(&mitem->list, &mode_list);
446 447 448 449 450 451
unlock:
	spin_unlock(&mode_list_lock);
	return err;
}
EXPORT_SYMBOL(team_mode_register);

J
Jiri Pirko 已提交
452
void team_mode_unregister(const struct team_mode *mode)
453
{
J
Jiri Pirko 已提交
454 455
	struct team_mode_item *mitem;

456
	spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
457 458 459 460 461
	mitem = __find_mode(mode->kind);
	if (mitem) {
		list_del_init(&mitem->list);
		kfree(mitem);
	}
462 463 464 465
	spin_unlock(&mode_list_lock);
}
EXPORT_SYMBOL(team_mode_unregister);

J
Jiri Pirko 已提交
466
static const struct team_mode *team_mode_get(const char *kind)
467
{
J
Jiri Pirko 已提交
468 469
	struct team_mode_item *mitem;
	const struct team_mode *mode = NULL;
470 471

	spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
472 473
	mitem = __find_mode(kind);
	if (!mitem) {
474 475 476
		spin_unlock(&mode_list_lock);
		request_module("team-mode-%s", kind);
		spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
477
		mitem = __find_mode(kind);
478
	}
J
Jiri Pirko 已提交
479 480
	if (mitem) {
		mode = mitem->mode;
481 482
		if (!try_module_get(mode->owner))
			mode = NULL;
J
Jiri Pirko 已提交
483
	}
484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499

	spin_unlock(&mode_list_lock);
	return mode;
}

static void team_mode_put(const struct team_mode *mode)
{
	module_put(mode->owner);
}

static bool team_dummy_transmit(struct team *team, struct sk_buff *skb)
{
	dev_kfree_skb_any(skb);
	return false;
}

500 501 502
static rx_handler_result_t team_dummy_receive(struct team *team,
					      struct team_port *port,
					      struct sk_buff *skb)
503 504 505 506
{
	return RX_HANDLER_ANOTHER;
}

507 508 509 510 511 512 513 514 515 516 517
static const struct team_mode __team_no_mode = {
	.kind		= "*NOMODE*",
};

static bool team_is_mode_set(struct team *team)
{
	return team->mode != &__team_no_mode;
}

static void team_set_no_mode(struct team *team)
{
518
	team->user_carrier_enabled = false;
519 520 521
	team->mode = &__team_no_mode;
}

522
static void team_adjust_ops(struct team *team)
523 524 525 526 527 528
{
	/*
	 * To avoid checks in rx/tx skb paths, ensure here that non-null and
	 * correct ops are always set.
	 */

529
	if (!team->en_port_count || !team_is_mode_set(team) ||
530
	    !team->mode->ops->transmit)
531 532 533 534
		team->ops.transmit = team_dummy_transmit;
	else
		team->ops.transmit = team->mode->ops->transmit;

535
	if (!team->en_port_count || !team_is_mode_set(team) ||
536
	    !team->mode->ops->receive)
537 538 539 540 541 542 543 544 545 546 547 548 549 550
		team->ops.receive = team_dummy_receive;
	else
		team->ops.receive = team->mode->ops->receive;
}

/*
 * We can benefit from the fact that it's ensured no port is present
 * at the time of mode change. Therefore no packets are in fly so there's no
 * need to set mode operations in any special way.
 */
static int __team_change_mode(struct team *team,
			      const struct team_mode *new_mode)
{
	/* Check if mode was previously set and do cleanup if so */
551
	if (team_is_mode_set(team)) {
552 553 554 555 556 557 558 559 560
		void (*exit_op)(struct team *team) = team->ops.exit;

		/* Clear ops area so no callback is called any longer */
		memset(&team->ops, 0, sizeof(struct team_mode_ops));
		team_adjust_ops(team);

		if (exit_op)
			exit_op(team);
		team_mode_put(team->mode);
561
		team_set_no_mode(team);
562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
		/* zero private data area */
		memset(&team->mode_priv, 0,
		       sizeof(struct team) - offsetof(struct team, mode_priv));
	}

	if (!new_mode)
		return 0;

	if (new_mode->ops->init) {
		int err;

		err = new_mode->ops->init(team);
		if (err)
			return err;
	}

	team->mode = new_mode;
	memcpy(&team->ops, new_mode->ops, sizeof(struct team_mode_ops));
	team_adjust_ops(team);

	return 0;
}

static int team_change_mode(struct team *team, const char *kind)
{
J
Jiri Pirko 已提交
587
	const struct team_mode *new_mode;
588 589 590 591 592 593 594 595
	struct net_device *dev = team->dev;
	int err;

	if (!list_empty(&team->port_list)) {
		netdev_err(dev, "No ports can be present during mode change\n");
		return -EBUSY;
	}

596
	if (team_is_mode_set(team) && strcmp(team->mode->kind, kind) == 0) {
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618
		netdev_err(dev, "Unable to change to the same mode the team is in\n");
		return -EINVAL;
	}

	new_mode = team_mode_get(kind);
	if (!new_mode) {
		netdev_err(dev, "Mode \"%s\" not found\n", kind);
		return -EINVAL;
	}

	err = __team_change_mode(team, new_mode);
	if (err) {
		netdev_err(dev, "Failed to change to mode \"%s\"\n", kind);
		team_mode_put(new_mode);
		return err;
	}

	netdev_info(dev, "Mode changed to \"%s\"\n", kind);
	return 0;
}


J
Jiri Pirko 已提交
619 620 621 622 623 624 625
/*********************
 * Peers notification
 *********************/

static void team_notify_peers_work(struct work_struct *work)
{
	struct team *team;
626
	int val;
J
Jiri Pirko 已提交
627 628 629 630 631 632 633

	team = container_of(work, struct team, notify_peers.dw.work);

	if (!rtnl_trylock()) {
		schedule_delayed_work(&team->notify_peers.dw, 0);
		return;
	}
634 635 636 637 638
	val = atomic_dec_if_positive(&team->notify_peers.count_pending);
	if (val < 0) {
		rtnl_unlock();
		return;
	}
J
Jiri Pirko 已提交
639 640
	call_netdevice_notifiers(NETDEV_NOTIFY_PEERS, team->dev);
	rtnl_unlock();
641
	if (val)
J
Jiri Pirko 已提交
642 643 644 645 646 647 648 649
		schedule_delayed_work(&team->notify_peers.dw,
				      msecs_to_jiffies(team->notify_peers.interval));
}

static void team_notify_peers(struct team *team)
{
	if (!team->notify_peers.count || !netif_running(team->dev))
		return;
650
	atomic_add(team->notify_peers.count, &team->notify_peers.count_pending);
J
Jiri Pirko 已提交
651 652 653 654 655 656 657 658 659 660 661 662 663 664
	schedule_delayed_work(&team->notify_peers.dw, 0);
}

static void team_notify_peers_init(struct team *team)
{
	INIT_DELAYED_WORK(&team->notify_peers.dw, team_notify_peers_work);
}

static void team_notify_peers_fini(struct team *team)
{
	cancel_delayed_work_sync(&team->notify_peers.dw);
}


665 666 667 668 669 670 671
/*******************************
 * Send multicast group rejoins
 *******************************/

static void team_mcast_rejoin_work(struct work_struct *work)
{
	struct team *team;
672
	int val;
673 674 675 676 677 678 679

	team = container_of(work, struct team, mcast_rejoin.dw.work);

	if (!rtnl_trylock()) {
		schedule_delayed_work(&team->mcast_rejoin.dw, 0);
		return;
	}
680 681 682 683 684
	val = atomic_dec_if_positive(&team->mcast_rejoin.count_pending);
	if (val < 0) {
		rtnl_unlock();
		return;
	}
685 686
	call_netdevice_notifiers(NETDEV_RESEND_IGMP, team->dev);
	rtnl_unlock();
687
	if (val)
688 689 690 691 692 693 694 695
		schedule_delayed_work(&team->mcast_rejoin.dw,
				      msecs_to_jiffies(team->mcast_rejoin.interval));
}

static void team_mcast_rejoin(struct team *team)
{
	if (!team->mcast_rejoin.count || !netif_running(team->dev))
		return;
696
	atomic_add(team->mcast_rejoin.count, &team->mcast_rejoin.count_pending);
697 698 699 700 701 702 703 704 705 706 707 708 709 710
	schedule_delayed_work(&team->mcast_rejoin.dw, 0);
}

static void team_mcast_rejoin_init(struct team *team)
{
	INIT_DELAYED_WORK(&team->mcast_rejoin.dw, team_mcast_rejoin_work);
}

static void team_mcast_rejoin_fini(struct team *team)
{
	cancel_delayed_work_sync(&team->mcast_rejoin.dw);
}


711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730
/************************
 * Rx path frame handler
 ************************/

/* note: already called with rcu_read_lock */
static rx_handler_result_t team_handle_frame(struct sk_buff **pskb)
{
	struct sk_buff *skb = *pskb;
	struct team_port *port;
	struct team *team;
	rx_handler_result_t res;

	skb = skb_share_check(skb, GFP_ATOMIC);
	if (!skb)
		return RX_HANDLER_CONSUMED;

	*pskb = skb;

	port = team_port_get_rcu(skb->dev);
	team = port->team;
J
Jiri Pirko 已提交
731 732 733 734 735 736
	if (!team_port_enabled(port)) {
		/* allow exact match delivery for disabled ports */
		res = RX_HANDLER_EXACT;
	} else {
		res = team->ops.receive(team, port, skb);
	}
737 738 739 740 741 742 743 744 745 746 747 748
	if (res == RX_HANDLER_ANOTHER) {
		struct team_pcpu_stats *pcpu_stats;

		pcpu_stats = this_cpu_ptr(team->pcpu_stats);
		u64_stats_update_begin(&pcpu_stats->syncp);
		pcpu_stats->rx_packets++;
		pcpu_stats->rx_bytes += skb->len;
		if (skb->pkt_type == PACKET_MULTICAST)
			pcpu_stats->rx_multicast++;
		u64_stats_update_end(&pcpu_stats->syncp);

		skb->dev = team->dev;
749 750
	} else if (res == RX_HANDLER_EXACT) {
		this_cpu_inc(team->pcpu_stats->rx_nohandler);
751 752 753 754 755 756 757 758
	} else {
		this_cpu_inc(team->pcpu_stats->rx_dropped);
	}

	return res;
}


759 760 761 762 763 764 765 766 767 768 769 770
/*************************************
 * Multiqueue Tx port select override
 *************************************/

static int team_queue_override_init(struct team *team)
{
	struct list_head *listarr;
	unsigned int queue_cnt = team->dev->num_tx_queues - 1;
	unsigned int i;

	if (!queue_cnt)
		return 0;
771 772
	listarr = kmalloc_array(queue_cnt, sizeof(struct list_head),
				GFP_KERNEL);
773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811
	if (!listarr)
		return -ENOMEM;
	team->qom_lists = listarr;
	for (i = 0; i < queue_cnt; i++)
		INIT_LIST_HEAD(listarr++);
	return 0;
}

static void team_queue_override_fini(struct team *team)
{
	kfree(team->qom_lists);
}

static struct list_head *__team_get_qom_list(struct team *team, u16 queue_id)
{
	return &team->qom_lists[queue_id - 1];
}

/*
 * note: already called with rcu_read_lock
 */
static bool team_queue_override_transmit(struct team *team, struct sk_buff *skb)
{
	struct list_head *qom_list;
	struct team_port *port;

	if (!team->queue_override_enabled || !skb->queue_mapping)
		return false;
	qom_list = __team_get_qom_list(team, skb->queue_mapping);
	list_for_each_entry_rcu(port, qom_list, qom_list) {
		if (!team_dev_queue_xmit(team, port, skb))
			return true;
	}
	return false;
}

static void __team_queue_override_port_del(struct team *team,
					   struct team_port *port)
{
812 813
	if (!port->queue_id)
		return;
814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835
	list_del_rcu(&port->qom_list);
}

static bool team_queue_override_port_has_gt_prio_than(struct team_port *port,
						      struct team_port *cur)
{
	if (port->priority < cur->priority)
		return true;
	if (port->priority > cur->priority)
		return false;
	if (port->index < cur->index)
		return true;
	return false;
}

static void __team_queue_override_port_add(struct team *team,
					   struct team_port *port)
{
	struct team_port *cur;
	struct list_head *qom_list;
	struct list_head *node;

836
	if (!port->queue_id)
837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853
		return;
	qom_list = __team_get_qom_list(team, port->queue_id);
	node = qom_list;
	list_for_each_entry(cur, qom_list, qom_list) {
		if (team_queue_override_port_has_gt_prio_than(port, cur))
			break;
		node = &cur->qom_list;
	}
	list_add_tail_rcu(&port->qom_list, node);
}

static void __team_queue_override_enabled_check(struct team *team)
{
	struct team_port *port;
	bool enabled = false;

	list_for_each_entry(port, &team->port_list, list) {
854
		if (port->queue_id) {
855 856 857 858 859 860 861 862 863 864 865
			enabled = true;
			break;
		}
	}
	if (enabled == team->queue_override_enabled)
		return;
	netdev_dbg(team->dev, "%s queue override\n",
		   enabled ? "Enabling" : "Disabling");
	team->queue_override_enabled = enabled;
}

866 867
static void team_queue_override_port_prio_changed(struct team *team,
						  struct team_port *port)
868
{
869 870
	if (!port->queue_id || team_port_enabled(port))
		return;
871 872 873 874 875
	__team_queue_override_port_del(team, port);
	__team_queue_override_port_add(team, port);
	__team_queue_override_enabled_check(team);
}

876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903
static void team_queue_override_port_change_queue_id(struct team *team,
						     struct team_port *port,
						     u16 new_queue_id)
{
	if (team_port_enabled(port)) {
		__team_queue_override_port_del(team, port);
		port->queue_id = new_queue_id;
		__team_queue_override_port_add(team, port);
		__team_queue_override_enabled_check(team);
	} else {
		port->queue_id = new_queue_id;
	}
}

static void team_queue_override_port_add(struct team *team,
					 struct team_port *port)
{
	__team_queue_override_port_add(team, port);
	__team_queue_override_enabled_check(team);
}

static void team_queue_override_port_del(struct team *team,
					 struct team_port *port)
{
	__team_queue_override_port_del(team, port);
	__team_queue_override_enabled_check(team);
}

904

905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920
/****************
 * Port handling
 ****************/

static bool team_port_find(const struct team *team,
			   const struct team_port *port)
{
	struct team_port *cur;

	list_for_each_entry(cur, &team->port_list, list)
		if (cur == port)
			return true;
	return false;
}

/*
J
Jiri Pirko 已提交
921 922 923 924
 * Enable/disable port by adding to enabled port hashlist and setting
 * port->index (Might be racy so reader could see incorrect ifindex when
 * processing a flying packet, but that is not a problem). Write guarded
 * by team->lock.
925
 */
J
Jiri Pirko 已提交
926 927
static void team_port_enable(struct team *team,
			     struct team_port *port)
928
{
J
Jiri Pirko 已提交
929 930 931
	if (team_port_enabled(port))
		return;
	port->index = team->en_port_count++;
932 933
	hlist_add_head_rcu(&port->hlist,
			   team_port_index_hash(team, port->index));
934
	team_adjust_ops(team);
935
	team_queue_override_port_add(team, port);
936 937
	if (team->ops.port_enabled)
		team->ops.port_enabled(team, port);
J
Jiri Pirko 已提交
938
	team_notify_peers(team);
939
	team_mcast_rejoin(team);
940
	team_lower_state_changed(port);
941 942 943 944 945 946 947
}

static void __reconstruct_port_hlist(struct team *team, int rm_index)
{
	int i;
	struct team_port *port;

J
Jiri Pirko 已提交
948
	for (i = rm_index + 1; i < team->en_port_count; i++) {
949 950 951 952 953 954 955 956
		port = team_get_port_by_index(team, i);
		hlist_del_rcu(&port->hlist);
		port->index--;
		hlist_add_head_rcu(&port->hlist,
				   team_port_index_hash(team, port->index));
	}
}

J
Jiri Pirko 已提交
957 958
static void team_port_disable(struct team *team,
			      struct team_port *port)
959
{
J
Jiri Pirko 已提交
960 961
	if (!team_port_enabled(port))
		return;
962 963
	if (team->ops.port_disabled)
		team->ops.port_disabled(team, port);
964
	hlist_del_rcu(&port->hlist);
965
	__reconstruct_port_hlist(team, port->index);
J
Jiri Pirko 已提交
966
	port->index = -1;
967
	team->en_port_count--;
968 969
	team_queue_override_port_del(team, port);
	team_adjust_ops(team);
970
	team_lower_state_changed(port);
971 972
}

973
#define TEAM_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
974 975 976
			    NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
			    NETIF_F_HIGHDMA | NETIF_F_LRO)

977 978 979
#define TEAM_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
				 NETIF_F_RXCSUM | NETIF_F_ALL_TSO)

980
static void __team_compute_features(struct team *team)
981 982
{
	struct team_port *port;
983 984
	netdev_features_t vlan_features = TEAM_VLAN_FEATURES &
					  NETIF_F_ALL_FOR_ALL;
985
	netdev_features_t enc_features  = TEAM_ENC_FEATURES;
986
	unsigned short max_hard_header_len = ETH_HLEN;
987 988
	unsigned int dst_release_flag = IFF_XMIT_DST_RELEASE |
					IFF_XMIT_DST_RELEASE_PERM;
989 990 991 992 993

	list_for_each_entry(port, &team->port_list, list) {
		vlan_features = netdev_increment_features(vlan_features,
					port->dev->vlan_features,
					TEAM_VLAN_FEATURES);
994 995 996 997 998
		enc_features =
			netdev_increment_features(enc_features,
						  port->dev->hw_enc_features,
						  TEAM_ENC_FEATURES);

999

1000
		dst_release_flag &= port->dev->priv_flags;
1001 1002 1003 1004 1005
		if (port->dev->hard_header_len > max_hard_header_len)
			max_hard_header_len = port->dev->hard_header_len;
	}

	team->dev->vlan_features = vlan_features;
1006
	team->dev->hw_enc_features = enc_features | NETIF_F_GSO_ENCAP_ALL |
1007 1008
				     NETIF_F_HW_VLAN_CTAG_TX |
				     NETIF_F_HW_VLAN_STAG_TX |
1009
				     NETIF_F_GSO_UDP_L4;
1010 1011
	team->dev->hard_header_len = max_hard_header_len;

1012 1013 1014
	team->dev->priv_flags &= ~IFF_XMIT_DST_RELEASE;
	if (dst_release_flag == (IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM))
		team->dev->priv_flags |= IFF_XMIT_DST_RELEASE;
1015
}
1016

1017 1018
static void team_compute_features(struct team *team)
{
1019
	mutex_lock(&team->lock);
1020
	__team_compute_features(team);
1021
	mutex_unlock(&team->lock);
1022
	netdev_change_features(team->dev);
1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053
}

static int team_port_enter(struct team *team, struct team_port *port)
{
	int err = 0;

	dev_hold(team->dev);
	if (team->ops.port_enter) {
		err = team->ops.port_enter(team, port);
		if (err) {
			netdev_err(team->dev, "Device %s failed to enter team mode\n",
				   port->dev->name);
			goto err_port_enter;
		}
	}

	return 0;

err_port_enter:
	dev_put(team->dev);

	return err;
}

static void team_port_leave(struct team *team, struct team_port *port)
{
	if (team->ops.port_leave)
		team->ops.port_leave(team, port);
	dev_put(team->dev);
}

J
Jiri Pirko 已提交
1054
#ifdef CONFIG_NET_POLL_CONTROLLER
X
Xin Long 已提交
1055
static int __team_port_enable_netpoll(struct team_port *port)
J
Jiri Pirko 已提交
1056 1057 1058 1059
{
	struct netpoll *np;
	int err;

1060
	np = kzalloc(sizeof(*np), GFP_KERNEL);
J
Jiri Pirko 已提交
1061 1062 1063
	if (!np)
		return -ENOMEM;

1064
	err = __netpoll_setup(np, port->dev);
J
Jiri Pirko 已提交
1065 1066 1067 1068 1069 1070 1071 1072
	if (err) {
		kfree(np);
		return err;
	}
	port->np = np;
	return err;
}

X
Xin Long 已提交
1073 1074 1075 1076 1077 1078 1079 1080
static int team_port_enable_netpoll(struct team_port *port)
{
	if (!port->team->dev->npinfo)
		return 0;

	return __team_port_enable_netpoll(port);
}

J
Jiri Pirko 已提交
1081 1082 1083 1084 1085 1086 1087 1088
static void team_port_disable_netpoll(struct team_port *port)
{
	struct netpoll *np = port->np;

	if (!np)
		return;
	port->np = NULL;

1089
	__netpoll_free(np);
J
Jiri Pirko 已提交
1090 1091
}
#else
X
Xin Long 已提交
1092
static int team_port_enable_netpoll(struct team_port *port)
J
Jiri Pirko 已提交
1093 1094 1095 1096 1097 1098 1099 1100
{
	return 0;
}
static void team_port_disable_netpoll(struct team_port *port)
{
}
#endif

1101 1102
static int team_upper_dev_link(struct team *team, struct team_port *port,
			       struct netlink_ext_ack *extack)
1103
{
1104
	struct netdev_lag_upper_info lag_upper_info;
1105 1106
	int err;

1107
	lag_upper_info.tx_type = team->mode->lag_tx_type;
1108
	lag_upper_info.hash_type = NETDEV_LAG_HASH_UNKNOWN;
1109
	err = netdev_master_upper_dev_link(port->dev, team->dev, NULL,
1110
					   &lag_upper_info, extack);
1111 1112
	if (err)
		return err;
1113
	port->dev->priv_flags |= IFF_TEAM_PORT;
1114 1115 1116
	return 0;
}

1117
static void team_upper_dev_unlink(struct team *team, struct team_port *port)
1118
{
1119 1120
	netdev_upper_dev_unlink(port->dev, team->dev);
	port->dev->priv_flags &= ~IFF_TEAM_PORT;
1121 1122
}

J
Jiri Pirko 已提交
1123
static void __team_port_change_port_added(struct team_port *port, bool linkup);
1124 1125
static int team_dev_type_check_change(struct net_device *dev,
				      struct net_device *port_dev);
1126

1127 1128
static int team_port_add(struct team *team, struct net_device *port_dev,
			 struct netlink_ext_ack *extack)
1129 1130 1131 1132 1133 1134
{
	struct net_device *dev = team->dev;
	struct team_port *port;
	char *portname = port_dev->name;
	int err;

1135
	if (port_dev->flags & IFF_LOOPBACK) {
1136
		NL_SET_ERR_MSG(extack, "Loopback device can't be added as a team port");
1137
		netdev_err(dev, "Device %s is loopback device. Loopback devices can't be added as a team port\n",
1138 1139 1140 1141
			   portname);
		return -EINVAL;
	}

J
Julian Wiedmann 已提交
1142
	if (netif_is_team_port(port_dev)) {
1143
		NL_SET_ERR_MSG(extack, "Device is already a port of a team device");
1144 1145 1146 1147 1148
		netdev_err(dev, "Device %s is already a port "
				"of a team device\n", portname);
		return -EBUSY;
	}

1149 1150 1151 1152 1153 1154
	if (dev == port_dev) {
		NL_SET_ERR_MSG(extack, "Cannot enslave team device to itself");
		netdev_err(dev, "Cannot enslave team device to itself\n");
		return -EINVAL;
	}

1155 1156 1157 1158 1159 1160 1161
	if (netdev_has_upper_dev(dev, port_dev)) {
		NL_SET_ERR_MSG(extack, "Device is already an upper device of the team interface");
		netdev_err(dev, "Device %s is already an upper device of the team interface\n",
			   portname);
		return -EBUSY;
	}

1162 1163
	if (port_dev->features & NETIF_F_VLAN_CHALLENGED &&
	    vlan_uses_dev(dev)) {
1164
		NL_SET_ERR_MSG(extack, "Device is VLAN challenged and team device has VLAN set up");
1165 1166 1167 1168 1169
		netdev_err(dev, "Device %s is VLAN challenged and team device has VLAN set up\n",
			   portname);
		return -EPERM;
	}

1170 1171 1172 1173
	err = team_dev_type_check_change(dev, port_dev);
	if (err)
		return err;

1174
	if (port_dev->flags & IFF_UP) {
1175
		NL_SET_ERR_MSG(extack, "Device is up. Set it down before adding it as a team port");
1176 1177 1178 1179 1180
		netdev_err(dev, "Device %s is up. Set it down before adding it as a team port\n",
			   portname);
		return -EBUSY;
	}

J
Jiri Pirko 已提交
1181 1182
	port = kzalloc(sizeof(struct team_port) + team->mode->port_priv_size,
		       GFP_KERNEL);
1183 1184 1185 1186 1187
	if (!port)
		return -ENOMEM;

	port->dev = port_dev;
	port->team = team;
1188
	INIT_LIST_HEAD(&port->qom_list);
1189 1190 1191 1192 1193 1194 1195 1196

	port->orig.mtu = port_dev->mtu;
	err = dev_set_mtu(port_dev, dev->mtu);
	if (err) {
		netdev_dbg(dev, "Error %d calling dev_set_mtu\n", err);
		goto err_set_mtu;
	}

1197
	memcpy(port->orig.dev_addr, port_dev->dev_addr, port_dev->addr_len);
1198 1199 1200 1201 1202 1203 1204 1205

	err = team_port_enter(team, port);
	if (err) {
		netdev_err(dev, "Device %s failed to enter team mode\n",
			   portname);
		goto err_port_enter;
	}

1206
	err = dev_open(port_dev, extack);
1207 1208 1209 1210 1211 1212
	if (err) {
		netdev_dbg(dev, "Device %s opening failed\n",
			   portname);
		goto err_dev_open;
	}

J
Jiri Pirko 已提交
1213 1214 1215 1216 1217 1218 1219
	err = vlan_vids_add_by_dev(port_dev, dev);
	if (err) {
		netdev_err(dev, "Failed to add vlan ids to device %s\n",
				portname);
		goto err_vids_add;
	}

X
Xin Long 已提交
1220
	err = team_port_enable_netpoll(port);
S
stephen hemminger 已提交
1221 1222 1223 1224
	if (err) {
		netdev_err(dev, "Failed to enable netpoll on device %s\n",
			   portname);
		goto err_enable_netpoll;
J
Jiri Pirko 已提交
1225 1226
	}

1227 1228 1229
	if (!(dev->features & NETIF_F_LRO))
		dev_disable_lro(port_dev);

1230 1231 1232 1233 1234 1235 1236 1237
	err = netdev_rx_handler_register(port_dev, team_handle_frame,
					 port);
	if (err) {
		netdev_err(dev, "Device %s failed to register rx_handler\n",
			   portname);
		goto err_handler_register;
	}

1238
	err = team_upper_dev_link(team, port, extack);
1239 1240 1241 1242 1243 1244
	if (err) {
		netdev_err(dev, "Device %s failed to set upper link\n",
			   portname);
		goto err_set_upper_link;
	}

1245
	err = __team_option_inst_add_port(team, port);
1246 1247 1248 1249 1250 1251
	if (err) {
		netdev_err(dev, "Device %s failed to add per-port options\n",
			   portname);
		goto err_option_port_add;
	}

1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268
	/* set promiscuity level to new slave */
	if (dev->flags & IFF_PROMISC) {
		err = dev_set_promiscuity(port_dev, 1);
		if (err)
			goto err_set_slave_promisc;
	}

	/* set allmulti level to new slave */
	if (dev->flags & IFF_ALLMULTI) {
		err = dev_set_allmulti(port_dev, 1);
		if (err) {
			if (dev->flags & IFF_PROMISC)
				dev_set_promiscuity(port_dev, -1);
			goto err_set_slave_promisc;
		}
	}

1269 1270 1271 1272 1273
	netif_addr_lock_bh(dev);
	dev_uc_sync_multiple(port_dev, dev);
	dev_mc_sync_multiple(port_dev, dev);
	netif_addr_unlock_bh(dev);

J
Jiri Pirko 已提交
1274 1275
	port->index = -1;
	list_add_tail_rcu(&port->list, &team->port_list);
1276
	team_port_enable(team, port);
1277
	__team_compute_features(team);
1278
	__team_port_change_port_added(port, !!netif_oper_up(port_dev));
1279
	__team_options_change_check(team);
1280 1281 1282 1283 1284

	netdev_info(dev, "Port device %s added\n", portname);

	return 0;

1285 1286 1287
err_set_slave_promisc:
	__team_option_inst_del_port(team, port);

1288
err_option_port_add:
1289
	team_upper_dev_unlink(team, port);
1290 1291

err_set_upper_link:
1292 1293
	netdev_rx_handler_unregister(port_dev);

1294
err_handler_register:
J
Jiri Pirko 已提交
1295 1296 1297
	team_port_disable_netpoll(port);

err_enable_netpoll:
J
Jiri Pirko 已提交
1298 1299 1300
	vlan_vids_del_by_dev(port_dev, dev);

err_vids_add:
1301 1302 1303 1304
	dev_close(port_dev);

err_dev_open:
	team_port_leave(team, port);
1305
	team_port_set_orig_dev_addr(port);
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315

err_port_enter:
	dev_set_mtu(port_dev, port->orig.mtu);

err_set_mtu:
	kfree(port);

	return err;
}

J
Jiri Pirko 已提交
1316 1317
static void __team_port_change_port_removed(struct team_port *port);

1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330
static int team_port_del(struct team *team, struct net_device *port_dev)
{
	struct net_device *dev = team->dev;
	struct team_port *port;
	char *portname = port_dev->name;

	port = team_port_get_rtnl(port_dev);
	if (!port || !team_port_find(team, port)) {
		netdev_err(dev, "Device %s does not act as a port of this team\n",
			   portname);
		return -ENOENT;
	}

J
Jiri Pirko 已提交
1331 1332
	team_port_disable(team, port);
	list_del_rcu(&port->list);
1333 1334 1335 1336 1337 1338

	if (dev->flags & IFF_PROMISC)
		dev_set_promiscuity(port_dev, -1);
	if (dev->flags & IFF_ALLMULTI)
		dev_set_allmulti(port_dev, -1);

1339
	team_upper_dev_unlink(team, port);
1340
	netdev_rx_handler_unregister(port_dev);
J
Jiri Pirko 已提交
1341
	team_port_disable_netpoll(port);
J
Jiri Pirko 已提交
1342
	vlan_vids_del_by_dev(port_dev, dev);
1343 1344
	dev_uc_unsync(port_dev, dev);
	dev_mc_unsync(port_dev, dev);
1345 1346
	dev_close(port_dev);
	team_port_leave(team, port);
1347 1348 1349 1350 1351 1352

	__team_option_inst_mark_removed_port(team, port);
	__team_options_change_check(team);
	__team_option_inst_del_port(team, port);
	__team_port_change_port_removed(port);

1353
	team_port_set_orig_dev_addr(port);
1354
	dev_set_mtu(port_dev, port->orig.mtu);
1355
	kfree_rcu(port, rcu);
1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366
	netdev_info(dev, "Port device %s removed\n", portname);
	__team_compute_features(team);

	return 0;
}


/*****************
 * Net device ops
 *****************/

1367
static int team_mode_option_get(struct team *team, struct team_gsetter_ctx *ctx)
1368
{
1369
	ctx->data.str_val = team->mode->kind;
1370 1371 1372
	return 0;
}

1373
static int team_mode_option_set(struct team *team, struct team_gsetter_ctx *ctx)
1374
{
1375
	return team_change_mode(team, ctx->data.str_val);
1376 1377
}

J
Jiri Pirko 已提交
1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405
static int team_notify_peers_count_get(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
	ctx->data.u32_val = team->notify_peers.count;
	return 0;
}

static int team_notify_peers_count_set(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
	team->notify_peers.count = ctx->data.u32_val;
	return 0;
}

static int team_notify_peers_interval_get(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
	ctx->data.u32_val = team->notify_peers.interval;
	return 0;
}

static int team_notify_peers_interval_set(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
	team->notify_peers.interval = ctx->data.u32_val;
	return 0;
}

1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433
static int team_mcast_rejoin_count_get(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
	ctx->data.u32_val = team->mcast_rejoin.count;
	return 0;
}

static int team_mcast_rejoin_count_set(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
	team->mcast_rejoin.count = ctx->data.u32_val;
	return 0;
}

static int team_mcast_rejoin_interval_get(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
	ctx->data.u32_val = team->mcast_rejoin.interval;
	return 0;
}

static int team_mcast_rejoin_interval_set(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
	team->mcast_rejoin.interval = ctx->data.u32_val;
	return 0;
}

1434 1435 1436
static int team_port_en_option_get(struct team *team,
				   struct team_gsetter_ctx *ctx)
{
1437 1438 1439
	struct team_port *port = ctx->info->port;

	ctx->data.bool_val = team_port_enabled(port);
1440 1441 1442 1443 1444 1445
	return 0;
}

static int team_port_en_option_set(struct team *team,
				   struct team_gsetter_ctx *ctx)
{
1446 1447
	struct team_port *port = ctx->info->port;

1448
	if (ctx->data.bool_val)
1449
		team_port_enable(team, port);
1450
	else
1451
		team_port_disable(team, port);
1452 1453 1454
	return 0;
}

1455 1456 1457
static int team_user_linkup_option_get(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
1458 1459 1460
	struct team_port *port = ctx->info->port;

	ctx->data.bool_val = port->user.linkup;
1461 1462 1463
	return 0;
}

1464 1465
static void __team_carrier_check(struct team *team);

1466 1467 1468
static int team_user_linkup_option_set(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
1469 1470 1471 1472
	struct team_port *port = ctx->info->port;

	port->user.linkup = ctx->data.bool_val;
	team_refresh_port_linkup(port);
1473
	__team_carrier_check(port->team);
1474 1475 1476 1477 1478 1479
	return 0;
}

static int team_user_linkup_en_option_get(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
1480
	struct team_port *port = ctx->info->port;
1481 1482 1483 1484 1485 1486 1487 1488

	ctx->data.bool_val = port->user.linkup_enabled;
	return 0;
}

static int team_user_linkup_en_option_set(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
1489
	struct team_port *port = ctx->info->port;
1490 1491

	port->user.linkup_enabled = ctx->data.bool_val;
1492
	team_refresh_port_linkup(port);
1493
	__team_carrier_check(port->team);
1494 1495 1496
	return 0;
}

J
Jiri Pirko 已提交
1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509
static int team_priority_option_get(struct team *team,
				    struct team_gsetter_ctx *ctx)
{
	struct team_port *port = ctx->info->port;

	ctx->data.s32_val = port->priority;
	return 0;
}

static int team_priority_option_set(struct team *team,
				    struct team_gsetter_ctx *ctx)
{
	struct team_port *port = ctx->info->port;
1510
	s32 priority = ctx->data.s32_val;
J
Jiri Pirko 已提交
1511

1512 1513 1514 1515
	if (port->priority == priority)
		return 0;
	port->priority = priority;
	team_queue_override_port_prio_changed(team, port);
1516 1517 1518 1519 1520 1521 1522 1523 1524
	return 0;
}

static int team_queue_id_option_get(struct team *team,
				    struct team_gsetter_ctx *ctx)
{
	struct team_port *port = ctx->info->port;

	ctx->data.u32_val = port->queue_id;
J
Jiri Pirko 已提交
1525 1526 1527
	return 0;
}

1528 1529 1530 1531
static int team_queue_id_option_set(struct team *team,
				    struct team_gsetter_ctx *ctx)
{
	struct team_port *port = ctx->info->port;
1532
	u16 new_queue_id = ctx->data.u32_val;
1533

1534
	if (port->queue_id == new_queue_id)
1535
		return 0;
1536
	if (new_queue_id >= team->dev->real_num_tx_queues)
1537
		return -EINVAL;
1538
	team_queue_override_port_change_queue_id(team, port, new_queue_id);
1539 1540 1541
	return 0;
}

J
Jiri Pirko 已提交
1542
static const struct team_option team_options[] = {
1543 1544 1545 1546 1547 1548
	{
		.name = "mode",
		.type = TEAM_OPTION_TYPE_STRING,
		.getter = team_mode_option_get,
		.setter = team_mode_option_set,
	},
J
Jiri Pirko 已提交
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
	{
		.name = "notify_peers_count",
		.type = TEAM_OPTION_TYPE_U32,
		.getter = team_notify_peers_count_get,
		.setter = team_notify_peers_count_set,
	},
	{
		.name = "notify_peers_interval",
		.type = TEAM_OPTION_TYPE_U32,
		.getter = team_notify_peers_interval_get,
		.setter = team_notify_peers_interval_set,
	},
1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572
	{
		.name = "mcast_rejoin_count",
		.type = TEAM_OPTION_TYPE_U32,
		.getter = team_mcast_rejoin_count_get,
		.setter = team_mcast_rejoin_count_set,
	},
	{
		.name = "mcast_rejoin_interval",
		.type = TEAM_OPTION_TYPE_U32,
		.getter = team_mcast_rejoin_interval_get,
		.setter = team_mcast_rejoin_interval_set,
	},
1573 1574 1575 1576 1577 1578 1579
	{
		.name = "enabled",
		.type = TEAM_OPTION_TYPE_BOOL,
		.per_port = true,
		.getter = team_port_en_option_get,
		.setter = team_port_en_option_set,
	},
1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
	{
		.name = "user_linkup",
		.type = TEAM_OPTION_TYPE_BOOL,
		.per_port = true,
		.getter = team_user_linkup_option_get,
		.setter = team_user_linkup_option_set,
	},
	{
		.name = "user_linkup_enabled",
		.type = TEAM_OPTION_TYPE_BOOL,
		.per_port = true,
		.getter = team_user_linkup_en_option_get,
		.setter = team_user_linkup_en_option_set,
	},
J
Jiri Pirko 已提交
1594 1595 1596 1597 1598 1599 1600
	{
		.name = "priority",
		.type = TEAM_OPTION_TYPE_S32,
		.per_port = true,
		.getter = team_priority_option_get,
		.setter = team_priority_option_set,
	},
1601 1602 1603 1604 1605 1606 1607
	{
		.name = "queue_id",
		.type = TEAM_OPTION_TYPE_U32,
		.per_port = true,
		.getter = team_queue_id_option_get,
		.setter = team_queue_id_option_set,
	},
1608 1609
};

J
Jiri Pirko 已提交
1610

1611 1612 1613 1614
static int team_init(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);
	int i;
J
Jiri Pirko 已提交
1615
	int err;
1616 1617

	team->dev = dev;
1618
	team_set_no_mode(team);
1619

1620
	team->pcpu_stats = netdev_alloc_pcpu_stats(struct team_pcpu_stats);
1621 1622 1623 1624
	if (!team->pcpu_stats)
		return -ENOMEM;

	for (i = 0; i < TEAM_PORT_HASHENTRIES; i++)
J
Jiri Pirko 已提交
1625
		INIT_HLIST_HEAD(&team->en_port_hlist[i]);
1626
	INIT_LIST_HEAD(&team->port_list);
1627 1628 1629
	err = team_queue_override_init(team);
	if (err)
		goto err_team_queue_override_init;
1630 1631 1632 1633

	team_adjust_ops(team);

	INIT_LIST_HEAD(&team->option_list);
1634
	INIT_LIST_HEAD(&team->option_inst_list);
J
Jiri Pirko 已提交
1635 1636

	team_notify_peers_init(team);
1637
	team_mcast_rejoin_init(team);
J
Jiri Pirko 已提交
1638

J
Jiri Pirko 已提交
1639 1640 1641
	err = team_options_register(team, team_options, ARRAY_SIZE(team_options));
	if (err)
		goto err_options_register;
1642 1643
	netif_carrier_off(dev);

1644 1645
	lockdep_register_key(&team->team_lock_key);
	__mutex_init(&team->lock, "team->team_lock_key", &team->team_lock_key);
J
Jiri Pirko 已提交
1646

1647
	return 0;
J
Jiri Pirko 已提交
1648 1649

err_options_register:
1650
	team_mcast_rejoin_fini(team);
J
Jiri Pirko 已提交
1651
	team_notify_peers_fini(team);
1652 1653
	team_queue_override_fini(team);
err_team_queue_override_init:
J
Jiri Pirko 已提交
1654 1655 1656
	free_percpu(team->pcpu_stats);

	return err;
1657 1658 1659 1660 1661 1662 1663 1664
}

static void team_uninit(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
	struct team_port *tmp;

1665
	mutex_lock(&team->lock);
1666 1667 1668 1669 1670
	list_for_each_entry_safe(port, tmp, &team->port_list, list)
		team_port_del(team, port->dev);

	__team_change_mode(team, NULL); /* cleanup */
	__team_options_unregister(team, team_options, ARRAY_SIZE(team_options));
1671
	team_mcast_rejoin_fini(team);
J
Jiri Pirko 已提交
1672
	team_notify_peers_fini(team);
1673
	team_queue_override_fini(team);
1674
	mutex_unlock(&team->lock);
1675
	netdev_change_features(dev);
1676
	lockdep_unregister_key(&team->team_lock_key);
1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701
}

static void team_destructor(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);

	free_percpu(team->pcpu_stats);
}

static int team_open(struct net_device *dev)
{
	return 0;
}

static int team_close(struct net_device *dev)
{
	return 0;
}

/*
 * note: already called with rcu_read_lock
 */
static netdev_tx_t team_xmit(struct sk_buff *skb, struct net_device *dev)
{
	struct team *team = netdev_priv(dev);
1702
	bool tx_success;
1703 1704
	unsigned int len = skb->len;

1705 1706 1707
	tx_success = team_queue_override_transmit(team, skb);
	if (!tx_success)
		tx_success = team->ops.transmit(team, skb);
1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722
	if (tx_success) {
		struct team_pcpu_stats *pcpu_stats;

		pcpu_stats = this_cpu_ptr(team->pcpu_stats);
		u64_stats_update_begin(&pcpu_stats->syncp);
		pcpu_stats->tx_packets++;
		pcpu_stats->tx_bytes += len;
		u64_stats_update_end(&pcpu_stats->syncp);
	} else {
		this_cpu_inc(team->pcpu_stats->tx_dropped);
	}

	return NETDEV_TX_OK;
}

1723
static u16 team_select_queue(struct net_device *dev, struct sk_buff *skb,
1724
			     struct net_device *sb_dev)
J
Jiri Pirko 已提交
1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746
{
	/*
	 * This helper function exists to help dev_pick_tx get the correct
	 * destination queue.  Using a helper function skips a call to
	 * skb_tx_hash and will put the skbs in the queue we expect on their
	 * way down to the team driver.
	 */
	u16 txq = skb_rx_queue_recorded(skb) ? skb_get_rx_queue(skb) : 0;

	/*
	 * Save the original txq to restore before passing to the driver
	 */
	qdisc_skb_cb(skb)->slave_dev_queue_mapping = skb->queue_mapping;

	if (unlikely(txq >= dev->real_num_tx_queues)) {
		do {
			txq -= dev->real_num_tx_queues;
		} while (txq >= dev->real_num_tx_queues);
	}
	return txq;
}

1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773
static void team_change_rx_flags(struct net_device *dev, int change)
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
	int inc;

	rcu_read_lock();
	list_for_each_entry_rcu(port, &team->port_list, list) {
		if (change & IFF_PROMISC) {
			inc = dev->flags & IFF_PROMISC ? 1 : -1;
			dev_set_promiscuity(port->dev, inc);
		}
		if (change & IFF_ALLMULTI) {
			inc = dev->flags & IFF_ALLMULTI ? 1 : -1;
			dev_set_allmulti(port->dev, inc);
		}
	}
	rcu_read_unlock();
}

static void team_set_rx_mode(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;

	rcu_read_lock();
	list_for_each_entry_rcu(port, &team->port_list, list) {
1774 1775
		dev_uc_sync_multiple(port->dev, dev);
		dev_mc_sync_multiple(port->dev, dev);
1776 1777 1778 1779 1780 1781
	}
	rcu_read_unlock();
}

static int team_set_mac_address(struct net_device *dev, void *p)
{
1782
	struct sockaddr *addr = p;
1783 1784 1785
	struct team *team = netdev_priv(dev);
	struct team_port *port;

1786 1787 1788
	if (dev->type == ARPHRD_ETHER && !is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;
	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
1789 1790
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list)
1791 1792
		if (team->ops.port_change_dev_addr)
			team->ops.port_change_dev_addr(team, port);
1793
	mutex_unlock(&team->lock);
1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806
	return 0;
}

static int team_change_mtu(struct net_device *dev, int new_mtu)
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
	int err;

	/*
	 * Alhough this is reader, it's guarded by team lock. It's not possible
	 * to traverse list in reverse under rcu_read_lock
	 */
1807
	mutex_lock(&team->lock);
J
Jiri Pirko 已提交
1808
	team->port_mtu_change_allowed = true;
1809 1810 1811 1812 1813 1814 1815 1816
	list_for_each_entry(port, &team->port_list, list) {
		err = dev_set_mtu(port->dev, new_mtu);
		if (err) {
			netdev_err(dev, "Device %s failed to change mtu",
				   port->dev->name);
			goto unwind;
		}
	}
J
Jiri Pirko 已提交
1817
	team->port_mtu_change_allowed = false;
1818
	mutex_unlock(&team->lock);
1819 1820 1821 1822 1823 1824 1825 1826

	dev->mtu = new_mtu;

	return 0;

unwind:
	list_for_each_entry_continue_reverse(port, &team->port_list, list)
		dev_set_mtu(port->dev, dev->mtu);
J
Jiri Pirko 已提交
1827
	team->port_mtu_change_allowed = false;
1828
	mutex_unlock(&team->lock);
1829 1830 1831 1832

	return err;
}

1833
static void
1834 1835 1836 1837 1838
team_get_stats64(struct net_device *dev, struct rtnl_link_stats64 *stats)
{
	struct team *team = netdev_priv(dev);
	struct team_pcpu_stats *p;
	u64 rx_packets, rx_bytes, rx_multicast, tx_packets, tx_bytes;
1839
	u32 rx_dropped = 0, tx_dropped = 0, rx_nohandler = 0;
1840 1841 1842 1843 1844 1845
	unsigned int start;
	int i;

	for_each_possible_cpu(i) {
		p = per_cpu_ptr(team->pcpu_stats, i);
		do {
1846
			start = u64_stats_fetch_begin_irq(&p->syncp);
1847 1848 1849 1850 1851
			rx_packets	= p->rx_packets;
			rx_bytes	= p->rx_bytes;
			rx_multicast	= p->rx_multicast;
			tx_packets	= p->tx_packets;
			tx_bytes	= p->tx_bytes;
1852
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
1853 1854 1855 1856 1857 1858 1859

		stats->rx_packets	+= rx_packets;
		stats->rx_bytes		+= rx_bytes;
		stats->multicast	+= rx_multicast;
		stats->tx_packets	+= tx_packets;
		stats->tx_bytes		+= tx_bytes;
		/*
1860 1861
		 * rx_dropped, tx_dropped & rx_nohandler are u32,
		 * updated without syncp protection.
1862 1863 1864
		 */
		rx_dropped	+= p->rx_dropped;
		tx_dropped	+= p->tx_dropped;
1865
		rx_nohandler	+= p->rx_nohandler;
1866 1867 1868
	}
	stats->rx_dropped	= rx_dropped;
	stats->tx_dropped	= tx_dropped;
1869
	stats->rx_nohandler	= rx_nohandler;
1870 1871
}

1872
static int team_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
1873 1874 1875
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
1876
	int err;
1877

1878 1879 1880 1881 1882 1883
	/*
	 * Alhough this is reader, it's guarded by team lock. It's not possible
	 * to traverse list in reverse under rcu_read_lock
	 */
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list) {
1884
		err = vlan_vid_add(port->dev, proto, vid);
1885 1886
		if (err)
			goto unwind;
1887
	}
1888
	mutex_unlock(&team->lock);
1889 1890

	return 0;
1891 1892 1893

unwind:
	list_for_each_entry_continue_reverse(port, &team->port_list, list)
1894
		vlan_vid_del(port->dev, proto, vid);
1895 1896 1897
	mutex_unlock(&team->lock);

	return err;
1898 1899
}

1900
static int team_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
1901 1902 1903 1904
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;

1905 1906
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list)
1907
		vlan_vid_del(port->dev, proto, vid);
1908
	mutex_unlock(&team->lock);
1909 1910

	return 0;
1911 1912
}

J
Jiri Pirko 已提交
1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935
#ifdef CONFIG_NET_POLL_CONTROLLER
static void team_poll_controller(struct net_device *dev)
{
}

static void __team_netpoll_cleanup(struct team *team)
{
	struct team_port *port;

	list_for_each_entry(port, &team->port_list, list)
		team_port_disable_netpoll(port);
}

static void team_netpoll_cleanup(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);

	mutex_lock(&team->lock);
	__team_netpoll_cleanup(team);
	mutex_unlock(&team->lock);
}

static int team_netpoll_setup(struct net_device *dev,
1936
			      struct netpoll_info *npifo)
J
Jiri Pirko 已提交
1937 1938 1939
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
1940
	int err = 0;
J
Jiri Pirko 已提交
1941 1942 1943

	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list) {
X
Xin Long 已提交
1944
		err = __team_port_enable_netpoll(port);
J
Jiri Pirko 已提交
1945 1946 1947 1948 1949 1950 1951 1952 1953 1954
		if (err) {
			__team_netpoll_cleanup(team);
			break;
		}
	}
	mutex_unlock(&team->lock);
	return err;
}
#endif

D
David Ahern 已提交
1955 1956
static int team_add_slave(struct net_device *dev, struct net_device *port_dev,
			  struct netlink_ext_ack *extack)
1957 1958 1959 1960
{
	struct team *team = netdev_priv(dev);
	int err;

1961
	mutex_lock(&team->lock);
1962
	err = team_port_add(team, port_dev, extack);
1963
	mutex_unlock(&team->lock);
1964 1965 1966 1967

	if (!err)
		netdev_change_features(dev);

1968 1969 1970 1971 1972 1973 1974 1975
	return err;
}

static int team_del_slave(struct net_device *dev, struct net_device *port_dev)
{
	struct team *team = netdev_priv(dev);
	int err;

1976
	mutex_lock(&team->lock);
1977
	err = team_port_del(team, port_dev);
1978
	mutex_unlock(&team->lock);
1979

1980 1981 1982 1983 1984 1985 1986 1987 1988
	if (err)
		return err;

	if (netif_is_team_master(port_dev)) {
		lockdep_unregister_key(&team->team_lock_key);
		lockdep_register_key(&team->team_lock_key);
		lockdep_set_class(&team->lock, &team->team_lock_key);
	}
	netdev_change_features(dev);
1989

1990 1991 1992
	return err;
}

J
Jiri Pirko 已提交
1993 1994 1995 1996 1997 1998 1999
static netdev_features_t team_fix_features(struct net_device *dev,
					   netdev_features_t features)
{
	struct team_port *port;
	struct team *team = netdev_priv(dev);
	netdev_features_t mask;

2000
	mask = features;
J
Jiri Pirko 已提交
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
	features &= ~NETIF_F_ONE_FOR_ALL;
	features |= NETIF_F_ALL_FOR_ALL;

	rcu_read_lock();
	list_for_each_entry_rcu(port, &team->port_list, list) {
		features = netdev_increment_features(features,
						     port->dev->features,
						     mask);
	}
	rcu_read_unlock();
J
Jiri Pirko 已提交
2011 2012 2013

	features = netdev_add_tso_features(features, mask);

J
Jiri Pirko 已提交
2014 2015 2016
	return features;
}

F
Flavio Leitner 已提交
2017 2018
static int team_change_carrier(struct net_device *dev, bool new_carrier)
{
2019 2020 2021 2022
	struct team *team = netdev_priv(dev);

	team->user_carrier_enabled = true;

F
Flavio Leitner 已提交
2023 2024 2025 2026 2027 2028 2029
	if (new_carrier)
		netif_carrier_on(dev);
	else
		netif_carrier_off(dev);
	return 0;
}

2030 2031 2032 2033 2034 2035
static const struct net_device_ops team_netdev_ops = {
	.ndo_init		= team_init,
	.ndo_uninit		= team_uninit,
	.ndo_open		= team_open,
	.ndo_stop		= team_close,
	.ndo_start_xmit		= team_xmit,
J
Jiri Pirko 已提交
2036
	.ndo_select_queue	= team_select_queue,
2037 2038 2039 2040 2041 2042 2043
	.ndo_change_rx_flags	= team_change_rx_flags,
	.ndo_set_rx_mode	= team_set_rx_mode,
	.ndo_set_mac_address	= team_set_mac_address,
	.ndo_change_mtu		= team_change_mtu,
	.ndo_get_stats64	= team_get_stats64,
	.ndo_vlan_rx_add_vid	= team_vlan_rx_add_vid,
	.ndo_vlan_rx_kill_vid	= team_vlan_rx_kill_vid,
J
Jiri Pirko 已提交
2044 2045 2046 2047 2048
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= team_poll_controller,
	.ndo_netpoll_setup	= team_netpoll_setup,
	.ndo_netpoll_cleanup	= team_netpoll_cleanup,
#endif
2049 2050
	.ndo_add_slave		= team_add_slave,
	.ndo_del_slave		= team_del_slave,
J
Jiri Pirko 已提交
2051
	.ndo_fix_features	= team_fix_features,
F
Flavio Leitner 已提交
2052
	.ndo_change_carrier     = team_change_carrier,
2053
	.ndo_features_check	= passthru_features_check,
2054 2055
};

F
Flavio Leitner 已提交
2056 2057 2058 2059 2060 2061 2062
/***********************
 * ethtool interface
 ***********************/

static void team_ethtool_get_drvinfo(struct net_device *dev,
				     struct ethtool_drvinfo *drvinfo)
{
2063 2064
	strlcpy(drvinfo->driver, DRV_NAME, sizeof(drvinfo->driver));
	strlcpy(drvinfo->version, UTS_RELEASE, sizeof(drvinfo->version));
F
Flavio Leitner 已提交
2065 2066
}

2067 2068 2069 2070 2071 2072 2073 2074 2075 2076
static int team_ethtool_get_link_ksettings(struct net_device *dev,
					   struct ethtool_link_ksettings *cmd)
{
	struct team *team= netdev_priv(dev);
	unsigned long speed = 0;
	struct team_port *port;

	cmd->base.duplex = DUPLEX_UNKNOWN;
	cmd->base.port = PORT_OTHER;

2077 2078
	rcu_read_lock();
	list_for_each_entry_rcu(port, &team->port_list, list) {
2079 2080 2081 2082 2083 2084 2085 2086
		if (team_port_txable(port)) {
			if (port->state.speed != SPEED_UNKNOWN)
				speed += port->state.speed;
			if (cmd->base.duplex == DUPLEX_UNKNOWN &&
			    port->state.duplex != DUPLEX_UNKNOWN)
				cmd->base.duplex = port->state.duplex;
		}
	}
2087 2088
	rcu_read_unlock();

2089 2090 2091 2092 2093
	cmd->base.speed = speed ? : SPEED_UNKNOWN;

	return 0;
}

F
Flavio Leitner 已提交
2094 2095 2096
static const struct ethtool_ops team_ethtool_ops = {
	.get_drvinfo		= team_ethtool_get_drvinfo,
	.get_link		= ethtool_op_get_link,
2097
	.get_link_ksettings	= team_ethtool_get_link_ksettings,
F
Flavio Leitner 已提交
2098
};
2099 2100 2101 2102 2103

/***********************
 * rt netlink interface
 ***********************/

2104 2105 2106 2107 2108 2109 2110 2111 2112
static void team_setup_by_port(struct net_device *dev,
			       struct net_device *port_dev)
{
	dev->header_ops	= port_dev->header_ops;
	dev->type = port_dev->type;
	dev->hard_header_len = port_dev->hard_header_len;
	dev->addr_len = port_dev->addr_len;
	dev->mtu = port_dev->mtu;
	memcpy(dev->broadcast, port_dev->broadcast, port_dev->addr_len);
2113
	eth_hw_addr_inherit(dev, port_dev);
2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
}

static int team_dev_type_check_change(struct net_device *dev,
				      struct net_device *port_dev)
{
	struct team *team = netdev_priv(dev);
	char *portname = port_dev->name;
	int err;

	if (dev->type == port_dev->type)
		return 0;
	if (!list_empty(&team->port_list)) {
		netdev_err(dev, "Device %s is of different type\n", portname);
		return -EBUSY;
	}
	err = call_netdevice_notifiers(NETDEV_PRE_TYPE_CHANGE, dev);
	err = notifier_to_errno(err);
	if (err) {
		netdev_err(dev, "Refused to change device type\n");
		return err;
	}
	dev_uc_flush(dev);
	dev_mc_flush(dev);
	team_setup_by_port(dev, port_dev);
	call_netdevice_notifiers(NETDEV_POST_TYPE_CHANGE, dev);
	return 0;
}

2142 2143 2144
static void team_setup(struct net_device *dev)
{
	ether_setup(dev);
J
Jarod Wilson 已提交
2145
	dev->max_mtu = ETH_MAX_MTU;
2146 2147

	dev->netdev_ops = &team_netdev_ops;
F
Flavio Leitner 已提交
2148
	dev->ethtool_ops = &team_ethtool_ops;
2149 2150
	dev->needs_free_netdev = true;
	dev->priv_destructor = team_destructor;
2151
	dev->priv_flags &= ~(IFF_XMIT_DST_RELEASE | IFF_TX_SKB_SHARING);
2152
	dev->priv_flags |= IFF_NO_QUEUE;
J
Jiri Pirko 已提交
2153
	dev->priv_flags |= IFF_TEAM;
2154 2155 2156 2157 2158 2159

	/*
	 * Indicate we support unicast address filtering. That way core won't
	 * bring us to promisc mode in case a unicast addr is added.
	 * Let this up to underlay drivers.
	 */
2160
	dev->priv_flags |= IFF_UNICAST_FLT | IFF_LIVE_ADDR_CHANGE;
2161 2162 2163

	dev->features |= NETIF_F_LLTX;
	dev->features |= NETIF_F_GRO;
2164 2165 2166 2167

	/* Don't allow team devices to change network namespaces. */
	dev->features |= NETIF_F_NETNS_LOCAL;

J
Jiri Pirko 已提交
2168
	dev->hw_features = TEAM_VLAN_FEATURES |
2169 2170
			   NETIF_F_HW_VLAN_CTAG_RX |
			   NETIF_F_HW_VLAN_CTAG_FILTER;
2171

2172
	dev->hw_features |= NETIF_F_GSO_ENCAP_ALL | NETIF_F_GSO_UDP_L4;
2173
	dev->features |= dev->hw_features;
Y
YueHaibing 已提交
2174
	dev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX;
2175 2176 2177
}

static int team_newlink(struct net *src_net, struct net_device *dev,
2178 2179
			struct nlattr *tb[], struct nlattr *data[],
			struct netlink_ext_ack *extack)
2180 2181
{
	if (tb[IFLA_ADDRESS] == NULL)
2182
		eth_hw_addr_random(dev);
2183

2184
	return register_netdevice(dev);
2185 2186
}

2187 2188
static int team_validate(struct nlattr *tb[], struct nlattr *data[],
			 struct netlink_ext_ack *extack)
2189 2190 2191 2192 2193 2194 2195 2196 2197 2198
{
	if (tb[IFLA_ADDRESS]) {
		if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
			return -EINVAL;
		if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
			return -EADDRNOTAVAIL;
	}
	return 0;
}

J
Jiri Pirko 已提交
2199 2200 2201 2202 2203 2204 2205 2206 2207 2208
static unsigned int team_get_num_tx_queues(void)
{
	return TEAM_DEFAULT_NUM_TX_QUEUES;
}

static unsigned int team_get_num_rx_queues(void)
{
	return TEAM_DEFAULT_NUM_RX_QUEUES;
}

2209
static struct rtnl_link_ops team_link_ops __read_mostly = {
J
Jiri Pirko 已提交
2210 2211 2212 2213 2214 2215 2216
	.kind			= DRV_NAME,
	.priv_size		= sizeof(struct team),
	.setup			= team_setup,
	.newlink		= team_newlink,
	.validate		= team_validate,
	.get_num_tx_queues	= team_get_num_tx_queues,
	.get_num_rx_queues	= team_get_num_rx_queues,
2217 2218 2219 2220 2221 2222 2223
};


/***********************************
 * Generic netlink custom interface
 ***********************************/

2224
static struct genl_family team_nl_family;
2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241

static const struct nla_policy team_nl_policy[TEAM_ATTR_MAX + 1] = {
	[TEAM_ATTR_UNSPEC]			= { .type = NLA_UNSPEC, },
	[TEAM_ATTR_TEAM_IFINDEX]		= { .type = NLA_U32 },
	[TEAM_ATTR_LIST_OPTION]			= { .type = NLA_NESTED },
	[TEAM_ATTR_LIST_PORT]			= { .type = NLA_NESTED },
};

static const struct nla_policy
team_nl_option_policy[TEAM_ATTR_OPTION_MAX + 1] = {
	[TEAM_ATTR_OPTION_UNSPEC]		= { .type = NLA_UNSPEC, },
	[TEAM_ATTR_OPTION_NAME] = {
		.type = NLA_STRING,
		.len = TEAM_STRING_MAX_LEN,
	},
	[TEAM_ATTR_OPTION_CHANGED]		= { .type = NLA_FLAG },
	[TEAM_ATTR_OPTION_TYPE]			= { .type = NLA_U8 },
J
Jiri Pirko 已提交
2242
	[TEAM_ATTR_OPTION_DATA]			= { .type = NLA_BINARY },
2243
	[TEAM_ATTR_OPTION_PORT_IFINDEX]		= { .type = NLA_U32 },
2244
	[TEAM_ATTR_OPTION_ARRAY_INDEX]		= { .type = NLA_U32 },
2245 2246 2247 2248 2249 2250 2251 2252
};

static int team_nl_cmd_noop(struct sk_buff *skb, struct genl_info *info)
{
	struct sk_buff *msg;
	void *hdr;
	int err;

2253
	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
2254 2255 2256
	if (!msg)
		return -ENOMEM;

2257
	hdr = genlmsg_put(msg, info->snd_portid, info->snd_seq,
2258
			  &team_nl_family, 0, TEAM_CMD_NOOP);
W
Wei Yongjun 已提交
2259 2260
	if (!hdr) {
		err = -EMSGSIZE;
2261 2262 2263 2264 2265
		goto err_msg_put;
	}

	genlmsg_end(msg, hdr);

2266
	return genlmsg_unicast(genl_info_net(info), msg, info->snd_portid);
2267 2268 2269 2270 2271 2272 2273 2274 2275

err_msg_put:
	nlmsg_free(msg);

	return err;
}

/*
 * Netlink cmd functions should be locked by following two functions.
2276
 * Since dev gets held here, that ensures dev won't disappear in between.
2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288
 */
static struct team *team_nl_team_get(struct genl_info *info)
{
	struct net *net = genl_info_net(info);
	int ifindex;
	struct net_device *dev;
	struct team *team;

	if (!info->attrs[TEAM_ATTR_TEAM_IFINDEX])
		return NULL;

	ifindex = nla_get_u32(info->attrs[TEAM_ATTR_TEAM_IFINDEX]);
2289
	dev = dev_get_by_index(net, ifindex);
2290
	if (!dev || dev->netdev_ops != &team_netdev_ops) {
2291 2292
		if (dev)
			dev_put(dev);
2293 2294 2295 2296
		return NULL;
	}

	team = netdev_priv(dev);
2297
	mutex_lock(&team->lock);
2298 2299 2300 2301 2302
	return team;
}

static void team_nl_team_put(struct team *team)
{
2303
	mutex_unlock(&team->lock);
2304
	dev_put(team->dev);
2305 2306
}

2307
typedef int team_nl_send_func_t(struct sk_buff *skb,
2308
				struct team *team, u32 portid);
2309

2310
static int team_nl_send_unicast(struct sk_buff *skb, struct team *team, u32 portid)
2311
{
2312
	return genlmsg_unicast(dev_net(team->dev), skb, portid);
2313 2314
}

2315 2316 2317 2318 2319
static int team_nl_fill_one_option_get(struct sk_buff *skb, struct team *team,
				       struct team_option_inst *opt_inst)
{
	struct nlattr *option_item;
	struct team_option *option = opt_inst->option;
2320
	struct team_option_inst_info *opt_inst_info = &opt_inst->info;
2321 2322 2323
	struct team_gsetter_ctx ctx;
	int err;

2324 2325 2326 2327 2328
	ctx.info = opt_inst_info;
	err = team_option_get(team, opt_inst, &ctx);
	if (err)
		return err;

2329
	option_item = nla_nest_start_noflag(skb, TEAM_ATTR_ITEM_OPTION);
2330
	if (!option_item)
2331
		return -EMSGSIZE;
2332

2333 2334
	if (nla_put_string(skb, TEAM_ATTR_OPTION_NAME, option->name))
		goto nest_cancel;
2335 2336 2337
	if (opt_inst_info->port &&
	    nla_put_u32(skb, TEAM_ATTR_OPTION_PORT_IFINDEX,
			opt_inst_info->port->dev->ifindex))
2338
		goto nest_cancel;
2339 2340 2341
	if (opt_inst->option->array_size &&
	    nla_put_u32(skb, TEAM_ATTR_OPTION_ARRAY_INDEX,
			opt_inst_info->array_index))
2342
		goto nest_cancel;
2343 2344 2345 2346

	switch (option->type) {
	case TEAM_OPTION_TYPE_U32:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_U32))
2347
			goto nest_cancel;
2348
		if (nla_put_u32(skb, TEAM_ATTR_OPTION_DATA, ctx.data.u32_val))
2349
			goto nest_cancel;
2350 2351 2352
		break;
	case TEAM_OPTION_TYPE_STRING:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_STRING))
2353
			goto nest_cancel;
2354 2355
		if (nla_put_string(skb, TEAM_ATTR_OPTION_DATA,
				   ctx.data.str_val))
2356
			goto nest_cancel;
2357 2358 2359
		break;
	case TEAM_OPTION_TYPE_BINARY:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_BINARY))
2360
			goto nest_cancel;
2361 2362
		if (nla_put(skb, TEAM_ATTR_OPTION_DATA, ctx.data.bin_val.len,
			    ctx.data.bin_val.ptr))
2363
			goto nest_cancel;
2364 2365 2366
		break;
	case TEAM_OPTION_TYPE_BOOL:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_FLAG))
2367
			goto nest_cancel;
2368 2369
		if (ctx.data.bool_val &&
		    nla_put_flag(skb, TEAM_ATTR_OPTION_DATA))
2370
			goto nest_cancel;
2371
		break;
2372 2373 2374 2375 2376 2377
	case TEAM_OPTION_TYPE_S32:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_S32))
			goto nest_cancel;
		if (nla_put_s32(skb, TEAM_ATTR_OPTION_DATA, ctx.data.s32_val))
			goto nest_cancel;
		break;
2378 2379 2380
	default:
		BUG();
	}
2381 2382 2383 2384 2385 2386 2387
	if (opt_inst->removed && nla_put_flag(skb, TEAM_ATTR_OPTION_REMOVED))
		goto nest_cancel;
	if (opt_inst->changed) {
		if (nla_put_flag(skb, TEAM_ATTR_OPTION_CHANGED))
			goto nest_cancel;
		opt_inst->changed = false;
	}
2388 2389 2390
	nla_nest_end(skb, option_item);
	return 0;

2391 2392 2393 2394 2395 2396
nest_cancel:
	nla_nest_cancel(skb, option_item);
	return -EMSGSIZE;
}

static int __send_and_alloc_skb(struct sk_buff **pskb,
2397
				struct team *team, u32 portid,
2398 2399 2400 2401 2402
				team_nl_send_func_t *send_func)
{
	int err;

	if (*pskb) {
2403
		err = send_func(*pskb, team, portid);
2404 2405 2406
		if (err)
			return err;
	}
2407
	*pskb = genlmsg_new(GENLMSG_DEFAULT_SIZE, GFP_KERNEL);
2408 2409 2410
	if (!*pskb)
		return -ENOMEM;
	return 0;
2411 2412
}

2413
static int team_nl_send_options_get(struct team *team, u32 portid, u32 seq,
2414
				    int flags, team_nl_send_func_t *send_func,
2415
				    struct list_head *sel_opt_inst_list)
2416 2417
{
	struct nlattr *option_list;
2418
	struct nlmsghdr *nlh;
2419
	void *hdr;
2420 2421
	struct team_option_inst *opt_inst;
	int err;
2422 2423 2424
	struct sk_buff *skb = NULL;
	bool incomplete;
	int i;
2425

2426 2427 2428 2429
	opt_inst = list_first_entry(sel_opt_inst_list,
				    struct team_option_inst, tmp_list);

start_again:
2430
	err = __send_and_alloc_skb(&skb, team, portid, send_func);
2431 2432 2433
	if (err)
		return err;

2434
	hdr = genlmsg_put(skb, portid, seq, &team_nl_family, flags | NLM_F_MULTI,
2435
			  TEAM_CMD_OPTIONS_GET);
P
Pan Bian 已提交
2436 2437
	if (!hdr) {
		nlmsg_free(skb);
W
Wei Yongjun 已提交
2438
		return -EMSGSIZE;
P
Pan Bian 已提交
2439
	}
2440

D
David S. Miller 已提交
2441 2442
	if (nla_put_u32(skb, TEAM_ATTR_TEAM_IFINDEX, team->dev->ifindex))
		goto nla_put_failure;
2443
	option_list = nla_nest_start_noflag(skb, TEAM_ATTR_LIST_OPTION);
2444
	if (!option_list)
2445
		goto nla_put_failure;
2446

2447 2448 2449
	i = 0;
	incomplete = false;
	list_for_each_entry_from(opt_inst, sel_opt_inst_list, tmp_list) {
2450
		err = team_nl_fill_one_option_get(skb, team, opt_inst);
2451 2452 2453 2454 2455 2456 2457
		if (err) {
			if (err == -EMSGSIZE) {
				if (!i)
					goto errout;
				incomplete = true;
				break;
			}
2458
			goto errout;
2459 2460
		}
		i++;
2461 2462 2463
	}

	nla_nest_end(skb, option_list);
2464 2465 2466 2467 2468
	genlmsg_end(skb, hdr);
	if (incomplete)
		goto start_again;

send_done:
2469
	nlh = nlmsg_put(skb, portid, seq, NLMSG_DONE, 0, flags | NLM_F_MULTI);
2470
	if (!nlh) {
2471
		err = __send_and_alloc_skb(&skb, team, portid, send_func);
2472
		if (err)
2473
			return err;
2474 2475 2476
		goto send_done;
	}

2477
	return send_func(skb, team, portid);
2478 2479

nla_put_failure:
2480 2481
	err = -EMSGSIZE;
errout:
2482
	nlmsg_free(skb);
2483
	return err;
2484 2485 2486 2487 2488
}

static int team_nl_cmd_options_get(struct sk_buff *skb, struct genl_info *info)
{
	struct team *team;
2489
	struct team_option_inst *opt_inst;
2490
	int err;
2491
	LIST_HEAD(sel_opt_inst_list);
2492 2493 2494 2495 2496

	team = team_nl_team_get(info);
	if (!team)
		return -EINVAL;

2497 2498
	list_for_each_entry(opt_inst, &team->option_inst_list, list)
		list_add_tail(&opt_inst->tmp_list, &sel_opt_inst_list);
2499
	err = team_nl_send_options_get(team, info->snd_portid, info->snd_seq,
2500 2501
				       NLM_F_ACK, team_nl_send_unicast,
				       &sel_opt_inst_list);
2502 2503 2504 2505 2506 2507

	team_nl_team_put(team);

	return err;
}

2508 2509 2510
static int team_nl_send_event_options_get(struct team *team,
					  struct list_head *sel_opt_inst_list);

2511 2512 2513 2514 2515 2516 2517
static int team_nl_cmd_options_set(struct sk_buff *skb, struct genl_info *info)
{
	struct team *team;
	int err = 0;
	int i;
	struct nlattr *nl_option;

J
Jiri Pirko 已提交
2518 2519
	rtnl_lock();

2520
	team = team_nl_team_get(info);
J
Jiri Pirko 已提交
2521 2522 2523 2524
	if (!team) {
		err = -EINVAL;
		goto rtnl_unlock;
	}
2525 2526 2527 2528 2529 2530 2531 2532

	err = -EINVAL;
	if (!info->attrs[TEAM_ATTR_LIST_OPTION]) {
		err = -EINVAL;
		goto team_put;
	}

	nla_for_each_nested(nl_option, info->attrs[TEAM_ATTR_LIST_OPTION], i) {
2533
		struct nlattr *opt_attrs[TEAM_ATTR_OPTION_MAX + 1];
J
Jiri Pirko 已提交
2534
		struct nlattr *attr;
J
Jiri Pirko 已提交
2535
		struct nlattr *attr_data;
2536
		LIST_HEAD(opt_inst_list);
2537
		enum team_option_type opt_type;
2538
		int opt_port_ifindex = 0; /* != 0 for per-port options */
J
Jiri Pirko 已提交
2539 2540
		u32 opt_array_index = 0;
		bool opt_is_array = false;
2541
		struct team_option_inst *opt_inst;
2542 2543 2544 2545 2546 2547 2548
		char *opt_name;
		bool opt_found = false;

		if (nla_type(nl_option) != TEAM_ATTR_ITEM_OPTION) {
			err = -EINVAL;
			goto team_put;
		}
2549 2550 2551 2552 2553
		err = nla_parse_nested_deprecated(opt_attrs,
						  TEAM_ATTR_OPTION_MAX,
						  nl_option,
						  team_nl_option_policy,
						  info->extack);
2554 2555
		if (err)
			goto team_put;
2556
		if (!opt_attrs[TEAM_ATTR_OPTION_NAME] ||
J
Jiri Pirko 已提交
2557
		    !opt_attrs[TEAM_ATTR_OPTION_TYPE]) {
2558 2559 2560
			err = -EINVAL;
			goto team_put;
		}
2561
		switch (nla_get_u8(opt_attrs[TEAM_ATTR_OPTION_TYPE])) {
2562 2563 2564 2565 2566 2567
		case NLA_U32:
			opt_type = TEAM_OPTION_TYPE_U32;
			break;
		case NLA_STRING:
			opt_type = TEAM_OPTION_TYPE_STRING;
			break;
J
Jiri Pirko 已提交
2568 2569 2570
		case NLA_BINARY:
			opt_type = TEAM_OPTION_TYPE_BINARY;
			break;
J
Jiri Pirko 已提交
2571 2572 2573
		case NLA_FLAG:
			opt_type = TEAM_OPTION_TYPE_BOOL;
			break;
2574 2575 2576
		case NLA_S32:
			opt_type = TEAM_OPTION_TYPE_S32;
			break;
2577 2578 2579 2580
		default:
			goto team_put;
		}

J
Jiri Pirko 已提交
2581 2582 2583 2584 2585 2586
		attr_data = opt_attrs[TEAM_ATTR_OPTION_DATA];
		if (opt_type != TEAM_OPTION_TYPE_BOOL && !attr_data) {
			err = -EINVAL;
			goto team_put;
		}

2587
		opt_name = nla_data(opt_attrs[TEAM_ATTR_OPTION_NAME]);
J
Jiri Pirko 已提交
2588 2589 2590 2591 2592 2593 2594 2595 2596
		attr = opt_attrs[TEAM_ATTR_OPTION_PORT_IFINDEX];
		if (attr)
			opt_port_ifindex = nla_get_u32(attr);

		attr = opt_attrs[TEAM_ATTR_OPTION_ARRAY_INDEX];
		if (attr) {
			opt_is_array = true;
			opt_array_index = nla_get_u32(attr);
		}
2597 2598 2599 2600

		list_for_each_entry(opt_inst, &team->option_inst_list, list) {
			struct team_option *option = opt_inst->option;
			struct team_gsetter_ctx ctx;
2601
			struct team_option_inst_info *opt_inst_info;
2602
			int tmp_ifindex;
2603

2604 2605 2606
			opt_inst_info = &opt_inst->info;
			tmp_ifindex = opt_inst_info->port ?
				      opt_inst_info->port->dev->ifindex : 0;
2607
			if (option->type != opt_type ||
2608
			    strcmp(option->name, opt_name) ||
J
Jiri Pirko 已提交
2609 2610
			    tmp_ifindex != opt_port_ifindex ||
			    (option->array_size && !opt_is_array) ||
2611
			    opt_inst_info->array_index != opt_array_index)
2612 2613
				continue;
			opt_found = true;
2614
			ctx.info = opt_inst_info;
2615 2616
			switch (opt_type) {
			case TEAM_OPTION_TYPE_U32:
J
Jiri Pirko 已提交
2617
				ctx.data.u32_val = nla_get_u32(attr_data);
2618 2619
				break;
			case TEAM_OPTION_TYPE_STRING:
J
Jiri Pirko 已提交
2620
				if (nla_len(attr_data) > TEAM_STRING_MAX_LEN) {
J
Jiri Pirko 已提交
2621 2622 2623
					err = -EINVAL;
					goto team_put;
				}
J
Jiri Pirko 已提交
2624
				ctx.data.str_val = nla_data(attr_data);
2625
				break;
J
Jiri Pirko 已提交
2626
			case TEAM_OPTION_TYPE_BINARY:
J
Jiri Pirko 已提交
2627 2628 2629 2630 2631
				ctx.data.bin_val.len = nla_len(attr_data);
				ctx.data.bin_val.ptr = nla_data(attr_data);
				break;
			case TEAM_OPTION_TYPE_BOOL:
				ctx.data.bool_val = attr_data ? true : false;
J
Jiri Pirko 已提交
2632
				break;
2633 2634 2635
			case TEAM_OPTION_TYPE_S32:
				ctx.data.s32_val = nla_get_s32(attr_data);
				break;
2636 2637 2638
			default:
				BUG();
			}
2639
			err = team_option_set(team, opt_inst, &ctx);
2640 2641
			if (err)
				goto team_put;
2642 2643
			opt_inst->changed = true;
			list_add(&opt_inst->tmp_list, &opt_inst_list);
2644 2645 2646 2647 2648 2649
		}
		if (!opt_found) {
			err = -ENOENT;
			goto team_put;
		}

2650 2651 2652 2653
		err = team_nl_send_event_options_get(team, &opt_inst_list);
		if (err)
			break;
	}
2654

2655 2656
team_put:
	team_nl_team_put(team);
J
Jiri Pirko 已提交
2657 2658
rtnl_unlock:
	rtnl_unlock();
2659 2660 2661
	return err;
}

2662 2663 2664 2665 2666
static int team_nl_fill_one_port_get(struct sk_buff *skb,
				     struct team_port *port)
{
	struct nlattr *port_item;

2667
	port_item = nla_nest_start_noflag(skb, TEAM_ATTR_ITEM_PORT);
2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694
	if (!port_item)
		goto nest_cancel;
	if (nla_put_u32(skb, TEAM_ATTR_PORT_IFINDEX, port->dev->ifindex))
		goto nest_cancel;
	if (port->changed) {
		if (nla_put_flag(skb, TEAM_ATTR_PORT_CHANGED))
			goto nest_cancel;
		port->changed = false;
	}
	if ((port->removed &&
	     nla_put_flag(skb, TEAM_ATTR_PORT_REMOVED)) ||
	    (port->state.linkup &&
	     nla_put_flag(skb, TEAM_ATTR_PORT_LINKUP)) ||
	    nla_put_u32(skb, TEAM_ATTR_PORT_SPEED, port->state.speed) ||
	    nla_put_u8(skb, TEAM_ATTR_PORT_DUPLEX, port->state.duplex))
		goto nest_cancel;
	nla_nest_end(skb, port_item);
	return 0;

nest_cancel:
	nla_nest_cancel(skb, port_item);
	return -EMSGSIZE;
}

static int team_nl_send_port_list_get(struct team *team, u32 portid, u32 seq,
				      int flags, team_nl_send_func_t *send_func,
				      struct team_port *one_port)
2695 2696
{
	struct nlattr *port_list;
2697
	struct nlmsghdr *nlh;
2698 2699
	void *hdr;
	struct team_port *port;
2700 2701 2702 2703
	int err;
	struct sk_buff *skb = NULL;
	bool incomplete;
	int i;
2704

2705 2706
	port = list_first_entry_or_null(&team->port_list,
					struct team_port, list);
2707 2708 2709 2710 2711 2712 2713

start_again:
	err = __send_and_alloc_skb(&skb, team, portid, send_func);
	if (err)
		return err;

	hdr = genlmsg_put(skb, portid, seq, &team_nl_family, flags | NLM_F_MULTI,
2714
			  TEAM_CMD_PORT_LIST_GET);
P
Pan Bian 已提交
2715 2716
	if (!hdr) {
		nlmsg_free(skb);
W
Wei Yongjun 已提交
2717
		return -EMSGSIZE;
P
Pan Bian 已提交
2718
	}
2719

D
David S. Miller 已提交
2720 2721
	if (nla_put_u32(skb, TEAM_ATTR_TEAM_IFINDEX, team->dev->ifindex))
		goto nla_put_failure;
2722
	port_list = nla_nest_start_noflag(skb, TEAM_ATTR_LIST_PORT);
2723
	if (!port_list)
2724
		goto nla_put_failure;
2725

2726 2727
	i = 0;
	incomplete = false;
2728

2729 2730 2731 2732 2733 2734 2735
	/* If one port is selected, called wants to send port list containing
	 * only this port. Otherwise go through all listed ports and send all
	 */
	if (one_port) {
		err = team_nl_fill_one_port_get(skb, one_port);
		if (err)
			goto errout;
2736 2737
	} else if (port) {
		list_for_each_entry_from(port, &team->port_list, list) {
2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748
			err = team_nl_fill_one_port_get(skb, port);
			if (err) {
				if (err == -EMSGSIZE) {
					if (!i)
						goto errout;
					incomplete = true;
					break;
				}
				goto errout;
			}
			i++;
2749
		}
2750 2751 2752
	}

	nla_nest_end(skb, port_list);
2753 2754 2755 2756 2757 2758 2759 2760 2761
	genlmsg_end(skb, hdr);
	if (incomplete)
		goto start_again;

send_done:
	nlh = nlmsg_put(skb, portid, seq, NLMSG_DONE, 0, flags | NLM_F_MULTI);
	if (!nlh) {
		err = __send_and_alloc_skb(&skb, team, portid, send_func);
		if (err)
2762
			return err;
2763 2764 2765 2766
		goto send_done;
	}

	return send_func(skb, team, portid);
2767 2768

nla_put_failure:
2769 2770 2771 2772
	err = -EMSGSIZE;
errout:
	nlmsg_free(skb);
	return err;
2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784
}

static int team_nl_cmd_port_list_get(struct sk_buff *skb,
				     struct genl_info *info)
{
	struct team *team;
	int err;

	team = team_nl_team_get(info);
	if (!team)
		return -EINVAL;

2785 2786
	err = team_nl_send_port_list_get(team, info->snd_portid, info->snd_seq,
					 NLM_F_ACK, team_nl_send_unicast, NULL);
2787 2788 2789 2790 2791 2792

	team_nl_team_put(team);

	return err;
}

2793
static const struct genl_ops team_nl_ops[] = {
2794 2795
	{
		.cmd = TEAM_CMD_NOOP,
2796
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2797 2798 2799 2800
		.doit = team_nl_cmd_noop,
	},
	{
		.cmd = TEAM_CMD_OPTIONS_SET,
2801
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2802 2803 2804 2805 2806
		.doit = team_nl_cmd_options_set,
		.flags = GENL_ADMIN_PERM,
	},
	{
		.cmd = TEAM_CMD_OPTIONS_GET,
2807
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2808 2809 2810 2811 2812
		.doit = team_nl_cmd_options_get,
		.flags = GENL_ADMIN_PERM,
	},
	{
		.cmd = TEAM_CMD_PORT_LIST_GET,
2813
		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
2814 2815 2816 2817 2818
		.doit = team_nl_cmd_port_list_get,
		.flags = GENL_ADMIN_PERM,
	},
};

2819 2820
static const struct genl_multicast_group team_nl_mcgrps[] = {
	{ .name = TEAM_GENL_CHANGE_EVENT_MC_GRP_NAME, },
2821 2822
};

2823
static struct genl_family team_nl_family __ro_after_init = {
2824 2825 2826
	.name		= TEAM_GENL_NAME,
	.version	= TEAM_GENL_VERSION,
	.maxattr	= TEAM_ATTR_MAX,
2827
	.policy = team_nl_policy,
2828 2829 2830 2831 2832 2833 2834 2835
	.netnsok	= true,
	.module		= THIS_MODULE,
	.ops		= team_nl_ops,
	.n_ops		= ARRAY_SIZE(team_nl_ops),
	.mcgrps		= team_nl_mcgrps,
	.n_mcgrps	= ARRAY_SIZE(team_nl_mcgrps),
};

2836
static int team_nl_send_multicast(struct sk_buff *skb,
2837
				  struct team *team, u32 portid)
2838
{
2839
	return genlmsg_multicast_netns(&team_nl_family, dev_net(team->dev),
2840
				       skb, 0, 0, GFP_KERNEL);
2841 2842
}

2843 2844
static int team_nl_send_event_options_get(struct team *team,
					  struct list_head *sel_opt_inst_list)
2845
{
2846 2847
	return team_nl_send_options_get(team, 0, 0, 0, team_nl_send_multicast,
					sel_opt_inst_list);
2848 2849
}

2850 2851
static int team_nl_send_event_port_get(struct team *team,
				       struct team_port *port)
2852
{
2853 2854
	return team_nl_send_port_list_get(team, 0, 0, 0, team_nl_send_multicast,
					  port);
2855 2856
}

2857
static int __init team_nl_init(void)
2858
{
2859
	return genl_register_family(&team_nl_family);
2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871
}

static void team_nl_fini(void)
{
	genl_unregister_family(&team_nl_family);
}


/******************
 * Change checkers
 ******************/

2872
static void __team_options_change_check(struct team *team)
2873 2874
{
	int err;
2875 2876
	struct team_option_inst *opt_inst;
	LIST_HEAD(sel_opt_inst_list);
2877

2878 2879 2880 2881 2882
	list_for_each_entry(opt_inst, &team->option_inst_list, list) {
		if (opt_inst->changed)
			list_add_tail(&opt_inst->tmp_list, &sel_opt_inst_list);
	}
	err = team_nl_send_event_options_get(team, &sel_opt_inst_list);
2883
	if (err && err != -ESRCH)
2884 2885
		netdev_warn(team->dev, "Failed to send options change via netlink (err %d)\n",
			    err);
2886 2887 2888
}

/* rtnl lock is held */
J
Jiri Pirko 已提交
2889 2890

static void __team_port_change_send(struct team_port *port, bool linkup)
2891 2892 2893
{
	int err;

2894
	port->changed = true;
2895 2896
	port->state.linkup = linkup;
	team_refresh_port_linkup(port);
2897
	if (linkup) {
2898
		struct ethtool_link_ksettings ecmd;
2899

2900
		err = __ethtool_get_link_ksettings(port->dev, &ecmd);
2901
		if (!err) {
2902 2903
			port->state.speed = ecmd.base.speed;
			port->state.duplex = ecmd.base.duplex;
2904 2905 2906
			goto send_event;
		}
	}
2907 2908
	port->state.speed = 0;
	port->state.duplex = 0;
2909 2910

send_event:
2911
	err = team_nl_send_event_port_get(port->team, port);
2912 2913 2914
	if (err && err != -ESRCH)
		netdev_warn(port->team->dev, "Failed to send port change of device %s via netlink (err %d)\n",
			    port->dev->name, err);
2915 2916 2917

}

2918 2919 2920 2921 2922
static void __team_carrier_check(struct team *team)
{
	struct team_port *port;
	bool team_linkup;

2923 2924 2925
	if (team->user_carrier_enabled)
		return;

2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939
	team_linkup = false;
	list_for_each_entry(port, &team->port_list, list) {
		if (port->linkup) {
			team_linkup = true;
			break;
		}
	}

	if (team_linkup)
		netif_carrier_on(team->dev);
	else
		netif_carrier_off(team->dev);
}

J
Jiri Pirko 已提交
2940 2941 2942 2943
static void __team_port_change_check(struct team_port *port, bool linkup)
{
	if (port->state.linkup != linkup)
		__team_port_change_send(port, linkup);
2944
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2945 2946 2947 2948 2949
}

static void __team_port_change_port_added(struct team_port *port, bool linkup)
{
	__team_port_change_send(port, linkup);
2950
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2951 2952 2953 2954 2955 2956
}

static void __team_port_change_port_removed(struct team_port *port)
{
	port->removed = true;
	__team_port_change_send(port, false);
2957
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2958 2959
}

2960 2961 2962 2963
static void team_port_change_check(struct team_port *port, bool linkup)
{
	struct team *team = port->team;

2964
	mutex_lock(&team->lock);
2965
	__team_port_change_check(port, linkup);
2966
	mutex_unlock(&team->lock);
2967 2968
}

J
Jiri Pirko 已提交
2969

2970 2971 2972 2973 2974 2975 2976
/************************************
 * Net device notifier event handler
 ************************************/

static int team_device_event(struct notifier_block *unused,
			     unsigned long event, void *ptr)
{
2977
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
2978 2979 2980 2981 2982 2983 2984 2985
	struct team_port *port;

	port = team_port_get_rtnl(dev);
	if (!port)
		return NOTIFY_DONE;

	switch (event) {
	case NETDEV_UP:
2986
		if (netif_oper_up(dev))
2987
			team_port_change_check(port, true);
2988
		break;
2989 2990
	case NETDEV_DOWN:
		team_port_change_check(port, false);
2991
		break;
2992 2993 2994
	case NETDEV_CHANGE:
		if (netif_running(port->dev))
			team_port_change_check(port,
G
George Wilkie 已提交
2995
					       !!netif_oper_up(port->dev));
2996 2997 2998 2999 3000 3001 3002
		break;
	case NETDEV_UNREGISTER:
		team_del_slave(port->team->dev, dev);
		break;
	case NETDEV_FEAT_CHANGE:
		team_compute_features(port->team);
		break;
3003
	case NETDEV_PRECHANGEMTU:
3004
		/* Forbid to change mtu of underlaying device */
J
Jiri Pirko 已提交
3005 3006 3007
		if (!port->team->port_mtu_change_allowed)
			return NOTIFY_BAD;
		break;
3008 3009 3010
	case NETDEV_PRE_TYPE_CHANGE:
		/* Forbid to change type of underlaying device */
		return NOTIFY_BAD;
3011 3012 3013 3014
	case NETDEV_RESEND_IGMP:
		/* Propagate to master device */
		call_netdevice_notifiers(event, port->team->dev);
		break;
3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066
	}
	return NOTIFY_DONE;
}

static struct notifier_block team_notifier_block __read_mostly = {
	.notifier_call = team_device_event,
};


/***********************
 * Module init and exit
 ***********************/

static int __init team_module_init(void)
{
	int err;

	register_netdevice_notifier(&team_notifier_block);

	err = rtnl_link_register(&team_link_ops);
	if (err)
		goto err_rtnl_reg;

	err = team_nl_init();
	if (err)
		goto err_nl_init;

	return 0;

err_nl_init:
	rtnl_link_unregister(&team_link_ops);

err_rtnl_reg:
	unregister_netdevice_notifier(&team_notifier_block);

	return err;
}

static void __exit team_module_exit(void)
{
	team_nl_fini();
	rtnl_link_unregister(&team_link_ops);
	unregister_netdevice_notifier(&team_notifier_block);
}

module_init(team_module_init);
module_exit(team_module_exit);

MODULE_LICENSE("GPL v2");
MODULE_AUTHOR("Jiri Pirko <jpirko@redhat.com>");
MODULE_DESCRIPTION("Ethernet team device driver");
MODULE_ALIAS_RTNL_LINK(DRV_NAME);