team.c 71.3 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>
F
Flavio Leitner 已提交
31
#include <generated/utsrelease.h>
32 33 34 35 36 37 38 39 40 41 42 43 44
#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
45
	return netif_is_team_port(dev) ? port : NULL;
46 47 48
}

/*
49
 * Since the ability to change device address for open port device is tested in
50 51
 * team_port_add, this function can be called without control of return value
 */
52 53
static int __set_port_dev_addr(struct net_device *port_dev,
			       const unsigned char *dev_addr)
54
{
55
	struct sockaddr_storage addr;
56

57 58
	memcpy(addr.__data, dev_addr, port_dev->addr_len);
	addr.ss_family = port_dev->type;
59
	return dev_set_mac_address(port_dev, (struct sockaddr *)&addr, NULL);
60 61
}

62
static int team_port_set_orig_dev_addr(struct team_port *port)
63
{
64
	return __set_port_dev_addr(port->dev, port->orig.dev_addr);
65 66
}

67 68
static int team_port_set_team_dev_addr(struct team *team,
				       struct team_port *port)
69
{
70
	return __set_port_dev_addr(port->dev, team->dev->dev_addr);
71
}
72 73 74 75 76 77 78 79 80 81 82 83 84

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);
85

86 87 88 89 90 91 92 93 94
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);
}

95 96
static void team_refresh_port_linkup(struct team_port *port)
{
97 98 99 100 101 102 103
	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);
	}
104
}
105

J
Jiri Pirko 已提交
106

107 108 109 110
/*******************
 * Options handling
 *******************/

111 112
struct team_option_inst { /* One for each option instance */
	struct list_head list;
113
	struct list_head tmp_list;
114
	struct team_option *option;
115
	struct team_option_inst_info info;
116 117 118 119 120 121
	bool changed;
	bool removed;
};

static struct team_option *__team_find_option(struct team *team,
					      const char *opt_name)
J
Jiri Pirko 已提交
122 123 124 125 126 127 128 129 130 131
{
	struct team_option *option;

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

132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
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
149 150 151 152 153 154
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;
155
	int err;
J
Jiri Pirko 已提交
156 157 158 159 160 161 162 163 164 165

	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;
166 167
		opt_inst->info.port = port;
		opt_inst->info.array_index = i;
J
Jiri Pirko 已提交
168 169 170
		opt_inst->changed = true;
		opt_inst->removed = false;
		list_add_tail(&opt_inst->list, &team->option_inst_list);
171 172 173 174 175 176
		if (option->init) {
			err = option->init(team, &opt_inst->info);
			if (err)
				return err;
		}

J
Jiri Pirko 已提交
177 178 179 180
	}
	return 0;
}

181 182 183 184 185
static int __team_option_inst_add_option(struct team *team,
					 struct team_option *option)
{
	int err;

J
Jiri Pirko 已提交
186
	if (!option->per_port) {
187
		err = __team_option_inst_add(team, option, NULL);
J
Jiri Pirko 已提交
188 189 190
		if (err)
			goto inst_del_option;
	}
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217
	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 &&
218
		    opt_inst->info.port == port)
219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
			__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) {
249
		if (opt_inst->info.port == port) {
250 251 252 253 254 255 256 257 258
			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)
259 260
{
	int i;
261
	struct team_option **dst_opts;
J
Jiri Pirko 已提交
262
	int err;
263

K
Kees Cook 已提交
264
	dst_opts = kcalloc(option_count, sizeof(struct team_option *),
265 266 267
			   GFP_KERNEL);
	if (!dst_opts)
		return -ENOMEM;
J
Jiri Pirko 已提交
268 269 270
	for (i = 0; i < option_count; i++, option++) {
		if (__team_find_option(team, option->name)) {
			err = -EEXIST;
271
			goto alloc_rollback;
J
Jiri Pirko 已提交
272
		}
273 274
		dst_opts[i] = kmemdup(option, sizeof(*option), GFP_KERNEL);
		if (!dst_opts[i]) {
J
Jiri Pirko 已提交
275
			err = -ENOMEM;
276
			goto alloc_rollback;
J
Jiri Pirko 已提交
277 278 279
		}
	}

280
	for (i = 0; i < option_count; i++) {
281 282 283
		err = __team_option_inst_add_option(team, dst_opts[i]);
		if (err)
			goto inst_rollback;
J
Jiri Pirko 已提交
284
		list_add_tail(&dst_opts[i]->list, &team->option_list);
285
	}
J
Jiri Pirko 已提交
286

287
	kfree(dst_opts);
J
Jiri Pirko 已提交
288 289
	return 0;

290 291 292 293 294 295 296
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
297 298
		kfree(dst_opts[i]);

299
	kfree(dst_opts);
J
Jiri Pirko 已提交
300
	return err;
301
}
J
Jiri Pirko 已提交
302

303 304 305 306 307 308 309 310
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;
311

312
		del_opt = __team_find_option(team, option->name);
313 314
		if (del_opt)
			__team_option_inst_mark_removed_option(team, del_opt);
315 316
	}
}
317 318

static void __team_options_unregister(struct team *team,
J
Jiri Pirko 已提交
319
				      const struct team_option *option,
320 321 322 323
				      size_t option_count)
{
	int i;

J
Jiri Pirko 已提交
324 325 326 327 328
	for (i = 0; i < option_count; i++, option++) {
		struct team_option *del_opt;

		del_opt = __team_find_option(team, option->name);
		if (del_opt) {
329
			__team_option_inst_del_option(team, del_opt);
J
Jiri Pirko 已提交
330 331 332 333
			list_del(&del_opt->list);
			kfree(del_opt);
		}
	}
334 335
}

336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
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
352 353
void team_options_unregister(struct team *team,
			     const struct team_option *option,
354 355
			     size_t option_count)
{
356 357
	__team_options_mark_removed(team, option, option_count);
	__team_options_change_check(team);
358 359 360 361
	__team_options_unregister(team, option, option_count);
}
EXPORT_SYMBOL(team_options_unregister);

362 363 364 365
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
366 367
	if (!opt_inst->option->getter)
		return -EOPNOTSUPP;
368 369 370 371 372 373
	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)
374
{
J
Jiri Pirko 已提交
375 376
	if (!opt_inst->option->setter)
		return -EOPNOTSUPP;
377
	return opt_inst->option->setter(team, ctx);
378 379
}

J
Jiri Pirko 已提交
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395
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);


396 397 398 399 400 401 402
/****************
 * Mode handling
 ****************/

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

J
Jiri Pirko 已提交
403 404 405 406 407 408
struct team_mode_item {
	struct list_head list;
	const struct team_mode *mode;
};

static struct team_mode_item *__find_mode(const char *kind)
409
{
J
Jiri Pirko 已提交
410
	struct team_mode_item *mitem;
411

J
Jiri Pirko 已提交
412 413 414
	list_for_each_entry(mitem, &mode_list, list) {
		if (strcmp(mitem->mode->kind, kind) == 0)
			return mitem;
415 416 417 418 419 420 421 422 423 424 425 426 427 428
	}
	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
429
int team_mode_register(const struct team_mode *mode)
430 431
{
	int err = 0;
J
Jiri Pirko 已提交
432
	struct team_mode_item *mitem;
433 434 435 436

	if (!is_good_mode_name(mode->kind) ||
	    mode->priv_size > TEAM_MODE_PRIV_SIZE)
		return -EINVAL;
J
Jiri Pirko 已提交
437 438 439 440 441

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

442 443 444
	spin_lock(&mode_list_lock);
	if (__find_mode(mode->kind)) {
		err = -EEXIST;
J
Jiri Pirko 已提交
445
		kfree(mitem);
446 447
		goto unlock;
	}
J
Jiri Pirko 已提交
448 449
	mitem->mode = mode;
	list_add_tail(&mitem->list, &mode_list);
450 451 452 453 454 455
unlock:
	spin_unlock(&mode_list_lock);
	return err;
}
EXPORT_SYMBOL(team_mode_register);

J
Jiri Pirko 已提交
456
void team_mode_unregister(const struct team_mode *mode)
457
{
J
Jiri Pirko 已提交
458 459
	struct team_mode_item *mitem;

460
	spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
461 462 463 464 465
	mitem = __find_mode(mode->kind);
	if (mitem) {
		list_del_init(&mitem->list);
		kfree(mitem);
	}
466 467 468 469
	spin_unlock(&mode_list_lock);
}
EXPORT_SYMBOL(team_mode_unregister);

J
Jiri Pirko 已提交
470
static const struct team_mode *team_mode_get(const char *kind)
471
{
J
Jiri Pirko 已提交
472 473
	struct team_mode_item *mitem;
	const struct team_mode *mode = NULL;
474 475

	spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
476 477
	mitem = __find_mode(kind);
	if (!mitem) {
478 479 480
		spin_unlock(&mode_list_lock);
		request_module("team-mode-%s", kind);
		spin_lock(&mode_list_lock);
J
Jiri Pirko 已提交
481
		mitem = __find_mode(kind);
482
	}
J
Jiri Pirko 已提交
483 484
	if (mitem) {
		mode = mitem->mode;
485 486
		if (!try_module_get(mode->owner))
			mode = NULL;
J
Jiri Pirko 已提交
487
	}
488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503

	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;
}

504 505 506
static rx_handler_result_t team_dummy_receive(struct team *team,
					      struct team_port *port,
					      struct sk_buff *skb)
507 508 509 510
{
	return RX_HANDLER_ANOTHER;
}

511 512 513 514 515 516 517 518 519 520 521
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)
{
522
	team->user_carrier_enabled = false;
523 524 525
	team->mode = &__team_no_mode;
}

526
static void team_adjust_ops(struct team *team)
527 528 529 530 531 532
{
	/*
	 * To avoid checks in rx/tx skb paths, ensure here that non-null and
	 * correct ops are always set.
	 */

533
	if (!team->en_port_count || !team_is_mode_set(team) ||
534
	    !team->mode->ops->transmit)
535 536 537 538
		team->ops.transmit = team_dummy_transmit;
	else
		team->ops.transmit = team->mode->ops->transmit;

539
	if (!team->en_port_count || !team_is_mode_set(team) ||
540
	    !team->mode->ops->receive)
541 542 543 544 545 546 547 548 549 550 551 552 553 554
		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 */
555
	if (team_is_mode_set(team)) {
556 557 558 559 560 561 562 563 564
		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);
565
		team_set_no_mode(team);
566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590
		/* 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
591
	const struct team_mode *new_mode;
592 593 594 595 596 597 598 599
	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;
	}

600
	if (team_is_mode_set(team) && strcmp(team->mode->kind, kind) == 0) {
601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622
		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
623 624 625 626 627 628 629
/*********************
 * Peers notification
 *********************/

static void team_notify_peers_work(struct work_struct *work)
{
	struct team *team;
630
	int val;
J
Jiri Pirko 已提交
631 632 633 634 635 636 637

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

	if (!rtnl_trylock()) {
		schedule_delayed_work(&team->notify_peers.dw, 0);
		return;
	}
638 639 640 641 642
	val = atomic_dec_if_positive(&team->notify_peers.count_pending);
	if (val < 0) {
		rtnl_unlock();
		return;
	}
J
Jiri Pirko 已提交
643 644
	call_netdevice_notifiers(NETDEV_NOTIFY_PEERS, team->dev);
	rtnl_unlock();
645
	if (val)
J
Jiri Pirko 已提交
646 647 648 649 650 651 652 653
		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;
654
	atomic_add(team->notify_peers.count, &team->notify_peers.count_pending);
J
Jiri Pirko 已提交
655 656 657 658 659 660 661 662 663 664 665 666 667 668
	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);
}


669 670 671 672 673 674 675
/*******************************
 * Send multicast group rejoins
 *******************************/

static void team_mcast_rejoin_work(struct work_struct *work)
{
	struct team *team;
676
	int val;
677 678 679 680 681 682 683

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

	if (!rtnl_trylock()) {
		schedule_delayed_work(&team->mcast_rejoin.dw, 0);
		return;
	}
684 685 686 687 688
	val = atomic_dec_if_positive(&team->mcast_rejoin.count_pending);
	if (val < 0) {
		rtnl_unlock();
		return;
	}
689 690
	call_netdevice_notifiers(NETDEV_RESEND_IGMP, team->dev);
	rtnl_unlock();
691
	if (val)
692 693 694 695 696 697 698 699
		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;
700
	atomic_add(team->mcast_rejoin.count, &team->mcast_rejoin.count_pending);
701 702 703 704 705 706 707 708 709 710 711 712 713 714
	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);
}


715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734
/************************
 * 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
735 736 737 738 739 740
	if (!team_port_enabled(port)) {
		/* allow exact match delivery for disabled ports */
		res = RX_HANDLER_EXACT;
	} else {
		res = team->ops.receive(team, port, skb);
	}
741 742 743 744 745 746 747 748 749 750 751 752
	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;
753 754
	} else if (res == RX_HANDLER_EXACT) {
		this_cpu_inc(team->pcpu_stats->rx_nohandler);
755 756 757 758 759 760 761 762
	} else {
		this_cpu_inc(team->pcpu_stats->rx_dropped);
	}

	return res;
}


763 764 765 766 767 768 769 770 771 772 773 774
/*************************************
 * 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;
775 776
	listarr = kmalloc_array(queue_cnt, sizeof(struct list_head),
				GFP_KERNEL);
777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815
	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)
{
816 817
	if (!port->queue_id)
		return;
818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839
	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;

840
	if (!port->queue_id)
841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857
		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) {
858
		if (port->queue_id) {
859 860 861 862 863 864 865 866 867 868 869
			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;
}

870 871
static void team_queue_override_port_prio_changed(struct team *team,
						  struct team_port *port)
872
{
873 874
	if (!port->queue_id || team_port_enabled(port))
		return;
875 876 877 878 879
	__team_queue_override_port_del(team, port);
	__team_queue_override_port_add(team, port);
	__team_queue_override_enabled_check(team);
}

880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907
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);
}

908

909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924
/****************
 * 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
925 926 927 928
 * 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.
929
 */
J
Jiri Pirko 已提交
930 931
static void team_port_enable(struct team *team,
			     struct team_port *port)
932
{
J
Jiri Pirko 已提交
933 934 935
	if (team_port_enabled(port))
		return;
	port->index = team->en_port_count++;
936 937
	hlist_add_head_rcu(&port->hlist,
			   team_port_index_hash(team, port->index));
938
	team_adjust_ops(team);
939
	team_queue_override_port_add(team, port);
940 941
	if (team->ops.port_enabled)
		team->ops.port_enabled(team, port);
J
Jiri Pirko 已提交
942
	team_notify_peers(team);
943
	team_mcast_rejoin(team);
944
	team_lower_state_changed(port);
945 946 947 948 949 950 951
}

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

J
Jiri Pirko 已提交
952
	for (i = rm_index + 1; i < team->en_port_count; i++) {
953 954 955 956 957 958 959 960
		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
961 962
static void team_port_disable(struct team *team,
			      struct team_port *port)
963
{
J
Jiri Pirko 已提交
964 965
	if (!team_port_enabled(port))
		return;
966 967
	if (team->ops.port_disabled)
		team->ops.port_disabled(team, port);
968
	hlist_del_rcu(&port->hlist);
969
	__reconstruct_port_hlist(team, port->index);
J
Jiri Pirko 已提交
970
	port->index = -1;
971
	team->en_port_count--;
972 973
	team_queue_override_port_del(team, port);
	team_adjust_ops(team);
974
	team_lower_state_changed(port);
975 976
}

977
#define TEAM_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
978 979 980
			    NETIF_F_FRAGLIST | NETIF_F_ALL_TSO | \
			    NETIF_F_HIGHDMA | NETIF_F_LRO)

981 982 983
#define TEAM_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
				 NETIF_F_RXCSUM | NETIF_F_ALL_TSO)

984
static void __team_compute_features(struct team *team)
985 986
{
	struct team_port *port;
987 988
	netdev_features_t vlan_features = TEAM_VLAN_FEATURES &
					  NETIF_F_ALL_FOR_ALL;
989
	netdev_features_t enc_features  = TEAM_ENC_FEATURES;
990
	unsigned short max_hard_header_len = ETH_HLEN;
991 992
	unsigned int dst_release_flag = IFF_XMIT_DST_RELEASE |
					IFF_XMIT_DST_RELEASE_PERM;
993 994 995 996 997

	list_for_each_entry(port, &team->port_list, list) {
		vlan_features = netdev_increment_features(vlan_features,
					port->dev->vlan_features,
					TEAM_VLAN_FEATURES);
998 999 1000 1001 1002
		enc_features =
			netdev_increment_features(enc_features,
						  port->dev->hw_enc_features,
						  TEAM_ENC_FEATURES);

1003

1004
		dst_release_flag &= port->dev->priv_flags;
1005 1006 1007 1008 1009
		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;
1010 1011
	team->dev->hw_enc_features = enc_features | NETIF_F_GSO_ENCAP_ALL |
				     NETIF_F_GSO_UDP_L4;
1012 1013
	team->dev->hard_header_len = max_hard_header_len;

1014 1015 1016
	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;
1017
}
1018

1019 1020
static void team_compute_features(struct team *team)
{
1021
	mutex_lock(&team->lock);
1022
	__team_compute_features(team);
1023
	mutex_unlock(&team->lock);
1024
	netdev_change_features(team->dev);
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 1054 1055
}

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
1056
#ifdef CONFIG_NET_POLL_CONTROLLER
X
Xin Long 已提交
1057
static int __team_port_enable_netpoll(struct team_port *port)
J
Jiri Pirko 已提交
1058 1059 1060 1061
{
	struct netpoll *np;
	int err;

1062
	np = kzalloc(sizeof(*np), GFP_KERNEL);
J
Jiri Pirko 已提交
1063 1064 1065
	if (!np)
		return -ENOMEM;

1066
	err = __netpoll_setup(np, port->dev);
J
Jiri Pirko 已提交
1067 1068 1069 1070 1071 1072 1073 1074
	if (err) {
		kfree(np);
		return err;
	}
	port->np = np;
	return err;
}

X
Xin Long 已提交
1075 1076 1077 1078 1079 1080 1081 1082
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
1083 1084 1085 1086 1087 1088 1089 1090
static void team_port_disable_netpoll(struct team_port *port)
{
	struct netpoll *np = port->np;

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

1091
	__netpoll_free(np);
J
Jiri Pirko 已提交
1092 1093
}
#else
X
Xin Long 已提交
1094
static int team_port_enable_netpoll(struct team_port *port)
J
Jiri Pirko 已提交
1095 1096 1097 1098 1099 1100 1101 1102
{
	return 0;
}
static void team_port_disable_netpoll(struct team_port *port)
{
}
#endif

1103 1104
static int team_upper_dev_link(struct team *team, struct team_port *port,
			       struct netlink_ext_ack *extack)
1105
{
1106
	struct netdev_lag_upper_info lag_upper_info;
1107 1108
	int err;

1109
	lag_upper_info.tx_type = team->mode->lag_tx_type;
1110
	lag_upper_info.hash_type = NETDEV_LAG_HASH_UNKNOWN;
1111
	err = netdev_master_upper_dev_link(port->dev, team->dev, NULL,
1112
					   &lag_upper_info, extack);
1113 1114
	if (err)
		return err;
1115
	port->dev->priv_flags |= IFF_TEAM_PORT;
1116 1117 1118
	return 0;
}

1119
static void team_upper_dev_unlink(struct team *team, struct team_port *port)
1120
{
1121 1122
	netdev_upper_dev_unlink(port->dev, team->dev);
	port->dev->priv_flags &= ~IFF_TEAM_PORT;
1123 1124
}

J
Jiri Pirko 已提交
1125
static void __team_port_change_port_added(struct team_port *port, bool linkup);
1126 1127
static int team_dev_type_check_change(struct net_device *dev,
				      struct net_device *port_dev);
1128

1129 1130
static int team_port_add(struct team *team, struct net_device *port_dev,
			 struct netlink_ext_ack *extack)
1131 1132 1133 1134 1135 1136
{
	struct net_device *dev = team->dev;
	struct team_port *port;
	char *portname = port_dev->name;
	int err;

1137
	if (port_dev->flags & IFF_LOOPBACK) {
1138
		NL_SET_ERR_MSG(extack, "Loopback device can't be added as a team port");
1139
		netdev_err(dev, "Device %s is loopback device. Loopback devices can't be added as a team port\n",
1140 1141 1142 1143
			   portname);
		return -EINVAL;
	}

J
Julian Wiedmann 已提交
1144
	if (netif_is_team_port(port_dev)) {
1145
		NL_SET_ERR_MSG(extack, "Device is already a port of a team device");
1146 1147 1148 1149 1150
		netdev_err(dev, "Device %s is already a port "
				"of a team device\n", portname);
		return -EBUSY;
	}

1151 1152 1153 1154 1155 1156
	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;
	}

1157 1158 1159 1160 1161 1162 1163
	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;
	}

1164 1165
	if (port_dev->features & NETIF_F_VLAN_CHALLENGED &&
	    vlan_uses_dev(dev)) {
1166
		NL_SET_ERR_MSG(extack, "Device is VLAN challenged and team device has VLAN set up");
1167 1168 1169 1170 1171
		netdev_err(dev, "Device %s is VLAN challenged and team device has VLAN set up\n",
			   portname);
		return -EPERM;
	}

1172 1173 1174 1175
	err = team_dev_type_check_change(dev, port_dev);
	if (err)
		return err;

1176
	if (port_dev->flags & IFF_UP) {
1177
		NL_SET_ERR_MSG(extack, "Device is up. Set it down before adding it as a team port");
1178 1179 1180 1181 1182
		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
1183 1184
	port = kzalloc(sizeof(struct team_port) + team->mode->port_priv_size,
		       GFP_KERNEL);
1185 1186 1187 1188 1189
	if (!port)
		return -ENOMEM;

	port->dev = port_dev;
	port->team = team;
1190
	INIT_LIST_HEAD(&port->qom_list);
1191 1192 1193 1194 1195 1196 1197 1198

	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;
	}

1199
	memcpy(port->orig.dev_addr, port_dev->dev_addr, port_dev->addr_len);
1200 1201 1202 1203 1204 1205 1206 1207

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

1208
	err = dev_open(port_dev, extack);
1209 1210 1211 1212 1213 1214
	if (err) {
		netdev_dbg(dev, "Device %s opening failed\n",
			   portname);
		goto err_dev_open;
	}

J
Jiri Pirko 已提交
1215 1216 1217 1218 1219 1220 1221
	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
1222
	err = team_port_enable_netpoll(port);
S
stephen hemminger 已提交
1223 1224 1225 1226
	if (err) {
		netdev_err(dev, "Failed to enable netpoll on device %s\n",
			   portname);
		goto err_enable_netpoll;
J
Jiri Pirko 已提交
1227 1228
	}

1229 1230 1231
	if (!(dev->features & NETIF_F_LRO))
		dev_disable_lro(port_dev);

1232 1233 1234 1235 1236 1237 1238 1239
	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;
	}

1240
	err = team_upper_dev_link(team, port, extack);
1241 1242 1243 1244 1245 1246
	if (err) {
		netdev_err(dev, "Device %s failed to set upper link\n",
			   portname);
		goto err_set_upper_link;
	}

1247
	err = __team_option_inst_add_port(team, port);
1248 1249 1250 1251 1252 1253
	if (err) {
		netdev_err(dev, "Device %s failed to add per-port options\n",
			   portname);
		goto err_option_port_add;
	}

1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270
	/* 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;
		}
	}

1271 1272 1273 1274 1275
	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
1276 1277
	port->index = -1;
	list_add_tail_rcu(&port->list, &team->port_list);
1278
	team_port_enable(team, port);
1279
	__team_compute_features(team);
1280
	__team_port_change_port_added(port, !!netif_oper_up(port_dev));
1281
	__team_options_change_check(team);
1282 1283 1284 1285 1286

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

	return 0;

1287 1288 1289
err_set_slave_promisc:
	__team_option_inst_del_port(team, port);

1290
err_option_port_add:
1291
	team_upper_dev_unlink(team, port);
1292 1293

err_set_upper_link:
1294 1295
	netdev_rx_handler_unregister(port_dev);

1296
err_handler_register:
J
Jiri Pirko 已提交
1297 1298 1299
	team_port_disable_netpoll(port);

err_enable_netpoll:
J
Jiri Pirko 已提交
1300 1301 1302
	vlan_vids_del_by_dev(port_dev, dev);

err_vids_add:
1303 1304 1305 1306
	dev_close(port_dev);

err_dev_open:
	team_port_leave(team, port);
1307
	team_port_set_orig_dev_addr(port);
1308 1309 1310 1311 1312 1313 1314 1315 1316 1317

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

err_set_mtu:
	kfree(port);

	return err;
}

J
Jiri Pirko 已提交
1318 1319
static void __team_port_change_port_removed(struct team_port *port);

1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332
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
1333 1334
	team_port_disable(team, port);
	list_del_rcu(&port->list);
1335 1336 1337 1338 1339 1340

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

1341
	team_upper_dev_unlink(team, port);
1342
	netdev_rx_handler_unregister(port_dev);
J
Jiri Pirko 已提交
1343
	team_port_disable_netpoll(port);
J
Jiri Pirko 已提交
1344
	vlan_vids_del_by_dev(port_dev, dev);
1345 1346
	dev_uc_unsync(port_dev, dev);
	dev_mc_unsync(port_dev, dev);
1347 1348
	dev_close(port_dev);
	team_port_leave(team, port);
1349 1350 1351 1352 1353 1354

	__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);

1355
	team_port_set_orig_dev_addr(port);
1356
	dev_set_mtu(port_dev, port->orig.mtu);
1357
	kfree_rcu(port, rcu);
1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368
	netdev_info(dev, "Port device %s removed\n", portname);
	__team_compute_features(team);

	return 0;
}


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

1369
static int team_mode_option_get(struct team *team, struct team_gsetter_ctx *ctx)
1370
{
1371
	ctx->data.str_val = team->mode->kind;
1372 1373 1374
	return 0;
}

1375
static int team_mode_option_set(struct team *team, struct team_gsetter_ctx *ctx)
1376
{
1377
	return team_change_mode(team, ctx->data.str_val);
1378 1379
}

J
Jiri Pirko 已提交
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 1406 1407
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;
}

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 1434 1435
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;
}

1436 1437 1438
static int team_port_en_option_get(struct team *team,
				   struct team_gsetter_ctx *ctx)
{
1439 1440 1441
	struct team_port *port = ctx->info->port;

	ctx->data.bool_val = team_port_enabled(port);
1442 1443 1444 1445 1446 1447
	return 0;
}

static int team_port_en_option_set(struct team *team,
				   struct team_gsetter_ctx *ctx)
{
1448 1449
	struct team_port *port = ctx->info->port;

1450
	if (ctx->data.bool_val)
1451
		team_port_enable(team, port);
1452
	else
1453
		team_port_disable(team, port);
1454 1455 1456
	return 0;
}

1457 1458 1459
static int team_user_linkup_option_get(struct team *team,
				       struct team_gsetter_ctx *ctx)
{
1460 1461 1462
	struct team_port *port = ctx->info->port;

	ctx->data.bool_val = port->user.linkup;
1463 1464 1465
	return 0;
}

1466 1467
static void __team_carrier_check(struct team *team);

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

	port->user.linkup = ctx->data.bool_val;
	team_refresh_port_linkup(port);
1475
	__team_carrier_check(port->team);
1476 1477 1478 1479 1480 1481
	return 0;
}

static int team_user_linkup_en_option_get(struct team *team,
					  struct team_gsetter_ctx *ctx)
{
1482
	struct team_port *port = ctx->info->port;
1483 1484 1485 1486 1487 1488 1489 1490

	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)
{
1491
	struct team_port *port = ctx->info->port;
1492 1493

	port->user.linkup_enabled = ctx->data.bool_val;
1494
	team_refresh_port_linkup(port);
1495
	__team_carrier_check(port->team);
1496 1497 1498
	return 0;
}

J
Jiri Pirko 已提交
1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511
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;
1512
	s32 priority = ctx->data.s32_val;
J
Jiri Pirko 已提交
1513

1514 1515 1516 1517
	if (port->priority == priority)
		return 0;
	port->priority = priority;
	team_queue_override_port_prio_changed(team, port);
1518 1519 1520 1521 1522 1523 1524 1525 1526
	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
1527 1528 1529
	return 0;
}

1530 1531 1532 1533
static int team_queue_id_option_set(struct team *team,
				    struct team_gsetter_ctx *ctx)
{
	struct team_port *port = ctx->info->port;
1534
	u16 new_queue_id = ctx->data.u32_val;
1535

1536
	if (port->queue_id == new_queue_id)
1537
		return 0;
1538
	if (new_queue_id >= team->dev->real_num_tx_queues)
1539
		return -EINVAL;
1540
	team_queue_override_port_change_queue_id(team, port, new_queue_id);
1541 1542 1543
	return 0;
}

J
Jiri Pirko 已提交
1544
static const struct team_option team_options[] = {
1545 1546 1547 1548 1549 1550
	{
		.name = "mode",
		.type = TEAM_OPTION_TYPE_STRING,
		.getter = team_mode_option_get,
		.setter = team_mode_option_set,
	},
J
Jiri Pirko 已提交
1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562
	{
		.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,
	},
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574
	{
		.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,
	},
1575 1576 1577 1578 1579 1580 1581
	{
		.name = "enabled",
		.type = TEAM_OPTION_TYPE_BOOL,
		.per_port = true,
		.getter = team_port_en_option_get,
		.setter = team_port_en_option_set,
	},
1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595
	{
		.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
1596 1597 1598 1599 1600 1601 1602
	{
		.name = "priority",
		.type = TEAM_OPTION_TYPE_S32,
		.per_port = true,
		.getter = team_priority_option_get,
		.setter = team_priority_option_set,
	},
1603 1604 1605 1606 1607 1608 1609
	{
		.name = "queue_id",
		.type = TEAM_OPTION_TYPE_U32,
		.per_port = true,
		.getter = team_queue_id_option_get,
		.setter = team_queue_id_option_set,
	},
1610 1611
};

J
Jiri Pirko 已提交
1612

1613 1614 1615 1616
static int team_init(struct net_device *dev)
{
	struct team *team = netdev_priv(dev);
	int i;
J
Jiri Pirko 已提交
1617
	int err;
1618 1619

	team->dev = dev;
1620
	mutex_init(&team->lock);
1621
	team_set_no_mode(team);
1622

1623
	team->pcpu_stats = netdev_alloc_pcpu_stats(struct team_pcpu_stats);
1624 1625 1626 1627
	if (!team->pcpu_stats)
		return -ENOMEM;

	for (i = 0; i < TEAM_PORT_HASHENTRIES; i++)
J
Jiri Pirko 已提交
1628
		INIT_HLIST_HEAD(&team->en_port_hlist[i]);
1629
	INIT_LIST_HEAD(&team->port_list);
1630 1631 1632
	err = team_queue_override_init(team);
	if (err)
		goto err_team_queue_override_init;
1633 1634 1635 1636

	team_adjust_ops(team);

	INIT_LIST_HEAD(&team->option_list);
1637
	INIT_LIST_HEAD(&team->option_inst_list);
J
Jiri Pirko 已提交
1638 1639

	team_notify_peers_init(team);
1640
	team_mcast_rejoin_init(team);
J
Jiri Pirko 已提交
1641

J
Jiri Pirko 已提交
1642 1643 1644
	err = team_options_register(team, team_options, ARRAY_SIZE(team_options));
	if (err)
		goto err_options_register;
1645 1646
	netif_carrier_off(dev);

1647
	netdev_lockdep_set_classes(dev);
J
Jiri Pirko 已提交
1648

1649
	return 0;
J
Jiri Pirko 已提交
1650 1651

err_options_register:
1652
	team_mcast_rejoin_fini(team);
J
Jiri Pirko 已提交
1653
	team_notify_peers_fini(team);
1654 1655
	team_queue_override_fini(team);
err_team_queue_override_init:
J
Jiri Pirko 已提交
1656 1657 1658
	free_percpu(team->pcpu_stats);

	return err;
1659 1660 1661 1662 1663 1664 1665 1666
}

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

1667
	mutex_lock(&team->lock);
1668 1669 1670 1671 1672
	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));
1673
	team_mcast_rejoin_fini(team);
J
Jiri Pirko 已提交
1674
	team_notify_peers_fini(team);
1675
	team_queue_override_fini(team);
1676
	mutex_unlock(&team->lock);
1677
	netdev_change_features(dev);
1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702
}

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);
1703
	bool tx_success;
1704 1705
	unsigned int len = skb->len;

1706 1707 1708
	tx_success = team_queue_override_transmit(team, skb);
	if (!tx_success)
		tx_success = team->ops.transmit(team, skb);
1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723
	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;
}

1724
static u16 team_select_queue(struct net_device *dev, struct sk_buff *skb,
1725
			     struct net_device *sb_dev)
J
Jiri Pirko 已提交
1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747
{
	/*
	 * 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;
}

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 1774
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) {
1775 1776
		dev_uc_sync_multiple(port->dev, dev);
		dev_mc_sync_multiple(port->dev, dev);
1777 1778 1779 1780 1781 1782
	}
	rcu_read_unlock();
}

static int team_set_mac_address(struct net_device *dev, void *p)
{
1783
	struct sockaddr *addr = p;
1784 1785 1786
	struct team *team = netdev_priv(dev);
	struct team_port *port;

1787 1788 1789
	if (dev->type == ARPHRD_ETHER && !is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;
	memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
1790 1791
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list)
1792 1793
		if (team->ops.port_change_dev_addr)
			team->ops.port_change_dev_addr(team, port);
1794
	mutex_unlock(&team->lock);
1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807
	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
	 */
1808
	mutex_lock(&team->lock);
J
Jiri Pirko 已提交
1809
	team->port_mtu_change_allowed = true;
1810 1811 1812 1813 1814 1815 1816 1817
	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
1818
	team->port_mtu_change_allowed = false;
1819
	mutex_unlock(&team->lock);
1820 1821 1822 1823 1824 1825 1826 1827

	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
1828
	team->port_mtu_change_allowed = false;
1829
	mutex_unlock(&team->lock);
1830 1831 1832 1833

	return err;
}

1834
static void
1835 1836 1837 1838 1839
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;
1840
	u32 rx_dropped = 0, tx_dropped = 0, rx_nohandler = 0;
1841 1842 1843 1844 1845 1846
	unsigned int start;
	int i;

	for_each_possible_cpu(i) {
		p = per_cpu_ptr(team->pcpu_stats, i);
		do {
1847
			start = u64_stats_fetch_begin_irq(&p->syncp);
1848 1849 1850 1851 1852
			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;
1853
		} while (u64_stats_fetch_retry_irq(&p->syncp, start));
1854 1855 1856 1857 1858 1859 1860

		stats->rx_packets	+= rx_packets;
		stats->rx_bytes		+= rx_bytes;
		stats->multicast	+= rx_multicast;
		stats->tx_packets	+= tx_packets;
		stats->tx_bytes		+= tx_bytes;
		/*
1861 1862
		 * rx_dropped, tx_dropped & rx_nohandler are u32,
		 * updated without syncp protection.
1863 1864 1865
		 */
		rx_dropped	+= p->rx_dropped;
		tx_dropped	+= p->tx_dropped;
1866
		rx_nohandler	+= p->rx_nohandler;
1867 1868 1869
	}
	stats->rx_dropped	= rx_dropped;
	stats->tx_dropped	= tx_dropped;
1870
	stats->rx_nohandler	= rx_nohandler;
1871 1872
}

1873
static int team_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
1874 1875 1876
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
1877
	int err;
1878

1879 1880 1881 1882 1883 1884
	/*
	 * 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) {
1885
		err = vlan_vid_add(port->dev, proto, vid);
1886 1887
		if (err)
			goto unwind;
1888
	}
1889
	mutex_unlock(&team->lock);
1890 1891

	return 0;
1892 1893 1894

unwind:
	list_for_each_entry_continue_reverse(port, &team->port_list, list)
1895
		vlan_vid_del(port->dev, proto, vid);
1896 1897 1898
	mutex_unlock(&team->lock);

	return err;
1899 1900
}

1901
static int team_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
1902 1903 1904 1905
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;

1906 1907
	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list)
1908
		vlan_vid_del(port->dev, proto, vid);
1909
	mutex_unlock(&team->lock);
1910 1911

	return 0;
1912 1913
}

J
Jiri Pirko 已提交
1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936
#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,
1937
			      struct netpoll_info *npifo)
J
Jiri Pirko 已提交
1938 1939 1940
{
	struct team *team = netdev_priv(dev);
	struct team_port *port;
1941
	int err = 0;
J
Jiri Pirko 已提交
1942 1943 1944

	mutex_lock(&team->lock);
	list_for_each_entry(port, &team->port_list, list) {
X
Xin Long 已提交
1945
		err = __team_port_enable_netpoll(port);
J
Jiri Pirko 已提交
1946 1947 1948 1949 1950 1951 1952 1953 1954 1955
		if (err) {
			__team_netpoll_cleanup(team);
			break;
		}
	}
	mutex_unlock(&team->lock);
	return err;
}
#endif

D
David Ahern 已提交
1956 1957
static int team_add_slave(struct net_device *dev, struct net_device *port_dev,
			  struct netlink_ext_ack *extack)
1958 1959 1960 1961
{
	struct team *team = netdev_priv(dev);
	int err;

1962
	mutex_lock(&team->lock);
1963
	err = team_port_add(team, port_dev, extack);
1964
	mutex_unlock(&team->lock);
1965 1966 1967 1968

	if (!err)
		netdev_change_features(dev);

1969 1970 1971 1972 1973 1974 1975 1976
	return err;
}

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

1977
	mutex_lock(&team->lock);
1978
	err = team_port_del(team, port_dev);
1979
	mutex_unlock(&team->lock);
1980 1981 1982 1983

	if (!err)
		netdev_change_features(dev);

1984 1985 1986
	return err;
}

J
Jiri Pirko 已提交
1987 1988 1989 1990 1991 1992 1993
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;

1994
	mask = features;
J
Jiri Pirko 已提交
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
	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
2005 2006 2007

	features = netdev_add_tso_features(features, mask);

J
Jiri Pirko 已提交
2008 2009 2010
	return features;
}

F
Flavio Leitner 已提交
2011 2012
static int team_change_carrier(struct net_device *dev, bool new_carrier)
{
2013 2014 2015 2016
	struct team *team = netdev_priv(dev);

	team->user_carrier_enabled = true;

F
Flavio Leitner 已提交
2017 2018 2019 2020 2021 2022 2023
	if (new_carrier)
		netif_carrier_on(dev);
	else
		netif_carrier_off(dev);
	return 0;
}

2024 2025 2026 2027 2028 2029
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
2030
	.ndo_select_queue	= team_select_queue,
2031 2032 2033 2034 2035 2036 2037
	.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
2038 2039 2040 2041 2042
#ifdef CONFIG_NET_POLL_CONTROLLER
	.ndo_poll_controller	= team_poll_controller,
	.ndo_netpoll_setup	= team_netpoll_setup,
	.ndo_netpoll_cleanup	= team_netpoll_cleanup,
#endif
2043 2044
	.ndo_add_slave		= team_add_slave,
	.ndo_del_slave		= team_del_slave,
J
Jiri Pirko 已提交
2045
	.ndo_fix_features	= team_fix_features,
F
Flavio Leitner 已提交
2046
	.ndo_change_carrier     = team_change_carrier,
2047
	.ndo_features_check	= passthru_features_check,
2048 2049
};

F
Flavio Leitner 已提交
2050 2051 2052 2053 2054 2055 2056
/***********************
 * ethtool interface
 ***********************/

static void team_ethtool_get_drvinfo(struct net_device *dev,
				     struct ethtool_drvinfo *drvinfo)
{
2057 2058
	strlcpy(drvinfo->driver, DRV_NAME, sizeof(drvinfo->driver));
	strlcpy(drvinfo->version, UTS_RELEASE, sizeof(drvinfo->version));
F
Flavio Leitner 已提交
2059 2060 2061 2062 2063 2064
}

static const struct ethtool_ops team_ethtool_ops = {
	.get_drvinfo		= team_ethtool_get_drvinfo,
	.get_link		= ethtool_op_get_link,
};
2065 2066 2067 2068 2069

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

2070 2071 2072 2073 2074 2075 2076 2077 2078
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);
2079
	eth_hw_addr_inherit(dev, port_dev);
2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107
}

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;
}

2108 2109 2110
static void team_setup(struct net_device *dev)
{
	ether_setup(dev);
J
Jarod Wilson 已提交
2111
	dev->max_mtu = ETH_MAX_MTU;
2112 2113

	dev->netdev_ops = &team_netdev_ops;
F
Flavio Leitner 已提交
2114
	dev->ethtool_ops = &team_ethtool_ops;
2115 2116
	dev->needs_free_netdev = true;
	dev->priv_destructor = team_destructor;
2117
	dev->priv_flags &= ~(IFF_XMIT_DST_RELEASE | IFF_TX_SKB_SHARING);
2118
	dev->priv_flags |= IFF_NO_QUEUE;
J
Jiri Pirko 已提交
2119
	dev->priv_flags |= IFF_TEAM;
2120 2121 2122 2123 2124 2125

	/*
	 * 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.
	 */
2126
	dev->priv_flags |= IFF_UNICAST_FLT | IFF_LIVE_ADDR_CHANGE;
2127 2128 2129

	dev->features |= NETIF_F_LLTX;
	dev->features |= NETIF_F_GRO;
2130 2131 2132 2133

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

J
Jiri Pirko 已提交
2134
	dev->hw_features = TEAM_VLAN_FEATURES |
2135 2136 2137
			   NETIF_F_HW_VLAN_CTAG_TX |
			   NETIF_F_HW_VLAN_CTAG_RX |
			   NETIF_F_HW_VLAN_CTAG_FILTER;
2138

2139
	dev->hw_features |= NETIF_F_GSO_ENCAP_ALL | NETIF_F_GSO_UDP_L4;
2140 2141 2142 2143
	dev->features |= dev->hw_features;
}

static int team_newlink(struct net *src_net, struct net_device *dev,
2144 2145
			struct nlattr *tb[], struct nlattr *data[],
			struct netlink_ext_ack *extack)
2146 2147
{
	if (tb[IFLA_ADDRESS] == NULL)
2148
		eth_hw_addr_random(dev);
2149

2150
	return register_netdevice(dev);
2151 2152
}

2153 2154
static int team_validate(struct nlattr *tb[], struct nlattr *data[],
			 struct netlink_ext_ack *extack)
2155 2156 2157 2158 2159 2160 2161 2162 2163 2164
{
	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
2165 2166 2167 2168 2169 2170 2171 2172 2173 2174
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;
}

2175
static struct rtnl_link_ops team_link_ops __read_mostly = {
J
Jiri Pirko 已提交
2176 2177 2178 2179 2180 2181 2182
	.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,
2183 2184 2185 2186 2187 2188 2189
};


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

2190
static struct genl_family team_nl_family;
2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207

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
2208
	[TEAM_ATTR_OPTION_DATA]			= { .type = NLA_BINARY },
2209 2210 2211 2212 2213 2214 2215 2216
};

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

2217
	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
2218 2219 2220
	if (!msg)
		return -ENOMEM;

2221
	hdr = genlmsg_put(msg, info->snd_portid, info->snd_seq,
2222
			  &team_nl_family, 0, TEAM_CMD_NOOP);
W
Wei Yongjun 已提交
2223 2224
	if (!hdr) {
		err = -EMSGSIZE;
2225 2226 2227 2228 2229
		goto err_msg_put;
	}

	genlmsg_end(msg, hdr);

2230
	return genlmsg_unicast(genl_info_net(info), msg, info->snd_portid);
2231 2232 2233 2234 2235 2236 2237 2238 2239

err_msg_put:
	nlmsg_free(msg);

	return err;
}

/*
 * Netlink cmd functions should be locked by following two functions.
2240
 * Since dev gets held here, that ensures dev won't disappear in between.
2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252
 */
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]);
2253
	dev = dev_get_by_index(net, ifindex);
2254
	if (!dev || dev->netdev_ops != &team_netdev_ops) {
2255 2256
		if (dev)
			dev_put(dev);
2257 2258 2259 2260
		return NULL;
	}

	team = netdev_priv(dev);
2261
	mutex_lock(&team->lock);
2262 2263 2264 2265 2266
	return team;
}

static void team_nl_team_put(struct team *team)
{
2267
	mutex_unlock(&team->lock);
2268
	dev_put(team->dev);
2269 2270
}

2271
typedef int team_nl_send_func_t(struct sk_buff *skb,
2272
				struct team *team, u32 portid);
2273

2274
static int team_nl_send_unicast(struct sk_buff *skb, struct team *team, u32 portid)
2275
{
2276
	return genlmsg_unicast(dev_net(team->dev), skb, portid);
2277 2278
}

2279 2280 2281 2282 2283
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;
2284
	struct team_option_inst_info *opt_inst_info = &opt_inst->info;
2285 2286 2287
	struct team_gsetter_ctx ctx;
	int err;

2288 2289 2290 2291 2292
	ctx.info = opt_inst_info;
	err = team_option_get(team, opt_inst, &ctx);
	if (err)
		return err;

2293
	option_item = nla_nest_start_noflag(skb, TEAM_ATTR_ITEM_OPTION);
2294
	if (!option_item)
2295
		return -EMSGSIZE;
2296

2297 2298
	if (nla_put_string(skb, TEAM_ATTR_OPTION_NAME, option->name))
		goto nest_cancel;
2299 2300 2301
	if (opt_inst_info->port &&
	    nla_put_u32(skb, TEAM_ATTR_OPTION_PORT_IFINDEX,
			opt_inst_info->port->dev->ifindex))
2302
		goto nest_cancel;
2303 2304 2305
	if (opt_inst->option->array_size &&
	    nla_put_u32(skb, TEAM_ATTR_OPTION_ARRAY_INDEX,
			opt_inst_info->array_index))
2306
		goto nest_cancel;
2307 2308 2309 2310

	switch (option->type) {
	case TEAM_OPTION_TYPE_U32:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_U32))
2311
			goto nest_cancel;
2312
		if (nla_put_u32(skb, TEAM_ATTR_OPTION_DATA, ctx.data.u32_val))
2313
			goto nest_cancel;
2314 2315 2316
		break;
	case TEAM_OPTION_TYPE_STRING:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_STRING))
2317
			goto nest_cancel;
2318 2319
		if (nla_put_string(skb, TEAM_ATTR_OPTION_DATA,
				   ctx.data.str_val))
2320
			goto nest_cancel;
2321 2322 2323
		break;
	case TEAM_OPTION_TYPE_BINARY:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_BINARY))
2324
			goto nest_cancel;
2325 2326
		if (nla_put(skb, TEAM_ATTR_OPTION_DATA, ctx.data.bin_val.len,
			    ctx.data.bin_val.ptr))
2327
			goto nest_cancel;
2328 2329 2330
		break;
	case TEAM_OPTION_TYPE_BOOL:
		if (nla_put_u8(skb, TEAM_ATTR_OPTION_TYPE, NLA_FLAG))
2331
			goto nest_cancel;
2332 2333
		if (ctx.data.bool_val &&
		    nla_put_flag(skb, TEAM_ATTR_OPTION_DATA))
2334
			goto nest_cancel;
2335
		break;
2336 2337 2338 2339 2340 2341
	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;
2342 2343 2344
	default:
		BUG();
	}
2345 2346 2347 2348 2349 2350 2351
	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;
	}
2352 2353 2354
	nla_nest_end(skb, option_item);
	return 0;

2355 2356 2357 2358 2359 2360
nest_cancel:
	nla_nest_cancel(skb, option_item);
	return -EMSGSIZE;
}

static int __send_and_alloc_skb(struct sk_buff **pskb,
2361
				struct team *team, u32 portid,
2362 2363 2364 2365 2366
				team_nl_send_func_t *send_func)
{
	int err;

	if (*pskb) {
2367
		err = send_func(*pskb, team, portid);
2368 2369 2370
		if (err)
			return err;
	}
2371
	*pskb = genlmsg_new(GENLMSG_DEFAULT_SIZE, GFP_KERNEL);
2372 2373 2374
	if (!*pskb)
		return -ENOMEM;
	return 0;
2375 2376
}

2377
static int team_nl_send_options_get(struct team *team, u32 portid, u32 seq,
2378
				    int flags, team_nl_send_func_t *send_func,
2379
				    struct list_head *sel_opt_inst_list)
2380 2381
{
	struct nlattr *option_list;
2382
	struct nlmsghdr *nlh;
2383
	void *hdr;
2384 2385
	struct team_option_inst *opt_inst;
	int err;
2386 2387 2388
	struct sk_buff *skb = NULL;
	bool incomplete;
	int i;
2389

2390 2391 2392 2393
	opt_inst = list_first_entry(sel_opt_inst_list,
				    struct team_option_inst, tmp_list);

start_again:
2394
	err = __send_and_alloc_skb(&skb, team, portid, send_func);
2395 2396 2397
	if (err)
		return err;

2398
	hdr = genlmsg_put(skb, portid, seq, &team_nl_family, flags | NLM_F_MULTI,
2399
			  TEAM_CMD_OPTIONS_GET);
P
Pan Bian 已提交
2400 2401
	if (!hdr) {
		nlmsg_free(skb);
W
Wei Yongjun 已提交
2402
		return -EMSGSIZE;
P
Pan Bian 已提交
2403
	}
2404

D
David S. Miller 已提交
2405 2406
	if (nla_put_u32(skb, TEAM_ATTR_TEAM_IFINDEX, team->dev->ifindex))
		goto nla_put_failure;
2407
	option_list = nla_nest_start_noflag(skb, TEAM_ATTR_LIST_OPTION);
2408
	if (!option_list)
2409
		goto nla_put_failure;
2410

2411 2412 2413
	i = 0;
	incomplete = false;
	list_for_each_entry_from(opt_inst, sel_opt_inst_list, tmp_list) {
2414
		err = team_nl_fill_one_option_get(skb, team, opt_inst);
2415 2416 2417 2418 2419 2420 2421
		if (err) {
			if (err == -EMSGSIZE) {
				if (!i)
					goto errout;
				incomplete = true;
				break;
			}
2422
			goto errout;
2423 2424
		}
		i++;
2425 2426 2427
	}

	nla_nest_end(skb, option_list);
2428 2429 2430 2431 2432
	genlmsg_end(skb, hdr);
	if (incomplete)
		goto start_again;

send_done:
2433
	nlh = nlmsg_put(skb, portid, seq, NLMSG_DONE, 0, flags | NLM_F_MULTI);
2434
	if (!nlh) {
2435
		err = __send_and_alloc_skb(&skb, team, portid, send_func);
2436
		if (err)
2437
			return err;
2438 2439 2440
		goto send_done;
	}

2441
	return send_func(skb, team, portid);
2442 2443

nla_put_failure:
2444 2445
	err = -EMSGSIZE;
errout:
2446
	nlmsg_free(skb);
2447
	return err;
2448 2449 2450 2451 2452
}

static int team_nl_cmd_options_get(struct sk_buff *skb, struct genl_info *info)
{
	struct team *team;
2453
	struct team_option_inst *opt_inst;
2454
	int err;
2455
	LIST_HEAD(sel_opt_inst_list);
2456 2457 2458 2459 2460

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

2461 2462
	list_for_each_entry(opt_inst, &team->option_inst_list, list)
		list_add_tail(&opt_inst->tmp_list, &sel_opt_inst_list);
2463
	err = team_nl_send_options_get(team, info->snd_portid, info->snd_seq,
2464 2465
				       NLM_F_ACK, team_nl_send_unicast,
				       &sel_opt_inst_list);
2466 2467 2468 2469 2470 2471

	team_nl_team_put(team);

	return err;
}

2472 2473 2474
static int team_nl_send_event_options_get(struct team *team,
					  struct list_head *sel_opt_inst_list);

2475 2476 2477 2478 2479 2480 2481
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
2482 2483
	rtnl_lock();

2484
	team = team_nl_team_get(info);
J
Jiri Pirko 已提交
2485 2486 2487 2488
	if (!team) {
		err = -EINVAL;
		goto rtnl_unlock;
	}
2489 2490 2491 2492 2493 2494 2495 2496

	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) {
2497
		struct nlattr *opt_attrs[TEAM_ATTR_OPTION_MAX + 1];
J
Jiri Pirko 已提交
2498
		struct nlattr *attr;
J
Jiri Pirko 已提交
2499
		struct nlattr *attr_data;
2500
		LIST_HEAD(opt_inst_list);
2501
		enum team_option_type opt_type;
2502
		int opt_port_ifindex = 0; /* != 0 for per-port options */
J
Jiri Pirko 已提交
2503 2504
		u32 opt_array_index = 0;
		bool opt_is_array = false;
2505
		struct team_option_inst *opt_inst;
2506 2507 2508 2509 2510 2511 2512
		char *opt_name;
		bool opt_found = false;

		if (nla_type(nl_option) != TEAM_ATTR_ITEM_OPTION) {
			err = -EINVAL;
			goto team_put;
		}
2513
		err = nla_parse_nested(opt_attrs, TEAM_ATTR_OPTION_MAX,
2514 2515
				       nl_option, team_nl_option_policy,
				       info->extack);
2516 2517
		if (err)
			goto team_put;
2518
		if (!opt_attrs[TEAM_ATTR_OPTION_NAME] ||
J
Jiri Pirko 已提交
2519
		    !opt_attrs[TEAM_ATTR_OPTION_TYPE]) {
2520 2521 2522
			err = -EINVAL;
			goto team_put;
		}
2523
		switch (nla_get_u8(opt_attrs[TEAM_ATTR_OPTION_TYPE])) {
2524 2525 2526 2527 2528 2529
		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
2530 2531 2532
		case NLA_BINARY:
			opt_type = TEAM_OPTION_TYPE_BINARY;
			break;
J
Jiri Pirko 已提交
2533 2534 2535
		case NLA_FLAG:
			opt_type = TEAM_OPTION_TYPE_BOOL;
			break;
2536 2537 2538
		case NLA_S32:
			opt_type = TEAM_OPTION_TYPE_S32;
			break;
2539 2540 2541 2542
		default:
			goto team_put;
		}

J
Jiri Pirko 已提交
2543 2544 2545 2546 2547 2548
		attr_data = opt_attrs[TEAM_ATTR_OPTION_DATA];
		if (opt_type != TEAM_OPTION_TYPE_BOOL && !attr_data) {
			err = -EINVAL;
			goto team_put;
		}

2549
		opt_name = nla_data(opt_attrs[TEAM_ATTR_OPTION_NAME]);
J
Jiri Pirko 已提交
2550 2551 2552 2553 2554 2555 2556 2557 2558
		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);
		}
2559 2560 2561 2562

		list_for_each_entry(opt_inst, &team->option_inst_list, list) {
			struct team_option *option = opt_inst->option;
			struct team_gsetter_ctx ctx;
2563
			struct team_option_inst_info *opt_inst_info;
2564
			int tmp_ifindex;
2565

2566 2567 2568
			opt_inst_info = &opt_inst->info;
			tmp_ifindex = opt_inst_info->port ?
				      opt_inst_info->port->dev->ifindex : 0;
2569
			if (option->type != opt_type ||
2570
			    strcmp(option->name, opt_name) ||
J
Jiri Pirko 已提交
2571 2572
			    tmp_ifindex != opt_port_ifindex ||
			    (option->array_size && !opt_is_array) ||
2573
			    opt_inst_info->array_index != opt_array_index)
2574 2575
				continue;
			opt_found = true;
2576
			ctx.info = opt_inst_info;
2577 2578
			switch (opt_type) {
			case TEAM_OPTION_TYPE_U32:
J
Jiri Pirko 已提交
2579
				ctx.data.u32_val = nla_get_u32(attr_data);
2580 2581
				break;
			case TEAM_OPTION_TYPE_STRING:
J
Jiri Pirko 已提交
2582
				if (nla_len(attr_data) > TEAM_STRING_MAX_LEN) {
J
Jiri Pirko 已提交
2583 2584 2585
					err = -EINVAL;
					goto team_put;
				}
J
Jiri Pirko 已提交
2586
				ctx.data.str_val = nla_data(attr_data);
2587
				break;
J
Jiri Pirko 已提交
2588
			case TEAM_OPTION_TYPE_BINARY:
J
Jiri Pirko 已提交
2589 2590 2591 2592 2593
				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
2594
				break;
2595 2596 2597
			case TEAM_OPTION_TYPE_S32:
				ctx.data.s32_val = nla_get_s32(attr_data);
				break;
2598 2599 2600
			default:
				BUG();
			}
2601
			err = team_option_set(team, opt_inst, &ctx);
2602 2603
			if (err)
				goto team_put;
2604 2605
			opt_inst->changed = true;
			list_add(&opt_inst->tmp_list, &opt_inst_list);
2606 2607 2608 2609 2610 2611
		}
		if (!opt_found) {
			err = -ENOENT;
			goto team_put;
		}

2612 2613 2614 2615
		err = team_nl_send_event_options_get(team, &opt_inst_list);
		if (err)
			break;
	}
2616

2617 2618
team_put:
	team_nl_team_put(team);
J
Jiri Pirko 已提交
2619 2620
rtnl_unlock:
	rtnl_unlock();
2621 2622 2623
	return err;
}

2624 2625 2626 2627 2628
static int team_nl_fill_one_port_get(struct sk_buff *skb,
				     struct team_port *port)
{
	struct nlattr *port_item;

2629
	port_item = nla_nest_start_noflag(skb, TEAM_ATTR_ITEM_PORT);
2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656
	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)
2657 2658
{
	struct nlattr *port_list;
2659
	struct nlmsghdr *nlh;
2660 2661
	void *hdr;
	struct team_port *port;
2662 2663 2664 2665
	int err;
	struct sk_buff *skb = NULL;
	bool incomplete;
	int i;
2666

2667 2668
	port = list_first_entry_or_null(&team->port_list,
					struct team_port, list);
2669 2670 2671 2672 2673 2674 2675

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,
2676
			  TEAM_CMD_PORT_LIST_GET);
P
Pan Bian 已提交
2677 2678
	if (!hdr) {
		nlmsg_free(skb);
W
Wei Yongjun 已提交
2679
		return -EMSGSIZE;
P
Pan Bian 已提交
2680
	}
2681

D
David S. Miller 已提交
2682 2683
	if (nla_put_u32(skb, TEAM_ATTR_TEAM_IFINDEX, team->dev->ifindex))
		goto nla_put_failure;
2684
	port_list = nla_nest_start_noflag(skb, TEAM_ATTR_LIST_PORT);
2685
	if (!port_list)
2686
		goto nla_put_failure;
2687

2688 2689
	i = 0;
	incomplete = false;
2690

2691 2692 2693 2694 2695 2696 2697
	/* 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;
2698 2699
	} else if (port) {
		list_for_each_entry_from(port, &team->port_list, list) {
2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710
			err = team_nl_fill_one_port_get(skb, port);
			if (err) {
				if (err == -EMSGSIZE) {
					if (!i)
						goto errout;
					incomplete = true;
					break;
				}
				goto errout;
			}
			i++;
2711
		}
2712 2713 2714
	}

	nla_nest_end(skb, port_list);
2715 2716 2717 2718 2719 2720 2721 2722 2723
	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)
2724
			return err;
2725 2726 2727 2728
		goto send_done;
	}

	return send_func(skb, team, portid);
2729 2730

nla_put_failure:
2731 2732 2733 2734
	err = -EMSGSIZE;
errout:
	nlmsg_free(skb);
	return err;
2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746
}

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;

2747 2748
	err = team_nl_send_port_list_get(team, info->snd_portid, info->snd_seq,
					 NLM_F_ACK, team_nl_send_unicast, NULL);
2749 2750 2751 2752 2753 2754

	team_nl_team_put(team);

	return err;
}

2755
static const struct genl_ops team_nl_ops[] = {
2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776
	{
		.cmd = TEAM_CMD_NOOP,
		.doit = team_nl_cmd_noop,
	},
	{
		.cmd = TEAM_CMD_OPTIONS_SET,
		.doit = team_nl_cmd_options_set,
		.flags = GENL_ADMIN_PERM,
	},
	{
		.cmd = TEAM_CMD_OPTIONS_GET,
		.doit = team_nl_cmd_options_get,
		.flags = GENL_ADMIN_PERM,
	},
	{
		.cmd = TEAM_CMD_PORT_LIST_GET,
		.doit = team_nl_cmd_port_list_get,
		.flags = GENL_ADMIN_PERM,
	},
};

2777 2778
static const struct genl_multicast_group team_nl_mcgrps[] = {
	{ .name = TEAM_GENL_CHANGE_EVENT_MC_GRP_NAME, },
2779 2780
};

2781
static struct genl_family team_nl_family __ro_after_init = {
2782 2783 2784
	.name		= TEAM_GENL_NAME,
	.version	= TEAM_GENL_VERSION,
	.maxattr	= TEAM_ATTR_MAX,
2785
	.policy = team_nl_policy,
2786 2787 2788 2789 2790 2791 2792 2793
	.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),
};

2794
static int team_nl_send_multicast(struct sk_buff *skb,
2795
				  struct team *team, u32 portid)
2796
{
2797
	return genlmsg_multicast_netns(&team_nl_family, dev_net(team->dev),
2798
				       skb, 0, 0, GFP_KERNEL);
2799 2800
}

2801 2802
static int team_nl_send_event_options_get(struct team *team,
					  struct list_head *sel_opt_inst_list)
2803
{
2804 2805
	return team_nl_send_options_get(team, 0, 0, 0, team_nl_send_multicast,
					sel_opt_inst_list);
2806 2807
}

2808 2809
static int team_nl_send_event_port_get(struct team *team,
				       struct team_port *port)
2810
{
2811 2812
	return team_nl_send_port_list_get(team, 0, 0, 0, team_nl_send_multicast,
					  port);
2813 2814
}

2815
static int __init team_nl_init(void)
2816
{
2817
	return genl_register_family(&team_nl_family);
2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829
}

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


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

2830
static void __team_options_change_check(struct team *team)
2831 2832
{
	int err;
2833 2834
	struct team_option_inst *opt_inst;
	LIST_HEAD(sel_opt_inst_list);
2835

2836 2837 2838 2839 2840
	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);
2841
	if (err && err != -ESRCH)
2842 2843
		netdev_warn(team->dev, "Failed to send options change via netlink (err %d)\n",
			    err);
2844 2845 2846
}

/* rtnl lock is held */
J
Jiri Pirko 已提交
2847 2848

static void __team_port_change_send(struct team_port *port, bool linkup)
2849 2850 2851
{
	int err;

2852
	port->changed = true;
2853 2854
	port->state.linkup = linkup;
	team_refresh_port_linkup(port);
2855
	if (linkup) {
2856
		struct ethtool_link_ksettings ecmd;
2857

2858
		err = __ethtool_get_link_ksettings(port->dev, &ecmd);
2859
		if (!err) {
2860 2861
			port->state.speed = ecmd.base.speed;
			port->state.duplex = ecmd.base.duplex;
2862 2863 2864
			goto send_event;
		}
	}
2865 2866
	port->state.speed = 0;
	port->state.duplex = 0;
2867 2868

send_event:
2869
	err = team_nl_send_event_port_get(port->team, port);
2870 2871 2872
	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);
2873 2874 2875

}

2876 2877 2878 2879 2880
static void __team_carrier_check(struct team *team)
{
	struct team_port *port;
	bool team_linkup;

2881 2882 2883
	if (team->user_carrier_enabled)
		return;

2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897
	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
2898 2899 2900 2901
static void __team_port_change_check(struct team_port *port, bool linkup)
{
	if (port->state.linkup != linkup)
		__team_port_change_send(port, linkup);
2902
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2903 2904 2905 2906 2907
}

static void __team_port_change_port_added(struct team_port *port, bool linkup)
{
	__team_port_change_send(port, linkup);
2908
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2909 2910 2911 2912 2913 2914
}

static void __team_port_change_port_removed(struct team_port *port)
{
	port->removed = true;
	__team_port_change_send(port, false);
2915
	__team_carrier_check(port->team);
J
Jiri Pirko 已提交
2916 2917
}

2918 2919 2920 2921
static void team_port_change_check(struct team_port *port, bool linkup)
{
	struct team *team = port->team;

2922
	mutex_lock(&team->lock);
2923
	__team_port_change_check(port, linkup);
2924
	mutex_unlock(&team->lock);
2925 2926
}

J
Jiri Pirko 已提交
2927

2928 2929 2930 2931 2932 2933 2934
/************************************
 * Net device notifier event handler
 ************************************/

static int team_device_event(struct notifier_block *unused,
			     unsigned long event, void *ptr)
{
2935
	struct net_device *dev = netdev_notifier_info_to_dev(ptr);
2936 2937 2938 2939 2940 2941 2942 2943
	struct team_port *port;

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

	switch (event) {
	case NETDEV_UP:
2944
		if (netif_oper_up(dev))
2945
			team_port_change_check(port, true);
2946
		break;
2947 2948
	case NETDEV_DOWN:
		team_port_change_check(port, false);
2949
		break;
2950 2951 2952
	case NETDEV_CHANGE:
		if (netif_running(port->dev))
			team_port_change_check(port,
G
George Wilkie 已提交
2953
					       !!netif_oper_up(port->dev));
2954 2955 2956 2957 2958 2959 2960
		break;
	case NETDEV_UNREGISTER:
		team_del_slave(port->team->dev, dev);
		break;
	case NETDEV_FEAT_CHANGE:
		team_compute_features(port->team);
		break;
2961
	case NETDEV_PRECHANGEMTU:
2962
		/* Forbid to change mtu of underlaying device */
J
Jiri Pirko 已提交
2963 2964 2965
		if (!port->team->port_mtu_change_allowed)
			return NOTIFY_BAD;
		break;
2966 2967 2968
	case NETDEV_PRE_TYPE_CHANGE:
		/* Forbid to change type of underlaying device */
		return NOTIFY_BAD;
2969 2970 2971 2972
	case NETDEV_RESEND_IGMP:
		/* Propagate to master device */
		call_netdevice_notifiers(event, port->team->dev);
		break;
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 3021 3022 3023 3024
	}
	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);