team.c 71.2 KB
Newer Older
1
/*
2
 * drivers/net/team/team.c - Network team device driver
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
 * Copyright (c) 2011 Jiri Pirko <jpirko@redhat.com>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 */

#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
21
#include <linux/netpoll.h>
22
#include <linux/if_vlan.h>
23 24 25 26 27 28 29
#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
30
#include <net/sch_generic.h>
31
#include <net/switchdev.h>
F
Flavio Leitner 已提交
32
#include <generated/utsrelease.h>
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
#include <linux/if_team.h>

#define DRV_NAME "team"


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

#define team_port_exists(dev) (dev->priv_flags & IFF_TEAM_PORT)

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

	return team_port_exists(dev) ? port : NULL;
}

/*
52
 * Since the ability to change device address for open port device is tested in
53 54
 * team_port_add, this function can be called without control of return value
 */
55 56
static int __set_port_dev_addr(struct net_device *port_dev,
			       const unsigned char *dev_addr)
57
{
58
	struct sockaddr_storage addr;
59

60 61 62
	memcpy(addr.__data, dev_addr, port_dev->addr_len);
	addr.ss_family = port_dev->type;
	return dev_set_mac_address(port_dev, (struct sockaddr *)&addr);
63 64
}

65
static int team_port_set_orig_dev_addr(struct team_port *port)
66
{
67
	return __set_port_dev_addr(port->dev, port->orig.dev_addr);
68 69
}

70 71
static int team_port_set_team_dev_addr(struct team *team,
				       struct team_port *port)
72
{
73
	return __set_port_dev_addr(port->dev, team->dev->dev_addr);
74
}
75 76 77 78 79 80 81 82 83 84 85 86 87

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);
88

89 90 91 92 93 94 95 96 97
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);
}

98 99
static void team_refresh_port_linkup(struct team_port *port)
{
100 101 102 103 104 105 106
	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);
	}
107
}
108

J
Jiri Pirko 已提交
109

110 111 112 113
/*******************
 * Options handling
 *******************/

114 115
struct team_option_inst { /* One for each option instance */
	struct list_head list;
116
	struct list_head tmp_list;
117
	struct team_option *option;
118
	struct team_option_inst_info info;
119 120 121 122 123 124
	bool changed;
	bool removed;
};

static struct team_option *__team_find_option(struct team *team,
					      const char *opt_name)
J
Jiri Pirko 已提交
125 126 127 128 129 130 131 132 133 134
{
	struct team_option *option;

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

135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151
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
152 153 154 155 156 157
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;
158
	int err;
J
Jiri Pirko 已提交
159 160 161 162 163 164 165 166 167 168

	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;
169 170
		opt_inst->info.port = port;
		opt_inst->info.array_index = i;
J
Jiri Pirko 已提交
171 172 173
		opt_inst->changed = true;
		opt_inst->removed = false;
		list_add_tail(&opt_inst->list, &team->option_inst_list);
174 175 176 177 178 179
		if (option->init) {
			err = option->init(team, &opt_inst->info);
			if (err)
				return err;
		}

J
Jiri Pirko 已提交
180 181 182 183
	}
	return 0;
}

184 185 186 187 188
static int __team_option_inst_add_option(struct team *team,
					 struct team_option *option)
{
	int err;

J
Jiri Pirko 已提交
189
	if (!option->per_port) {
190
		err = __team_option_inst_add(team, option, NULL);
J
Jiri Pirko 已提交
191 192 193
		if (err)
			goto inst_del_option;
	}
194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220
	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 &&
221
		    opt_inst->info.port == port)
222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251
			__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) {
252
		if (opt_inst->info.port == port) {
253 254 255 256 257 258
			opt_inst->changed = true;
			opt_inst->removed = true;
		}
	}
}

259 260 261 262 263 264 265 266 267 268 269
static bool __team_option_inst_tmp_find(const struct list_head *opts,
					const struct team_option_inst *needle)
{
	struct team_option_inst *opt_inst;

	list_for_each_entry(opt_inst, opts, tmp_list)
		if (opt_inst == needle)
			return true;
	return false;
}

270 271 272
static int __team_options_register(struct team *team,
				   const struct team_option *option,
				   size_t option_count)
273 274
{
	int i;
275
	struct team_option **dst_opts;
J
Jiri Pirko 已提交
276
	int err;
277

K
Kees Cook 已提交
278
	dst_opts = kcalloc(option_count, sizeof(struct team_option *),
279 280 281
			   GFP_KERNEL);
	if (!dst_opts)
		return -ENOMEM;
J
Jiri Pirko 已提交
282 283 284
	for (i = 0; i < option_count; i++, option++) {
		if (__team_find_option(team, option->name)) {
			err = -EEXIST;
285
			goto alloc_rollback;
J
Jiri Pirko 已提交
286
		}
287 288
		dst_opts[i] = kmemdup(option, sizeof(*option), GFP_KERNEL);
		if (!dst_opts[i]) {
J
Jiri Pirko 已提交
289
			err = -ENOMEM;
290
			goto alloc_rollback;
J
Jiri Pirko 已提交
291 292 293
		}
	}

294
	for (i = 0; i < option_count; i++) {
295 296 297
		err = __team_option_inst_add_option(team, dst_opts[i]);
		if (err)
			goto inst_rollback;
J
Jiri Pirko 已提交
298
		list_add_tail(&dst_opts[i]->list, &team->option_list);
299
	}
J
Jiri Pirko 已提交
300

301
	kfree(dst_opts);
J
Jiri Pirko 已提交
302 303
	return 0;

304 305 306 307 308 309 310
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
311 312
		kfree(dst_opts[i]);

313
	kfree(dst_opts);
J
Jiri Pirko 已提交
314
	return err;
315
}
J
Jiri Pirko 已提交
316

317 318 319 320 321 322 323 324
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;
325

326
		del_opt = __team_find_option(team, option->name);
327 328
		if (del_opt)
			__team_option_inst_mark_removed_option(team, del_opt);
329 330
	}
}
331 332

static void __team_options_unregister(struct team *team,
J
Jiri Pirko 已提交
333
				      const struct team_option *option,
334 335 336 337
				      size_t option_count)
{
	int i;

J
Jiri Pirko 已提交
338 339 340 341 342
	for (i = 0; i < option_count; i++, option++) {
		struct team_option *del_opt;

		del_opt = __team_find_option(team, option->name);
		if (del_opt) {
343
			__team_option_inst_del_option(team, del_opt);
J
Jiri Pirko 已提交
344 345 346 347
			list_del(&del_opt->list);
			kfree(del_opt);
		}
	}
348 349
}

350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365
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
366 367
void team_options_unregister(struct team *team,
			     const struct team_option *option,
368 369
			     size_t option_count)
{
370 371
	__team_options_mark_removed(team, option, option_count);
	__team_options_change_check(team);
372 373 374 375
	__team_options_unregister(team, option, option_count);
}
EXPORT_SYMBOL(team_options_unregister);

376 377 378 379
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
380 381
	if (!opt_inst->option->getter)
		return -EOPNOTSUPP;
382 383 384 385 386 387
	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)
388
{
J
Jiri Pirko 已提交
389 390
	if (!opt_inst->option->setter)
		return -EOPNOTSUPP;
391
	return opt_inst->option->setter(team, ctx);
392 393
}

J
Jiri Pirko 已提交
394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409
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);


410 411 412 413 414 415 416
/****************
 * Mode handling
 ****************/

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

J
Jiri Pirko 已提交
417 418 419 420 421 422
struct team_mode_item {
	struct list_head list;
	const struct team_mode *mode;
};

static struct team_mode_item *__find_mode(const char *kind)
423
{
J
Jiri Pirko 已提交
424
	struct team_mode_item *mitem;
425

J
Jiri Pirko 已提交
426 427 428
	list_for_each_entry(mitem, &mode_list, list) {
		if (strcmp(mitem->mode->kind, kind) == 0)
			return mitem;
429 430 431 432 433 434 435 436 437 438 439 440 441 442
	}
	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
443
int team_mode_register(const struct team_mode *mode)
444 445
{
	int err = 0;
J
Jiri Pirko 已提交
446
	struct team_mode_item *mitem;
447 448 449 450

	if (!is_good_mode_name(mode->kind) ||
	    mode->priv_size > TEAM_MODE_PRIV_SIZE)
		return -EINVAL;
J
Jiri Pirko 已提交
451 452 453 454 455

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

456 457 458
	spin_lock(&mode_list_lock);
	if (__find_mode(mode->kind)) {
		err = -EEXIST;
J
Jiri Pirko 已提交
459
		kfree(mitem);
460 461
		goto unlock;
	}
J
Jiri Pirko 已提交
462 463
	mitem->mode = mode;
	list_add_tail(&mitem->list, &mode_list);
464 465 466 467 468 469
unlock:
	spin_unlock(&mode_list_lock);
	return err;
}
EXPORT_SYMBOL(team_mode_register);

J
Jiri Pirko 已提交
470
void team_mode_unregister(const struct team_mode *mode)
471
{
J
Jiri Pirko 已提交
472 473
	struct team_mode_item *mitem;

474
	spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
475 476 477 478 479
	mitem = __find_mode(mode->kind);
	if (mitem) {
		list_del_init(&mitem->list);
		kfree(mitem);
	}
480 481 482 483
	spin_unlock(&mode_list_lock);
}
EXPORT_SYMBOL(team_mode_unregister);

J
Jiri Pirko 已提交
484
static const struct team_mode *team_mode_get(const char *kind)
485
{
J
Jiri Pirko 已提交
486 487
	struct team_mode_item *mitem;
	const struct team_mode *mode = NULL;
488 489

	spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
490 491
	mitem = __find_mode(kind);
	if (!mitem) {
492 493 494
		spin_unlock(&mode_list_lock);
		request_module("team-mode-%s", kind);
		spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
495
		mitem = __find_mode(kind);
496
	}
J
Jiri Pirko 已提交
497 498
	if (mitem) {
		mode = mitem->mode;
499 500
		if (!try_module_get(mode->owner))
			mode = NULL;
J
Jiri Pirko 已提交
501
	}
502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517

	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;
}

518 519 520
static rx_handler_result_t team_dummy_receive(struct team *team,
					      struct team_port *port,
					      struct sk_buff *skb)
521 522 523 524
{
	return RX_HANDLER_ANOTHER;
}

525 526 527 528 529 530 531 532 533 534 535
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)
{
536
	team->user_carrier_enabled = false;
537 538 539
	team->mode = &__team_no_mode;
}

540
static void team_adjust_ops(struct team *team)
541 542 543 544 545 546
{
	/*
	 * To avoid checks in rx/tx skb paths, ensure here that non-null and
	 * correct ops are always set.
	 */

547
	if (!team->en_port_count || !team_is_mode_set(team) ||
548
	    !team->mode->ops->transmit)
549 550 551 552
		team->ops.transmit = team_dummy_transmit;
	else
		team->ops.transmit = team->mode->ops->transmit;

553
	if (!team->en_port_count || !team_is_mode_set(team) ||
554
	    !team->mode->ops->receive)
555 556 557 558 559 560 561 562 563 564 565 566 567 568
		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 */
569
	if (team_is_mode_set(team)) {
570 571 572 573 574 575 576 577 578
		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);
579
		team_set_no_mode(team);
580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604
		/* 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
605
	const struct team_mode *new_mode;
606 607 608 609 610 611 612 613
	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;
	}

614
	if (team_is_mode_set(team) && strcmp(team->mode->kind, kind) == 0) {
615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
		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
637 638 639 640 641 642 643
/*********************
 * Peers notification
 *********************/

static void team_notify_peers_work(struct work_struct *work)
{
	struct team *team;
644
	int val;
J
Jiri Pirko 已提交
645 646 647 648 649 650 651

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

	if (!rtnl_trylock()) {
		schedule_delayed_work(&team->notify_peers.dw, 0);
		return;
	}
652 653 654 655 656
	val = atomic_dec_if_positive(&team->notify_peers.count_pending);
	if (val < 0) {
		rtnl_unlock();
		return;
	}
J
Jiri Pirko 已提交
657 658
	call_netdevice_notifiers(NETDEV_NOTIFY_PEERS, team->dev);
	rtnl_unlock();
659
	if (val)
J
Jiri Pirko 已提交
660 661 662 663 664 665 666 667
		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;
668
	atomic_add(team->notify_peers.count, &team->notify_peers.count_pending);
J
Jiri Pirko 已提交
669 670 671 672 673 674 675 676 677 678 679 680 681 682
	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);
}


683 684 685 686 687 688 689
/*******************************
 * Send multicast group rejoins
 *******************************/

static void team_mcast_rejoin_work(struct work_struct *work)
{
	struct team *team;
690
	int val;
691 692 693 694 695 696 697

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

	if (!rtnl_trylock()) {
		schedule_delayed_work(&team->mcast_rejoin.dw, 0);
		return;
	}
698 699 700 701 702
	val = atomic_dec_if_positive(&team->mcast_rejoin.count_pending);
	if (val < 0) {
		rtnl_unlock();
		return;
	}
703 704
	call_netdevice_notifiers(NETDEV_RESEND_IGMP, team->dev);
	rtnl_unlock();
705
	if (val)
706 707 708 709 710 711 712 713
		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;
714
	atomic_add(team->mcast_rejoin.count, &team->mcast_rejoin.count_pending);
715 716 717 718 719 720 721 722 723 724 725 726 727 728
	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);
}


729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748
/************************
 * 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
749 750 751 752 753 754
	if (!team_port_enabled(port)) {
		/* allow exact match delivery for disabled ports */
		res = RX_HANDLER_EXACT;
	} else {
		res = team->ops.receive(team, port, skb);
	}
755 756 757 758 759 760 761 762 763 764 765 766
	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;
767 768
	} else if (res == RX_HANDLER_EXACT) {
		this_cpu_inc(team->pcpu_stats->rx_nohandler);
769 770 771 772 773 774 775 776
	} else {
		this_cpu_inc(team->pcpu_stats->rx_dropped);
	}

	return res;
}


777 778 779 780 781 782 783 784 785 786 787 788
/*************************************
 * 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;
789 790
	listarr = kmalloc_array(queue_cnt, sizeof(struct list_head),
				GFP_KERNEL);
791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829
	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)
{
830 831
	if (!port->queue_id)
		return;
832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853
	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;

854
	if (!port->queue_id)
855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871
		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) {
872
		if (port->queue_id) {
873 874 875 876 877 878 879 880 881 882 883
			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;
}

884 885
static void team_queue_override_port_prio_changed(struct team *team,
						  struct team_port *port)
886
{
887 888
	if (!port->queue_id || team_port_enabled(port))
		return;
889 890 891 892 893
	__team_queue_override_port_del(team, port);
	__team_queue_override_port_add(team, port);
	__team_queue_override_enabled_check(team);
}

894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921
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);
}

922

923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938
/****************
 * 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
939 940 941 942
 * 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.
943
 */
J
Jiri Pirko 已提交
944 945
static void team_port_enable(struct team *team,
			     struct team_port *port)
946
{
J
Jiri Pirko 已提交
947 948 949
	if (team_port_enabled(port))
		return;
	port->index = team->en_port_count++;
950 951
	hlist_add_head_rcu(&port->hlist,
			   team_port_index_hash(team, port->index));
952
	team_adjust_ops(team);
953
	team_queue_override_port_add(team, port);
954 955
	if (team->ops.port_enabled)
		team->ops.port_enabled(team, port);
J
Jiri Pirko 已提交
956
	team_notify_peers(team);
957
	team_mcast_rejoin(team);
958
	team_lower_state_changed(port);
959 960 961 962 963 964 965
}

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

J
Jiri Pirko 已提交
966
	for (i = rm_index + 1; i < team->en_port_count; i++) {
967 968 969 970 971 972 973 974
		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
975 976
static void team_port_disable(struct team *team,
			      struct team_port *port)
977
{
J
Jiri Pirko 已提交
978 979
	if (!team_port_enabled(port))
		return;
980 981
	if (team->ops.port_disabled)
		team->ops.port_disabled(team, port);
982
	hlist_del_rcu(&port->hlist);
983
	__reconstruct_port_hlist(team, port->index);
J
Jiri Pirko 已提交
984
	port->index = -1;
985
	team->en_port_count--;
986 987
	team_queue_override_port_del(team, port);
	team_adjust_ops(team);
J
Jiri Pirko 已提交
988
	team_notify_peers(team);
989
	team_mcast_rejoin(team);
990
	team_lower_state_changed(port);
991 992
}

993
#define TEAM_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
994 995 996
			    NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
			    NETIF_F_HIGHDMA | NETIF_F_LRO)

997 998 999
#define TEAM_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
				 NETIF_F_RXCSUM | NETIF_F_ALL_TSO)

1000
static void __team_compute_features(struct team *team)
1001 1002
{
	struct team_port *port;
1003 1004
	netdev_features_t vlan_features = TEAM_VLAN_FEATURES &
					  NETIF_F_ALL_FOR_ALL;
1005
	netdev_features_t enc_features  = TEAM_ENC_FEATURES;
1006
	unsigned short max_hard_header_len = ETH_HLEN;
1007 1008
	unsigned int dst_release_flag = IFF_XMIT_DST_RELEASE |
					IFF_XMIT_DST_RELEASE_PERM;
1009 1010 1011 1012 1013

	list_for_each_entry(port, &team->port_list, list) {
		vlan_features = netdev_increment_features(vlan_features,
					port->dev->vlan_features,
					TEAM_VLAN_FEATURES);
1014 1015 1016 1017 1018
		enc_features =
			netdev_increment_features(enc_features,
						  port->dev->hw_enc_features,
						  TEAM_ENC_FEATURES);

1019

1020
		dst_release_flag &= port->dev->priv_flags;
1021 1022 1023 1024 1025
		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;
1026 1027
	team->dev->hw_enc_features = enc_features | NETIF_F_GSO_ENCAP_ALL |
				     NETIF_F_GSO_UDP_L4;
1028 1029
	team->dev->hard_header_len = max_hard_header_len;

1030 1031 1032
	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;
1033
}
1034

1035 1036
static void team_compute_features(struct team *team)
{
1037
	mutex_lock(&team->lock);
1038
	__team_compute_features(team);
1039
	mutex_unlock(&team->lock);
1040
	netdev_change_features(team->dev);
1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071
}

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
1072
#ifdef CONFIG_NET_POLL_CONTROLLER
X
Xin Long 已提交
1073
static int __team_port_enable_netpoll(struct team_port *port)
J
Jiri Pirko 已提交
1074 1075 1076 1077
{
	struct netpoll *np;
	int err;

1078
	np = kzalloc(sizeof(*np), GFP_KERNEL);
J
Jiri Pirko 已提交
1079 1080 1081
	if (!np)
		return -ENOMEM;

1082
	err = __netpoll_setup(np, port->dev);
J
Jiri Pirko 已提交
1083 1084 1085 1086 1087 1088 1089 1090
	if (err) {
		kfree(np);
		return err;
	}
	port->np = np;
	return err;
}

X
Xin Long 已提交
1091 1092 1093 1094 1095 1096 1097 1098
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
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112
static void team_port_disable_netpoll(struct team_port *port)
{
	struct netpoll *np = port->np;

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

	/* Wait for transmitting packets to finish before freeing. */
	synchronize_rcu_bh();
	__netpoll_cleanup(np);
	kfree(np);
}
#else
X
Xin Long 已提交
1113
static int team_port_enable_netpoll(struct team_port *port)
J
Jiri Pirko 已提交
1114 1115 1116 1117 1118 1119 1120 1121
{
	return 0;
}
static void team_port_disable_netpoll(struct team_port *port)
{
}
#endif

1122 1123
static int team_upper_dev_link(struct team *team, struct team_port *port,
			       struct netlink_ext_ack *extack)
1124
{
1125
	struct netdev_lag_upper_info lag_upper_info;
1126 1127
	int err;

1128
	lag_upper_info.tx_type = team->mode->lag_tx_type;
1129
	lag_upper_info.hash_type = NETDEV_LAG_HASH_UNKNOWN;
1130
	err = netdev_master_upper_dev_link(port->dev, team->dev, NULL,
1131
					   &lag_upper_info, extack);
1132 1133
	if (err)
		return err;
1134
	port->dev->priv_flags |= IFF_TEAM_PORT;
1135 1136 1137
	return 0;
}

1138
static void team_upper_dev_unlink(struct team *team, struct team_port *port)
1139
{
1140 1141
	netdev_upper_dev_unlink(port->dev, team->dev);
	port->dev->priv_flags &= ~IFF_TEAM_PORT;
1142 1143
}

J
Jiri Pirko 已提交
1144
static void __team_port_change_port_added(struct team_port *port, bool linkup);
1145 1146
static int team_dev_type_check_change(struct net_device *dev,
				      struct net_device *port_dev);
1147

1148 1149
static int team_port_add(struct team *team, struct net_device *port_dev,
			 struct netlink_ext_ack *extack)
1150 1151 1152 1153 1154 1155
{
	struct net_device *dev = team->dev;
	struct team_port *port;
	char *portname = port_dev->name;
	int err;

1156
	if (port_dev->flags & IFF_LOOPBACK) {
1157
		NL_SET_ERR_MSG(extack, "Loopback device can't be added as a team port");
1158
		netdev_err(dev, "Device %s is loopback device. Loopback devices can't be added as a team port\n",
1159 1160 1161 1162 1163
			   portname);
		return -EINVAL;
	}

	if (team_port_exists(port_dev)) {
1164
		NL_SET_ERR_MSG(extack, "Device is already a port of a team device");
1165 1166 1167 1168 1169
		netdev_err(dev, "Device %s is already a port "
				"of a team device\n", portname);
		return -EBUSY;
	}

1170 1171 1172 1173 1174 1175
	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;
	}

1176 1177
	if (port_dev->features & NETIF_F_VLAN_CHALLENGED &&
	    vlan_uses_dev(dev)) {
1178
		NL_SET_ERR_MSG(extack, "Device is VLAN challenged and team device has VLAN set up");
1179 1180 1181 1182 1183
		netdev_err(dev, "Device %s is VLAN challenged and team device has VLAN set up\n",
			   portname);
		return -EPERM;
	}

1184 1185 1186 1187
	err = team_dev_type_check_change(dev, port_dev);
	if (err)
		return err;

1188
	if (port_dev->flags & IFF_UP) {
1189
		NL_SET_ERR_MSG(extack, "Device is up. Set it down before adding it as a team port");
1190 1191 1192 1193 1194
		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
1195 1196
	port = kzalloc(sizeof(struct team_port) + team->mode->port_priv_size,
		       GFP_KERNEL);
1197 1198 1199 1200 1201
	if (!port)
		return -ENOMEM;

	port->dev = port_dev;
	port->team = team;
1202
	INIT_LIST_HEAD(&port->qom_list);
1203 1204 1205 1206 1207 1208 1209 1210

	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;
	}

1211
	memcpy(port->orig.dev_addr, port_dev->dev_addr, port_dev->addr_len);
1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226

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

	err = dev_open(port_dev);
	if (err) {
		netdev_dbg(dev, "Device %s opening failed\n",
			   portname);
		goto err_dev_open;
	}

J
Jiri Pirko 已提交
1227 1228 1229 1230 1231 1232 1233
	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
1234
	err = team_port_enable_netpoll(port);
S
stephen hemminger 已提交
1235 1236 1237 1238
	if (err) {
		netdev_err(dev, "Failed to enable netpoll on device %s\n",
			   portname);
		goto err_enable_netpoll;
J
Jiri Pirko 已提交
1239 1240
	}

1241 1242 1243
	if (!(dev->features & NETIF_F_LRO))
		dev_disable_lro(port_dev);

1244 1245 1246 1247 1248 1249 1250 1251
	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;
	}

1252
	err = team_upper_dev_link(team, port, extack);
1253 1254 1255 1256 1257 1258
	if (err) {
		netdev_err(dev, "Device %s failed to set upper link\n",
			   portname);
		goto err_set_upper_link;
	}

1259
	err = __team_option_inst_add_port(team, port);
1260 1261 1262 1263 1264 1265
	if (err) {
		netdev_err(dev, "Device %s failed to add per-port options\n",
			   portname);
		goto err_option_port_add;
	}

1266 1267 1268 1269 1270
	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
1271 1272
	port->index = -1;
	list_add_tail_rcu(&port->list, &team->port_list);
1273
	team_port_enable(team, port);
1274
	__team_compute_features(team);
J
Jiri Pirko 已提交
1275
	__team_port_change_port_added(port, !!netif_carrier_ok(port_dev));
1276
	__team_options_change_check(team);
1277 1278 1279 1280 1281

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

	return 0;

1282
err_option_port_add:
1283
	team_upper_dev_unlink(team, port);
1284 1285

err_set_upper_link:
1286 1287
	netdev_rx_handler_unregister(port_dev);

1288
err_handler_register:
J
Jiri Pirko 已提交
1289 1290 1291
	team_port_disable_netpoll(port);

err_enable_netpoll:
J
Jiri Pirko 已提交
1292 1293 1294
	vlan_vids_del_by_dev(port_dev, dev);

err_vids_add:
1295 1296 1297 1298
	dev_close(port_dev);

err_dev_open:
	team_port_leave(team, port);
1299
	team_port_set_orig_dev_addr(port);
1300 1301 1302 1303 1304 1305 1306 1307 1308 1309

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

err_set_mtu:
	kfree(port);

	return err;
}

J
Jiri Pirko 已提交
1310 1311
static void __team_port_change_port_removed(struct team_port *port);

1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324
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
1325 1326
	team_port_disable(team, port);
	list_del_rcu(&port->list);
1327
	team_upper_dev_unlink(team, port);
1328
	netdev_rx_handler_unregister(port_dev);
J
Jiri Pirko 已提交
1329
	team_port_disable_netpoll(port);
J
Jiri Pirko 已提交
1330
	vlan_vids_del_by_dev(port_dev, dev);
1331 1332
	dev_uc_unsync(port_dev, dev);
	dev_mc_unsync(port_dev, dev);
1333 1334
	dev_close(port_dev);
	team_port_leave(team, port);
1335 1336 1337 1338 1339 1340

	__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);

1341
	team_port_set_orig_dev_addr(port);
1342
	dev_set_mtu(port_dev, port->orig.mtu);
1343
	kfree_rcu(port, rcu);
1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354
	netdev_info(dev, "Port device %s removed\n", portname);
	__team_compute_features(team);

	return 0;
}


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

1355
static int team_mode_option_get(struct team *team, struct team_gsetter_ctx *ctx)
1356
{
1357
	ctx->data.str_val = team->mode->kind;
1358 1359 1360
	return 0;
}

1361
static int team_mode_option_set(struct team *team, struct team_gsetter_ctx *ctx)
1362
{
1363
	return team_change_mode(team, ctx->data.str_val);
1364 1365
}

J
Jiri Pirko 已提交
1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393
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;
}

1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421
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;
}

1422 1423 1424
static int team_port_en_option_get(struct team *team,
				   struct team_gsetter_ctx *ctx)
{
1425 1426 1427
	struct team_port *port = ctx->info->port;

	ctx->data.bool_val = team_port_enabled(port);
1428 1429 1430 1431 1432 1433
	return 0;
}

static int team_port_en_option_set(struct team *team,
				   struct team_gsetter_ctx *ctx)
{
1434 1435
	struct team_port *port = ctx->info->port;

1436
	if (ctx->data.bool_val)
1437
		team_port_enable(team, port);
1438
	else
1439
		team_port_disable(team, port);
1440 1441 1442
	return 0;
}

1443 1444 1445
static int team_user_linkup_option_get(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
1446 1447 1448
	struct team_port *port = ctx->info->port;

	ctx->data.bool_val = port->user.linkup;
1449 1450 1451
	return 0;
}

1452 1453
static void __team_carrier_check(struct team *team);

1454 1455 1456
static int team_user_linkup_option_set(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
1457 1458 1459 1460
	struct team_port *port = ctx->info->port;

	port->user.linkup = ctx->data.bool_val;
	team_refresh_port_linkup(port);
1461
	__team_carrier_check(port->team);
1462 1463 1464 1465 1466 1467
	return 0;
}

static int team_user_linkup_en_option_get(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
1468
	struct team_port *port = ctx->info->port;
1469 1470 1471 1472 1473 1474 1475 1476

	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)
{
1477
	struct team_port *port = ctx->info->port;
1478 1479

	port->user.linkup_enabled = ctx->data.bool_val;
1480
	team_refresh_port_linkup(port);
1481
	__team_carrier_check(port->team);
1482 1483 1484
	return 0;
}

J
Jiri Pirko 已提交
1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497
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;
1498
	s32 priority = ctx->data.s32_val;
J
Jiri Pirko 已提交
1499

1500 1501 1502 1503
	if (port->priority == priority)
		return 0;
	port->priority = priority;
	team_queue_override_port_prio_changed(team, port);
1504 1505 1506 1507 1508 1509 1510 1511 1512
	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
1513 1514 1515
	return 0;
}

1516 1517 1518 1519
static int team_queue_id_option_set(struct team *team,
				    struct team_gsetter_ctx *ctx)
{
	struct team_port *port = ctx->info->port;
1520
	u16 new_queue_id = ctx->data.u32_val;
1521

1522
	if (port->queue_id == new_queue_id)
1523
		return 0;
1524
	if (new_queue_id >= team->dev->real_num_tx_queues)
1525
		return -EINVAL;
1526
	team_queue_override_port_change_queue_id(team, port, new_queue_id);
1527 1528 1529
	return 0;
}

J
Jiri Pirko 已提交
1530
static const struct team_option team_options[] = {
1531 1532 1533 1534 1535 1536
	{
		.name = "mode",
		.type = TEAM_OPTION_TYPE_STRING,
		.getter = team_mode_option_get,
		.setter = team_mode_option_set,
	},
J
Jiri Pirko 已提交
1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548
	{
		.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,
	},
1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
	{
		.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,
	},
1561 1562 1563 1564 1565 1566 1567
	{
		.name = "enabled",
		.type = TEAM_OPTION_TYPE_BOOL,
		.per_port = true,
		.getter = team_port_en_option_get,
		.setter = team_port_en_option_set,
	},
1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581
	{
		.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
1582 1583 1584 1585 1586 1587 1588
	{
		.name = "priority",
		.type = TEAM_OPTION_TYPE_S32,
		.per_port = true,
		.getter = team_priority_option_get,
		.setter = team_priority_option_set,
	},
1589 1590 1591 1592 1593 1594 1595
	{
		.name = "queue_id",
		.type = TEAM_OPTION_TYPE_U32,
		.per_port = true,
		.getter = team_queue_id_option_get,
		.setter = team_queue_id_option_set,
	},
1596 1597
};

J
Jiri Pirko 已提交
1598

1599 1600 1601 1602
static int team_init(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);
	int i;
J
Jiri Pirko 已提交
1603
	int err;
1604 1605

	team->dev = dev;
1606
	mutex_init(&team->lock);
1607
	team_set_no_mode(team);
1608

1609
	team->pcpu_stats = netdev_alloc_pcpu_stats(struct team_pcpu_stats);
1610 1611 1612 1613
	if (!team->pcpu_stats)
		return -ENOMEM;

	for (i = 0; i < TEAM_PORT_HASHENTRIES; i++)
J
Jiri Pirko 已提交
1614
		INIT_HLIST_HEAD(&team->en_port_hlist[i]);
1615
	INIT_LIST_HEAD(&team->port_list);
1616 1617 1618
	err = team_queue_override_init(team);
	if (err)
		goto err_team_queue_override_init;
1619 1620 1621 1622

	team_adjust_ops(team);

	INIT_LIST_HEAD(&team->option_list);
1623
	INIT_LIST_HEAD(&team->option_inst_list);
J
Jiri Pirko 已提交
1624 1625

	team_notify_peers_init(team);
1626
	team_mcast_rejoin_init(team);
J
Jiri Pirko 已提交
1627

J
Jiri Pirko 已提交
1628 1629 1630
	err = team_options_register(team, team_options, ARRAY_SIZE(team_options));
	if (err)
		goto err_options_register;
1631 1632
	netif_carrier_off(dev);

1633
	netdev_lockdep_set_classes(dev);
J
Jiri Pirko 已提交
1634

1635
	return 0;
J
Jiri Pirko 已提交
1636 1637

err_options_register:
1638
	team_mcast_rejoin_fini(team);
J
Jiri Pirko 已提交
1639
	team_notify_peers_fini(team);
1640 1641
	team_queue_override_fini(team);
err_team_queue_override_init:
J
Jiri Pirko 已提交
1642 1643 1644
	free_percpu(team->pcpu_stats);

	return err;
1645 1646 1647 1648 1649 1650 1651 1652
}

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

1653
	mutex_lock(&team->lock);
1654 1655 1656 1657 1658
	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));
1659
	team_mcast_rejoin_fini(team);
J
Jiri Pirko 已提交
1660
	team_notify_peers_fini(team);
1661
	team_queue_override_fini(team);
1662
	mutex_unlock(&team->lock);
1663
	netdev_change_features(dev);
1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688
}

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);
1689
	bool tx_success;
1690 1691
	unsigned int len = skb->len;

1692 1693 1694
	tx_success = team_queue_override_transmit(team, skb);
	if (!tx_success)
		tx_success = team->ops.transmit(team, skb);
1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709
	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;
}

1710
static u16 team_select_queue(struct net_device *dev, struct sk_buff *skb,
1711 1712
			     struct net_device *sb_dev,
			     select_queue_fallback_t fallback)
J
Jiri Pirko 已提交
1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734
{
	/*
	 * 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;
}

1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761
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) {
1762 1763
		dev_uc_sync_multiple(port->dev, dev);
		dev_mc_sync_multiple(port->dev, dev);
1764 1765 1766 1767 1768 1769
	}
	rcu_read_unlock();
}

static int team_set_mac_address(struct net_device *dev, void *p)
{
1770
	struct sockaddr *addr = p;
1771 1772 1773
	struct team *team = netdev_priv(dev);
	struct team_port *port;

1774 1775 1776
	if (dev->type == ARPHRD_ETHER && !is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;
	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
1777 1778
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list)
1779 1780
		if (team->ops.port_change_dev_addr)
			team->ops.port_change_dev_addr(team, port);
1781
	mutex_unlock(&team->lock);
1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794
	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
	 */
1795
	mutex_lock(&team->lock);
J
Jiri Pirko 已提交
1796
	team->port_mtu_change_allowed = true;
1797 1798 1799 1800 1801 1802 1803 1804
	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
1805
	team->port_mtu_change_allowed = false;
1806
	mutex_unlock(&team->lock);
1807 1808 1809 1810 1811 1812 1813 1814

	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
1815
	team->port_mtu_change_allowed = false;
1816
	mutex_unlock(&team->lock);
1817 1818 1819 1820

	return err;
}

1821
static void
1822 1823 1824 1825 1826
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;
1827
	u32 rx_dropped = 0, tx_dropped = 0, rx_nohandler = 0;
1828 1829 1830 1831 1832 1833
	unsigned int start;
	int i;

	for_each_possible_cpu(i) {
		p = per_cpu_ptr(team->pcpu_stats, i);
		do {
1834
			start = u64_stats_fetch_begin_irq(&p->syncp);
1835 1836 1837 1838 1839
			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;
1840
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
1841 1842 1843 1844 1845 1846 1847

		stats->rx_packets	+= rx_packets;
		stats->rx_bytes		+= rx_bytes;
		stats->multicast	+= rx_multicast;
		stats->tx_packets	+= tx_packets;
		stats->tx_bytes		+= tx_bytes;
		/*
1848 1849
		 * rx_dropped, tx_dropped & rx_nohandler are u32,
		 * updated without syncp protection.
1850 1851 1852
		 */
		rx_dropped	+= p->rx_dropped;
		tx_dropped	+= p->tx_dropped;
1853
		rx_nohandler	+= p->rx_nohandler;
1854 1855 1856
	}
	stats->rx_dropped	= rx_dropped;
	stats->tx_dropped	= tx_dropped;
1857
	stats->rx_nohandler	= rx_nohandler;
1858 1859
}

1860
static int team_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
1861 1862 1863
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
1864
	int err;
1865

1866 1867 1868 1869 1870 1871
	/*
	 * 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) {
1872
		err = vlan_vid_add(port->dev, proto, vid);
1873 1874
		if (err)
			goto unwind;
1875
	}
1876
	mutex_unlock(&team->lock);
1877 1878

	return 0;
1879 1880 1881

unwind:
	list_for_each_entry_continue_reverse(port, &team->port_list, list)
1882
		vlan_vid_del(port->dev, proto, vid);
1883 1884 1885
	mutex_unlock(&team->lock);

	return err;
1886 1887
}

1888
static int team_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
1889 1890 1891 1892
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;

1893 1894
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list)
1895
		vlan_vid_del(port->dev, proto, vid);
1896
	mutex_unlock(&team->lock);
1897 1898

	return 0;
1899 1900
}

J
Jiri Pirko 已提交
1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923
#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,
1924
			      struct netpoll_info *npifo)
J
Jiri Pirko 已提交
1925 1926 1927
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
1928
	int err = 0;
J
Jiri Pirko 已提交
1929 1930 1931

	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list) {
X
Xin Long 已提交
1932
		err = __team_port_enable_netpoll(port);
J
Jiri Pirko 已提交
1933 1934 1935 1936 1937 1938 1939 1940 1941 1942
		if (err) {
			__team_netpoll_cleanup(team);
			break;
		}
	}
	mutex_unlock(&team->lock);
	return err;
}
#endif

D
David Ahern 已提交
1943 1944
static int team_add_slave(struct net_device *dev, struct net_device *port_dev,
			  struct netlink_ext_ack *extack)
1945 1946 1947 1948
{
	struct team *team = netdev_priv(dev);
	int err;

1949
	mutex_lock(&team->lock);
1950
	err = team_port_add(team, port_dev, extack);
1951
	mutex_unlock(&team->lock);
1952 1953 1954 1955

	if (!err)
		netdev_change_features(dev);

1956 1957 1958 1959 1960 1961 1962 1963
	return err;
}

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

1964
	mutex_lock(&team->lock);
1965
	err = team_port_del(team, port_dev);
1966
	mutex_unlock(&team->lock);
1967 1968 1969 1970

	if (!err)
		netdev_change_features(dev);

1971 1972 1973
	return err;
}

J
Jiri Pirko 已提交
1974 1975 1976 1977 1978 1979 1980
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;

1981
	mask = features;
J
Jiri Pirko 已提交
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991
	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
1992 1993 1994

	features = netdev_add_tso_features(features, mask);

J
Jiri Pirko 已提交
1995 1996 1997
	return features;
}

F
Flavio Leitner 已提交
1998 1999
static int team_change_carrier(struct net_device *dev, bool new_carrier)
{
2000 2001 2002 2003
	struct team *team = netdev_priv(dev);

	team->user_carrier_enabled = true;

F
Flavio Leitner 已提交
2004 2005 2006 2007 2008 2009 2010
	if (new_carrier)
		netif_carrier_on(dev);
	else
		netif_carrier_off(dev);
	return 0;
}

2011 2012 2013 2014 2015 2016
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
2017
	.ndo_select_queue	= team_select_queue,
2018 2019 2020 2021 2022 2023 2024
	.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
2025 2026 2027 2028 2029
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= team_poll_controller,
	.ndo_netpoll_setup	= team_netpoll_setup,
	.ndo_netpoll_cleanup	= team_netpoll_cleanup,
#endif
2030 2031
	.ndo_add_slave		= team_add_slave,
	.ndo_del_slave		= team_del_slave,
J
Jiri Pirko 已提交
2032
	.ndo_fix_features	= team_fix_features,
F
Flavio Leitner 已提交
2033
	.ndo_change_carrier     = team_change_carrier,
2034
	.ndo_features_check	= passthru_features_check,
2035 2036
};

F
Flavio Leitner 已提交
2037 2038 2039 2040 2041 2042 2043
/***********************
 * ethtool interface
 ***********************/

static void team_ethtool_get_drvinfo(struct net_device *dev,
				     struct ethtool_drvinfo *drvinfo)
{
2044 2045
	strlcpy(drvinfo->driver, DRV_NAME, sizeof(drvinfo->driver));
	strlcpy(drvinfo->version, UTS_RELEASE, sizeof(drvinfo->version));
F
Flavio Leitner 已提交
2046 2047 2048 2049 2050 2051
}

static const struct ethtool_ops team_ethtool_ops = {
	.get_drvinfo		= team_ethtool_get_drvinfo,
	.get_link		= ethtool_op_get_link,
};
2052 2053 2054 2055 2056

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

2057 2058 2059 2060 2061 2062 2063 2064 2065
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);
2066
	eth_hw_addr_inherit(dev, port_dev);
2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094
}

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;
}

2095 2096 2097
static void team_setup(struct net_device *dev)
{
	ether_setup(dev);
J
Jarod Wilson 已提交
2098
	dev->max_mtu = ETH_MAX_MTU;
2099 2100

	dev->netdev_ops = &team_netdev_ops;
F
Flavio Leitner 已提交
2101
	dev->ethtool_ops = &team_ethtool_ops;
2102 2103
	dev->needs_free_netdev = true;
	dev->priv_destructor = team_destructor;
2104
	dev->priv_flags &= ~(IFF_XMIT_DST_RELEASE | IFF_TX_SKB_SHARING);
2105
	dev->priv_flags |= IFF_NO_QUEUE;
J
Jiri Pirko 已提交
2106
	dev->priv_flags |= IFF_TEAM;
2107 2108 2109 2110 2111 2112

	/*
	 * 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.
	 */
2113
	dev->priv_flags |= IFF_UNICAST_FLT | IFF_LIVE_ADDR_CHANGE;
2114 2115 2116

	dev->features |= NETIF_F_LLTX;
	dev->features |= NETIF_F_GRO;
2117 2118 2119 2120

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

J
Jiri Pirko 已提交
2121
	dev->hw_features = TEAM_VLAN_FEATURES |
2122 2123 2124
			   NETIF_F_HW_VLAN_CTAG_TX |
			   NETIF_F_HW_VLAN_CTAG_RX |
			   NETIF_F_HW_VLAN_CTAG_FILTER;
2125

2126
	dev->hw_features |= NETIF_F_GSO_ENCAP_ALL | NETIF_F_GSO_UDP_L4;
2127 2128 2129 2130
	dev->features |= dev->hw_features;
}

static int team_newlink(struct net *src_net, struct net_device *dev,
2131 2132
			struct nlattr *tb[], struct nlattr *data[],
			struct netlink_ext_ack *extack)
2133 2134
{
	if (tb[IFLA_ADDRESS] == NULL)
2135
		eth_hw_addr_random(dev);
2136

2137
	return register_netdevice(dev);
2138 2139
}

2140 2141
static int team_validate(struct nlattr *tb[], struct nlattr *data[],
			 struct netlink_ext_ack *extack)
2142 2143 2144 2145 2146 2147 2148 2149 2150 2151
{
	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
2152 2153 2154 2155 2156 2157 2158 2159 2160 2161
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;
}

2162
static struct rtnl_link_ops team_link_ops __read_mostly = {
J
Jiri Pirko 已提交
2163 2164 2165 2166 2167 2168 2169
	.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,
2170 2171 2172 2173 2174 2175 2176
};


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

2177
static struct genl_family team_nl_family;
2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194

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
2195
	[TEAM_ATTR_OPTION_DATA]			= { .type = NLA_BINARY },
2196 2197 2198 2199 2200 2201 2202 2203
};

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

2204
	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
2205 2206 2207
	if (!msg)
		return -ENOMEM;

2208
	hdr = genlmsg_put(msg, info->snd_portid, info->snd_seq,
2209
			  &team_nl_family, 0, TEAM_CMD_NOOP);
W
Wei Yongjun 已提交
2210 2211
	if (!hdr) {
		err = -EMSGSIZE;
2212 2213 2214 2215 2216
		goto err_msg_put;
	}

	genlmsg_end(msg, hdr);

2217
	return genlmsg_unicast(genl_info_net(info), msg, info->snd_portid);
2218 2219 2220 2221 2222 2223 2224 2225 2226

err_msg_put:
	nlmsg_free(msg);

	return err;
}

/*
 * Netlink cmd functions should be locked by following two functions.
2227
 * Since dev gets held here, that ensures dev won't disappear in between.
2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239
 */
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]);
2240
	dev = dev_get_by_index(net, ifindex);
2241
	if (!dev || dev->netdev_ops != &team_netdev_ops) {
2242 2243
		if (dev)
			dev_put(dev);
2244 2245 2246 2247
		return NULL;
	}

	team = netdev_priv(dev);
2248
	mutex_lock(&team->lock);
2249 2250 2251 2252 2253
	return team;
}

static void team_nl_team_put(struct team *team)
{
2254
	mutex_unlock(&team->lock);
2255
	dev_put(team->dev);
2256 2257
}

2258
typedef int team_nl_send_func_t(struct sk_buff *skb,
2259
				struct team *team, u32 portid);
2260

2261
static int team_nl_send_unicast(struct sk_buff *skb, struct team *team, u32 portid)
2262
{
2263
	return genlmsg_unicast(dev_net(team->dev), skb, portid);
2264 2265
}

2266 2267 2268 2269 2270
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;
2271
	struct team_option_inst_info *opt_inst_info = &opt_inst->info;
2272 2273 2274
	struct team_gsetter_ctx ctx;
	int err;

2275 2276 2277 2278 2279
	ctx.info = opt_inst_info;
	err = team_option_get(team, opt_inst, &ctx);
	if (err)
		return err;

2280 2281
	option_item = nla_nest_start(skb, TEAM_ATTR_ITEM_OPTION);
	if (!option_item)
2282
		return -EMSGSIZE;
2283

2284 2285
	if (nla_put_string(skb, TEAM_ATTR_OPTION_NAME, option->name))
		goto nest_cancel;
2286 2287 2288
	if (opt_inst_info->port &&
	    nla_put_u32(skb, TEAM_ATTR_OPTION_PORT_IFINDEX,
			opt_inst_info->port->dev->ifindex))
2289
		goto nest_cancel;
2290 2291 2292
	if (opt_inst->option->array_size &&
	    nla_put_u32(skb, TEAM_ATTR_OPTION_ARRAY_INDEX,
			opt_inst_info->array_index))
2293
		goto nest_cancel;
2294 2295 2296 2297

	switch (option->type) {
	case TEAM_OPTION_TYPE_U32:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_U32))
2298
			goto nest_cancel;
2299
		if (nla_put_u32(skb, TEAM_ATTR_OPTION_DATA, ctx.data.u32_val))
2300
			goto nest_cancel;
2301 2302 2303
		break;
	case TEAM_OPTION_TYPE_STRING:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_STRING))
2304
			goto nest_cancel;
2305 2306
		if (nla_put_string(skb, TEAM_ATTR_OPTION_DATA,
				   ctx.data.str_val))
2307
			goto nest_cancel;
2308 2309 2310
		break;
	case TEAM_OPTION_TYPE_BINARY:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_BINARY))
2311
			goto nest_cancel;
2312 2313
		if (nla_put(skb, TEAM_ATTR_OPTION_DATA, ctx.data.bin_val.len,
			    ctx.data.bin_val.ptr))
2314
			goto nest_cancel;
2315 2316 2317
		break;
	case TEAM_OPTION_TYPE_BOOL:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_FLAG))
2318
			goto nest_cancel;
2319 2320
		if (ctx.data.bool_val &&
		    nla_put_flag(skb, TEAM_ATTR_OPTION_DATA))
2321
			goto nest_cancel;
2322
		break;
2323 2324 2325 2326 2327 2328
	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;
2329 2330 2331
	default:
		BUG();
	}
2332 2333 2334 2335 2336 2337 2338
	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;
	}
2339 2340 2341
	nla_nest_end(skb, option_item);
	return 0;

2342 2343 2344 2345 2346 2347
nest_cancel:
	nla_nest_cancel(skb, option_item);
	return -EMSGSIZE;
}

static int __send_and_alloc_skb(struct sk_buff **pskb,
2348
				struct team *team, u32 portid,
2349 2350 2351 2352 2353
				team_nl_send_func_t *send_func)
{
	int err;

	if (*pskb) {
2354
		err = send_func(*pskb, team, portid);
2355 2356 2357
		if (err)
			return err;
	}
2358
	*pskb = genlmsg_new(GENLMSG_DEFAULT_SIZE, GFP_KERNEL);
2359 2360 2361
	if (!*pskb)
		return -ENOMEM;
	return 0;
2362 2363
}

2364
static int team_nl_send_options_get(struct team *team, u32 portid, u32 seq,
2365
				    int flags, team_nl_send_func_t *send_func,
2366
				    struct list_head *sel_opt_inst_list)
2367 2368
{
	struct nlattr *option_list;
2369
	struct nlmsghdr *nlh;
2370
	void *hdr;
2371 2372
	struct team_option_inst *opt_inst;
	int err;
2373 2374 2375
	struct sk_buff *skb = NULL;
	bool incomplete;
	int i;
2376

2377 2378 2379 2380
	opt_inst = list_first_entry(sel_opt_inst_list,
				    struct team_option_inst, tmp_list);

start_again:
2381
	err = __send_and_alloc_skb(&skb, team, portid, send_func);
2382 2383 2384
	if (err)
		return err;

2385
	hdr = genlmsg_put(skb, portid, seq, &team_nl_family, flags | NLM_F_MULTI,
2386
			  TEAM_CMD_OPTIONS_GET);
P
Pan Bian 已提交
2387 2388
	if (!hdr) {
		nlmsg_free(skb);
W
Wei Yongjun 已提交
2389
		return -EMSGSIZE;
P
Pan Bian 已提交
2390
	}
2391

D
David S. Miller 已提交
2392 2393
	if (nla_put_u32(skb, TEAM_ATTR_TEAM_IFINDEX, team->dev->ifindex))
		goto nla_put_failure;
2394 2395
	option_list = nla_nest_start(skb, TEAM_ATTR_LIST_OPTION);
	if (!option_list)
2396
		goto nla_put_failure;
2397

2398 2399 2400
	i = 0;
	incomplete = false;
	list_for_each_entry_from(opt_inst, sel_opt_inst_list, tmp_list) {
2401
		err = team_nl_fill_one_option_get(skb, team, opt_inst);
2402 2403 2404 2405 2406 2407 2408
		if (err) {
			if (err == -EMSGSIZE) {
				if (!i)
					goto errout;
				incomplete = true;
				break;
			}
2409
			goto errout;
2410 2411
		}
		i++;
2412 2413 2414
	}

	nla_nest_end(skb, option_list);
2415 2416 2417 2418 2419
	genlmsg_end(skb, hdr);
	if (incomplete)
		goto start_again;

send_done:
2420
	nlh = nlmsg_put(skb, portid, seq, NLMSG_DONE, 0, flags | NLM_F_MULTI);
2421
	if (!nlh) {
2422
		err = __send_and_alloc_skb(&skb, team, portid, send_func);
2423
		if (err)
2424
			return err;
2425 2426 2427
		goto send_done;
	}

2428
	return send_func(skb, team, portid);
2429 2430

nla_put_failure:
2431 2432
	err = -EMSGSIZE;
errout:
2433
	nlmsg_free(skb);
2434
	return err;
2435 2436 2437 2438 2439
}

static int team_nl_cmd_options_get(struct sk_buff *skb, struct genl_info *info)
{
	struct team *team;
2440
	struct team_option_inst *opt_inst;
2441
	int err;
2442
	LIST_HEAD(sel_opt_inst_list);
2443 2444 2445 2446 2447

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

2448 2449
	list_for_each_entry(opt_inst, &team->option_inst_list, list)
		list_add_tail(&opt_inst->tmp_list, &sel_opt_inst_list);
2450
	err = team_nl_send_options_get(team, info->snd_portid, info->snd_seq,
2451 2452
				       NLM_F_ACK, team_nl_send_unicast,
				       &sel_opt_inst_list);
2453 2454 2455 2456 2457 2458

	team_nl_team_put(team);

	return err;
}

2459 2460 2461
static int team_nl_send_event_options_get(struct team *team,
					  struct list_head *sel_opt_inst_list);

2462 2463 2464 2465 2466 2467
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;
2468
	LIST_HEAD(opt_inst_list);
2469

J
Jiri Pirko 已提交
2470 2471
	rtnl_lock();

2472
	team = team_nl_team_get(info);
J
Jiri Pirko 已提交
2473 2474 2475 2476
	if (!team) {
		err = -EINVAL;
		goto rtnl_unlock;
	}
2477 2478 2479 2480 2481 2482 2483 2484

	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) {
2485
		struct nlattr *opt_attrs[TEAM_ATTR_OPTION_MAX + 1];
J
Jiri Pirko 已提交
2486
		struct nlattr *attr;
J
Jiri Pirko 已提交
2487
		struct nlattr *attr_data;
2488
		enum team_option_type opt_type;
2489
		int opt_port_ifindex = 0; /* != 0 for per-port options */
J
Jiri Pirko 已提交
2490 2491
		u32 opt_array_index = 0;
		bool opt_is_array = false;
2492
		struct team_option_inst *opt_inst;
2493 2494 2495 2496 2497 2498 2499
		char *opt_name;
		bool opt_found = false;

		if (nla_type(nl_option) != TEAM_ATTR_ITEM_OPTION) {
			err = -EINVAL;
			goto team_put;
		}
2500
		err = nla_parse_nested(opt_attrs, TEAM_ATTR_OPTION_MAX,
2501 2502
				       nl_option, team_nl_option_policy,
				       info->extack);
2503 2504
		if (err)
			goto team_put;
2505
		if (!opt_attrs[TEAM_ATTR_OPTION_NAME] ||
J
Jiri Pirko 已提交
2506
		    !opt_attrs[TEAM_ATTR_OPTION_TYPE]) {
2507 2508 2509
			err = -EINVAL;
			goto team_put;
		}
2510
		switch (nla_get_u8(opt_attrs[TEAM_ATTR_OPTION_TYPE])) {
2511 2512 2513 2514 2515 2516
		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
2517 2518 2519
		case NLA_BINARY:
			opt_type = TEAM_OPTION_TYPE_BINARY;
			break;
J
Jiri Pirko 已提交
2520 2521 2522
		case NLA_FLAG:
			opt_type = TEAM_OPTION_TYPE_BOOL;
			break;
2523 2524 2525
		case NLA_S32:
			opt_type = TEAM_OPTION_TYPE_S32;
			break;
2526 2527 2528 2529
		default:
			goto team_put;
		}

J
Jiri Pirko 已提交
2530 2531 2532 2533 2534 2535
		attr_data = opt_attrs[TEAM_ATTR_OPTION_DATA];
		if (opt_type != TEAM_OPTION_TYPE_BOOL && !attr_data) {
			err = -EINVAL;
			goto team_put;
		}

2536
		opt_name = nla_data(opt_attrs[TEAM_ATTR_OPTION_NAME]);
J
Jiri Pirko 已提交
2537 2538 2539 2540 2541 2542 2543 2544 2545
		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);
		}
2546 2547 2548 2549

		list_for_each_entry(opt_inst, &team->option_inst_list, list) {
			struct team_option *option = opt_inst->option;
			struct team_gsetter_ctx ctx;
2550
			struct team_option_inst_info *opt_inst_info;
2551
			int tmp_ifindex;
2552

2553 2554 2555
			opt_inst_info = &opt_inst->info;
			tmp_ifindex = opt_inst_info->port ?
				      opt_inst_info->port->dev->ifindex : 0;
2556
			if (option->type != opt_type ||
2557
			    strcmp(option->name, opt_name) ||
J
Jiri Pirko 已提交
2558 2559
			    tmp_ifindex != opt_port_ifindex ||
			    (option->array_size && !opt_is_array) ||
2560
			    opt_inst_info->array_index != opt_array_index)
2561 2562
				continue;
			opt_found = true;
2563
			ctx.info = opt_inst_info;
2564 2565
			switch (opt_type) {
			case TEAM_OPTION_TYPE_U32:
J
Jiri Pirko 已提交
2566
				ctx.data.u32_val = nla_get_u32(attr_data);
2567 2568
				break;
			case TEAM_OPTION_TYPE_STRING:
J
Jiri Pirko 已提交
2569
				if (nla_len(attr_data) > TEAM_STRING_MAX_LEN) {
J
Jiri Pirko 已提交
2570 2571 2572
					err = -EINVAL;
					goto team_put;
				}
J
Jiri Pirko 已提交
2573
				ctx.data.str_val = nla_data(attr_data);
2574
				break;
J
Jiri Pirko 已提交
2575
			case TEAM_OPTION_TYPE_BINARY:
J
Jiri Pirko 已提交
2576 2577 2578 2579 2580
				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
2581
				break;
2582 2583 2584
			case TEAM_OPTION_TYPE_S32:
				ctx.data.s32_val = nla_get_s32(attr_data);
				break;
2585 2586 2587
			default:
				BUG();
			}
2588
			err = team_option_set(team, opt_inst, &ctx);
2589 2590
			if (err)
				goto team_put;
2591
			opt_inst->changed = true;
2592 2593 2594 2595 2596 2597 2598 2599

			/* dumb/evil user-space can send us duplicate opt,
			 * keep only the last one
			 */
			if (__team_option_inst_tmp_find(&opt_inst_list,
							opt_inst))
				continue;

2600
			list_add(&opt_inst->tmp_list, &opt_inst_list);
2601 2602 2603 2604 2605 2606 2607
		}
		if (!opt_found) {
			err = -ENOENT;
			goto team_put;
		}
	}

2608 2609
	err = team_nl_send_event_options_get(team, &opt_inst_list);

2610 2611
team_put:
	team_nl_team_put(team);
J
Jiri Pirko 已提交
2612 2613
rtnl_unlock:
	rtnl_unlock();
2614 2615 2616
	return err;
}

2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649
static int team_nl_fill_one_port_get(struct sk_buff *skb,
				     struct team_port *port)
{
	struct nlattr *port_item;

	port_item = nla_nest_start(skb, TEAM_ATTR_ITEM_PORT);
	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)
2650 2651
{
	struct nlattr *port_list;
2652
	struct nlmsghdr *nlh;
2653 2654
	void *hdr;
	struct team_port *port;
2655 2656 2657 2658
	int err;
	struct sk_buff *skb = NULL;
	bool incomplete;
	int i;
2659

2660 2661
	port = list_first_entry_or_null(&team->port_list,
					struct team_port, list);
2662 2663 2664 2665 2666 2667 2668

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,
2669
			  TEAM_CMD_PORT_LIST_GET);
P
Pan Bian 已提交
2670 2671
	if (!hdr) {
		nlmsg_free(skb);
W
Wei Yongjun 已提交
2672
		return -EMSGSIZE;
P
Pan Bian 已提交
2673
	}
2674

D
David S. Miller 已提交
2675 2676
	if (nla_put_u32(skb, TEAM_ATTR_TEAM_IFINDEX, team->dev->ifindex))
		goto nla_put_failure;
2677 2678
	port_list = nla_nest_start(skb, TEAM_ATTR_LIST_PORT);
	if (!port_list)
2679
		goto nla_put_failure;
2680

2681 2682
	i = 0;
	incomplete = false;
2683

2684 2685 2686 2687 2688 2689 2690
	/* 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;
2691 2692
	} else if (port) {
		list_for_each_entry_from(port, &team->port_list, list) {
2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703
			err = team_nl_fill_one_port_get(skb, port);
			if (err) {
				if (err == -EMSGSIZE) {
					if (!i)
						goto errout;
					incomplete = true;
					break;
				}
				goto errout;
			}
			i++;
2704
		}
2705 2706 2707
	}

	nla_nest_end(skb, port_list);
2708 2709 2710 2711 2712 2713 2714 2715 2716
	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)
2717
			return err;
2718 2719 2720 2721
		goto send_done;
	}

	return send_func(skb, team, portid);
2722 2723

nla_put_failure:
2724 2725 2726 2727
	err = -EMSGSIZE;
errout:
	nlmsg_free(skb);
	return err;
2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739
}

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;

2740 2741
	err = team_nl_send_port_list_get(team, info->snd_portid, info->snd_seq,
					 NLM_F_ACK, team_nl_send_unicast, NULL);
2742 2743 2744 2745 2746 2747

	team_nl_team_put(team);

	return err;
}

2748
static const struct genl_ops team_nl_ops[] = {
2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773
	{
		.cmd = TEAM_CMD_NOOP,
		.doit = team_nl_cmd_noop,
		.policy = team_nl_policy,
	},
	{
		.cmd = TEAM_CMD_OPTIONS_SET,
		.doit = team_nl_cmd_options_set,
		.policy = team_nl_policy,
		.flags = GENL_ADMIN_PERM,
	},
	{
		.cmd = TEAM_CMD_OPTIONS_GET,
		.doit = team_nl_cmd_options_get,
		.policy = team_nl_policy,
		.flags = GENL_ADMIN_PERM,
	},
	{
		.cmd = TEAM_CMD_PORT_LIST_GET,
		.doit = team_nl_cmd_port_list_get,
		.policy = team_nl_policy,
		.flags = GENL_ADMIN_PERM,
	},
};

2774 2775
static const struct genl_multicast_group team_nl_mcgrps[] = {
	{ .name = TEAM_GENL_CHANGE_EVENT_MC_GRP_NAME, },
2776 2777
};

2778
static struct genl_family team_nl_family __ro_after_init = {
2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789
	.name		= TEAM_GENL_NAME,
	.version	= TEAM_GENL_VERSION,
	.maxattr	= TEAM_ATTR_MAX,
	.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),
};

2790
static int team_nl_send_multicast(struct sk_buff *skb,
2791
				  struct team *team, u32 portid)
2792
{
2793
	return genlmsg_multicast_netns(&team_nl_family, dev_net(team->dev),
2794
				       skb, 0, 0, GFP_KERNEL);
2795 2796
}

2797 2798
static int team_nl_send_event_options_get(struct team *team,
					  struct list_head *sel_opt_inst_list)
2799
{
2800 2801
	return team_nl_send_options_get(team, 0, 0, 0, team_nl_send_multicast,
					sel_opt_inst_list);
2802 2803
}

2804 2805
static int team_nl_send_event_port_get(struct team *team,
				       struct team_port *port)
2806
{
2807 2808
	return team_nl_send_port_list_get(team, 0, 0, 0, team_nl_send_multicast,
					  port);
2809 2810
}

2811
static int __init team_nl_init(void)
2812
{
2813
	return genl_register_family(&team_nl_family);
2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825
}

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


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

2826
static void __team_options_change_check(struct team *team)
2827 2828
{
	int err;
2829 2830
	struct team_option_inst *opt_inst;
	LIST_HEAD(sel_opt_inst_list);
2831

2832 2833 2834 2835 2836
	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);
2837
	if (err && err != -ESRCH)
2838 2839
		netdev_warn(team->dev, "Failed to send options change via netlink (err %d)\n",
			    err);
2840 2841 2842
}

/* rtnl lock is held */
J
Jiri Pirko 已提交
2843 2844

static void __team_port_change_send(struct team_port *port, bool linkup)
2845 2846 2847
{
	int err;

2848
	port->changed = true;
2849 2850
	port->state.linkup = linkup;
	team_refresh_port_linkup(port);
2851
	if (linkup) {
2852
		struct ethtool_link_ksettings ecmd;
2853

2854
		err = __ethtool_get_link_ksettings(port->dev, &ecmd);
2855
		if (!err) {
2856 2857
			port->state.speed = ecmd.base.speed;
			port->state.duplex = ecmd.base.duplex;
2858 2859 2860
			goto send_event;
		}
	}
2861 2862
	port->state.speed = 0;
	port->state.duplex = 0;
2863 2864

send_event:
2865
	err = team_nl_send_event_port_get(port->team, port);
2866 2867 2868
	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);
2869 2870 2871

}

2872 2873 2874 2875 2876
static void __team_carrier_check(struct team *team)
{
	struct team_port *port;
	bool team_linkup;

2877 2878 2879
	if (team->user_carrier_enabled)
		return;

2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893
	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
2894 2895 2896 2897
static void __team_port_change_check(struct team_port *port, bool linkup)
{
	if (port->state.linkup != linkup)
		__team_port_change_send(port, linkup);
2898
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2899 2900 2901 2902 2903
}

static void __team_port_change_port_added(struct team_port *port, bool linkup)
{
	__team_port_change_send(port, linkup);
2904
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2905 2906 2907 2908 2909 2910
}

static void __team_port_change_port_removed(struct team_port *port)
{
	port->removed = true;
	__team_port_change_send(port, false);
2911
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2912 2913
}

2914 2915 2916 2917
static void team_port_change_check(struct team_port *port, bool linkup)
{
	struct team *team = port->team;

2918
	mutex_lock(&team->lock);
2919
	__team_port_change_check(port, linkup);
2920
	mutex_unlock(&team->lock);
2921 2922
}

J
Jiri Pirko 已提交
2923

2924 2925 2926 2927 2928 2929 2930
/************************************
 * Net device notifier event handler
 ************************************/

static int team_device_event(struct notifier_block *unused,
			     unsigned long event, void *ptr)
{
2931
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
2932 2933 2934 2935 2936 2937 2938 2939 2940 2941
	struct team_port *port;

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

	switch (event) {
	case NETDEV_UP:
		if (netif_carrier_ok(dev))
			team_port_change_check(port, true);
2942
		break;
2943 2944
	case NETDEV_DOWN:
		team_port_change_check(port, false);
2945
		break;
2946 2947 2948
	case NETDEV_CHANGE:
		if (netif_running(port->dev))
			team_port_change_check(port,
G
George Wilkie 已提交
2949
					       !!netif_oper_up(port->dev));
2950 2951 2952 2953 2954 2955 2956
		break;
	case NETDEV_UNREGISTER:
		team_del_slave(port->team->dev, dev);
		break;
	case NETDEV_FEAT_CHANGE:
		team_compute_features(port->team);
		break;
2957
	case NETDEV_PRECHANGEMTU:
2958
		/* Forbid to change mtu of underlaying device */
J
Jiri Pirko 已提交
2959 2960 2961
		if (!port->team->port_mtu_change_allowed)
			return NOTIFY_BAD;
		break;
2962 2963 2964
	case NETDEV_PRE_TYPE_CHANGE:
		/* Forbid to change type of underlaying device */
		return NOTIFY_BAD;
2965 2966 2967 2968
	case NETDEV_RESEND_IGMP:
		/* Propagate to master device */
		call_netdevice_notifiers(event, port->team->dev);
		break;
2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020
	}
	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);