bond_alb.c 44.7 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
/*
 * Copyright(c) 1999 - 2004 Intel Corporation. All rights reserved.
 *
 * 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.
 *
 * This program is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
 * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 * for more details.
 *
 * You should have received a copy of the GNU General Public License along
 * with this program; if not, write to the Free Software Foundation, Inc.,
 * 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
 *
 * The full GNU General Public License is included in this distribution in the
 * file called LICENSE.
 *
 */

#include <linux/skbuff.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/pkt_sched.h>
#include <linux/spinlock.h>
#include <linux/slab.h>
#include <linux/timer.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/if_arp.h>
#include <linux/if_ether.h>
#include <linux/if_bonding.h>
#include <linux/if_vlan.h>
#include <linux/in.h>
#include <net/ipx.h>
#include <net/arp.h>
39
#include <net/ipv6.h>
L
Linus Torvalds 已提交
40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82
#include <asm/byteorder.h>
#include "bonding.h"
#include "bond_alb.h"


#define ALB_TIMER_TICKS_PER_SEC	    10	/* should be a divisor of HZ */
#define BOND_TLB_REBALANCE_INTERVAL 10	/* In seconds, periodic re-balancing.
					 * Used for division - never set
					 * to zero !!!
					 */
#define BOND_ALB_LP_INTERVAL	    1	/* In seconds, periodic send of
					 * learning packets to the switch
					 */

#define BOND_TLB_REBALANCE_TICKS (BOND_TLB_REBALANCE_INTERVAL \
				  * ALB_TIMER_TICKS_PER_SEC)

#define BOND_ALB_LP_TICKS (BOND_ALB_LP_INTERVAL \
			   * ALB_TIMER_TICKS_PER_SEC)

#define TLB_HASH_TABLE_SIZE 256	/* The size of the clients hash table.
				 * Note that this value MUST NOT be smaller
				 * because the key hash table is BYTE wide !
				 */


#define TLB_NULL_INDEX		0xffffffff
#define MAX_LP_BURST		3

/* rlb defs */
#define RLB_HASH_TABLE_SIZE	256
#define RLB_NULL_INDEX		0xffffffff
#define RLB_UPDATE_DELAY	2*ALB_TIMER_TICKS_PER_SEC /* 2 seconds */
#define RLB_ARP_BURST_SIZE	2
#define RLB_UPDATE_RETRY	3	/* 3-ticks - must be smaller than the rlb
					 * rebalance interval (5 min).
					 */
/* RLB_PROMISC_TIMEOUT = 10 sec equals the time that the current slave is
 * promiscuous after failover
 */
#define RLB_PROMISC_TIMEOUT	10*ALB_TIMER_TICKS_PER_SEC

static const u8 mac_bcast[ETH_ALEN] = {0xff,0xff,0xff,0xff,0xff,0xff};
83
static const u8 mac_v6_allmcast[ETH_ALEN] = {0x33,0x33,0x00,0x00,0x00,0x01};
L
Linus Torvalds 已提交
84 85 86 87 88 89
static const int alb_delta_in_ticks = HZ / ALB_TIMER_TICKS_PER_SEC;

#pragma pack(1)
struct learning_pkt {
	u8 mac_dst[ETH_ALEN];
	u8 mac_src[ETH_ALEN];
90
	__be16 type;
L
Linus Torvalds 已提交
91 92 93 94
	u8 padding[ETH_ZLEN - ETH_HLEN];
};

struct arp_pkt {
95 96
	__be16  hw_addr_space;
	__be16  prot_addr_space;
L
Linus Torvalds 已提交
97 98
	u8      hw_addr_len;
	u8      prot_addr_len;
99
	__be16  op_code;
L
Linus Torvalds 已提交
100
	u8      mac_src[ETH_ALEN];	/* sender hardware address */
101
	__be32  ip_src;			/* sender IP address */
L
Linus Torvalds 已提交
102
	u8      mac_dst[ETH_ALEN];	/* target hardware address */
103
	__be32  ip_dst;			/* target IP address */
L
Linus Torvalds 已提交
104 105 106
};
#pragma pack()

107 108
static inline struct arp_pkt *arp_pkt(const struct sk_buff *skb)
{
109
	return (struct arp_pkt *)skb_network_header(skb);
110 111
}

L
Linus Torvalds 已提交
112 113 114
/* Forward declaration */
static void alb_send_learning_packets(struct slave *slave, u8 mac_addr[]);

115
static inline u8 _simple_hash(const u8 *hash_start, int hash_size)
L
Linus Torvalds 已提交
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
{
	int i;
	u8 hash = 0;

	for (i = 0; i < hash_size; i++) {
		hash ^= hash_start[i];
	}

	return hash;
}

/*********************** tlb specific functions ***************************/

static inline void _lock_tx_hashtbl(struct bonding *bond)
{
131
	spin_lock_bh(&(BOND_ALB_INFO(bond).tx_hashtbl_lock));
L
Linus Torvalds 已提交
132 133 134 135
}

static inline void _unlock_tx_hashtbl(struct bonding *bond)
{
136
	spin_unlock_bh(&(BOND_ALB_INFO(bond).tx_hashtbl_lock));
L
Linus Torvalds 已提交
137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169
}

/* Caller must hold tx_hashtbl lock */
static inline void tlb_init_table_entry(struct tlb_client_info *entry, int save_load)
{
	if (save_load) {
		entry->load_history = 1 + entry->tx_bytes /
				      BOND_TLB_REBALANCE_INTERVAL;
		entry->tx_bytes = 0;
	}

	entry->tx_slave = NULL;
	entry->next = TLB_NULL_INDEX;
	entry->prev = TLB_NULL_INDEX;
}

static inline void tlb_init_slave(struct slave *slave)
{
	SLAVE_TLB_INFO(slave).load = 0;
	SLAVE_TLB_INFO(slave).head = TLB_NULL_INDEX;
}

/* Caller must hold bond lock for read */
static void tlb_clear_slave(struct bonding *bond, struct slave *slave, int save_load)
{
	struct tlb_client_info *tx_hash_table;
	u32 index;

	_lock_tx_hashtbl(bond);

	/* clear slave from tx_hashtbl */
	tx_hash_table = BOND_ALB_INFO(bond).tx_hashtbl;

170 171 172 173 174 175 176 177
	/* skip this if we've already freed the tx hash table */
	if (tx_hash_table) {
		index = SLAVE_TLB_INFO(slave).head;
		while (index != TLB_NULL_INDEX) {
			u32 next_index = tx_hash_table[index].next;
			tlb_init_table_entry(&tx_hash_table[index], save_load);
			index = next_index;
		}
L
Linus Torvalds 已提交
178 179 180
	}

	tlb_init_slave(slave);
181 182

	_unlock_tx_hashtbl(bond);
L
Linus Torvalds 已提交
183 184 185 186 187 188 189
}

/* Must be called before starting the monitor timer */
static int tlb_initialize(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	int size = TLB_HASH_TABLE_SIZE * sizeof(struct tlb_client_info);
190
	struct tlb_client_info *new_hashtbl;
L
Linus Torvalds 已提交
191 192 193 194
	int i;

	spin_lock_init(&(bond_info->tx_hashtbl_lock));

195
	new_hashtbl = kzalloc(size, GFP_KERNEL);
196
	if (!new_hashtbl) {
L
Linus Torvalds 已提交
197
		printk(KERN_ERR DRV_NAME
198
		       ": %s: Error: Failed to allocate TLB hash table\n",
L
Linus Torvalds 已提交
199 200 201
		       bond->dev->name);
		return -1;
	}
202 203 204
	_lock_tx_hashtbl(bond);

	bond_info->tx_hashtbl = new_hashtbl;
L
Linus Torvalds 已提交
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310

	for (i = 0; i < TLB_HASH_TABLE_SIZE; i++) {
		tlb_init_table_entry(&bond_info->tx_hashtbl[i], 1);
	}

	_unlock_tx_hashtbl(bond);

	return 0;
}

/* Must be called only after all slaves have been released */
static void tlb_deinitialize(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));

	_lock_tx_hashtbl(bond);

	kfree(bond_info->tx_hashtbl);
	bond_info->tx_hashtbl = NULL;

	_unlock_tx_hashtbl(bond);
}

/* Caller must hold bond lock for read */
static struct slave *tlb_get_least_loaded_slave(struct bonding *bond)
{
	struct slave *slave, *least_loaded;
	s64 max_gap;
	int i, found = 0;

	/* Find the first enabled slave */
	bond_for_each_slave(bond, slave, i) {
		if (SLAVE_IS_OK(slave)) {
			found = 1;
			break;
		}
	}

	if (!found) {
		return NULL;
	}

	least_loaded = slave;
	max_gap = (s64)(slave->speed << 20) - /* Convert to Megabit per sec */
			(s64)(SLAVE_TLB_INFO(slave).load << 3); /* Bytes to bits */

	/* Find the slave with the largest gap */
	bond_for_each_slave_from(bond, slave, i, least_loaded) {
		if (SLAVE_IS_OK(slave)) {
			s64 gap = (s64)(slave->speed << 20) -
					(s64)(SLAVE_TLB_INFO(slave).load << 3);
			if (max_gap < gap) {
				least_loaded = slave;
				max_gap = gap;
			}
		}
	}

	return least_loaded;
}

/* Caller must hold bond lock for read */
static struct slave *tlb_choose_channel(struct bonding *bond, u32 hash_index, u32 skb_len)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct tlb_client_info *hash_table;
	struct slave *assigned_slave;

	_lock_tx_hashtbl(bond);

	hash_table = bond_info->tx_hashtbl;
	assigned_slave = hash_table[hash_index].tx_slave;
	if (!assigned_slave) {
		assigned_slave = tlb_get_least_loaded_slave(bond);

		if (assigned_slave) {
			struct tlb_slave_info *slave_info =
				&(SLAVE_TLB_INFO(assigned_slave));
			u32 next_index = slave_info->head;

			hash_table[hash_index].tx_slave = assigned_slave;
			hash_table[hash_index].next = next_index;
			hash_table[hash_index].prev = TLB_NULL_INDEX;

			if (next_index != TLB_NULL_INDEX) {
				hash_table[next_index].prev = hash_index;
			}

			slave_info->head = hash_index;
			slave_info->load +=
				hash_table[hash_index].load_history;
		}
	}

	if (assigned_slave) {
		hash_table[hash_index].tx_bytes += skb_len;
	}

	_unlock_tx_hashtbl(bond);

	return assigned_slave;
}

/*********************** rlb specific functions ***************************/
static inline void _lock_rx_hashtbl(struct bonding *bond)
{
311
	spin_lock_bh(&(BOND_ALB_INFO(bond).rx_hashtbl_lock));
L
Linus Torvalds 已提交
312 313 314 315
}

static inline void _unlock_rx_hashtbl(struct bonding *bond)
{
316
	spin_unlock_bh(&(BOND_ALB_INFO(bond).rx_hashtbl_lock));
L
Linus Torvalds 已提交
317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344
}

/* when an ARP REPLY is received from a client update its info
 * in the rx_hashtbl
 */
static void rlb_update_entry_from_arp(struct bonding *bond, struct arp_pkt *arp)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct rlb_client_info *client_info;
	u32 hash_index;

	_lock_rx_hashtbl(bond);

	hash_index = _simple_hash((u8*)&(arp->ip_src), sizeof(arp->ip_src));
	client_info = &(bond_info->rx_hashtbl[hash_index]);

	if ((client_info->assigned) &&
	    (client_info->ip_src == arp->ip_dst) &&
	    (client_info->ip_dst == arp->ip_src)) {
		/* update the clients MAC address */
		memcpy(client_info->mac_dst, arp->mac_src, ETH_ALEN);
		client_info->ntt = 1;
		bond_info->rx_ntt = 1;
	}

	_unlock_rx_hashtbl(bond);
}

D
David S. Miller 已提交
345
static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct packet_type *ptype, struct net_device *orig_dev)
L
Linus Torvalds 已提交
346
{
347
	struct bonding *bond;
L
Linus Torvalds 已提交
348 349 350
	struct arp_pkt *arp = (struct arp_pkt *)skb->data;
	int res = NET_RX_DROP;

351
	if (dev_net(bond_dev) != &init_net)
352 353
		goto out;

354 355 356 357 358
	while (bond_dev->priv_flags & IFF_802_1Q_VLAN)
		bond_dev = vlan_dev_real_dev(bond_dev);

	if (!(bond_dev->priv_flags & IFF_BONDING) ||
	    !(bond_dev->flags & IFF_MASTER))
L
Linus Torvalds 已提交
359 360 361
		goto out;

	if (!arp) {
362
		pr_debug("Packet has no ARP data\n");
L
Linus Torvalds 已提交
363 364 365 366
		goto out;
	}

	if (skb->len < sizeof(struct arp_pkt)) {
367
		pr_debug("Packet is too small to be an ARP\n");
L
Linus Torvalds 已提交
368 369 370 371 372
		goto out;
	}

	if (arp->op_code == htons(ARPOP_REPLY)) {
		/* update rx hash table for this ARP */
373 374
		printk("rar: update orig %s bond_dev %s\n", orig_dev->name,
		       bond_dev->name);
375
		bond = netdev_priv(bond_dev);
L
Linus Torvalds 已提交
376
		rlb_update_entry_from_arp(bond, arp);
377
		pr_debug("Server received an ARP Reply from client\n");
L
Linus Torvalds 已提交
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431
	}

	res = NET_RX_SUCCESS;

out:
	dev_kfree_skb(skb);

	return res;
}

/* Caller must hold bond lock for read */
static struct slave *rlb_next_rx_slave(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct slave *rx_slave, *slave, *start_at;
	int i = 0;

	if (bond_info->next_rx_slave) {
		start_at = bond_info->next_rx_slave;
	} else {
		start_at = bond->first_slave;
	}

	rx_slave = NULL;

	bond_for_each_slave_from(bond, slave, i, start_at) {
		if (SLAVE_IS_OK(slave)) {
			if (!rx_slave) {
				rx_slave = slave;
			} else if (slave->speed > rx_slave->speed) {
				rx_slave = slave;
			}
		}
	}

	if (rx_slave) {
		bond_info->next_rx_slave = rx_slave->next;
	}

	return rx_slave;
}

/* teach the switch the mac of a disabled slave
 * on the primary for fault tolerance
 *
 * Caller must hold bond->curr_slave_lock for write or bond lock for write
 */
static void rlb_teach_disabled_mac_on_primary(struct bonding *bond, u8 addr[])
{
	if (!bond->curr_active_slave) {
		return;
	}

	if (!bond->alb_info.primary_is_promisc) {
432 433 434 435
		if (!dev_set_promiscuity(bond->curr_active_slave->dev, 1))
			bond->alb_info.primary_is_promisc = 1;
		else
			bond->alb_info.primary_is_promisc = 0;
L
Linus Torvalds 已提交
436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486
	}

	bond->alb_info.rlb_promisc_timeout_counter = 0;

	alb_send_learning_packets(bond->curr_active_slave, addr);
}

/* slave being removed should not be active at this point
 *
 * Caller must hold bond lock for read
 */
static void rlb_clear_slave(struct bonding *bond, struct slave *slave)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct rlb_client_info *rx_hash_table;
	u32 index, next_index;

	/* clear slave from rx_hashtbl */
	_lock_rx_hashtbl(bond);

	rx_hash_table = bond_info->rx_hashtbl;
	index = bond_info->rx_hashtbl_head;
	for (; index != RLB_NULL_INDEX; index = next_index) {
		next_index = rx_hash_table[index].next;
		if (rx_hash_table[index].slave == slave) {
			struct slave *assigned_slave = rlb_next_rx_slave(bond);

			if (assigned_slave) {
				rx_hash_table[index].slave = assigned_slave;
				if (memcmp(rx_hash_table[index].mac_dst,
					   mac_bcast, ETH_ALEN)) {
					bond_info->rx_hashtbl[index].ntt = 1;
					bond_info->rx_ntt = 1;
					/* A slave has been removed from the
					 * table because it is either disabled
					 * or being released. We must retry the
					 * update to avoid clients from not
					 * being updated & disconnecting when
					 * there is stress
					 */
					bond_info->rlb_update_retry_counter =
						RLB_UPDATE_RETRY;
				}
			} else {  /* there is no active slave */
				rx_hash_table[index].slave = NULL;
			}
		}
	}

	_unlock_rx_hashtbl(bond);

487
	write_lock_bh(&bond->curr_slave_lock);
L
Linus Torvalds 已提交
488 489 490 491 492

	if (slave != bond->curr_active_slave) {
		rlb_teach_disabled_mac_on_primary(bond, slave->dev->dev_addr);
	}

493
	write_unlock_bh(&bond->curr_slave_lock);
L
Linus Torvalds 已提交
494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
}

static void rlb_update_client(struct rlb_client_info *client_info)
{
	int i;

	if (!client_info->slave) {
		return;
	}

	for (i = 0; i < RLB_ARP_BURST_SIZE; i++) {
		struct sk_buff *skb;

		skb = arp_create(ARPOP_REPLY, ETH_P_ARP,
				 client_info->ip_dst,
				 client_info->slave->dev,
				 client_info->ip_src,
				 client_info->mac_dst,
				 client_info->slave->dev->dev_addr,
				 client_info->mac_dst);
		if (!skb) {
			printk(KERN_ERR DRV_NAME
516 517
			       ": %s: Error: failed to create an ARP packet\n",
			       client_info->slave->dev->master->name);
L
Linus Torvalds 已提交
518 519 520 521 522 523 524 525 526
			continue;
		}

		skb->dev = client_info->slave->dev;

		if (client_info->tag) {
			skb = vlan_put_tag(skb, client_info->vlan_id);
			if (!skb) {
				printk(KERN_ERR DRV_NAME
527 528
				       ": %s: Error: failed to insert VLAN tag\n",
				       client_info->slave->dev->master->name);
L
Linus Torvalds 已提交
529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
				continue;
			}
		}

		arp_xmit(skb);
	}
}

/* sends ARP REPLIES that update the clients that need updating */
static void rlb_update_rx_clients(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct rlb_client_info *client_info;
	u32 hash_index;

	_lock_rx_hashtbl(bond);

	hash_index = bond_info->rx_hashtbl_head;
	for (; hash_index != RLB_NULL_INDEX; hash_index = client_info->next) {
		client_info = &(bond_info->rx_hashtbl[hash_index]);
		if (client_info->ntt) {
			rlb_update_client(client_info);
			if (bond_info->rlb_update_retry_counter == 0) {
				client_info->ntt = 0;
			}
		}
	}

	/* do not update the entries again untill this counter is zero so that
	 * not to confuse the clients.
	 */
	bond_info->rlb_update_delay_counter = RLB_UPDATE_DELAY;

	_unlock_rx_hashtbl(bond);
}

/* The slave was assigned a new mac address - update the clients */
static void rlb_req_update_slave_clients(struct bonding *bond, struct slave *slave)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct rlb_client_info *client_info;
	int ntt = 0;
	u32 hash_index;

	_lock_rx_hashtbl(bond);

	hash_index = bond_info->rx_hashtbl_head;
	for (; hash_index != RLB_NULL_INDEX; hash_index = client_info->next) {
		client_info = &(bond_info->rx_hashtbl[hash_index]);

		if ((client_info->slave == slave) &&
		    memcmp(client_info->mac_dst, mac_bcast, ETH_ALEN)) {
			client_info->ntt = 1;
			ntt = 1;
		}
	}

	// update the team's flag only after the whole iteration
	if (ntt) {
		bond_info->rx_ntt = 1;
		//fasten the change
		bond_info->rlb_update_retry_counter = RLB_UPDATE_RETRY;
	}

	_unlock_rx_hashtbl(bond);
}

/* mark all clients using src_ip to be updated */
597
static void rlb_req_update_subnet_clients(struct bonding *bond, __be32 src_ip)
L
Linus Torvalds 已提交
598 599 600 601 602 603 604 605 606 607 608 609 610
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct rlb_client_info *client_info;
	u32 hash_index;

	_lock_rx_hashtbl(bond);

	hash_index = bond_info->rx_hashtbl_head;
	for (; hash_index != RLB_NULL_INDEX; hash_index = client_info->next) {
		client_info = &(bond_info->rx_hashtbl[hash_index]);

		if (!client_info->slave) {
			printk(KERN_ERR DRV_NAME
611 612 613
			       ": %s: Error: found a client with no channel in "
			       "the client's hash table\n",
			       bond->dev->name);
L
Linus Torvalds 已提交
614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635
			continue;
		}
		/*update all clients using this src_ip, that are not assigned
		 * to the team's address (curr_active_slave) and have a known
		 * unicast mac address.
		 */
		if ((client_info->ip_src == src_ip) &&
		    memcmp(client_info->slave->dev->dev_addr,
			   bond->dev->dev_addr, ETH_ALEN) &&
		    memcmp(client_info->mac_dst, mac_bcast, ETH_ALEN)) {
			client_info->ntt = 1;
			bond_info->rx_ntt = 1;
		}
	}

	_unlock_rx_hashtbl(bond);
}

/* Caller must hold both bond and ptr locks for read */
static struct slave *rlb_choose_channel(struct sk_buff *skb, struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
636
	struct arp_pkt *arp = arp_pkt(skb);
L
Linus Torvalds 已提交
637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692
	struct slave *assigned_slave;
	struct rlb_client_info *client_info;
	u32 hash_index = 0;

	_lock_rx_hashtbl(bond);

	hash_index = _simple_hash((u8 *)&arp->ip_dst, sizeof(arp->ip_src));
	client_info = &(bond_info->rx_hashtbl[hash_index]);

	if (client_info->assigned) {
		if ((client_info->ip_src == arp->ip_src) &&
		    (client_info->ip_dst == arp->ip_dst)) {
			/* the entry is already assigned to this client */
			if (memcmp(arp->mac_dst, mac_bcast, ETH_ALEN)) {
				/* update mac address from arp */
				memcpy(client_info->mac_dst, arp->mac_dst, ETH_ALEN);
			}

			assigned_slave = client_info->slave;
			if (assigned_slave) {
				_unlock_rx_hashtbl(bond);
				return assigned_slave;
			}
		} else {
			/* the entry is already assigned to some other client,
			 * move the old client to primary (curr_active_slave) so
			 * that the new client can be assigned to this entry.
			 */
			if (bond->curr_active_slave &&
			    client_info->slave != bond->curr_active_slave) {
				client_info->slave = bond->curr_active_slave;
				rlb_update_client(client_info);
			}
		}
	}
	/* assign a new slave */
	assigned_slave = rlb_next_rx_slave(bond);

	if (assigned_slave) {
		client_info->ip_src = arp->ip_src;
		client_info->ip_dst = arp->ip_dst;
		/* arp->mac_dst is broadcast for arp reqeusts.
		 * will be updated with clients actual unicast mac address
		 * upon receiving an arp reply.
		 */
		memcpy(client_info->mac_dst, arp->mac_dst, ETH_ALEN);
		client_info->slave = assigned_slave;

		if (memcmp(client_info->mac_dst, mac_bcast, ETH_ALEN)) {
			client_info->ntt = 1;
			bond->alb_info.rx_ntt = 1;
		} else {
			client_info->ntt = 0;
		}

		if (!list_empty(&bond->vlan_list)) {
693
			if (!vlan_get_tag(skb, &client_info->vlan_id))
L
Linus Torvalds 已提交
694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719
				client_info->tag = 1;
		}

		if (!client_info->assigned) {
			u32 prev_tbl_head = bond_info->rx_hashtbl_head;
			bond_info->rx_hashtbl_head = hash_index;
			client_info->next = prev_tbl_head;
			if (prev_tbl_head != RLB_NULL_INDEX) {
				bond_info->rx_hashtbl[prev_tbl_head].prev =
					hash_index;
			}
			client_info->assigned = 1;
		}
	}

	_unlock_rx_hashtbl(bond);

	return assigned_slave;
}

/* chooses (and returns) transmit channel for arp reply
 * does not choose channel for other arp types since they are
 * sent on the curr_active_slave
 */
static struct slave *rlb_arp_xmit(struct sk_buff *skb, struct bonding *bond)
{
720
	struct arp_pkt *arp = arp_pkt(skb);
L
Linus Torvalds 已提交
721 722
	struct slave *tx_slave = NULL;

723
	if (arp->op_code == htons(ARPOP_REPLY)) {
L
Linus Torvalds 已提交
724 725 726 727 728 729 730
		/* the arp must be sent on the selected
		* rx channel
		*/
		tx_slave = rlb_choose_channel(skb, bond);
		if (tx_slave) {
			memcpy(arp->mac_src,tx_slave->dev->dev_addr, ETH_ALEN);
		}
731
		pr_debug("Server sent ARP Reply packet\n");
732
	} else if (arp->op_code == htons(ARPOP_REQUEST)) {
L
Linus Torvalds 已提交
733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750
		/* Create an entry in the rx_hashtbl for this client as a
		 * place holder.
		 * When the arp reply is received the entry will be updated
		 * with the correct unicast address of the client.
		 */
		rlb_choose_channel(skb, bond);

		/* The ARP relpy packets must be delayed so that
		 * they can cancel out the influence of the ARP request.
		 */
		bond->alb_info.rlb_update_delay_counter = RLB_UPDATE_DELAY;

		/* arp requests are broadcast and are sent on the primary
		 * the arp request will collapse all clients on the subnet to
		 * the primary slave. We must register these clients to be
		 * updated with their assigned mac.
		 */
		rlb_req_update_subnet_clients(bond, arp->ip_src);
751
		pr_debug("Server sent ARP Request packet\n");
L
Linus Torvalds 已提交
752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798
	}

	return tx_slave;
}

/* Caller must hold bond lock for read */
static void rlb_rebalance(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct slave *assigned_slave;
	struct rlb_client_info *client_info;
	int ntt;
	u32 hash_index;

	_lock_rx_hashtbl(bond);

	ntt = 0;
	hash_index = bond_info->rx_hashtbl_head;
	for (; hash_index != RLB_NULL_INDEX; hash_index = client_info->next) {
		client_info = &(bond_info->rx_hashtbl[hash_index]);
		assigned_slave = rlb_next_rx_slave(bond);
		if (assigned_slave && (client_info->slave != assigned_slave)) {
			client_info->slave = assigned_slave;
			client_info->ntt = 1;
			ntt = 1;
		}
	}

	/* update the team's flag only after the whole iteration */
	if (ntt) {
		bond_info->rx_ntt = 1;
	}
	_unlock_rx_hashtbl(bond);
}

/* Caller must hold rx_hashtbl lock */
static void rlb_init_table_entry(struct rlb_client_info *entry)
{
	memset(entry, 0, sizeof(struct rlb_client_info));
	entry->next = RLB_NULL_INDEX;
	entry->prev = RLB_NULL_INDEX;
}

static int rlb_initialize(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct packet_type *pk_type = &(BOND_ALB_INFO(bond).rlb_pkt_type);
799
	struct rlb_client_info	*new_hashtbl;
L
Linus Torvalds 已提交
800 801 802 803 804
	int size = RLB_HASH_TABLE_SIZE * sizeof(struct rlb_client_info);
	int i;

	spin_lock_init(&(bond_info->rx_hashtbl_lock));

805 806
	new_hashtbl = kmalloc(size, GFP_KERNEL);
	if (!new_hashtbl) {
L
Linus Torvalds 已提交
807
		printk(KERN_ERR DRV_NAME
808
		       ": %s: Error: Failed to allocate RLB hash table\n",
L
Linus Torvalds 已提交
809 810 811
		       bond->dev->name);
		return -1;
	}
812 813 814
	_lock_rx_hashtbl(bond);

	bond_info->rx_hashtbl = new_hashtbl;
L
Linus Torvalds 已提交
815 816 817 818 819 820 821 822 823 824

	bond_info->rx_hashtbl_head = RLB_NULL_INDEX;

	for (i = 0; i < RLB_HASH_TABLE_SIZE; i++) {
		rlb_init_table_entry(bond_info->rx_hashtbl + i);
	}

	_unlock_rx_hashtbl(bond);

	/*initialize packet type*/
825
	pk_type->type = cpu_to_be16(ETH_P_ARP);
826
	pk_type->dev = NULL;
L
Linus Torvalds 已提交
827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
	pk_type->func = rlb_arp_recv;

	/* register to receive ARPs */
	dev_add_pack(pk_type);

	return 0;
}

static void rlb_deinitialize(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));

	dev_remove_pack(&(bond_info->rlb_pkt_type));

	_lock_rx_hashtbl(bond);

	kfree(bond_info->rx_hashtbl);
	bond_info->rx_hashtbl = NULL;
	bond_info->rx_hashtbl_head = RLB_NULL_INDEX;

	_unlock_rx_hashtbl(bond);
}

static void rlb_clear_vlan(struct bonding *bond, unsigned short vlan_id)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	u32 curr_index;

	_lock_rx_hashtbl(bond);

	curr_index = bond_info->rx_hashtbl_head;
	while (curr_index != RLB_NULL_INDEX) {
		struct rlb_client_info *curr = &(bond_info->rx_hashtbl[curr_index]);
		u32 next_index = bond_info->rx_hashtbl[curr_index].next;
		u32 prev_index = bond_info->rx_hashtbl[curr_index].prev;

		if (curr->tag && (curr->vlan_id == vlan_id)) {
			if (curr_index == bond_info->rx_hashtbl_head) {
				bond_info->rx_hashtbl_head = next_index;
			}
			if (prev_index != RLB_NULL_INDEX) {
				bond_info->rx_hashtbl[prev_index].next = next_index;
			}
			if (next_index != RLB_NULL_INDEX) {
				bond_info->rx_hashtbl[next_index].prev = prev_index;
			}

			rlb_init_table_entry(curr);
		}

		curr_index = next_index;
	}

	_unlock_rx_hashtbl(bond);
}

/*********************** tlb/rlb shared functions *********************/

static void alb_send_learning_packets(struct slave *slave, u8 mac_addr[])
{
	struct bonding *bond = bond_get_bond_by_slave(slave);
	struct learning_pkt pkt;
	int size = sizeof(struct learning_pkt);
	int i;

	memset(&pkt, 0, size);
	memcpy(pkt.mac_dst, mac_addr, ETH_ALEN);
	memcpy(pkt.mac_src, mac_addr, ETH_ALEN);
895
	pkt.type = cpu_to_be16(ETH_P_LOOP);
L
Linus Torvalds 已提交
896 897 898 899 900 901 902 903 904 905 906 907 908

	for (i = 0; i < MAX_LP_BURST; i++) {
		struct sk_buff *skb;
		char *data;

		skb = dev_alloc_skb(size);
		if (!skb) {
			return;
		}

		data = skb_put(skb, size);
		memcpy(data, &pkt, size);

909
		skb_reset_mac_header(skb);
910
		skb->network_header = skb->mac_header + ETH_HLEN;
L
Linus Torvalds 已提交
911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929
		skb->protocol = pkt.type;
		skb->priority = TC_PRIO_CONTROL;
		skb->dev = slave->dev;

		if (!list_empty(&bond->vlan_list)) {
			struct vlan_entry *vlan;

			vlan = bond_next_vlan(bond,
					      bond->alb_info.current_alb_vlan);

			bond->alb_info.current_alb_vlan = vlan;
			if (!vlan) {
				kfree_skb(skb);
				continue;
			}

			skb = vlan_put_tag(skb, vlan->vlan_id);
			if (!skb) {
				printk(KERN_ERR DRV_NAME
930 931
				       ": %s: Error: failed to insert VLAN tag\n",
				       bond->dev->name);
L
Linus Torvalds 已提交
932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959
				continue;
			}
		}

		dev_queue_xmit(skb);
	}
}

/* hw is a boolean parameter that determines whether we should try and
 * set the hw address of the device as well as the hw address of the
 * net_device
 */
static int alb_set_slave_mac_addr(struct slave *slave, u8 addr[], int hw)
{
	struct net_device *dev = slave->dev;
	struct sockaddr s_addr;

	if (!hw) {
		memcpy(dev->dev_addr, addr, dev->addr_len);
		return 0;
	}

	/* for rlb each slave must have a unique hw mac addresses so that */
	/* each slave will receive packets destined to a different mac */
	memcpy(s_addr.sa_data, addr, dev->addr_len);
	s_addr.sa_family = dev->type;
	if (dev_set_mac_address(dev, &s_addr)) {
		printk(KERN_ERR DRV_NAME
960
		       ": %s: Error: dev_set_mac_address of dev %s failed! ALB "
L
Linus Torvalds 已提交
961 962 963
		       "mode requires that the base driver support setting "
		       "the hw address also when the network device's "
		       "interface is open\n",
964
		       dev->master->name, dev->name);
L
Linus Torvalds 已提交
965 966 967 968 969
		return -EOPNOTSUPP;
	}
	return 0;
}

970 971 972 973 974 975 976
/*
 * Swap MAC addresses between two slaves.
 *
 * Called with RTNL held, and no other locks.
 *
 */

L
Linus Torvalds 已提交
977 978 979 980 981 982 983 984
static void alb_swap_mac_addr(struct bonding *bond, struct slave *slave1, struct slave *slave2)
{
	u8 tmp_mac_addr[ETH_ALEN];

	memcpy(tmp_mac_addr, slave1->dev->dev_addr, ETH_ALEN);
	alb_set_slave_mac_addr(slave1, slave2->dev->dev_addr, bond->alb_info.rlb_enabled);
	alb_set_slave_mac_addr(slave2, tmp_mac_addr, bond->alb_info.rlb_enabled);

985 986 987 988 989
}

/*
 * Send learning packets after MAC address swap.
 *
990
 * Called with RTNL and no other locks
991 992 993 994 995 996 997
 */
static void alb_fasten_mac_swap(struct bonding *bond, struct slave *slave1,
				struct slave *slave2)
{
	int slaves_state_differ = (SLAVE_IS_OK(slave1) != SLAVE_IS_OK(slave2));
	struct slave *disabled_slave = NULL;

998 999
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043
	/* fasten the change in the switch */
	if (SLAVE_IS_OK(slave1)) {
		alb_send_learning_packets(slave1, slave1->dev->dev_addr);
		if (bond->alb_info.rlb_enabled) {
			/* inform the clients that the mac address
			 * has changed
			 */
			rlb_req_update_slave_clients(bond, slave1);
		}
	} else {
		disabled_slave = slave1;
	}

	if (SLAVE_IS_OK(slave2)) {
		alb_send_learning_packets(slave2, slave2->dev->dev_addr);
		if (bond->alb_info.rlb_enabled) {
			/* inform the clients that the mac address
			 * has changed
			 */
			rlb_req_update_slave_clients(bond, slave2);
		}
	} else {
		disabled_slave = slave2;
	}

	if (bond->alb_info.rlb_enabled && slaves_state_differ) {
		/* A disabled slave was assigned an active mac addr */
		rlb_teach_disabled_mac_on_primary(bond,
						  disabled_slave->dev->dev_addr);
	}
}

/**
 * alb_change_hw_addr_on_detach
 * @bond: bonding we're working on
 * @slave: the slave that was just detached
 *
 * We assume that @slave was already detached from the slave list.
 *
 * If @slave's permanent hw address is different both from its current
 * address and from @bond's address, then somewhere in the bond there's
 * a slave that has @slave's permanet address as its current address.
 * We'll make sure that that slave no longer uses @slave's permanent address.
 *
1044
 * Caller must hold RTNL and no other locks
L
Linus Torvalds 已提交
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 void alb_change_hw_addr_on_detach(struct bonding *bond, struct slave *slave)
{
	int perm_curr_diff;
	int perm_bond_diff;

	perm_curr_diff = memcmp(slave->perm_hwaddr,
				slave->dev->dev_addr,
				ETH_ALEN);
	perm_bond_diff = memcmp(slave->perm_hwaddr,
				bond->dev->dev_addr,
				ETH_ALEN);

	if (perm_curr_diff && perm_bond_diff) {
		struct slave *tmp_slave;
		int i, found = 0;

		bond_for_each_slave(bond, tmp_slave, i) {
			if (!memcmp(slave->perm_hwaddr,
				    tmp_slave->dev->dev_addr,
				    ETH_ALEN)) {
				found = 1;
				break;
			}
		}

		if (found) {
1072
			/* locking: needs RTNL and nothing else */
L
Linus Torvalds 已提交
1073
			alb_swap_mac_addr(bond, slave, tmp_slave);
1074
			alb_fasten_mac_swap(bond, slave, tmp_slave);
L
Linus Torvalds 已提交
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127
		}
	}
}

/**
 * alb_handle_addr_collision_on_attach
 * @bond: bonding we're working on
 * @slave: the slave that was just attached
 *
 * checks uniqueness of slave's mac address and handles the case the
 * new slave uses the bonds mac address.
 *
 * If the permanent hw address of @slave is @bond's hw address, we need to
 * find a different hw address to give @slave, that isn't in use by any other
 * slave in the bond. This address must be, of course, one of the premanent
 * addresses of the other slaves.
 *
 * We go over the slave list, and for each slave there we compare its
 * permanent hw address with the current address of all the other slaves.
 * If no match was found, then we've found a slave with a permanent address
 * that isn't used by any other slave in the bond, so we can assign it to
 * @slave.
 *
 * assumption: this function is called before @slave is attached to the
 * 	       bond slave list.
 *
 * caller must hold the bond lock for write since the mac addresses are compared
 * and may be swapped.
 */
static int alb_handle_addr_collision_on_attach(struct bonding *bond, struct slave *slave)
{
	struct slave *tmp_slave1, *tmp_slave2, *free_mac_slave;
	struct slave *has_bond_addr = bond->curr_active_slave;
	int i, j, found = 0;

	if (bond->slave_cnt == 0) {
		/* this is the first slave */
		return 0;
	}

	/* if slave's mac address differs from bond's mac address
	 * check uniqueness of slave's mac address against the other
	 * slaves in the bond.
	 */
	if (memcmp(slave->perm_hwaddr, bond->dev->dev_addr, ETH_ALEN)) {
		bond_for_each_slave(bond, tmp_slave1, i) {
			if (!memcmp(tmp_slave1->dev->dev_addr, slave->dev->dev_addr,
				    ETH_ALEN)) {
				found = 1;
				break;
			}
		}

1128 1129
		if (!found)
			return 0;
L
Linus Torvalds 已提交
1130

1131 1132 1133 1134
		/* Try setting slave mac to bond address and fall-through
		   to code handling that situation below... */
		alb_set_slave_mac_addr(slave, bond->dev->dev_addr,
				       bond->alb_info.rlb_enabled);
L
Linus Torvalds 已提交
1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175
	}

	/* The slave's address is equal to the address of the bond.
	 * Search for a spare address in the bond for this slave.
	 */
	free_mac_slave = NULL;

	bond_for_each_slave(bond, tmp_slave1, i) {
		found = 0;
		bond_for_each_slave(bond, tmp_slave2, j) {
			if (!memcmp(tmp_slave1->perm_hwaddr,
				    tmp_slave2->dev->dev_addr,
				    ETH_ALEN)) {
				found = 1;
				break;
			}
		}

		if (!found) {
			/* no slave has tmp_slave1's perm addr
			 * as its curr addr
			 */
			free_mac_slave = tmp_slave1;
			break;
		}

		if (!has_bond_addr) {
			if (!memcmp(tmp_slave1->dev->dev_addr,
				    bond->dev->dev_addr,
				    ETH_ALEN)) {

				has_bond_addr = tmp_slave1;
			}
		}
	}

	if (free_mac_slave) {
		alb_set_slave_mac_addr(slave, free_mac_slave->perm_hwaddr,
				       bond->alb_info.rlb_enabled);

		printk(KERN_WARNING DRV_NAME
1176
		       ": %s: Warning: the hw address of slave %s is in use by "
L
Linus Torvalds 已提交
1177
		       "the bond; giving it the hw address of %s\n",
1178
		       bond->dev->name, slave->dev->name, free_mac_slave->dev->name);
L
Linus Torvalds 已提交
1179 1180 1181

	} else if (has_bond_addr) {
		printk(KERN_ERR DRV_NAME
1182
		       ": %s: Error: the hw address of slave %s is in use by the "
L
Linus Torvalds 已提交
1183 1184
		       "bond; couldn't find a slave with a free hw address to "
		       "give it (this should not have happened)\n",
1185
		       bond->dev->name, slave->dev->name);
L
Linus Torvalds 已提交
1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226
		return -EFAULT;
	}

	return 0;
}

/**
 * alb_set_mac_address
 * @bond:
 * @addr:
 *
 * In TLB mode all slaves are configured to the bond's hw address, but set
 * their dev_addr field to different addresses (based on their permanent hw
 * addresses).
 *
 * For each slave, this function sets the interface to the new address and then
 * changes its dev_addr field to its previous value.
 *
 * Unwinding assumes bond's mac address has not yet changed.
 */
static int alb_set_mac_address(struct bonding *bond, void *addr)
{
	struct sockaddr sa;
	struct slave *slave, *stop_at;
	char tmp_addr[ETH_ALEN];
	int res;
	int i;

	if (bond->alb_info.rlb_enabled) {
		return 0;
	}

	bond_for_each_slave(bond, slave, i) {
		/* save net_device's current hw address */
		memcpy(tmp_addr, slave->dev->dev_addr, ETH_ALEN);

		res = dev_set_mac_address(slave->dev, addr);

		/* restore net_device's hw address */
		memcpy(slave->dev->dev_addr, tmp_addr, ETH_ALEN);

1227
		if (res)
L
Linus Torvalds 已提交
1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266
			goto unwind;
	}

	return 0;

unwind:
	memcpy(sa.sa_data, bond->dev->dev_addr, bond->dev->addr_len);
	sa.sa_family = bond->dev->type;

	/* unwind from head to the slave that failed */
	stop_at = slave;
	bond_for_each_slave_from_to(bond, slave, i, bond->first_slave, stop_at) {
		memcpy(tmp_addr, slave->dev->dev_addr, ETH_ALEN);
		dev_set_mac_address(slave->dev, &sa);
		memcpy(slave->dev->dev_addr, tmp_addr, ETH_ALEN);
	}

	return res;
}

/************************ exported alb funcions ************************/

int bond_alb_initialize(struct bonding *bond, int rlb_enabled)
{
	int res;

	res = tlb_initialize(bond);
	if (res) {
		return res;
	}

	if (rlb_enabled) {
		bond->alb_info.rlb_enabled = 1;
		/* initialize rlb */
		res = rlb_initialize(bond);
		if (res) {
			tlb_deinitialize(bond);
			return res;
		}
1267 1268
	} else {
		bond->alb_info.rlb_enabled = 0;
L
Linus Torvalds 已提交
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286
	}

	return 0;
}

void bond_alb_deinitialize(struct bonding *bond)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));

	tlb_deinitialize(bond);

	if (bond_info->rlb_enabled) {
		rlb_deinitialize(bond);
	}
}

int bond_alb_xmit(struct sk_buff *skb, struct net_device *bond_dev)
{
1287
	struct bonding *bond = netdev_priv(bond_dev);
L
Linus Torvalds 已提交
1288 1289 1290
	struct ethhdr *eth_data;
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct slave *tx_slave = NULL;
1291
	static const __be32 ip_bcast = htonl(0xffffffff);
L
Linus Torvalds 已提交
1292 1293 1294
	int hash_size = 0;
	int do_tx_balance = 1;
	u32 hash_index = 0;
1295
	const u8 *hash_start = NULL;
L
Linus Torvalds 已提交
1296
	int res = 1;
1297
	struct ipv6hdr *ip6hdr;
L
Linus Torvalds 已提交
1298

1299
	skb_reset_mac_header(skb);
L
Linus Torvalds 已提交
1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
	eth_data = eth_hdr(skb);

	/* make sure that the curr_active_slave and the slaves list do
	 * not change during tx
	 */
	read_lock(&bond->lock);
	read_lock(&bond->curr_slave_lock);

	if (!BOND_IS_OK(bond)) {
		goto out;
	}

	switch (ntohs(skb->protocol)) {
1313 1314 1315
	case ETH_P_IP: {
		const struct iphdr *iph = ip_hdr(skb);

L
Linus Torvalds 已提交
1316
		if ((memcmp(eth_data->h_dest, mac_bcast, ETH_ALEN) == 0) ||
1317 1318
		    (iph->daddr == ip_bcast) ||
		    (iph->protocol == IPPROTO_IGMP)) {
L
Linus Torvalds 已提交
1319 1320 1321
			do_tx_balance = 0;
			break;
		}
1322 1323 1324
		hash_start = (char *)&(iph->daddr);
		hash_size = sizeof(iph->daddr);
	}
L
Linus Torvalds 已提交
1325 1326
		break;
	case ETH_P_IPV6:
1327 1328 1329
		/* IPv6 doesn't really use broadcast mac address, but leave
		 * that here just in case.
		 */
L
Linus Torvalds 已提交
1330 1331 1332
		if (memcmp(eth_data->h_dest, mac_bcast, ETH_ALEN) == 0) {
			do_tx_balance = 0;
			break;
1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350
		}

		/* IPv6 uses all-nodes multicast as an equivalent to
		 * broadcasts in IPv4.
		 */
		if (memcmp(eth_data->h_dest, mac_v6_allmcast, ETH_ALEN) == 0) {
			do_tx_balance = 0;
			break;
		}

		/* Additianally, DAD probes should not be tx-balanced as that
		 * will lead to false positives for duplicate addresses and
		 * prevent address configuration from working.
		 */
		ip6hdr = ipv6_hdr(skb);
		if (ipv6_addr_any(&ip6hdr->saddr)) {
			do_tx_balance = 0;
			break;
L
Linus Torvalds 已提交
1351 1352
		}

1353 1354
		hash_start = (char *)&(ipv6_hdr(skb)->daddr);
		hash_size = sizeof(ipv6_hdr(skb)->daddr);
L
Linus Torvalds 已提交
1355 1356
		break;
	case ETH_P_IPX:
1357
		if (ipx_hdr(skb)->ipx_checksum != IPX_NO_CHECKSUM) {
L
Linus Torvalds 已提交
1358 1359 1360 1361 1362 1363 1364 1365 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 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
			/* something is wrong with this packet */
			do_tx_balance = 0;
			break;
		}

		if (ipx_hdr(skb)->ipx_type != IPX_TYPE_NCP) {
			/* The only protocol worth balancing in
			 * this family since it has an "ARP" like
			 * mechanism
			 */
			do_tx_balance = 0;
			break;
		}

		hash_start = (char*)eth_data->h_dest;
		hash_size = ETH_ALEN;
		break;
	case ETH_P_ARP:
		do_tx_balance = 0;
		if (bond_info->rlb_enabled) {
			tx_slave = rlb_arp_xmit(skb, bond);
		}
		break;
	default:
		do_tx_balance = 0;
		break;
	}

	if (do_tx_balance) {
		hash_index = _simple_hash(hash_start, hash_size);
		tx_slave = tlb_choose_channel(bond, hash_index, skb->len);
	}

	if (!tx_slave) {
		/* unbalanced or unassigned, send through primary */
		tx_slave = bond->curr_active_slave;
		bond_info->unbalanced_load += skb->len;
	}

	if (tx_slave && SLAVE_IS_OK(tx_slave)) {
		if (tx_slave != bond->curr_active_slave) {
			memcpy(eth_data->h_source,
			       tx_slave->dev->dev_addr,
			       ETH_ALEN);
		}

		res = bond_dev_queue_xmit(bond, skb, tx_slave->dev);
	} else {
		if (tx_slave) {
			tlb_clear_slave(bond, tx_slave, 0);
		}
	}

out:
	if (res) {
		/* no suitable interface, frame not sent */
		dev_kfree_skb(skb);
	}
	read_unlock(&bond->curr_slave_lock);
	read_unlock(&bond->lock);
	return 0;
}

1421
void bond_alb_monitor(struct work_struct *work)
L
Linus Torvalds 已提交
1422
{
1423 1424
	struct bonding *bond = container_of(work, struct bonding,
					    alb_work.work);
L
Linus Torvalds 已提交
1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct slave *slave;
	int i;

	read_lock(&bond->lock);

	if (bond->kill_timers) {
		goto out;
	}

	if (bond->slave_cnt == 0) {
		bond_info->tx_rebalance_counter = 0;
		bond_info->lp_counter = 0;
		goto re_arm;
	}

	bond_info->tx_rebalance_counter++;
	bond_info->lp_counter++;

	/* send learning packets */
	if (bond_info->lp_counter >= BOND_ALB_LP_TICKS) {
		/* change of curr_active_slave involves swapping of mac addresses.
		 * in order to avoid this swapping from happening while
		 * sending the learning packets, the curr_slave_lock must be held for
		 * read.
		 */
		read_lock(&bond->curr_slave_lock);

		bond_for_each_slave(bond, slave, i) {
1454
			alb_send_learning_packets(slave, slave->dev->dev_addr);
L
Linus Torvalds 已提交
1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486
		}

		read_unlock(&bond->curr_slave_lock);

		bond_info->lp_counter = 0;
	}

	/* rebalance tx traffic */
	if (bond_info->tx_rebalance_counter >= BOND_TLB_REBALANCE_TICKS) {

		read_lock(&bond->curr_slave_lock);

		bond_for_each_slave(bond, slave, i) {
			tlb_clear_slave(bond, slave, 1);
			if (slave == bond->curr_active_slave) {
				SLAVE_TLB_INFO(slave).load =
					bond_info->unbalanced_load /
						BOND_TLB_REBALANCE_INTERVAL;
				bond_info->unbalanced_load = 0;
			}
		}

		read_unlock(&bond->curr_slave_lock);

		bond_info->tx_rebalance_counter = 0;
	}

	/* handle rlb stuff */
	if (bond_info->rlb_enabled) {
		if (bond_info->primary_is_promisc &&
		    (++bond_info->rlb_promisc_timeout_counter >= RLB_PROMISC_TIMEOUT)) {

1487 1488 1489 1490 1491 1492 1493
			/*
			 * dev_set_promiscuity requires rtnl and
			 * nothing else.
			 */
			read_unlock(&bond->lock);
			rtnl_lock();

L
Linus Torvalds 已提交
1494 1495 1496 1497 1498 1499 1500 1501 1502
			bond_info->rlb_promisc_timeout_counter = 0;

			/* If the primary was set to promiscuous mode
			 * because a slave was disabled then
			 * it can now leave promiscuous mode.
			 */
			dev_set_promiscuity(bond->curr_active_slave->dev, -1);
			bond_info->primary_is_promisc = 0;

1503 1504 1505
			rtnl_unlock();
			read_lock(&bond->lock);
		}
L
Linus Torvalds 已提交
1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527

		if (bond_info->rlb_rebalance) {
			bond_info->rlb_rebalance = 0;
			rlb_rebalance(bond);
		}

		/* check if clients need updating */
		if (bond_info->rx_ntt) {
			if (bond_info->rlb_update_delay_counter) {
				--bond_info->rlb_update_delay_counter;
			} else {
				rlb_update_rx_clients(bond);
				if (bond_info->rlb_update_retry_counter) {
					--bond_info->rlb_update_retry_counter;
				} else {
					bond_info->rx_ntt = 0;
				}
			}
		}
	}

re_arm:
1528
	queue_delayed_work(bond->wq, &bond->alb_work, alb_delta_in_ticks);
L
Linus Torvalds 已提交
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548
out:
	read_unlock(&bond->lock);
}

/* assumption: called before the slave is attached to the bond
 * and not locked by the bond lock
 */
int bond_alb_init_slave(struct bonding *bond, struct slave *slave)
{
	int res;

	res = alb_set_slave_mac_addr(slave, slave->perm_hwaddr,
				     bond->alb_info.rlb_enabled);
	if (res) {
		return res;
	}

	/* caller must hold the bond lock for write since the mac addresses
	 * are compared and may be swapped.
	 */
1549
	read_lock(&bond->lock);
L
Linus Torvalds 已提交
1550 1551 1552

	res = alb_handle_addr_collision_on_attach(bond, slave);

1553
	read_unlock(&bond->lock);
L
Linus Torvalds 已提交
1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570

	if (res) {
		return res;
	}

	tlb_init_slave(slave);

	/* order a rebalance ASAP */
	bond->alb_info.tx_rebalance_counter = BOND_TLB_REBALANCE_TICKS;

	if (bond->alb_info.rlb_enabled) {
		bond->alb_info.rlb_rebalance = 1;
	}

	return 0;
}

1571 1572 1573 1574 1575 1576
/*
 * Remove slave from tlb and rlb hash tables, and fix up MAC addresses
 * if necessary.
 *
 * Caller must hold RTNL and no other locks
 */
L
Linus Torvalds 已提交
1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622
void bond_alb_deinit_slave(struct bonding *bond, struct slave *slave)
{
	if (bond->slave_cnt > 1) {
		alb_change_hw_addr_on_detach(bond, slave);
	}

	tlb_clear_slave(bond, slave, 0);

	if (bond->alb_info.rlb_enabled) {
		bond->alb_info.next_rx_slave = NULL;
		rlb_clear_slave(bond, slave);
	}
}

/* Caller must hold bond lock for read */
void bond_alb_handle_link_change(struct bonding *bond, struct slave *slave, char link)
{
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));

	if (link == BOND_LINK_DOWN) {
		tlb_clear_slave(bond, slave, 0);
		if (bond->alb_info.rlb_enabled) {
			rlb_clear_slave(bond, slave);
		}
	} else if (link == BOND_LINK_UP) {
		/* order a rebalance ASAP */
		bond_info->tx_rebalance_counter = BOND_TLB_REBALANCE_TICKS;
		if (bond->alb_info.rlb_enabled) {
			bond->alb_info.rlb_rebalance = 1;
			/* If the updelay module parameter is smaller than the
			 * forwarding delay of the switch the rebalance will
			 * not work because the rebalance arp replies will
			 * not be forwarded to the clients..
			 */
		}
	}
}

/**
 * bond_alb_handle_active_change - assign new curr_active_slave
 * @bond: our bonding struct
 * @new_slave: new slave to assign
 *
 * Set the bond->curr_active_slave to @new_slave and handle
 * mac address swapping and promiscuity changes as needed.
 *
1623 1624 1625 1626 1627 1628
 * If new_slave is NULL, caller must hold curr_slave_lock or
 * bond->lock for write.
 *
 * If new_slave is not NULL, caller must hold RTNL, bond->lock for
 * read and curr_slave_lock for write.  Processing here may sleep, so
 * no other locks may be held.
L
Linus Torvalds 已提交
1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666
 */
void bond_alb_handle_active_change(struct bonding *bond, struct slave *new_slave)
{
	struct slave *swap_slave;
	int i;

	if (bond->curr_active_slave == new_slave) {
		return;
	}

	if (bond->curr_active_slave && bond->alb_info.primary_is_promisc) {
		dev_set_promiscuity(bond->curr_active_slave->dev, -1);
		bond->alb_info.primary_is_promisc = 0;
		bond->alb_info.rlb_promisc_timeout_counter = 0;
	}

	swap_slave = bond->curr_active_slave;
	bond->curr_active_slave = new_slave;

	if (!new_slave || (bond->slave_cnt == 0)) {
		return;
	}

	/* set the new curr_active_slave to the bonds mac address
	 * i.e. swap mac addresses of old curr_active_slave and new curr_active_slave
	 */
	if (!swap_slave) {
		struct slave *tmp_slave;
		/* find slave that is holding the bond's mac address */
		bond_for_each_slave(bond, tmp_slave, i) {
			if (!memcmp(tmp_slave->dev->dev_addr,
				    bond->dev->dev_addr, ETH_ALEN)) {
				swap_slave = tmp_slave;
				break;
			}
		}
	}

1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679
	/*
	 * Arrange for swap_slave and new_slave to temporarily be
	 * ignored so we can mess with their MAC addresses without
	 * fear of interference from transmit activity.
	 */
	if (swap_slave) {
		tlb_clear_slave(bond, swap_slave, 1);
	}
	tlb_clear_slave(bond, new_slave, 1);

	write_unlock_bh(&bond->curr_slave_lock);
	read_unlock(&bond->lock);

1680 1681
	ASSERT_RTNL();

L
Linus Torvalds 已提交
1682 1683 1684 1685 1686 1687 1688 1689
	/* curr_active_slave must be set before calling alb_swap_mac_addr */
	if (swap_slave) {
		/* swap mac address */
		alb_swap_mac_addr(bond, swap_slave, new_slave);
	} else {
		/* set the new_slave to the bond mac address */
		alb_set_slave_mac_addr(new_slave, bond->dev->dev_addr,
				       bond->alb_info.rlb_enabled);
1690 1691 1692 1693
	}

	if (swap_slave) {
		alb_fasten_mac_swap(bond, swap_slave, new_slave);
1694
		read_lock(&bond->lock);
1695
	} else {
1696
		read_lock(&bond->lock);
L
Linus Torvalds 已提交
1697 1698
		alb_send_learning_packets(new_slave, bond->dev->dev_addr);
	}
1699 1700

	write_lock_bh(&bond->curr_slave_lock);
L
Linus Torvalds 已提交
1701 1702
}

1703 1704 1705
/*
 * Called with RTNL
 */
L
Linus Torvalds 已提交
1706 1707
int bond_alb_set_mac_address(struct net_device *bond_dev, void *addr)
{
1708
	struct bonding *bond = netdev_priv(bond_dev);
L
Linus Torvalds 已提交
1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741
	struct sockaddr *sa = addr;
	struct slave *slave, *swap_slave;
	int res;
	int i;

	if (!is_valid_ether_addr(sa->sa_data)) {
		return -EADDRNOTAVAIL;
	}

	res = alb_set_mac_address(bond, addr);
	if (res) {
		return res;
	}

	memcpy(bond_dev->dev_addr, sa->sa_data, bond_dev->addr_len);

	/* If there is no curr_active_slave there is nothing else to do.
	 * Otherwise we'll need to pass the new address to it and handle
	 * duplications.
	 */
	if (!bond->curr_active_slave) {
		return 0;
	}

	swap_slave = NULL;

	bond_for_each_slave(bond, slave, i) {
		if (!memcmp(slave->dev->dev_addr, bond_dev->dev_addr, ETH_ALEN)) {
			swap_slave = slave;
			break;
		}
	}

1742 1743 1744
	write_unlock_bh(&bond->curr_slave_lock);
	read_unlock(&bond->lock);

L
Linus Torvalds 已提交
1745 1746
	if (swap_slave) {
		alb_swap_mac_addr(bond, swap_slave, bond->curr_active_slave);
1747
		alb_fasten_mac_swap(bond, swap_slave, bond->curr_active_slave);
L
Linus Torvalds 已提交
1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758
	} else {
		alb_set_slave_mac_addr(bond->curr_active_slave, bond_dev->dev_addr,
				       bond->alb_info.rlb_enabled);

		alb_send_learning_packets(bond->curr_active_slave, bond_dev->dev_addr);
		if (bond->alb_info.rlb_enabled) {
			/* inform clients mac address has changed */
			rlb_req_update_slave_clients(bond, bond->curr_active_slave);
		}
	}

1759 1760 1761
	read_lock(&bond->lock);
	write_lock_bh(&bond->curr_slave_lock);

L
Linus Torvalds 已提交
1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776
	return 0;
}

void bond_alb_clear_vlan(struct bonding *bond, unsigned short vlan_id)
{
	if (bond->alb_info.current_alb_vlan &&
	    (bond->alb_info.current_alb_vlan->vlan_id == vlan_id)) {
		bond->alb_info.current_alb_vlan = NULL;
	}

	if (bond->alb_info.rlb_enabled) {
		rlb_clear_vlan(bond, vlan_id);
	}
}