diff options
Diffstat (limited to 'drivers/net/bonding')
| -rw-r--r-- | drivers/net/bonding/Makefile | 3 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_3ad.c | 326 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_3ad.h | 10 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_alb.c | 13 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_ipv6.c | 218 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_main.c | 63 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_sysfs.c | 91 | ||||
| -rw-r--r-- | drivers/net/bonding/bonding.h | 35 |
8 files changed, 625 insertions, 134 deletions
diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile index 5cdae2bc055..6f9c6faef24 100644 --- a/drivers/net/bonding/Makefile +++ b/drivers/net/bonding/Makefile @@ -6,3 +6,6 @@ obj-$(CONFIG_BONDING) += bonding.o bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o +ipv6-$(subst m,y,$(CONFIG_IPV6)) += bond_ipv6.o +bonding-objs += $(ipv6-y) + diff --git a/drivers/net/bonding/bond_3ad.c b/drivers/net/bonding/bond_3ad.c index 6106660a4a4..ba1372f2f14 100644 --- a/drivers/net/bonding/bond_3ad.c +++ b/drivers/net/bonding/bond_3ad.c @@ -27,6 +27,7 @@ #include <linux/netdevice.h> #include <linux/spinlock.h> #include <linux/ethtool.h> +#include <linux/etherdevice.h> #include <linux/if_bonding.h> #include <linux/pkt_sched.h> #include <net/net_namespace.h> @@ -236,6 +237,17 @@ static inline struct aggregator *__get_next_agg(struct aggregator *aggregator) return &(SLAVE_AD_INFO(slave->next).aggregator); } +/* + * __agg_has_partner + * + * Return nonzero if aggregator has a partner (denoted by a non-zero ether + * address for the partner). Return 0 if not. + */ +static inline int __agg_has_partner(struct aggregator *agg) +{ + return !is_zero_ether_addr(agg->partner_system.mac_addr_value); +} + /** * __disable_port - disable the port's slave * @port: the port we're looking at @@ -274,14 +286,14 @@ static inline int __port_is_enabled(struct port *port) * __get_agg_selection_mode - get the aggregator selection mode * @port: the port we're looking at * - * Get the aggregator selection mode. Can be %BANDWIDTH or %COUNT. + * Get the aggregator selection mode. Can be %STABLE, %BANDWIDTH or %COUNT. */ static inline u32 __get_agg_selection_mode(struct port *port) { struct bonding *bond = __get_bond_by_port(port); if (bond == NULL) { - return AD_BANDWIDTH; + return BOND_AD_STABLE; } return BOND_AD_INFO(bond).agg_select_mode; @@ -1414,9 +1426,82 @@ static void ad_port_selection_logic(struct port *port) // else set ready=FALSE in all aggregator's ports __set_agg_ports_ready(port->aggregator, __agg_ports_are_ready(port->aggregator)); - if (!__check_agg_selection_timer(port) && (aggregator = __get_first_agg(port))) { - ad_agg_selection_logic(aggregator); + aggregator = __get_first_agg(port); + ad_agg_selection_logic(aggregator); +} + +/* + * Decide if "agg" is a better choice for the new active aggregator that + * the current best, according to the ad_select policy. + */ +static struct aggregator *ad_agg_selection_test(struct aggregator *best, + struct aggregator *curr) +{ + /* + * 0. If no best, select current. + * + * 1. If the current agg is not individual, and the best is + * individual, select current. + * + * 2. If current agg is individual and the best is not, keep best. + * + * 3. Therefore, current and best are both individual or both not + * individual, so: + * + * 3a. If current agg partner replied, and best agg partner did not, + * select current. + * + * 3b. If current agg partner did not reply and best agg partner + * did reply, keep best. + * + * 4. Therefore, current and best both have partner replies or + * both do not, so perform selection policy: + * + * BOND_AD_COUNT: Select by count of ports. If count is equal, + * select by bandwidth. + * + * BOND_AD_STABLE, BOND_AD_BANDWIDTH: Select by bandwidth. + */ + if (!best) + return curr; + + if (!curr->is_individual && best->is_individual) + return curr; + + if (curr->is_individual && !best->is_individual) + return best; + + if (__agg_has_partner(curr) && !__agg_has_partner(best)) + return curr; + + if (!__agg_has_partner(curr) && __agg_has_partner(best)) + return best; + + switch (__get_agg_selection_mode(curr->lag_ports)) { + case BOND_AD_COUNT: + if (curr->num_of_ports > best->num_of_ports) + return curr; + + if (curr->num_of_ports < best->num_of_ports) + return best; + + /*FALLTHROUGH*/ + case BOND_AD_STABLE: + case BOND_AD_BANDWIDTH: + if (__get_agg_bandwidth(curr) > __get_agg_bandwidth(best)) + return curr; + + break; + + default: + printk(KERN_WARNING DRV_NAME + ": %s: Impossible agg select mode %d\n", + curr->slave->dev->master->name, + __get_agg_selection_mode(curr->lag_ports)); + break; } + + return best; } /** @@ -1424,156 +1509,138 @@ static void ad_port_selection_logic(struct port *port) * @aggregator: the aggregator we're looking at * * It is assumed that only one aggregator may be selected for a team. - * The logic of this function is to select (at first time) the aggregator with - * the most ports attached to it, and to reselect the active aggregator only if - * the previous aggregator has no more ports related to it. + * + * The logic of this function is to select the aggregator according to + * the ad_select policy: + * + * BOND_AD_STABLE: select the aggregator with the most ports attached to + * it, and to reselect the active aggregator only if the previous + * aggregator has no more ports related to it. + * + * BOND_AD_BANDWIDTH: select the aggregator with the highest total + * bandwidth, and reselect whenever a link state change takes place or the + * set of slaves in the bond changes. + * + * BOND_AD_COUNT: select the aggregator with largest number of ports + * (slaves), and reselect whenever a link state change takes place or the + * set of slaves in the bond changes. * * FIXME: this function MUST be called with the first agg in the bond, or * __get_active_agg() won't work correctly. This function should be better * called with the bond itself, and retrieve the first agg from it. */ -static void ad_agg_selection_logic(struct aggregator *aggregator) +static void ad_agg_selection_logic(struct aggregator *agg) { - struct aggregator *best_aggregator = NULL, *active_aggregator = NULL; - struct aggregator *last_active_aggregator = NULL, *origin_aggregator; + struct aggregator *best, *active, *origin; struct port *port; - u16 num_of_aggs=0; - origin_aggregator = aggregator; + origin = agg; - //get current active aggregator - last_active_aggregator = __get_active_agg(aggregator); + active = __get_active_agg(agg); + best = active; - // search for the aggregator with the most ports attached to it. do { - // count how many candidate lag's we have - if (aggregator->lag_ports) { - num_of_aggs++; - } - if (aggregator->is_active && !aggregator->is_individual && // if current aggregator is the active aggregator - MAC_ADDRESS_COMPARE(&(aggregator->partner_system), &(null_mac_addr))) { // and partner answers to 802.3ad PDUs - if (aggregator->num_of_ports) { // if any ports attached to the current aggregator - best_aggregator=NULL; // disregard the best aggregator that was chosen by now - break; // stop the selection of other aggregator if there are any ports attached to this active aggregator - } else { // no ports attached to this active aggregator - aggregator->is_active = 0; // mark this aggregator as not active anymore + agg->is_active = 0; + + if (agg->num_of_ports) + best = ad_agg_selection_test(best, agg); + + } while ((agg = __get_next_agg(agg))); + + if (best && + __get_agg_selection_mode(best->lag_ports) == BOND_AD_STABLE) { + /* + * For the STABLE policy, don't replace the old active + * aggregator if it's still active (it has an answering + * partner) or if both the best and active don't have an + * answering partner. + */ + if (active && active->lag_ports && + active->lag_ports->is_enabled && + (__agg_has_partner(active) || + (!__agg_has_partner(active) && !__agg_has_partner(best)))) { + if (!(!active->actor_oper_aggregator_key && + best->actor_oper_aggregator_key)) { + best = NULL; + active->is_active = 1; } } - if (aggregator->num_of_ports) { // if any ports attached - if (best_aggregator) { // if there is a candidte aggregator - //The reasons for choosing new best aggregator: - // 1. if current agg is NOT individual and the best agg chosen so far is individual OR - // current and best aggs are both individual or both not individual, AND - // 2a. current agg partner reply but best agg partner do not reply OR - // 2b. current agg partner reply OR current agg partner do not reply AND best agg partner also do not reply AND - // current has more ports/bandwidth, or same amount of ports but current has faster ports, THEN - // current agg become best agg so far - - //if current agg is NOT individual and the best agg chosen so far is individual change best_aggregator - if (!aggregator->is_individual && best_aggregator->is_individual) { - best_aggregator=aggregator; - } - // current and best aggs are both individual or both not individual - else if ((aggregator->is_individual && best_aggregator->is_individual) || - (!aggregator->is_individual && !best_aggregator->is_individual)) { - // current and best aggs are both individual or both not individual AND - // current agg partner reply but best agg partner do not reply - if ((MAC_ADDRESS_COMPARE(&(aggregator->partner_system), &(null_mac_addr)) && - !MAC_ADDRESS_COMPARE(&(best_aggregator->partner_system), &(null_mac_addr)))) { - best_aggregator=aggregator; - } - // current agg partner reply OR current agg partner do not reply AND best agg partner also do not reply - else if (! (!MAC_ADDRESS_COMPARE(&(aggregator->partner_system), &(null_mac_addr)) && - MAC_ADDRESS_COMPARE(&(best_aggregator->partner_system), &(null_mac_addr)))) { - if ((__get_agg_selection_mode(aggregator->lag_ports) == AD_BANDWIDTH)&& - (__get_agg_bandwidth(aggregator) > __get_agg_bandwidth(best_aggregator))) { - best_aggregator=aggregator; - } else if (__get_agg_selection_mode(aggregator->lag_ports) == AD_COUNT) { - if (((aggregator->num_of_ports > best_aggregator->num_of_ports) && - (aggregator->actor_oper_aggregator_key & AD_SPEED_KEY_BITS))|| - ((aggregator->num_of_ports == best_aggregator->num_of_ports) && - ((u16)(aggregator->actor_oper_aggregator_key & AD_SPEED_KEY_BITS) > - (u16)(best_aggregator->actor_oper_aggregator_key & AD_SPEED_KEY_BITS)))) { - best_aggregator=aggregator; - } - } - } - } - } else { - best_aggregator=aggregator; - } - } - aggregator->is_active = 0; // mark all aggregators as not active anymore - } while ((aggregator = __get_next_agg(aggregator))); - - // if we have new aggregator selected, don't replace the old aggregator if it has an answering partner, - // or if both old aggregator and new aggregator don't have answering partner - if (best_aggregator) { - if (last_active_aggregator && last_active_aggregator->lag_ports && last_active_aggregator->lag_ports->is_enabled && - (MAC_ADDRESS_COMPARE(&(last_active_aggregator->partner_system), &(null_mac_addr)) || // partner answers OR - (!MAC_ADDRESS_COMPARE(&(last_active_aggregator->partner_system), &(null_mac_addr)) && // both old and new - !MAC_ADDRESS_COMPARE(&(best_aggregator->partner_system), &(null_mac_addr)))) // partner do not answer - ) { - // if new aggregator has link, and old aggregator does not, replace old aggregator.(do nothing) - // -> don't replace otherwise. - if (!(!last_active_aggregator->actor_oper_aggregator_key && best_aggregator->actor_oper_aggregator_key)) { - best_aggregator=NULL; - last_active_aggregator->is_active = 1; // don't replace good old aggregator + } - } - } + if (best && (best == active)) { + best = NULL; + active->is_active = 1; } // if there is new best aggregator, activate it - if (best_aggregator) { - for (aggregator = __get_first_agg(best_aggregator->lag_ports); - aggregator; - aggregator = __get_next_agg(aggregator)) { - - dprintk("Agg=%d; Ports=%d; a key=%d; p key=%d; Indiv=%d; Active=%d\n", - aggregator->aggregator_identifier, aggregator->num_of_ports, - aggregator->actor_oper_aggregator_key, aggregator->partner_oper_aggregator_key, - aggregator->is_individual, aggregator->is_active); + if (best) { + dprintk("best Agg=%d; P=%d; a k=%d; p k=%d; Ind=%d; Act=%d\n", + best->aggregator_identifier, best->num_of_ports, + best->actor_oper_aggregator_key, + best->partner_oper_aggregator_key, + best->is_individual, best->is_active); + dprintk("best ports %p slave %p %s\n", + best->lag_ports, best->slave, + best->slave ? best->slave->dev->name : "NULL"); + + for (agg = __get_first_agg(best->lag_ports); agg; + agg = __get_next_agg(agg)) { + + dprintk("Agg=%d; P=%d; a k=%d; p k=%d; Ind=%d; Act=%d\n", + agg->aggregator_identifier, agg->num_of_ports, + agg->actor_oper_aggregator_key, + agg->partner_oper_aggregator_key, + agg->is_individual, agg->is_active); } // check if any partner replys - if (best_aggregator->is_individual) { - printk(KERN_WARNING DRV_NAME ": %s: Warning: No 802.3ad response from " - "the link partner for any adapters in the bond\n", - best_aggregator->slave->dev->master->name); - } - - // check if there are more than one aggregator - if (num_of_aggs > 1) { - dprintk("Warning: More than one Link Aggregation Group was " - "found in the bond. Only one group will function in the bond\n"); + if (best->is_individual) { + printk(KERN_WARNING DRV_NAME ": %s: Warning: No 802.3ad" + " response from the link partner for any" + " adapters in the bond\n", + best->slave->dev->master->name); } - best_aggregator->is_active = 1; - dprintk("LAG %d choosed as the active LAG\n", best_aggregator->aggregator_identifier); - dprintk("Agg=%d; Ports=%d; a key=%d; p key=%d; Indiv=%d; Active=%d\n", - best_aggregator->aggregator_identifier, best_aggregator->num_of_ports, - best_aggregator->actor_oper_aggregator_key, best_aggregator->partner_oper_aggregator_key, - best_aggregator->is_individual, best_aggregator->is_active); + best->is_active = 1; + dprintk("LAG %d chosen as the active LAG\n", + best->aggregator_identifier); + dprintk("Agg=%d; P=%d; a k=%d; p k=%d; Ind=%d; Act=%d\n", + best->aggregator_identifier, best->num_of_ports, + best->actor_oper_aggregator_key, + best->partner_oper_aggregator_key, + best->is_individual, best->is_active); // disable the ports that were related to the former active_aggregator - if (last_active_aggregator) { - for (port=last_active_aggregator->lag_ports; port; port=port->next_port_in_aggregator) { + if (active) { + for (port = active->lag_ports; port; + port = port->next_port_in_aggregator) { __disable_port(port); } } } - // if the selected aggregator is of join individuals(partner_system is NULL), enable their ports - active_aggregator = __get_active_agg(origin_aggregator); + /* + * if the selected aggregator is of join individuals + * (partner_system is NULL), enable their ports + */ + active = __get_active_agg(origin); - if (active_aggregator) { - if (!MAC_ADDRESS_COMPARE(&(active_aggregator->partner_system), &(null_mac_addr))) { - for (port=active_aggregator->lag_ports; port; port=port->next_port_in_aggregator) { + if (active) { + if (!__agg_has_partner(active)) { + for (port = active->lag_ports; port; + port = port->next_port_in_aggregator) { __enable_port(port); } } } + + if (origin->slave) { + struct bonding *bond; + + bond = bond_get_bond_by_slave(origin->slave); + if (bond) + bond_3ad_set_carrier(bond); + } } /** @@ -1830,6 +1897,19 @@ static void ad_initialize_lacpdu(struct lacpdu *lacpdu) // Check aggregators status in team every T seconds #define AD_AGGREGATOR_SELECTION_TIMER 8 +/* + * bond_3ad_initiate_agg_selection(struct bonding *bond) + * + * Set the aggregation selection timer, to initiate an agg selection in + * the very near future. Called during first initialization, and during + * any down to up transitions of the bond. + */ +void bond_3ad_initiate_agg_selection(struct bonding *bond, int timeout) +{ + BOND_AD_INFO(bond).agg_select_timer = timeout; + BOND_AD_INFO(bond).agg_select_mode = bond->params.ad_select; +} + static u16 aggregator_identifier; /** @@ -1854,9 +1934,9 @@ void bond_3ad_initialize(struct bonding *bond, u16 tick_resolution, int lacp_fas // initialize how many times this module is called in one second(should be about every 100ms) ad_ticks_per_sec = tick_resolution; - // initialize the aggregator selection timer(to activate an aggregation selection after initialize) - BOND_AD_INFO(bond).agg_select_timer = (AD_AGGREGATOR_SELECTION_TIMER * ad_ticks_per_sec); - BOND_AD_INFO(bond).agg_select_mode = AD_BANDWIDTH; + bond_3ad_initiate_agg_selection(bond, + AD_AGGREGATOR_SELECTION_TIMER * + ad_ticks_per_sec); } } diff --git a/drivers/net/bonding/bond_3ad.h b/drivers/net/bonding/bond_3ad.h index b5ee45f6d55..a803fe05f63 100644 --- a/drivers/net/bonding/bond_3ad.h +++ b/drivers/net/bonding/bond_3ad.h @@ -42,10 +42,11 @@ typedef struct mac_addr { u8 mac_addr_value[ETH_ALEN]; } mac_addr_t; -typedef enum { - AD_BANDWIDTH = 0, - AD_COUNT -} agg_selection_t; +enum { + BOND_AD_STABLE = 0, + BOND_AD_BANDWIDTH = 1, + BOND_AD_COUNT = 2, +}; // rx machine states(43.4.11 in the 802.3ad standard) typedef enum { @@ -277,6 +278,7 @@ void bond_3ad_initialize(struct bonding *bond, u16 tick_resolution, int lacp_fas int bond_3ad_bind_slave(struct slave *slave); void bond_3ad_unbind_slave(struct slave *slave); void bond_3ad_state_machine_handler(struct work_struct *); +void bond_3ad_initiate_agg_selection(struct bonding *bond, int timeout); void bond_3ad_adapter_speed_changed(struct slave *slave); void bond_3ad_adapter_duplex_changed(struct slave *slave); void bond_3ad_handle_link_change(struct slave *slave, char link); diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index 87437c78847..e170fa2fa1a 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -346,14 +346,18 @@ static void rlb_update_entry_from_arp(struct bonding *bond, struct arp_pkt *arp) static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct packet_type *ptype, struct net_device *orig_dev) { - struct bonding *bond = bond_dev->priv; + struct bonding *bond; struct arp_pkt *arp = (struct arp_pkt *)skb->data; int res = NET_RX_DROP; if (dev_net(bond_dev) != &init_net) goto out; - if (!(bond_dev->flags & IFF_MASTER)) + 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)) goto out; if (!arp) { @@ -368,6 +372,9 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct if (arp->op_code == htons(ARPOP_REPLY)) { /* update rx hash table for this ARP */ + printk("rar: update orig %s bond_dev %s\n", orig_dev->name, + bond_dev->name); + bond = bond_dev->priv; rlb_update_entry_from_arp(bond, arp); dprintk("Server received an ARP Reply from client\n"); } @@ -818,7 +825,7 @@ static int rlb_initialize(struct bonding *bond) /*initialize packet type*/ pk_type->type = __constant_htons(ETH_P_ARP); - pk_type->dev = bond->dev; + pk_type->dev = NULL; pk_type->func = rlb_arp_recv; /* register to receive ARPs */ diff --git a/drivers/net/bonding/bond_ipv6.c b/drivers/net/bonding/bond_ipv6.c new file mode 100644 index 00000000000..7c78b7bf671 --- /dev/null +++ b/drivers/net/bonding/bond_ipv6.c @@ -0,0 +1,218 @@ +/* + * Copyright(c) 2008 Hewlett-Packard Development Company, L.P. + * + * 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. + * + */ + +//#define BONDING_DEBUG 1 + +#include <linux/types.h> +#include <linux/if_vlan.h> +#include <net/ipv6.h> +#include <net/ndisc.h> +#include <net/addrconf.h> +#include "bonding.h" + +/* + * Assign bond->master_ipv6 to the next IPv6 address in the list, or + * zero it out if there are none. + */ +static void bond_glean_dev_ipv6(struct net_device *dev, struct in6_addr *addr) +{ + struct inet6_dev *idev; + struct inet6_ifaddr *ifa; + + if (!dev) + return; + + idev = in6_dev_get(dev); + if (!idev) + return; + + read_lock_bh(&idev->lock); + ifa = idev->addr_list; + if (ifa) + ipv6_addr_copy(addr, &ifa->addr); + else + ipv6_addr_set(addr, 0, 0, 0, 0); + + read_unlock_bh(&idev->lock); + + in6_dev_put(idev); +} + +static void bond_na_send(struct net_device *slave_dev, + struct in6_addr *daddr, + int router, + unsigned short vlan_id) +{ + struct in6_addr mcaddr; + struct icmp6hdr icmp6h = { + .icmp6_type = NDISC_NEIGHBOUR_ADVERTISEMENT, + }; + struct sk_buff *skb; + + icmp6h.icmp6_router = router; + icmp6h.icmp6_solicited = 0; + icmp6h.icmp6_override = 1; + + addrconf_addr_solict_mult(daddr, &mcaddr); + + dprintk("ipv6 na on slave %s: dest %pI6, src %pI6\n", + slave->name, &mcaddr, daddr); + + skb = ndisc_build_skb(slave_dev, &mcaddr, daddr, &icmp6h, daddr, + ND_OPT_TARGET_LL_ADDR); + + if (!skb) { + printk(KERN_ERR DRV_NAME ": NA packet allocation failed\n"); + return; + } + + if (vlan_id) { + skb = vlan_put_tag(skb, vlan_id); + if (!skb) { + printk(KERN_ERR DRV_NAME ": failed to insert VLAN tag\n"); + return; + } + } + + ndisc_send_skb(skb, slave_dev, NULL, &mcaddr, daddr, &icmp6h); +} + +/* + * Kick out an unsolicited Neighbor Advertisement for an IPv6 address on + * the bonding master. This will help the switch learn our address + * if in active-backup mode. + * + * Caller must hold curr_slave_lock for read or better + */ +void bond_send_unsolicited_na(struct bonding *bond) +{ + struct slave *slave = bond->curr_active_slave; + struct vlan_entry *vlan; + struct inet6_dev *idev; + int is_router; + + dprintk("bond_send_unsol_na: bond %s slave %s\n", bond->dev->name, + slave ? slave->dev->name : "NULL"); + + if (!slave || !bond->send_unsol_na || + test_bit(__LINK_STATE_LINKWATCH_PENDING, &slave->dev->state)) + return; + + bond->send_unsol_na--; + + idev = in6_dev_get(bond->dev); + if (!idev) + return; + + is_router = !!idev->cnf.forwarding; + + in6_dev_put(idev); + + if (!ipv6_addr_any(&bond->master_ipv6)) + bond_na_send(slave->dev, &bond->master_ipv6, is_router, 0); + + list_for_each_entry(vlan, &bond->vlan_list, vlan_list) { + if (!ipv6_addr_any(&vlan->vlan_ipv6)) { + bond_na_send(slave->dev, &vlan->vlan_ipv6, is_router, + vlan->vlan_id); + } + } +} + +/* + * bond_inet6addr_event: handle inet6addr notifier chain events. + * + * We keep track of device IPv6 addresses primarily to use as source + * addresses in NS probes. + * + * We track one IPv6 for the main device (if it has one). + */ +static int bond_inet6addr_event(struct notifier_block *this, + unsigned long event, + void *ptr) +{ + struct inet6_ifaddr *ifa = ptr; + struct net_device *vlan_dev, *event_dev = ifa->idev->dev; + struct bonding *bond; + struct vlan_entry *vlan; + + if (dev_net(event_dev) != &init_net) + return NOTIFY_DONE; + + list_for_each_entry(bond, &bond_dev_list, bond_list) { + if (bond->dev == event_dev) { + switch (event) { + case NETDEV_UP: + if (ipv6_addr_any(&bond->master_ipv6)) + ipv6_addr_copy(&bond->master_ipv6, + &ifa->addr); + return NOTIFY_OK; + case NETDEV_DOWN: + if (ipv6_addr_equal(&bond->master_ipv6, + &ifa->addr)) + bond_glean_dev_ipv6(bond->dev, + &bond->master_ipv6); + return NOTIFY_OK; + default: + return NOTIFY_DONE; + } + } + + list_for_each_entry(vlan, &bond->vlan_list, vlan_list) { + vlan_dev = vlan_group_get_device(bond->vlgrp, + vlan->vlan_id); + if (vlan_dev == event_dev) { + switch (event) { + case NETDEV_UP: + if (ipv6_addr_any(&vlan->vlan_ipv6)) + ipv6_addr_copy(&vlan->vlan_ipv6, + &ifa->addr); + return NOTIFY_OK; + case NETDEV_DOWN: + if (ipv6_addr_equal(&vlan->vlan_ipv6, + &ifa->addr)) + bond_glean_dev_ipv6(vlan_dev, + &vlan->vlan_ipv6); + return NOTIFY_OK; + default: + return NOTIFY_DONE; + } + } + } + } + return NOTIFY_DONE; +} + +static struct notifier_block bond_inet6addr_notifier = { + .notifier_call = bond_inet6addr_event, +}; + +void bond_register_ipv6_notifier(void) +{ + register_inet6addr_notifier(&bond_inet6addr_notifier); +} + +void bond_unregister_ipv6_notifier(void) +{ + unregister_inet6addr_notifier(&bond_inet6addr_notifier); +} + diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 39575d76497..02de3e03123 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -89,6 +89,7 @@ static int max_bonds = BOND_DEFAULT_MAX_BONDS; static int num_grat_arp = 1; +static int num_unsol_na = 1; static int miimon = BOND_LINK_MON_INTERV; static int updelay = 0; static int downdelay = 0; @@ -96,6 +97,7 @@ static int use_carrier = 1; static char *mode = NULL; static char *primary = NULL; static char *lacp_rate = NULL; +static char *ad_select = NULL; static char *xmit_hash_policy = NULL; static int arp_interval = BOND_LINK_ARP_INTERV; static char *arp_ip_target[BOND_MAX_ARP_TARGETS] = { NULL, }; @@ -107,6 +109,8 @@ module_param(max_bonds, int, 0); MODULE_PARM_DESC(max_bonds, "Max number of bonded devices"); module_param(num_grat_arp, int, 0644); MODULE_PARM_DESC(num_grat_arp, "Number of gratuitous ARP packets to send on failover event"); +module_param(num_unsol_na, int, 0644); +MODULE_PARM_DESC(num_unsol_na, "Number of unsolicited IPv6 Neighbor Advertisements packets to send on failover event"); module_param(miimon, int, 0); MODULE_PARM_DESC(miimon, "Link check interval in milliseconds"); module_param(updelay, int, 0); @@ -127,6 +131,8 @@ MODULE_PARM_DESC(primary, "Primary network device to use"); module_param(lacp_rate, charp, 0); MODULE_PARM_DESC(lacp_rate, "LACPDU tx rate to request from 802.3ad partner " "(slow/fast)"); +module_param(ad_select, charp, 0); +MODULE_PARM_DESC(ad_select, "803.ad aggregation selection logic: stable (0, default), bandwidth (1), count (2)"); module_param(xmit_hash_policy, charp, 0); MODULE_PARM_DESC(xmit_hash_policy, "XOR hashing method: 0 for layer 2 (default)" ", 1 for layer 3+4"); @@ -197,6 +203,13 @@ struct bond_parm_tbl fail_over_mac_tbl[] = { { NULL, -1}, }; +struct bond_parm_tbl ad_select_tbl[] = { +{ "stable", BOND_AD_STABLE}, +{ "bandwidth", BOND_AD_BANDWIDTH}, +{ "count", BOND_AD_COUNT}, +{ NULL, -1}, +}; + /*-------------------------- Forward declarations ---------------------------*/ static void bond_send_gratuitous_arp(struct bonding *bond); @@ -242,14 +255,13 @@ static int bond_add_vlan(struct bonding *bond, unsigned short vlan_id) dprintk("bond: %s, vlan id %d\n", (bond ? bond->dev->name: "None"), vlan_id); - vlan = kmalloc(sizeof(struct vlan_entry), GFP_KERNEL); + vlan = kzalloc(sizeof(struct vlan_entry), GFP_KERNEL); if (!vlan) { return -ENOMEM; } INIT_LIST_HEAD(&vlan->vlan_list); vlan->vlan_id = vlan_id; - vlan->vlan_ip = 0; write_lock_bh(&bond->lock); @@ -1208,6 +1220,9 @@ void bond_change_active_slave(struct bonding *bond, struct slave *new_active) bond->send_grat_arp = bond->params.num_grat_arp; bond_send_gratuitous_arp(bond); + bond->send_unsol_na = bond->params.num_unsol_na; + bond_send_unsolicited_na(bond); + write_unlock_bh(&bond->curr_slave_lock); read_unlock(&bond->lock); @@ -2463,6 +2478,12 @@ void bond_mii_monitor(struct work_struct *work) read_unlock(&bond->curr_slave_lock); } + if (bond->send_unsol_na) { + read_lock(&bond->curr_slave_lock); + bond_send_unsolicited_na(bond); + read_unlock(&bond->curr_slave_lock); + } + if (bond_miimon_inspect(bond)) { read_unlock(&bond->lock); rtnl_lock(); @@ -3158,6 +3179,12 @@ void bond_activebackup_arp_mon(struct work_struct *work) read_unlock(&bond->curr_slave_lock); } + if (bond->send_unsol_na) { + read_lock(&bond->curr_slave_lock); + bond_send_unsolicited_na(bond); + read_unlock(&bond->curr_slave_lock); + } + if (bond_ab_arp_inspect(bond, delta_in_ticks)) { read_unlock(&bond->lock); rtnl_lock(); @@ -3301,6 +3328,8 @@ static void bond_info_show_master(struct seq_file *seq) seq_puts(seq, "\n802.3ad info\n"); seq_printf(seq, "LACP rate: %s\n", (bond->params.lacp_fast) ? "fast" : "slow"); + seq_printf(seq, "Aggregator selection policy (ad_select): %s\n", + ad_select_tbl[bond->params.ad_select].modename); if (bond_3ad_get_active_agg_info(bond, &ad_info)) { seq_printf(seq, "bond %s has no active aggregator\n", @@ -3807,6 +3836,7 @@ static int bond_open(struct net_device *bond_dev) queue_delayed_work(bond->wq, &bond->ad_work, 0); /* register to receive LACPDUs */ bond_register_lacpdu(bond); + bond_3ad_initiate_agg_selection(bond, 1); } return 0; @@ -3827,6 +3857,7 @@ static int bond_close(struct net_device *bond_dev) write_lock_bh(&bond->lock); bond->send_grat_arp = 0; + bond->send_unsol_na = 0; /* signal timers not to re-arm */ bond->kill_timers = 1; @@ -4542,6 +4573,7 @@ static int bond_init(struct net_device *bond_dev, struct bond_params *params) bond->primary_slave = NULL; bond->dev = bond_dev; bond->send_grat_arp = 0; + bond->send_unsol_na = 0; bond->setup_by_slave = 0; INIT_LIST_HEAD(&bond->vlan_list); @@ -4744,6 +4776,23 @@ static int bond_check_params(struct bond_params *params) } } + if (ad_select) { + params->ad_select = bond_parse_parm(ad_select, ad_select_tbl); + if (params->ad_select == -1) { + printk(KERN_ERR DRV_NAME + ": Error: Invalid ad_select \"%s\"\n", + ad_select == NULL ? "NULL" : ad_select); + return -EINVAL; + } + + if (bond_mode != BOND_MODE_8023AD) { + printk(KERN_WARNING DRV_NAME + ": ad_select param only affects 802.3ad mode\n"); + } + } else { + params->ad_select = BOND_AD_STABLE; + } + if (max_bonds < 0 || max_bonds > INT_MAX) { printk(KERN_WARNING DRV_NAME ": Warning: max_bonds (%d) not in range %d-%d, so it " @@ -4791,6 +4840,13 @@ static int bond_check_params(struct bond_params *params) num_grat_arp = 1; } + if (num_unsol_na < 0 || num_unsol_na > 255) { + printk(KERN_WARNING DRV_NAME + ": Warning: num_unsol_na (%d) not in range 0-255 so it " + "was reset to 1 \n", num_unsol_na); + num_unsol_na = 1; + } + /* reset values for 802.3ad */ if (bond_mode == BOND_MODE_8023AD) { if (!miimon) { @@ -4992,6 +5048,7 @@ static int bond_check_params(struct bond_params *params) params->xmit_policy = xmit_hashtype; params->miimon = miimon; params->num_grat_arp = num_grat_arp; + params->num_unsol_na = num_unsol_na; params->arp_interval = arp_interval; params->arp_validate = arp_validate_value; params->updelay = updelay; @@ -5144,6 +5201,7 @@ static int __init bonding_init(void) register_netdevice_notifier(&bond_netdev_notifier); register_inetaddr_notifier(&bond_inetaddr_notifier); + bond_register_ipv6_notifier(); goto out; err: @@ -5166,6 +5224,7 @@ static void __exit bonding_exit(void) { unregister_netdevice_notifier(&bond_netdev_notifier); unregister_inetaddr_notifier(&bond_inetaddr_notifier); + bond_unregister_ipv6_notifier(); bond_destroy_sysfs(); diff --git a/drivers/net/bonding/bond_sysfs.c b/drivers/net/bonding/bond_sysfs.c index e400d7dfdfc..aaf2927b5c3 100644 --- a/drivers/net/bonding/bond_sysfs.c +++ b/drivers/net/bonding/bond_sysfs.c @@ -48,6 +48,7 @@ extern struct list_head bond_dev_list; extern struct bond_params bonding_defaults; extern struct bond_parm_tbl bond_mode_tbl[]; extern struct bond_parm_tbl bond_lacp_tbl[]; +extern struct bond_parm_tbl ad_select_tbl[]; extern struct bond_parm_tbl xmit_hashtype_tbl[]; extern struct bond_parm_tbl arp_validate_tbl[]; extern struct bond_parm_tbl fail_over_mac_tbl[]; @@ -944,6 +945,53 @@ out: } static DEVICE_ATTR(lacp_rate, S_IRUGO | S_IWUSR, bonding_show_lacp, bonding_store_lacp); +static ssize_t bonding_show_ad_select(struct device *d, + struct device_attribute *attr, + char *buf) +{ + struct bonding *bond = to_bond(d); + + return sprintf(buf, "%s %d\n", + ad_select_tbl[bond->params.ad_select].modename, + bond->params.ad_select); +} + + +static ssize_t bonding_store_ad_select(struct device *d, + struct device_attribute *attr, + const char *buf, size_t count) +{ + int new_value, ret = count; + struct bonding *bond = to_bond(d); + + if (bond->dev->flags & IFF_UP) { + printk(KERN_ERR DRV_NAME + ": %s: Unable to update ad_select because interface " + "is up.\n", bond->dev->name); + ret = -EPERM; + goto out; + } + + new_value = bond_parse_parm(buf, ad_select_tbl); + + if (new_value != -1) { + bond->params.ad_select = new_value; + printk(KERN_INFO DRV_NAME + ": %s: Setting ad_select to %s (%d).\n", + bond->dev->name, ad_select_tbl[new_value].modename, + new_value); + } else { + printk(KERN_ERR DRV_NAME + ": %s: Ignoring invalid ad_select value %.*s.\n", + bond->dev->name, (int)strlen(buf) - 1, buf); + ret = -EINVAL; + } +out: + return ret; +} + +static DEVICE_ATTR(ad_select, S_IRUGO | S_IWUSR, bonding_show_ad_select, bonding_store_ad_select); + /* * Show and set the number of grat ARP to send after a failover event. */ @@ -983,6 +1031,47 @@ out: return ret; } static DEVICE_ATTR(num_grat_arp, S_IRUGO | S_IWUSR, bonding_show_n_grat_arp, bonding_store_n_grat_arp); + +/* + * Show and set the number of unsolicted NA's to send after a failover event. + */ +static ssize_t bonding_show_n_unsol_na(struct device *d, + struct device_attribute *attr, + char *buf) +{ + struct bonding *bond = to_bond(d); + + return sprintf(buf, "%d\n", bond->params.num_unsol_na); +} + +static ssize_t bonding_store_n_unsol_na(struct device *d, + struct device_attribute *attr, + const char *buf, size_t count) +{ + int new_value, ret = count; + struct bonding *bond = to_bond(d); + + if (sscanf(buf, "%d", &new_value) != 1) { + printk(KERN_ERR DRV_NAME + ": %s: no num_unsol_na value specified.\n", + bond->dev->name); + ret = -EINVAL; + goto out; + } + if (new_value < 0 || new_value > 255) { + printk(KERN_ERR DRV_NAME + ": %s: Invalid num_unsol_na value %d not in range 0-255; rejected.\n", + bond->dev->name, new_value); + ret = -EINVAL; + goto out; + } else { + bond->params.num_unsol_na = new_value; + } +out: + return ret; +} +static DEVICE_ATTR(num_unsol_na, S_IRUGO | S_IWUSR, bonding_show_n_unsol_na, bonding_store_n_unsol_na); + /* * Show and set the MII monitor interval. There are two tricky bits * here. First, if MII monitoring is activated, then we must disable @@ -1418,8 +1507,10 @@ static struct attribute *per_bond_attrs[] = { &dev_attr_downdelay.attr, &dev_attr_updelay.attr, &dev_attr_lacp_rate.attr, + &dev_attr_ad_select.attr, &dev_attr_xmit_hash_policy.attr, &dev_attr_num_grat_arp.attr, + &dev_attr_num_unsol_na.attr, &dev_attr_miimon.attr, &dev_attr_primary.attr, &dev_attr_use_carrier.attr, diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h index ffb668dd6d3..b5eb8e65b30 100644 --- a/drivers/net/bonding/bonding.h +++ b/drivers/net/bonding/bonding.h @@ -19,16 +19,19 @@ #include <linux/proc_fs.h> #include <linux/if_bonding.h> #include <linux/kobject.h> +#include <linux/in6.h> #include "bond_3ad.h" #include "bond_alb.h" -#define DRV_VERSION "3.3.0" -#define DRV_RELDATE "June 10, 2008" +#define DRV_VERSION "3.5.0" +#define DRV_RELDATE "November 4, 2008" #define DRV_NAME "bonding" #define DRV_DESCRIPTION "Ethernet Channel Bonding Driver" #define BOND_MAX_ARP_TARGETS 16 +extern struct list_head bond_dev_list; + #ifdef BONDING_DEBUG #define dprintk(fmt, args...) \ printk(KERN_DEBUG \ @@ -126,6 +129,7 @@ struct bond_params { int xmit_policy; int miimon; int num_grat_arp; + int num_unsol_na; int arp_interval; int arp_validate; int use_carrier; @@ -133,6 +137,7 @@ struct bond_params { int updelay; int downdelay; int lacp_fast; + int ad_select; char primary[IFNAMSIZ]; __be32 arp_targets[BOND_MAX_ARP_TARGETS]; }; @@ -148,6 +153,9 @@ struct vlan_entry { struct list_head vlan_list; __be32 vlan_ip; unsigned short vlan_id; +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) + struct in6_addr vlan_ipv6; +#endif }; struct slave { @@ -195,6 +203,7 @@ struct bonding { rwlock_t curr_slave_lock; s8 kill_timers; s8 send_grat_arp; + s8 send_unsol_na; s8 setup_by_slave; struct net_device_stats stats; #ifdef CONFIG_PROC_FS @@ -218,6 +227,9 @@ struct bonding { struct delayed_work arp_work; struct delayed_work alb_work; struct delayed_work ad_work; +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) + struct in6_addr master_ipv6; +#endif }; /** @@ -341,5 +353,24 @@ extern struct bond_parm_tbl xmit_hashtype_tbl[]; extern struct bond_parm_tbl arp_validate_tbl[]; extern struct bond_parm_tbl fail_over_mac_tbl[]; +#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) +void bond_send_unsolicited_na(struct bonding *bond); +void bond_register_ipv6_notifier(void); +void bond_unregister_ipv6_notifier(void); +#else +static inline void bond_send_unsolicited_na(struct bonding *bond) +{ + return; +} +static inline void bond_register_ipv6_notifier(void) +{ + return; +} +static inline void bond_unregister_ipv6_notifier(void) +{ + return; +} +#endif + #endif /* _LINUX_BONDING_H */ |