bond_alb.c

linux-2.6.15.6

				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
		       ": Warning: the hw address of slave %s is in use by "
		       "the bond; giving it the hw address of %s\n",
		       slave->dev->name, free_mac_slave->dev->name);

	} else if (has_bond_addr) {
		printk(KERN_ERR DRV_NAME
		       ": Error: the hw address of slave %s is in use by the "
		       "bond; couldn't find a slave with a free hw address to "
		       "give it (this should not have happened)\n",
		       slave->dev->name);
		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) {
		if (slave->dev->set_mac_address == NULL) {
			res = -EOPNOTSUPP;
			goto unwind;
		}

		/* 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);

		if (res) {
			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;
		}
	}

	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)
{
	struct bonding *bond = bond_dev->priv;
	struct ethhdr *eth_data;
	struct alb_bond_info *bond_info = &(BOND_ALB_INFO(bond));
	struct slave *tx_slave = NULL;
	static u32 ip_bcast = 0xffffffff;
	int hash_size = 0;
	int do_tx_balance = 1;
	u32 hash_index = 0;
	u8 *hash_start = NULL;
	int res = 1;

	skb->mac.raw = (unsigned char *)skb->data;
	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)) {
	case ETH_P_IP:
		if ((memcmp(eth_data->h_dest, mac_bcast, ETH_ALEN) == 0) ||
		    (skb->nh.iph->daddr == ip_bcast) ||
		    (skb->nh.iph->protocol == IPPROTO_IGMP)) {
			do_tx_balance = 0;
			break;
		}
		hash_start = (char*)&(skb->nh.iph->daddr);
		hash_size = sizeof(skb->nh.iph->daddr);
		break;
	case ETH_P_IPV6:
		if (memcmp(eth_data->h_dest, mac_bcast, ETH_ALEN) == 0) {
			do_tx_balance = 0;
			break;
		}

		hash_start = (char*)&(skb->nh.ipv6h->daddr);
		hash_size = sizeof(skb->nh.ipv6h->daddr);
		break;
	case ETH_P_IPX:
		if (ipx_hdr(skb)->ipx_checksum !=
		    __constant_htons(IPX_NO_CHECKSUM)) {
			/* 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;
}

void bond_alb_monitor(struct bonding *bond)
{
	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) {
			alb_send_learning_packets(slave,slave->dev->dev_addr);
		}

		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) {
		/* the following code changes the promiscuity of the
		 * the curr_active_slave. It needs to be locked with a
		 * write lock to protect from other code that also
		 * sets the promiscuity.
		 */
		write_lock(&bond->curr_slave_lock);

		if (bond_info->primary_is_promisc &&
		    (++bond_info->rlb_promisc_timeout_counter >= RLB_PROMISC_TIMEOUT)) {

			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;
		}

		write_unlock(&bond->curr_slave_lock);

		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:
	mod_timer(&(bond_info->alb_timer), jiffies + alb_delta_in_ticks);
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.
	 */
	write_lock_bh(&bond->lock);

	res = alb_handle_addr_collision_on_attach(bond, slave);

	write_unlock_bh(&bond->lock);

	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;
}

/* Caller must hold bond lock for write */
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.
 *
 * Caller must hold bond curr_slave_lock for write (or bond lock for write)
 */
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;
			}
		}
	}

	/* 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);
		/* fasten bond mac on new current slave */
		alb_send_learning_packets(new_slave, bond->dev->dev_addr);
	}
}

int bond_alb_set_mac_address(struct net_device *bond_dev, void *addr)
{
	struct bonding *bond = bond_dev->priv;
	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;
		}
	}

	if (swap_slave) {
		alb_swap_mac_addr(bond, swap_slave, bond->curr_active_slave);
	} 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);
		}
	}

	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);
	}
}