diff options
Diffstat (limited to 'drivers/net/bonding')
| -rw-r--r-- | drivers/net/bonding/Makefile | 8 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_3ad.c | 2451 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_3ad.h | 300 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_alb.c | 1696 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_alb.h | 141 | ||||
| -rw-r--r-- | drivers/net/bonding/bond_main.c | 4708 | ||||
| -rw-r--r-- | drivers/net/bonding/bonding.h | 252 |
7 files changed, 9556 insertions, 0 deletions
diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile new file mode 100644 index 00000000000..cf50384b469 --- /dev/null +++ b/drivers/net/bonding/Makefile @@ -0,0 +1,8 @@ +# +# Makefile for the Ethernet Bonding driver +# + +obj-$(CONFIG_BONDING) += bonding.o + +bonding-objs := bond_main.o bond_3ad.o bond_alb.o + diff --git a/drivers/net/bonding/bond_3ad.c b/drivers/net/bonding/bond_3ad.c new file mode 100644 index 00000000000..6233c4ffb80 --- /dev/null +++ b/drivers/net/bonding/bond_3ad.c @@ -0,0 +1,2451 @@ +/* + * 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. + * + * + * Changes: + * + * 2003/05/01 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Amir Noam <amir.noam at intel dot com> + * - Added support for lacp_rate module param. + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Based on discussion on mailing list, changed locking scheme + * to use lock/unlock or lock_bh/unlock_bh appropriately instead + * of lock_irqsave/unlock_irqrestore. The new scheme helps exposing + * hidden bugs and solves system hangs that occurred due to the fact + * that holding lock_irqsave doesn't prevent softirqs from running. + * This also increases total throughput since interrupts are not + * blocked on each transmitted packets or monitor timeout. + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Renamed bond_3ad_link_status_changed() to + * bond_3ad_handle_link_change() for compatibility with TLB. + * + * 2003/05/20 - Amir Noam <amir.noam at intel dot com> + * - Fix long fail over time when releasing last slave of an active + * aggregator - send LACPDU on unbind of slave to tell partner this + * port is no longer aggregatable. + * + * 2003/06/25 - Tsippy Mendelson <tsippy.mendelson at intel dot com> + * - Send LACPDU as highest priority packet to further fix the above + * problem on very high Tx traffic load where packets may get dropped + * by the slave. + * + * 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Code cleanup and style changes + */ + +//#define BONDING_DEBUG 1 + +#include <linux/skbuff.h> +#include <linux/if_ether.h> +#include <linux/netdevice.h> +#include <linux/spinlock.h> +#include <linux/ethtool.h> +#include <linux/if_bonding.h> +#include <linux/pkt_sched.h> +#include "bonding.h" +#include "bond_3ad.h" + +// General definitions +#define AD_SHORT_TIMEOUT 1 +#define AD_LONG_TIMEOUT 0 +#define AD_STANDBY 0x2 +#define AD_MAX_TX_IN_SECOND 3 +#define AD_COLLECTOR_MAX_DELAY 0 + +// Timer definitions(43.4.4 in the 802.3ad standard) +#define AD_FAST_PERIODIC_TIME 1 +#define AD_SLOW_PERIODIC_TIME 30 +#define AD_SHORT_TIMEOUT_TIME (3*AD_FAST_PERIODIC_TIME) +#define AD_LONG_TIMEOUT_TIME (3*AD_SLOW_PERIODIC_TIME) +#define AD_CHURN_DETECTION_TIME 60 +#define AD_AGGREGATE_WAIT_TIME 2 + +// Port state definitions(43.4.2.2 in the 802.3ad standard) +#define AD_STATE_LACP_ACTIVITY 0x1 +#define AD_STATE_LACP_TIMEOUT 0x2 +#define AD_STATE_AGGREGATION 0x4 +#define AD_STATE_SYNCHRONIZATION 0x8 +#define AD_STATE_COLLECTING 0x10 +#define AD_STATE_DISTRIBUTING 0x20 +#define AD_STATE_DEFAULTED 0x40 +#define AD_STATE_EXPIRED 0x80 + +// Port Variables definitions used by the State Machines(43.4.7 in the 802.3ad standard) +#define AD_PORT_BEGIN 0x1 +#define AD_PORT_LACP_ENABLED 0x2 +#define AD_PORT_ACTOR_CHURN 0x4 +#define AD_PORT_PARTNER_CHURN 0x8 +#define AD_PORT_READY 0x10 +#define AD_PORT_READY_N 0x20 +#define AD_PORT_MATCHED 0x40 +#define AD_PORT_STANDBY 0x80 +#define AD_PORT_SELECTED 0x100 +#define AD_PORT_MOVED 0x200 + +// Port Key definitions +// key is determined according to the link speed, duplex and +// user key(which is yet not supported) +// ------------------------------------------------------------ +// Port key : | User key | Speed |Duplex| +// ------------------------------------------------------------ +// 16 6 1 0 +#define AD_DUPLEX_KEY_BITS 0x1 +#define AD_SPEED_KEY_BITS 0x3E +#define AD_USER_KEY_BITS 0xFFC0 + +//dalloun +#define AD_LINK_SPEED_BITMASK_1MBPS 0x1 +#define AD_LINK_SPEED_BITMASK_10MBPS 0x2 +#define AD_LINK_SPEED_BITMASK_100MBPS 0x4 +#define AD_LINK_SPEED_BITMASK_1000MBPS 0x8 +//endalloun + +// compare MAC addresses +#define MAC_ADDRESS_COMPARE(A, B) memcmp(A, B, ETH_ALEN) + +static struct mac_addr null_mac_addr = {{0, 0, 0, 0, 0, 0}}; +static u16 ad_ticks_per_sec; +static const int ad_delta_in_ticks = (AD_TIMER_INTERVAL * HZ) / 1000; + +// ================= 3AD api to bonding and kernel code ================== +static u16 __get_link_speed(struct port *port); +static u8 __get_duplex(struct port *port); +static inline void __initialize_port_locks(struct port *port); +//conversions +static void __ntohs_lacpdu(struct lacpdu *lacpdu); +static u16 __ad_timer_to_ticks(u16 timer_type, u16 Par); + + +// ================= ad code helper functions ================== +//needed by ad_rx_machine(...) +static void __record_pdu(struct lacpdu *lacpdu, struct port *port); +static void __record_default(struct port *port); +static void __update_selected(struct lacpdu *lacpdu, struct port *port); +static void __update_default_selected(struct port *port); +static void __choose_matched(struct lacpdu *lacpdu, struct port *port); +static void __update_ntt(struct lacpdu *lacpdu, struct port *port); + +//needed for ad_mux_machine(..) +static void __attach_bond_to_agg(struct port *port); +static void __detach_bond_from_agg(struct port *port); +static int __agg_ports_are_ready(struct aggregator *aggregator); +static void __set_agg_ports_ready(struct aggregator *aggregator, int val); + +//needed for ad_agg_selection_logic(...) +static u32 __get_agg_bandwidth(struct aggregator *aggregator); +static struct aggregator *__get_active_agg(struct aggregator *aggregator); + + +// ================= main 802.3ad protocol functions ================== +static int ad_lacpdu_send(struct port *port); +static int ad_marker_send(struct port *port, struct marker *marker); +static void ad_mux_machine(struct port *port); +static void ad_rx_machine(struct lacpdu *lacpdu, struct port *port); +static void ad_tx_machine(struct port *port); +static void ad_periodic_machine(struct port *port); +static void ad_port_selection_logic(struct port *port); +static void ad_agg_selection_logic(struct aggregator *aggregator); +static void ad_clear_agg(struct aggregator *aggregator); +static void ad_initialize_agg(struct aggregator *aggregator); +static void ad_initialize_port(struct port *port, int lacp_fast); +static void ad_initialize_lacpdu(struct lacpdu *Lacpdu); +static void ad_enable_collecting_distributing(struct port *port); +static void ad_disable_collecting_distributing(struct port *port); +static void ad_marker_info_received(struct marker *marker_info, struct port *port); +static void ad_marker_response_received(struct marker *marker, struct port *port); + + +///////////////////////////////////////////////////////////////////////////////// +// ================= api to bonding and kernel code ================== +///////////////////////////////////////////////////////////////////////////////// + +/** + * __get_bond_by_port - get the port's bonding struct + * @port: the port we're looking at + * + * Return @port's bonding struct, or %NULL if it can't be found. + */ +static inline struct bonding *__get_bond_by_port(struct port *port) +{ + if (port->slave == NULL) { + return NULL; + } + + return bond_get_bond_by_slave(port->slave); +} + +/** + * __get_first_port - get the first port in the bond + * @bond: the bond we're looking at + * + * Return the port of the first slave in @bond, or %NULL if it can't be found. + */ +static inline struct port *__get_first_port(struct bonding *bond) +{ + if (bond->slave_cnt == 0) { + return NULL; + } + + return &(SLAVE_AD_INFO(bond->first_slave).port); +} + +/** + * __get_next_port - get the next port in the bond + * @port: the port we're looking at + * + * Return the port of the slave that is next in line of @port's slave in the + * bond, or %NULL if it can't be found. + */ +static inline struct port *__get_next_port(struct port *port) +{ + struct bonding *bond = __get_bond_by_port(port); + struct slave *slave = port->slave; + + // If there's no bond for this port, or this is the last slave + if ((bond == NULL) || (slave->next == bond->first_slave)) { + return NULL; + } + + return &(SLAVE_AD_INFO(slave->next).port); +} + +/** + * __get_first_agg - get the first aggregator in the bond + * @bond: the bond we're looking at + * + * Return the aggregator of the first slave in @bond, or %NULL if it can't be + * found. + */ +static inline struct aggregator *__get_first_agg(struct port *port) +{ + struct bonding *bond = __get_bond_by_port(port); + + // If there's no bond for this port, or bond has no slaves + if ((bond == NULL) || (bond->slave_cnt == 0)) { + return NULL; + } + + return &(SLAVE_AD_INFO(bond->first_slave).aggregator); +} + +/** + * __get_next_agg - get the next aggregator in the bond + * @aggregator: the aggregator we're looking at + * + * Return the aggregator of the slave that is next in line of @aggregator's + * slave in the bond, or %NULL if it can't be found. + */ +static inline struct aggregator *__get_next_agg(struct aggregator *aggregator) +{ + struct slave *slave = aggregator->slave; + struct bonding *bond = bond_get_bond_by_slave(slave); + + // If there's no bond for this aggregator, or this is the last slave + if ((bond == NULL) || (slave->next == bond->first_slave)) { + return NULL; + } + + return &(SLAVE_AD_INFO(slave->next).aggregator); +} + +/** + * __disable_port - disable the port's slave + * @port: the port we're looking at + * + */ +static inline void __disable_port(struct port *port) +{ + bond_set_slave_inactive_flags(port->slave); +} + +/** + * __enable_port - enable the port's slave, if it's up + * @port: the port we're looking at + * + */ +static inline void __enable_port(struct port *port) +{ + struct slave *slave = port->slave; + + if ((slave->link == BOND_LINK_UP) && IS_UP(slave->dev)) { + bond_set_slave_active_flags(slave); + } +} + +/** + * __port_is_enabled - check if the port's slave is in active state + * @port: the port we're looking at + * + */ +static inline int __port_is_enabled(struct port *port) +{ + return(port->slave->state == BOND_STATE_ACTIVE); +} + +/** + * __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. + */ +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_INFO(bond).agg_select_mode; +} + +/** + * __check_agg_selection_timer - check if the selection timer has expired + * @port: the port we're looking at + * + */ +static inline int __check_agg_selection_timer(struct port *port) +{ + struct bonding *bond = __get_bond_by_port(port); + + if (bond == NULL) { + return 0; + } + + return BOND_AD_INFO(bond).agg_select_timer ? 1 : 0; +} + +/** + * __get_rx_machine_lock - lock the port's RX machine + * @port: the port we're looking at + * + */ +static inline void __get_rx_machine_lock(struct port *port) +{ + spin_lock(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); +} + +/** + * __release_rx_machine_lock - unlock the port's RX machine + * @port: the port we're looking at + * + */ +static inline void __release_rx_machine_lock(struct port *port) +{ + spin_unlock(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); +} + +/** + * __get_link_speed - get a port's speed + * @port: the port we're looking at + * + * Return @port's speed in 802.3ad bitmask format. i.e. one of: + * 0, + * %AD_LINK_SPEED_BITMASK_10MBPS, + * %AD_LINK_SPEED_BITMASK_100MBPS, + * %AD_LINK_SPEED_BITMASK_1000MBPS + */ +static u16 __get_link_speed(struct port *port) +{ + struct slave *slave = port->slave; + u16 speed; + + /* this if covers only a special case: when the configuration starts with + * link down, it sets the speed to 0. + * This is done in spite of the fact that the e100 driver reports 0 to be + * compatible with MVT in the future.*/ + if (slave->link != BOND_LINK_UP) { + speed=0; + } else { + switch (slave->speed) { + case SPEED_10: + speed = AD_LINK_SPEED_BITMASK_10MBPS; + break; + + case SPEED_100: + speed = AD_LINK_SPEED_BITMASK_100MBPS; + break; + + case SPEED_1000: + speed = AD_LINK_SPEED_BITMASK_1000MBPS; + break; + + default: + speed = 0; // unknown speed value from ethtool. shouldn't happen + break; + } + } + + dprintk("Port %d Received link speed %d update from adapter\n", port->actor_port_number, speed); + return speed; +} + +/** + * __get_duplex - get a port's duplex + * @port: the port we're looking at + * + * Return @port's duplex in 802.3ad bitmask format. i.e.: + * 0x01 if in full duplex + * 0x00 otherwise + */ +static u8 __get_duplex(struct port *port) +{ + struct slave *slave = port->slave; + + u8 retval; + + // handling a special case: when the configuration starts with + // link down, it sets the duplex to 0. + if (slave->link != BOND_LINK_UP) { + retval=0x0; + } else { + switch (slave->duplex) { + case DUPLEX_FULL: + retval=0x1; + dprintk("Port %d Received status full duplex update from adapter\n", port->actor_port_number); + break; + case DUPLEX_HALF: + default: + retval=0x0; + dprintk("Port %d Received status NOT full duplex update from adapter\n", port->actor_port_number); + break; + } + } + return retval; +} + +/** + * __initialize_port_locks - initialize a port's RX machine spinlock + * @port: the port we're looking at + * + */ +static inline void __initialize_port_locks(struct port *port) +{ + // make sure it isn't called twice + spin_lock_init(&(SLAVE_AD_INFO(port->slave).rx_machine_lock)); +} + +//conversions +/** + * __ntohs_lacpdu - convert the contents of a LACPDU to host byte order + * @lacpdu: the speicifed lacpdu + * + * For each multi-byte field in the lacpdu, convert its content + */ +static void __ntohs_lacpdu(struct lacpdu *lacpdu) +{ + if (lacpdu) { + lacpdu->actor_system_priority = ntohs(lacpdu->actor_system_priority); + lacpdu->actor_key = ntohs(lacpdu->actor_key); + lacpdu->actor_port_priority = ntohs(lacpdu->actor_port_priority); + lacpdu->actor_port = ntohs(lacpdu->actor_port); + lacpdu->partner_system_priority = ntohs(lacpdu->partner_system_priority); + lacpdu->partner_key = ntohs(lacpdu->partner_key); + lacpdu->partner_port_priority = ntohs(lacpdu->partner_port_priority); + lacpdu->partner_port = ntohs(lacpdu->partner_port); + lacpdu->collector_max_delay = ntohs(lacpdu->collector_max_delay); + } +} + +/** + * __ad_timer_to_ticks - convert a given timer type to AD module ticks + * @timer_type: which timer to operate + * @par: timer parameter. see below + * + * If @timer_type is %current_while_timer, @par indicates long/short timer. + * If @timer_type is %periodic_timer, @par is one of %FAST_PERIODIC_TIME, + * %SLOW_PERIODIC_TIME. + */ +static u16 __ad_timer_to_ticks(u16 timer_type, u16 par) +{ + u16 retval=0; //to silence the compiler + + switch (timer_type) { + case AD_CURRENT_WHILE_TIMER: // for rx machine usage + if (par) { // for short or long timeout + retval = (AD_SHORT_TIMEOUT_TIME*ad_ticks_per_sec); // short timeout + } else { + retval = (AD_LONG_TIMEOUT_TIME*ad_ticks_per_sec); // long timeout + } + break; + case AD_ACTOR_CHURN_TIMER: // for local churn machine + retval = (AD_CHURN_DETECTION_TIME*ad_ticks_per_sec); + break; + case AD_PERIODIC_TIMER: // for periodic machine + retval = (par*ad_ticks_per_sec); // long timeout + break; + case AD_PARTNER_CHURN_TIMER: // for remote churn machine + retval = (AD_CHURN_DETECTION_TIME*ad_ticks_per_sec); + break; + case AD_WAIT_WHILE_TIMER: // for selection machine + retval = (AD_AGGREGATE_WAIT_TIME*ad_ticks_per_sec); + break; + } + return retval; +} + + +///////////////////////////////////////////////////////////////////////////////// +// ================= ad_rx_machine helper functions ================== +///////////////////////////////////////////////////////////////////////////////// + +/** + * __record_pdu - record parameters from a received lacpdu + * @lacpdu: the lacpdu we've received + * @port: the port we're looking at + * + * Record the parameter values for the Actor carried in a received lacpdu as + * the current partner operational parameter values and sets + * actor_oper_port_state.defaulted to FALSE. + */ +static void __record_pdu(struct lacpdu *lacpdu, struct port *port) +{ + // validate lacpdu and port + if (lacpdu && port) { + // record the new parameter values for the partner operational + port->partner_oper_port_number = lacpdu->actor_port; + port->partner_oper_port_priority = lacpdu->actor_port_priority; + port->partner_oper_system = lacpdu->actor_system; + port->partner_oper_system_priority = lacpdu->actor_system_priority; + port->partner_oper_key = lacpdu->actor_key; + // zero partener's lase states + port->partner_oper_port_state = 0; + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_LACP_ACTIVITY); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_LACP_TIMEOUT); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_AGGREGATION); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_SYNCHRONIZATION); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_COLLECTING); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_DISTRIBUTING); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_DEFAULTED); + port->partner_oper_port_state |= (lacpdu->actor_state & AD_STATE_EXPIRED); + + // set actor_oper_port_state.defaulted to FALSE + port->actor_oper_port_state &= ~AD_STATE_DEFAULTED; + + // set the partner sync. to on if the partner is sync. and the port is matched + if ((port->sm_vars & AD_PORT_MATCHED) && (lacpdu->actor_state & AD_STATE_SYNCHRONIZATION)) { + port->partner_oper_port_state |= AD_STATE_SYNCHRONIZATION; + } else { + port->partner_oper_port_state &= ~AD_STATE_SYNCHRONIZATION; + } + } +} + +/** + * __record_default - record default parameters + * @port: the port we're looking at + * + * This function records the default parameter values for the partner carried + * in the Partner Admin parameters as the current partner operational parameter + * values and sets actor_oper_port_state.defaulted to TRUE. + */ +static void __record_default(struct port *port) +{ + // validate the port + if (port) { + // record the partner admin parameters + port->partner_oper_port_number = port->partner_admin_port_number; + port->partner_oper_port_priority = port->partner_admin_port_priority; + port->partner_oper_system = port->partner_admin_system; + port->partner_oper_system_priority = port->partner_admin_system_priority; + port->partner_oper_key = port->partner_admin_key; + port->partner_oper_port_state = port->partner_admin_port_state; + + // set actor_oper_port_state.defaulted to true + port->actor_oper_port_state |= AD_STATE_DEFAULTED; + } +} + +/** + * __update_selected - update a port's Selected variable from a received lacpdu + * @lacpdu: the lacpdu we've received + * @port: the port we're looking at + * + * Update the value of the selected variable, using parameter values from a + * newly received lacpdu. The parameter values for the Actor carried in the + * received PDU are compared with the corresponding operational parameter + * values for the ports partner. If one or more of the comparisons shows that + * the value(s) received in the PDU differ from the current operational values, + * then selected is set to FALSE and actor_oper_port_state.synchronization is + * set to out_of_sync. Otherwise, selected remains unchanged. + */ +static void __update_selected(struct lacpdu *lacpdu, struct port *port) +{ + // validate lacpdu and port + if (lacpdu && port) { + // check if any parameter is different + if ((lacpdu->actor_port != port->partner_oper_port_number) || + (lacpdu->actor_port_priority != port->partner_oper_port_priority) || + MAC_ADDRESS_COMPARE(&(lacpdu->actor_system), &(port->partner_oper_system)) || + (lacpdu->actor_system_priority != port->partner_oper_system_priority) || + (lacpdu->actor_key != port->partner_oper_key) || + ((lacpdu->actor_state & AD_STATE_AGGREGATION) != (port->partner_oper_port_state & AD_STATE_AGGREGATION)) + ) { + // update the state machine Selected variable + port->sm_vars &= ~AD_PORT_SELECTED; + } + } +} + +/** + * __update_default_selected - update a port's Selected variable from Partner + * @port: the port we're looking at + * + * This function updates the value of the selected variable, using the partner + * administrative parameter values. The administrative values are compared with + * the corresponding operational parameter values for the partner. If one or + * more of the comparisons shows that the administrative value(s) differ from + * the current operational values, then Selected is set to FALSE and + * actor_oper_port_state.synchronization is set to OUT_OF_SYNC. Otherwise, + * Selected remains unchanged. + */ +static void __update_default_selected(struct port *port) +{ + // validate the port + if (port) { + // check if any parameter is different + if ((port->partner_admin_port_number != port->partner_oper_port_number) || + (port->partner_admin_port_priority != port->partner_oper_port_priority) || + MAC_ADDRESS_COMPARE(&(port->partner_admin_system), &(port->partner_oper_system)) || + (port->partner_admin_system_priority != port->partner_oper_system_priority) || + (port->partner_admin_key != port->partner_oper_key) || + ((port->partner_admin_port_state & AD_STATE_AGGREGATION) != (port->partner_oper_port_state & AD_STATE_AGGREGATION)) + ) { + // update the state machine Selected variable + port->sm_vars &= ~AD_PORT_SELECTED; + } + } +} + +/** + * __choose_matched - update a port's matched variable from a received lacpdu + * @lacpdu: the lacpdu we've received + * @port: the port we're looking at + * + * Update the value of the matched variable, using parameter values from a + * newly received lacpdu. Parameter values for the partner carried in the + * received PDU are compared with the corresponding operational parameter + * values for the actor. Matched is set to TRUE if all of these parameters + * match and the PDU parameter partner_state.aggregation has the same value as + * actor_oper_port_state.aggregation and lacp will actively maintain the link + * in the aggregation. Matched is also set to TRUE if the value of + * actor_state.aggregation in the received PDU is set to FALSE, i.e., indicates + * an individual link and lacp will actively maintain the link. Otherwise, + * matched is set to FALSE. LACP is considered to be actively maintaining the + * link if either the PDU's actor_state.lacp_activity variable is TRUE or both + * the actor's actor_oper_port_state.lacp_activity and the PDU's + * partner_state.lacp_activity variables are TRUE. + */ +static void __choose_matched(struct lacpdu *lacpdu, struct port *port) +{ + // validate lacpdu and port + if (lacpdu && port) { + // check if all parameters are alike + if (((lacpdu->partner_port == port->actor_port_number) && + (lacpdu->partner_port_priority == port->actor_port_priority) && + !MAC_ADDRESS_COMPARE(&(lacpdu->partner_system), &(port->actor_system)) && + (lacpdu->partner_system_priority == port->actor_system_priority) && + (lacpdu->partner_key == port->actor_oper_port_key) && + ((lacpdu->partner_state & AD_STATE_AGGREGATION) == (port->actor_oper_port_state & AD_STATE_AGGREGATION))) || + // or this is individual link(aggregation == FALSE) + ((lacpdu->actor_state & AD_STATE_AGGREGATION) == 0) + ) { + // update the state machine Matched variable + port->sm_vars |= AD_PORT_MATCHED; + } else { + port->sm_vars &= ~AD_PORT_MATCHED; + } + } +} + +/** + * __update_ntt - update a port's ntt variable from a received lacpdu + * @lacpdu: the lacpdu we've received + * @port: the port we're looking at + * + * Updates the value of the ntt variable, using parameter values from a newly + * received lacpdu. The parameter values for the partner carried in the + * received PDU are compared with the corresponding operational parameter + * values for the Actor. If one or more of the comparisons shows that the + * value(s) received in the PDU differ from the current operational values, + * then ntt is set to TRUE. Otherwise, ntt remains unchanged. + */ +static void __update_ntt(struct lacpdu *lacpdu, struct port *port) +{ + // validate lacpdu and port + if (lacpdu && port) { + // check if any parameter is different + if ((lacpdu->partner_port != port->actor_port_number) || + (lacpdu->partner_port_priority != port->actor_port_priority) || + MAC_ADDRESS_COMPARE(&(lacpdu->partner_system), &(port->actor_system)) || + (lacpdu->partner_system_priority != port->actor_system_priority) || + (lacpdu->partner_key != port->actor_oper_port_key) || + ((lacpdu->partner_state & AD_STATE_LACP_ACTIVITY) != (port->actor_oper_port_state & AD_STATE_LACP_ACTIVITY)) || + ((lacpdu->partner_state & AD_STATE_LACP_TIMEOUT) != (port->actor_oper_port_state & AD_STATE_LACP_TIMEOUT)) || + ((lacpdu->partner_state & AD_STATE_SYNCHRONIZATION) != (port->actor_oper_port_state & AD_STATE_SYNCHRONIZATION)) || + ((lacpdu->partner_state & AD_STATE_AGGREGATION) != (port->actor_oper_port_state & AD_STATE_AGGREGATION)) + ) { + // set ntt to be TRUE + port->ntt = 1; + } + } +} + +/** + * __attach_bond_to_agg + * @port: the port we're looking at + * + * Handle the attaching of the port's control parser/multiplexer and the + * aggregator. This function does nothing since the parser/multiplexer of the + * receive and the parser/multiplexer of the aggregator are already combined. + */ +static void __attach_bond_to_agg(struct port *port) +{ + port=NULL; // just to satisfy the compiler + // This function does nothing since the parser/multiplexer of the receive + // and the parser/multiplexer of the aggregator are already combined +} + +/** + * __detach_bond_from_agg + * @port: the port we're looking at + * + * Handle the detaching of the port's control parser/multiplexer from the + * aggregator. This function does nothing since the parser/multiplexer of the + * receive and the parser/multiplexer of the aggregator are already combined. + */ +static void __detach_bond_from_agg(struct port *port) +{ + port=NULL; // just to satisfy the compiler + // This function does nothing sience the parser/multiplexer of the receive + // and the parser/multiplexer of the aggregator are already combined +} + +/** + * __agg_ports_are_ready - check if all ports in an aggregator are ready + * @aggregator: the aggregator we're looking at + * + */ +static int __agg_ports_are_ready(struct aggregator *aggregator) +{ + struct port *port; + int retval = 1; + + if (aggregator) { + // scan all ports in this aggregator to verfy if they are all ready + for (port=aggregator->lag_ports; port; port=port->next_port_in_aggregator) { + if (!(port->sm_vars & AD_PORT_READY_N)) { + retval = 0; + break; + } + } + } + + return retval; +} + +/** + * __set_agg_ports_ready - set value of Ready bit in all ports of an aggregator + * @aggregator: the aggregator we're looking at + * @val: Should the ports' ready bit be set on or off + * + */ +static void __set_agg_ports_ready(struct aggregator *aggregator, int val) +{ + struct port *port; + + for (port=aggregator->lag_ports; port; port=port->next_port_in_aggregator) { + if (val) { + port->sm_vars |= AD_PORT_READY; + } else { + port->sm_vars &= ~AD_PORT_READY; + } + } +} + +/** + * __get_agg_bandwidth - get the total bandwidth of an aggregator + * @aggregator: the aggregator we're looking at + * + */ +static u32 __get_agg_bandwidth(struct aggregator *aggregator) +{ + u32 bandwidth=0; + u32 basic_speed; + + if (aggregator->num_of_ports) { + basic_speed = __get_link_speed(aggregator->lag_ports); + switch (basic_speed) { + case AD_LINK_SPEED_BITMASK_1MBPS: + bandwidth = aggregator->num_of_ports; + break; + case AD_LINK_SPEED_BITMASK_10MBPS: + bandwidth = aggregator->num_of_ports * 10; + break; + case AD_LINK_SPEED_BITMASK_100MBPS: + bandwidth = aggregator->num_of_ports * 100; + break; + case AD_LINK_SPEED_BITMASK_1000MBPS: + bandwidth = aggregator->num_of_ports * 1000; + break; + default: + bandwidth=0; // to silent the compilor .... + } + } + return bandwidth; +} + +/** + * __get_active_agg - get the current active aggregator + * @aggregator: the aggregator we're looking at + * + */ +static struct aggregator *__get_active_agg(struct aggregator *aggregator) +{ + struct aggregator *retval = NULL; + + for (; aggregator; aggregator = __get_next_agg(aggregator)) { + if (aggregator->is_active) { + retval = aggregator; + break; + } + } + + return retval; +} + +/** + * __update_lacpdu_from_port - update a port's lacpdu fields + * @port: the port we're looking at + * + */ +static inline void __update_lacpdu_from_port(struct port *port) +{ + struct lacpdu *lacpdu = &port->lacpdu; + + /* update current actual Actor parameters */ + /* lacpdu->subtype initialized + * lacpdu->version_number initialized + * lacpdu->tlv_type_actor_info initialized + * lacpdu->actor_information_length initialized + */ + + lacpdu->actor_system_priority = port->actor_system_priority; + lacpdu->actor_system = port->actor_system; + lacpdu->actor_key = port->actor_oper_port_key; + lacpdu->actor_port_priority = port->actor_port_priority; + lacpdu->actor_port = port->actor_port_number; + lacpdu->actor_state = port->actor_oper_port_state; + + /* lacpdu->reserved_3_1 initialized + * lacpdu->tlv_type_partner_info initialized + * lacpdu->partner_information_length initialized + */ + + lacpdu->partner_system_priority = port->partner_oper_system_priority; + lacpdu->partner_system = port->partner_oper_system; + lacpdu->partner_key = port->partner_oper_key; + lacpdu->partner_port_priority = port->partner_oper_port_priority; + lacpdu->partner_port = port->partner_oper_port_number; + lacpdu->partner_state = port->partner_oper_port_state; + + /* lacpdu->reserved_3_2 initialized + * lacpdu->tlv_type_collector_info initialized + * lacpdu->collector_information_length initialized + * collector_max_delay initialized + * reserved_12[12] initialized + * tlv_type_terminator initialized + * terminator_length initialized + * reserved_50[50] initialized + */ + + /* Convert all non u8 parameters to Big Endian for transmit */ + __ntohs_lacpdu(lacpdu); +} + +////////////////////////////////////////////////////////////////////////////////////// +// ================= main 802.3ad protocol code ====================================== +////////////////////////////////////////////////////////////////////////////////////// + +/** + * ad_lacpdu_send - send out a lacpdu packet on a given port + * @port: the port we're looking at + * + * Returns: 0 on success + * < 0 on error + */ +static int ad_lacpdu_send(struct port *port) +{ + struct slave *slave = port->slave; + struct sk_buff *skb; + struct lacpdu_header *lacpdu_header; + int length = sizeof(struct lacpdu_header); + struct mac_addr lacpdu_multicast_address = AD_MULTICAST_LACPDU_ADDR; + + skb = dev_alloc_skb(length); + if (!skb) { + return -ENOMEM; + } + + skb->dev = slave->dev; + skb->mac.raw = skb->data; + skb->nh.raw = skb->data + ETH_HLEN; + skb->protocol = PKT_TYPE_LACPDU; + skb->priority = TC_PRIO_CONTROL; + + lacpdu_header = (struct lacpdu_header *)skb_put(skb, length); + + lacpdu_header->ad_header.destination_address = lacpdu_multicast_address; + /* Note: source addres is set to be the member's PERMANENT address, because we use it + to identify loopback lacpdus in receive. */ + lacpdu_header->ad_header.source_address = *((struct mac_addr *)(slave->perm_hwaddr)); + lacpdu_header->ad_header.length_type = PKT_TYPE_LACPDU; + + lacpdu_header->lacpdu = port->lacpdu; // struct copy + + dev_queue_xmit(skb); + + return 0; +} + +/** + * ad_marker_send - send marker information/response on a given port + * @port: the port we're looking at + * @marker: marker data to send + * + * Returns: 0 on success + * < 0 on error + */ +static int ad_marker_send(struct port *port, struct marker *marker) +{ + struct slave *slave = port->slave; + struct sk_buff *skb; + struct marker_header *marker_header; + int length = sizeof(struct marker_header); + struct mac_addr lacpdu_multicast_address = AD_MULTICAST_LACPDU_ADDR; + + skb = dev_alloc_skb(length + 16); + if (!skb) { + return -ENOMEM; + } + + skb_reserve(skb, 16); + + skb->dev = slave->dev; + skb->mac.raw = skb->data; + skb->nh.raw = skb->data + ETH_HLEN; + skb->protocol = PKT_TYPE_LACPDU; + + marker_header = (struct marker_header *)skb_put(skb, length); + + marker_header->ad_header.destination_address = lacpdu_multicast_address; + /* Note: source addres is set to be the member's PERMANENT address, because we use it + to identify loopback MARKERs in receive. */ + marker_header->ad_header.source_address = *((struct mac_addr *)(slave->perm_hwaddr)); + marker_header->ad_header.length_type = PKT_TYPE_LACPDU; + + marker_header->marker = *marker; // struct copy + + dev_queue_xmit(skb); + + return 0; +} + +/** + * ad_mux_machine - handle a port's mux state machine + * @port: the port we're looking at + * + */ +static void ad_mux_machine(struct port *port) +{ + mux_states_t last_state; + + // keep current State Machine state to compare later if it was changed + last_state = port->sm_mux_state; + + if (port->sm_vars & AD_PORT_BEGIN) { + port->sm_mux_state = AD_MUX_DETACHED; // next state + } else { + switch (port->sm_mux_state) { + case AD_MUX_DETACHED: + if ((port->sm_vars & AD_PORT_SELECTED) || (port->sm_vars & AD_PORT_STANDBY)) { // if SELECTED or STANDBY + port->sm_mux_state = AD_MUX_WAITING; // next state + } + break; + case AD_MUX_WAITING: + // if SELECTED == FALSE return to DETACH state + if (!(port->sm_vars & AD_PORT_SELECTED)) { // if UNSELECTED + port->sm_vars &= ~AD_PORT_READY_N; + // in order to withhold the Selection Logic to check all ports READY_N value + // every callback cycle to update ready variable, we check READY_N and update READY here + __set_agg_ports_ready(port->aggregator, __agg_ports_are_ready(port->aggregator)); + port->sm_mux_state = AD_MUX_DETACHED; // next state + break; + } + + // check if the wait_while_timer expired + if (port->sm_mux_timer_counter && !(--port->sm_mux_timer_counter)) { + port->sm_vars |= AD_PORT_READY_N; + } + + // in order to withhold the selection logic to check all ports READY_N value + // every callback cycle to update ready variable, we check READY_N and update READY here + __set_agg_ports_ready(port->aggregator, __agg_ports_are_ready(port->aggregator)); + + // if the wait_while_timer expired, and the port is in READY state, move to ATTACHED state + if ((port->sm_vars & AD_PORT_READY) && !port->sm_mux_timer_counter) { + port->sm_mux_state = AD_MUX_ATTACHED; // next state + } + break; + case AD_MUX_ATTACHED: + // check also if agg_select_timer expired(so the edable port will take place only after this timer) + if ((port->sm_vars & AD_PORT_SELECTED) && (port->partner_oper_port_state & AD_STATE_SYNCHRONIZATION) && !__check_agg_selection_timer(port)) { + port->sm_mux_state = AD_MUX_COLLECTING_DISTRIBUTING;// next state + } else if (!(port->sm_vars & AD_PORT_SELECTED) || (port->sm_vars & AD_PORT_STANDBY)) { // if UNSELECTED or STANDBY + port->sm_vars &= ~AD_PORT_READY_N; + // in order to withhold the selection logic to check all ports READY_N value + // every callback cycle to update ready variable, we check READY_N and update READY here + __set_agg_ports_ready(port->aggregator, __agg_ports_are_ready(port->aggregator)); + port->sm_mux_state = AD_MUX_DETACHED;// next state + } + break; + case AD_MUX_COLLECTING_DISTRIBUTING: + if (!(port->sm_vars & AD_PORT_SELECTED) || (port->sm_vars & AD_PORT_STANDBY) || + !(port->partner_oper_port_state & AD_STATE_SYNCHRONIZATION) + ) { + port->sm_mux_state = AD_MUX_ATTACHED;// next state + + } else { + // if port state hasn't changed make + // sure that a collecting distributing + // port in an active aggregator is enabled + if (port->aggregator && + port->aggregator->is_active && + !__port_is_enabled(port)) { + + __enable_port(port); + } + } + break; + default: //to silence the compiler + break; + } + } + + // check if the state machine was changed + if (port->sm_mux_state != last_state) { + dprintk("Mux Machine: Port=%d, Last State=%d, Curr State=%d\n", port->actor_port_number, last_state, port->sm_mux_state); + switch (port->sm_mux_state) { + case AD_MUX_DETACHED: + __detach_bond_from_agg(port); + port->actor_oper_port_state &= ~AD_STATE_SYNCHRONIZATION; + ad_disable_collecting_distributing(port); + port->actor_oper_port_state &= ~AD_STATE_COLLECTING; + port->actor_oper_port_state &= ~AD_STATE_DISTRIBUTING; + port->ntt = 1; + break; + case AD_MUX_WAITING: + port->sm_mux_timer_counter = __ad_timer_to_ticks(AD_WAIT_WHILE_TIMER, 0); + break; + case AD_MUX_ATTACHED: + __attach_bond_to_agg(port); + port->actor_oper_port_state |= AD_STATE_SYNCHRONIZATION; + port->actor_oper_port_state &= ~AD_STATE_COLLECTING; + port->actor_oper_port_state &= ~AD_STATE_DISTRIBUTING; + ad_disable_collecting_distributing(port); + port->ntt = 1; + break; + case AD_MUX_COLLECTING_DISTRIBUTING: + port->actor_oper_port_state |= AD_STATE_COLLECTING; + port->actor_oper_port_state |= AD_STATE_DISTRIBUTING; + ad_enable_collecting_distributing(port); + port->ntt = 1; + break; + default: //to silence the compiler + break; + } + } +} + +/** + * ad_rx_machine - handle a port's rx State Machine + * @lacpdu: the lacpdu we've received + * @port: the port we're looking at + * + * If lacpdu arrived, stop previous timer (if exists) and set the next state as + * CURRENT. If timer expired set the state machine in the proper state. + * In other cases, this function checks if we need to switch to other state. + */ +static void ad_rx_machine(struct lacpdu *lacpdu, struct port *port) +{ + rx_states_t last_state; + + // Lock to prevent 2 instances of this function to run simultaneously(rx interrupt and periodic machine callback) + __get_rx_machine_lock(port); + + // keep current State Machine state to compare later if it was changed + last_state = port->sm_rx_state; + + // check if state machine should change state + // first, check if port was reinitialized + if (port->sm_vars & AD_PORT_BEGIN) { + port->sm_rx_state = AD_RX_INITIALIZE; // next state + } + // check if port is not enabled + else if (!(port->sm_vars & AD_PORT_BEGIN) && !port->is_enabled && !(port->sm_vars & AD_PORT_MOVED)) { + port->sm_rx_state = AD_RX_PORT_DISABLED; // next state + } + // check if new lacpdu arrived + else if (lacpdu && ((port->sm_rx_state == AD_RX_EXPIRED) || (port->sm_rx_state == AD_RX_DEFAULTED) || (port->sm_rx_state == AD_RX_CURRENT))) { + port->sm_rx_timer_counter = 0; // zero timer + port->sm_rx_state = AD_RX_CURRENT; + } else { + // if timer is on, and if it is expired + if (port->sm_rx_timer_counter && !(--port->sm_rx_timer_counter)) { + switch (port->sm_rx_state) { + case AD_RX_EXPIRED: + port->sm_rx_state = AD_RX_DEFAULTED; // next state + break; + case AD_RX_CURRENT: + port->sm_rx_state = AD_RX_EXPIRED; // next state + break; + default: //to silence the compiler + break; + } + } else { + // if no lacpdu arrived and no timer is on + switch (port->sm_rx_state) { + case AD_RX_PORT_DISABLED: + if (port->sm_vars & AD_PORT_MOVED) { + port->sm_rx_state = AD_RX_INITIALIZE; // next state + } else if (port->is_enabled && (port->sm_vars & AD_PORT_LACP_ENABLED)) { + port->sm_rx_state = AD_RX_EXPIRED; // next state + } else if (port->is_enabled && ((port->sm_vars & AD_PORT_LACP_ENABLED) == 0)) { + port->sm_rx_state = AD_RX_LACP_DISABLED; // next state + } + break; + default: //to silence the compiler + break; + + } + } + } + + // check if the State machine was changed or new lacpdu arrived + if ((port->sm_rx_state != last_state) || (lacpdu)) { + dprintk("Rx Machine: Port=%d, Last State=%d, Curr State=%d\n", port->actor_port_number, last_state, port->sm_rx_state); + switch (port->sm_rx_state) { + case AD_RX_INITIALIZE: + if (!(port->actor_oper_port_key & AD_DUPLEX_KEY_BITS)) { + port->sm_vars &= ~AD_PORT_LACP_ENABLED; + } else { + port->sm_vars |= AD_PORT_LACP_ENABLED; + } + port->sm_vars &= ~AD_PORT_SELECTED; + __record_default(port); + port->actor_oper_port_state &= ~AD_STATE_EXPIRED; + port->sm_vars &= ~AD_PORT_MOVED; + port->sm_rx_state = AD_RX_PORT_DISABLED; // next state + + /*- Fall Through -*/ + + case AD_RX_PORT_DISABLED: + port->sm_vars &= ~AD_PORT_MATCHED; + break; + case AD_RX_LACP_DISABLED: + port->sm_vars &= ~AD_PORT_SELECTED; + __record_default(port); + port->partner_oper_port_state &= ~AD_STATE_AGGREGATION; + port->sm_vars |= AD_PORT_MATCHED; + port->actor_oper_port_state &= ~AD_STATE_EXPIRED; + break; + case AD_RX_EXPIRED: + //Reset of the Synchronization flag. (Standard 43.4.12) + //This reset cause to disable this port in the COLLECTING_DISTRIBUTING state of the + //mux machine in case of EXPIRED even if LINK_DOWN didn't arrive for the port. + port->partner_oper_port_state &= ~AD_STATE_SYNCHRONIZATION; + port->sm_vars &= ~AD_PORT_MATCHED; + port->partner_oper_port_state |= AD_SHORT_TIMEOUT; + port->sm_rx_timer_counter = __ad_timer_to_ticks(AD_CURRENT_WHILE_TIMER, (u16)(AD_SHORT_TIMEOUT)); + port->actor_oper_port_state |= AD_STATE_EXPIRED; + break; + case AD_RX_DEFAULTED: + __update_default_selected(port); + __record_default(port); + port->sm_vars |= AD_PORT_MATCHED; + port->actor_oper_port_state &= ~AD_STATE_EXPIRED; + break; + case AD_RX_CURRENT: + // detect loopback situation + if (!MAC_ADDRESS_COMPARE(&(lacpdu->actor_system), &(port->actor_system))) { + // INFO_RECEIVED_LOOPBACK_FRAMES + printk(KERN_ERR DRV_NAME ": An illegal loopback occurred on adapter (%s)\n", + port->slave->dev->name); + printk(KERN_ERR "Check the configuration to verify that all Adapters " + "are connected to 802.3ad compliant switch ports\n"); + __release_rx_machine_lock(port); + return; + } + __update_selected(lacpdu, port); + __update_ntt(lacpdu, port); + __record_pdu(lacpdu, port); + __choose_matched(lacpdu, port); + port->sm_rx_timer_counter = __ad_timer_to_ticks(AD_CURRENT_WHILE_TIMER, (u16)(port->actor_oper_port_state & AD_STATE_LACP_TIMEOUT)); + port->actor_oper_port_state &= ~AD_STATE_EXPIRED; + // verify that if the aggregator is enabled, the port is enabled too. + //(because if the link goes down for a short time, the 802.3ad will not + // catch it, and the port will continue to be disabled) + if (port->aggregator && port->aggregator->is_active && !__port_is_enabled(port)) { + __enable_port(port); + } + break; + default: //to silence the compiler + break; + } + } + __release_rx_machine_lock(port); +} + +/** + * ad_tx_machine - handle a port's tx state machine + * @port: the port we're looking at + * + */ +static void ad_tx_machine(struct port *port) +{ + // check if tx timer expired, to verify that we do not send more than 3 packets per second + if (port->sm_tx_timer_counter && !(--port->sm_tx_timer_counter)) { + // check if there is something to send + if (port->ntt && (port->sm_vars & AD_PORT_LACP_ENABLED)) { + __update_lacpdu_from_port(port); + // send the lacpdu + if (ad_lacpdu_send(port) >= 0) { + dprintk("Sent LACPDU on port %d\n", port->actor_port_number); + // mark ntt as false, so it will not be sent again until demanded + port->ntt = 0; + } + } + // restart tx timer(to verify that we will not exceed AD_MAX_TX_IN_SECOND + port->sm_tx_timer_counter=ad_ticks_per_sec/AD_MAX_TX_IN_SECOND; + } +} + +/** + * ad_periodic_machine - handle a port's periodic state machine + * @port: the port we're looking at + * + * Turn ntt flag on priodically to perform periodic transmission of lacpdu's. + */ +static void ad_periodic_machine(struct port *port) +{ + periodic_states_t last_state; + + // keep current state machine state to compare later if it was changed + last_state = port->sm_periodic_state; + + // check if port was reinitialized + if (((port->sm_vars & AD_PORT_BEGIN) || !(port->sm_vars & AD_PORT_LACP_ENABLED) || !port->is_enabled) || + (!(port->actor_oper_port_state & AD_STATE_LACP_ACTIVITY) && !(port->partner_oper_port_state & AD_STATE_LACP_ACTIVITY)) + ) { + port->sm_periodic_state = AD_NO_PERIODIC; // next state + } + // check if state machine should change state + else if (port->sm_periodic_timer_counter) { + // check if periodic state machine expired + if (!(--port->sm_periodic_timer_counter)) { + // if expired then do tx + port->sm_periodic_state = AD_PERIODIC_TX; // next state + } else { + // If not expired, check if there is some new timeout parameter from the partner state + switch (port->sm_periodic_state) { + case AD_FAST_PERIODIC: + if (!(port->partner_oper_port_state & AD_STATE_LACP_TIMEOUT)) { + port->sm_periodic_state = AD_SLOW_PERIODIC; // next state + } + break; + case AD_SLOW_PERIODIC: + if ((port->partner_oper_port_state & AD_STATE_LACP_TIMEOUT)) { + // stop current timer + port->sm_periodic_timer_counter = 0; + port->sm_periodic_state = AD_PERIODIC_TX; // next state + } + break; + default: //to silence the compiler + break; + } + } + } else { + switch (port->sm_periodic_state) { + case AD_NO_PERIODIC: + port->sm_periodic_state = AD_FAST_PERIODIC; // next state + break; + case AD_PERIODIC_TX: + if (!(port->partner_oper_port_state & AD_STATE_LACP_TIMEOUT)) { + port->sm_periodic_state = AD_SLOW_PERIODIC; // next state + } else { + port->sm_periodic_state = AD_FAST_PERIODIC; // next state + } + break; + default: //to silence the compiler + break; + } + } + + // check if the state machine was changed + if (port->sm_periodic_state != last_state) { + dprintk("Periodic Machine: Port=%d, Last State=%d, Curr State=%d\n", port->actor_port_number, last_state, port->sm_periodic_state); + switch (port->sm_periodic_state) { + case AD_NO_PERIODIC: + port->sm_periodic_timer_counter = 0; // zero timer + break; + case AD_FAST_PERIODIC: + port->sm_periodic_timer_counter = __ad_timer_to_ticks(AD_PERIODIC_TIMER, (u16)(AD_FAST_PERIODIC_TIME))-1; // decrement 1 tick we lost in the PERIODIC_TX cycle + break; + case AD_SLOW_PERIODIC: + port->sm_periodic_timer_counter = __ad_timer_to_ticks(AD_PERIODIC_TIMER, (u16)(AD_SLOW_PERIODIC_TIME))-1; // decrement 1 tick we lost in the PERIODIC_TX cycle + break; + case AD_PERIODIC_TX: + port->ntt = 1; + break; + default: //to silence the compiler + break; + } + } +} + +/** + * ad_port_selection_logic - select aggregation groups + * @port: the port we're looking at + * + * Select aggregation groups, and assign each port for it's aggregetor. The + * selection logic is called in the inititalization (after all the handshkes), + * and after every lacpdu receive (if selected is off). + */ +static void ad_port_selection_logic(struct port *port) +{ + struct aggregator *aggregator, *free_aggregator = NULL, *temp_aggregator; + struct port *last_port = NULL, *curr_port; + int found = 0; + + // if the port is already Selected, do nothing + if (port->sm_vars & AD_PORT_SELECTED) { + return; + } + + // if the port is connected to other aggregator, detach it + if (port->aggregator) { + // detach the port from its former aggregator + temp_aggregator=port->aggregator; + for (curr_port=temp_aggregator->lag_ports; curr_port; last_port=curr_port, curr_port=curr_port->next_port_in_aggregator) { + if (curr_port == port) { + temp_aggregator->num_of_ports--; + if (!last_port) {// if it is the first port attached to the aggregator + temp_aggregator->lag_ports=port->next_port_in_aggregator; + } else {// not the first port attached to the aggregator + last_port->next_port_in_aggregator=port->next_port_in_aggregator; + } + + // clear the port's relations to this aggregator + port->aggregator = NULL; + port->next_port_in_aggregator=NULL; + port->actor_port_aggregator_identifier=0; + + dprintk("Port %d left LAG %d\n", port->actor_port_number, temp_aggregator->aggregator_identifier); + // if the aggregator is empty, clear its parameters, and set it ready to be attached + if (!temp_aggregator->lag_ports) { + ad_clear_agg(temp_aggregator); + } + break; + } + } + if (!curr_port) { // meaning: the port was related to an aggregator but was not on the aggregator port list + printk(KERN_WARNING DRV_NAME ": Warning: Port %d (on %s) was " + "related to aggregator %d but was not on its port list\n", + port->actor_port_number, port->slave->dev->name, + port->aggregator->aggregator_identifier); + } + } + // search on all aggregators for a suitable aggregator for this port + for (aggregator = __get_first_agg(port); aggregator; + aggregator = __get_next_agg(aggregator)) { + + // keep a free aggregator for later use(if needed) + if (!aggregator->lag_ports) { + if (!free_aggregator) { + free_aggregator=aggregator; + } + continue; + } + // check if current aggregator suits us + if (((aggregator->actor_oper_aggregator_key == port->actor_oper_port_key) && // if all parameters match AND + !MAC_ADDRESS_COMPARE(&(aggregator->partner_system), &(port->partner_oper_system)) && + (aggregator->partner_system_priority == port->partner_oper_system_priority) && + (aggregator->partner_oper_aggregator_key == port->partner_oper_key) + ) && + ((MAC_ADDRESS_COMPARE(&(port->partner_oper_system), &(null_mac_addr)) && // partner answers + !aggregator->is_individual) // but is not individual OR + ) + ) { + // attach to the founded aggregator + port->aggregator = aggregator; + port->actor_port_aggregator_identifier=port->aggregator->aggregator_identifier; + port->next_port_in_aggregator=aggregator->lag_ports; + port->aggregator->num_of_ports++; + aggregator->lag_ports=port; + dprintk("Port %d joined LAG %d(existing LAG)\n", port->actor_port_number, port->aggregator->aggregator_identifier); + + // mark this port as selected + port->sm_vars |= AD_PORT_SELECTED; + found = 1; + break; + } + } + + // the port couldn't find an aggregator - attach it to a new aggregator + if (!found) { + if (free_aggregator) { + // assign port a new aggregator + port->aggregator = free_aggregator; + port->actor_port_aggregator_identifier=port->aggregator->aggregator_identifier; + + // update the new aggregator's parameters + // if port was responsed from the end-user + if (port->actor_oper_port_key & AD_DUPLEX_KEY_BITS) {// if port is full duplex + port->aggregator->is_individual = 0; + } else { + port->aggregator->is_individual = 1; + } + + port->aggregator->actor_admin_aggregator_key = port->actor_admin_port_key; + port->aggregator->actor_oper_aggregator_key = port->actor_oper_port_key; + port->aggregator->partner_system=port->partner_oper_system; + port->aggregator->partner_system_priority = port->partner_oper_system_priority; + port->aggregator->partner_oper_aggregator_key = port->partner_oper_key; + port->aggregator->receive_state = 1; + port->aggregator->transmit_state = 1; + port->aggregator->lag_ports = port; + port->aggregator->num_of_ports++; + + // mark this port as selected + port->sm_vars |= AD_PORT_SELECTED; + + dprintk("Port %d joined LAG %d(new LAG)\n", port->actor_port_number, port->aggregator->aggregator_identifier); + } else { + printk(KERN_ERR DRV_NAME ": Port %d (on %s) did not find a suitable aggregator\n", + port->actor_port_number, port->slave->dev->name); + } + } + // if all aggregator's ports are READY_N == TRUE, set ready=TRUE in all aggregator's ports + // 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); + } +} + +/** + * ad_agg_selection_logic - select an aggregation group for a team + * @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. + * + * 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) +{ + struct aggregator *best_aggregator = NULL, *active_aggregator = NULL; + struct aggregator *last_active_aggregator = NULL, *origin_aggregator; + struct port *port; + u16 num_of_aggs=0; + + origin_aggregator = aggregator; + + //get current active aggregator + last_active_aggregator = __get_active_agg(aggregator); + + // 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 + } + } + 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 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); + } + + // check if any partner replys + if (best_aggregator->is_individual) { + printk(KERN_WARNING DRV_NAME ": Warning: No 802.3ad response from the link partner " + "for any adapters in the bond\n"); + } + + // 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"); + } + + 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); + + // 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) { + __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 (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) { + __enable_port(port); + } + } + } +} + +/** + * ad_clear_agg - clear a given aggregator's parameters + * @aggregator: the aggregator we're looking at + * + */ +static void ad_clear_agg(struct aggregator *aggregator) +{ + if (aggregator) { + aggregator->is_individual = 0; + aggregator->actor_admin_aggregator_key = 0; + aggregator->actor_oper_aggregator_key = 0; + aggregator->partner_system = null_mac_addr; + aggregator->partner_system_priority = 0; + aggregator->partner_oper_aggregator_key = 0; + aggregator->receive_state = 0; + aggregator->transmit_state = 0; + aggregator->lag_ports = NULL; + aggregator->is_active = 0; + aggregator->num_of_ports = 0; + dprintk("LAG %d was cleared\n", aggregator->aggregator_identifier); + } +} + +/** + * ad_initialize_agg - initialize a given aggregator's parameters + * @aggregator: the aggregator we're looking at + * + */ +static void ad_initialize_agg(struct aggregator *aggregator) +{ + if (aggregator) { + ad_clear_agg(aggregator); + + aggregator->aggregator_mac_address = null_mac_addr; + aggregator->aggregator_identifier = 0; + aggregator->slave = NULL; + } +} + +/** + * ad_initialize_port - initialize a given port's parameters + * @aggregator: the aggregator we're looking at + * @lacp_fast: boolean. whether fast periodic should be used + * + */ +static void ad_initialize_port(struct port *port, int lacp_fast) +{ + if (port) { + port->actor_port_number = 1; + port->actor_port_priority = 0xff; + port->actor_system = null_mac_addr; + port->actor_system_priority = 0xffff; + port->actor_port_aggregator_identifier = 0; + port->ntt = 0; + port->actor_admin_port_key = 1; + port->actor_oper_port_key = 1; + port->actor_admin_port_state = AD_STATE_AGGREGATION | AD_STATE_LACP_ACTIVITY; + port->actor_oper_port_state = AD_STATE_AGGREGATION | AD_STATE_LACP_ACTIVITY; + + if (lacp_fast) { + port->actor_oper_port_state |= AD_STATE_LACP_TIMEOUT; + } + + port->partner_admin_system = null_mac_addr; + port->partner_oper_system = null_mac_addr; + port->partner_admin_system_priority = 0xffff; + port->partner_oper_system_priority = 0xffff; + port->partner_admin_key = 1; + port->partner_oper_key = 1; + port->partner_admin_port_number = 1; + port->partner_oper_port_number = 1; + port->partner_admin_port_priority = 0xff; + port->partner_oper_port_priority = 0xff; + port->partner_admin_port_state = 1; + port->partner_oper_port_state = 1; + port->is_enabled = 1; + // ****** private parameters ****** + port->sm_vars = 0x3; + port->sm_rx_state = 0; + port->sm_rx_timer_counter = 0; + port->sm_periodic_state = 0; + port->sm_periodic_timer_counter = 0; + port->sm_mux_state = 0; + port->sm_mux_timer_counter = 0; + port->sm_tx_state = 0; + port->sm_tx_timer_counter = 0; + port->slave = NULL; + port->aggregator = NULL; + port->next_port_in_aggregator = NULL; + port->transaction_id = 0; + + ad_initialize_lacpdu(&(port->lacpdu)); + } +} + +/** + * ad_enable_collecting_distributing - enable a port's transmit/receive + * @port: the port we're looking at + * + * Enable @port if it's in an active aggregator + */ +static void ad_enable_collecting_distributing(struct port *port) +{ + if (port->aggregator->is_active) { + dprintk("Enabling port %d(LAG %d)\n", port->actor_port_number, port->aggregator->aggregator_identifier); + __enable_port(port); + } +} + +/** + * ad_disable_collecting_distributing - disable a port's transmit/receive + * @port: the port we're looking at + * + */ +static void ad_disable_collecting_distributing(struct port *port) +{ + if (port->aggregator && MAC_ADDRESS_COMPARE(&(port->aggregator->partner_system), &(null_mac_addr))) { + dprintk("Disabling port %d(LAG %d)\n", port->actor_port_number, port->aggregator->aggregator_identifier); + __disable_port(port); + } +} + +#if 0 +/** + * ad_marker_info_send - send a marker information frame + * @port: the port we're looking at + * + * This function does nothing since we decided not to implement send and handle + * response for marker PDU's, in this stage, but only to respond to marker + * information. + */ +static void ad_marker_info_send(struct port *port) +{ + struct marker marker; + u16 index; + + // fill the marker PDU with the appropriate values + marker.subtype = 0x02; + marker.version_number = 0x01; + marker.tlv_type = AD_MARKER_INFORMATION_SUBTYPE; + marker.marker_length = 0x16; + // convert requester_port to Big Endian + marker.requester_port = (((port->actor_port_number & 0xFF) << 8) |((u16)(port->actor_port_number & 0xFF00) >> 8)); + marker.requester_system = port->actor_system; + // convert requester_port(u32) to Big Endian + marker.requester_transaction_id = (((++port->transaction_id & 0xFF) << 24) |((port->transaction_id & 0xFF00) << 8) |((port->transaction_id & 0xFF0000) >> 8) |((port->transaction_id & 0xFF000000) >> 24)); + marker.pad = 0; + marker.tlv_type_terminator = 0x00; + marker.terminator_length = 0x00; + for (index=0; index<90; index++) { + marker.reserved_90[index]=0; + } + + // send the marker information + if (ad_marker_send(port, &marker) >= 0) { + dprintk("Sent Marker Information on port %d\n", port->actor_port_number); + } +} +#endif + +/** + * ad_marker_info_received - handle receive of a Marker information frame + * @marker_info: Marker info received + * @port: the port we're looking at + * + */ +static void ad_marker_info_received(struct marker *marker_info,struct port *port) +{ + struct marker marker; + + // copy the received marker data to the response marker + //marker = *marker_info; + memcpy(&marker, marker_info, sizeof(struct marker)); + // change the marker subtype to marker response + marker.tlv_type=AD_MARKER_RESPONSE_SUBTYPE; + // send the marker response + + if (ad_marker_send(port, &marker) >= 0) { + dprintk("Sent Marker Response on port %d\n", port->actor_port_number); + } +} + +/** + * ad_marker_response_received - handle receive of a marker response frame + * @marker: marker PDU received + * @port: the port we're looking at + * + * This function does nothing since we decided not to implement send and handle + * response for marker PDU's, in this stage, but only to respond to marker + * information. + */ +static void ad_marker_response_received(struct marker *marker, struct port *port) +{ + marker=NULL; // just to satisfy the compiler + port=NULL; // just to satisfy the compiler + // DO NOTHING, SINCE WE DECIDED NOT TO IMPLEMENT THIS FEATURE FOR NOW +} + +/** + * ad_initialize_lacpdu - initialize a given lacpdu structure + * @lacpdu: lacpdu structure to initialize + * + */ +static void ad_initialize_lacpdu(struct lacpdu *lacpdu) +{ + u16 index; + + // initialize lacpdu data + lacpdu->subtype = 0x01; + lacpdu->version_number = 0x01; + lacpdu->tlv_type_actor_info = 0x01; + lacpdu->actor_information_length = 0x14; + // lacpdu->actor_system_priority updated on send + // lacpdu->actor_system updated on send + // lacpdu->actor_key updated on send + // lacpdu->actor_port_priority updated on send + // lacpdu->actor_port updated on send + // lacpdu->actor_state updated on send + lacpdu->tlv_type_partner_info = 0x02; + lacpdu->partner_information_length = 0x14; + for (index=0; index<=2; index++) { + lacpdu->reserved_3_1[index]=0; + } + // lacpdu->partner_system_priority updated on send + // lacpdu->partner_system updated on send + // lacpdu->partner_key updated on send + // lacpdu->partner_port_priority updated on send + // lacpdu->partner_port updated on send + // lacpdu->partner_state updated on send + for (index=0; index<=2; index++) { + lacpdu->reserved_3_2[index]=0; + } + lacpdu->tlv_type_collector_info = 0x03; + lacpdu->collector_information_length= 0x10; + lacpdu->collector_max_delay = AD_COLLECTOR_MAX_DELAY; + for (index=0; index<=11; index++) { + lacpdu->reserved_12[index]=0; + } + lacpdu->tlv_type_terminator = 0x00; + lacpdu->terminator_length = 0; + for (index=0; index<=49; index++) { + lacpdu->reserved_50[index]=0; + } +} + +////////////////////////////////////////////////////////////////////////////////////// +// ================= AD exported functions to the main bonding code ================== +////////////////////////////////////////////////////////////////////////////////////// + +// Check aggregators status in team every T seconds +#define AD_AGGREGATOR_SELECTION_TIMER 8 + +static u16 aggregator_identifier; + +/** + * bond_3ad_initialize - initialize a bond's 802.3ad parameters and structures + * @bond: bonding struct to work on + * @tick_resolution: tick duration (millisecond resolution) + * @lacp_fast: boolean. whether fast periodic should be used + * + * Can be called only after the mac address of the bond is set. + */ +void bond_3ad_initialize(struct bonding *bond, u16 tick_resolution, int lacp_fast) +{ + // check that the bond is not initialized yet + if (MAC_ADDRESS_COMPARE(&(BOND_AD_INFO(bond).system.sys_mac_addr), &(bond->dev->dev_addr))) { + + aggregator_identifier = 0; + + BOND_AD_INFO(bond).lacp_fast = lacp_fast; + BOND_AD_INFO(bond).system.sys_priority = 0xFFFF; + BOND_AD_INFO(bond).system.sys_mac_addr = *((struct mac_addr *)bond->dev->dev_addr); + + // 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_bind_slave - initialize a slave's port + * @slave: slave struct to work on + * + * Returns: 0 on success + * < 0 on error + */ +int bond_3ad_bind_slave(struct slave *slave) +{ + struct bonding *bond = bond_get_bond_by_slave(slave); + struct port *port; + struct aggregator *aggregator; + + if (bond == NULL) { + printk(KERN_ERR "The slave %s is not attached to its bond\n", slave->dev->name); + return -1; + } + + //check that the slave has not been intialized yet. + if (SLAVE_AD_INFO(slave).port.slave != slave) { + + // port initialization + port = &(SLAVE_AD_INFO(slave).port); + + ad_initialize_port(port, BOND_AD_INFO(bond).lacp_fast); + + port->slave = slave; + port->actor_port_number = SLAVE_AD_INFO(slave).id; + // key is determined according to the link speed, duplex and user key(which is yet not supported) + // ------------------------------------------------------------ + // Port key : | User key | Speed |Duplex| + // ------------------------------------------------------------ + // 16 6 1 0 + port->actor_admin_port_key = 0; // initialize this parameter + port->actor_admin_port_key |= __get_duplex(port); + port->actor_admin_port_key |= (__get_link_speed(port) << 1); + port->actor_oper_port_key = port->actor_admin_port_key; + // if the port is not full duplex, then the port should be not lacp Enabled + if (!(port->actor_oper_port_key & AD_DUPLEX_KEY_BITS)) { + port->sm_vars &= ~AD_PORT_LACP_ENABLED; + } + // actor system is the bond's system + port->actor_system = BOND_AD_INFO(bond).system.sys_mac_addr; + // tx timer(to verify that no more than MAX_TX_IN_SECOND lacpdu's are sent in one second) + port->sm_tx_timer_counter = ad_ticks_per_sec/AD_MAX_TX_IN_SECOND; + port->aggregator = NULL; + port->next_port_in_aggregator = NULL; + + __disable_port(port); + __initialize_port_locks(port); + + + // aggregator initialization + aggregator = &(SLAVE_AD_INFO(slave).aggregator); + + ad_initialize_agg(aggregator); + + aggregator->aggregator_mac_address = *((struct mac_addr *)bond->dev->dev_addr); + aggregator->aggregator_identifier = (++aggregator_identifier); + aggregator->slave = slave; + aggregator->is_active = 0; + aggregator->num_of_ports = 0; + } + + return 0; +} + +/** + * bond_3ad_unbind_slave - deinitialize a slave's port + * @slave: slave struct to work on + * + * Search for the aggregator that is related to this port, remove the + * aggregator and assign another aggregator for other port related to it + * (if any), and remove the port. + */ +void bond_3ad_unbind_slave(struct slave *slave) +{ + struct port *port, *prev_port, *temp_port; + struct aggregator *aggregator, *new_aggregator, *temp_aggregator; + int select_new_active_agg = 0; + + // find the aggregator related to this slave + aggregator = &(SLAVE_AD_INFO(slave).aggregator); + + // find the port related to this slave + port = &(SLAVE_AD_INFO(slave).port); + + // if slave is null, the whole port is not initialized + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Trying to unbind an uninitialized port on %s\n", slave->dev->name); + return; + } + + dprintk("Unbinding Link Aggregation Group %d\n", aggregator->aggregator_identifier); + + /* Tell the partner that this port is not suitable for aggregation */ + port->actor_oper_port_state &= ~AD_STATE_AGGREGATION; + __update_lacpdu_from_port(port); + ad_lacpdu_send(port); + + // check if this aggregator is occupied + if (aggregator->lag_ports) { + // check if there are other ports related to this aggregator except + // the port related to this slave(thats ensure us that there is a + // reason to search for new aggregator, and that we will find one + if ((aggregator->lag_ports != port) || (aggregator->lag_ports->next_port_in_aggregator)) { + // find new aggregator for the related port(s) + new_aggregator = __get_first_agg(port); + for (; new_aggregator; new_aggregator = __get_next_agg(new_aggregator)) { + // if the new aggregator is empty, or it connected to to our port only + if (!new_aggregator->lag_ports || ((new_aggregator->lag_ports == port) && !new_aggregator->lag_ports->next_port_in_aggregator)) { + break; + } + } + // if new aggregator found, copy the aggregator's parameters + // and connect the related lag_ports to the new aggregator + if ((new_aggregator) && ((!new_aggregator->lag_ports) || ((new_aggregator->lag_ports == port) && !new_aggregator->lag_ports->next_port_in_aggregator))) { + dprintk("Some port(s) related to LAG %d - replaceing with LAG %d\n", aggregator->aggregator_identifier, new_aggregator->aggregator_identifier); + + if ((new_aggregator->lag_ports == port) && new_aggregator->is_active) { + printk(KERN_INFO DRV_NAME ": Removing an active aggregator\n"); + // select new active aggregator + select_new_active_agg = 1; + } + + new_aggregator->is_individual = aggregator->is_individual; + new_aggregator->actor_admin_aggregator_key = aggregator->actor_admin_aggregator_key; + new_aggregator->actor_oper_aggregator_key = aggregator->actor_oper_aggregator_key; + new_aggregator->partner_system = aggregator->partner_system; + new_aggregator->partner_system_priority = aggregator->partner_system_priority; + new_aggregator->partner_oper_aggregator_key = aggregator->partner_oper_aggregator_key; + new_aggregator->receive_state = aggregator->receive_state; + new_aggregator->transmit_state = aggregator->transmit_state; + new_aggregator->lag_ports = aggregator->lag_ports; + new_aggregator->is_active = aggregator->is_active; + new_aggregator->num_of_ports = aggregator->num_of_ports; + + // update the information that is written on the ports about the aggregator + for (temp_port=aggregator->lag_ports; temp_port; temp_port=temp_port->next_port_in_aggregator) { + temp_port->aggregator=new_aggregator; + temp_port->actor_port_aggregator_identifier = new_aggregator->aggregator_identifier; + } + + // clear the aggregator + ad_clear_agg(aggregator); + + if (select_new_active_agg) { + ad_agg_selection_logic(__get_first_agg(port)); + } + } else { + printk(KERN_WARNING DRV_NAME ": Warning: unbinding aggregator, " + "and could not find a new aggregator for its ports\n"); + } + } else { // in case that the only port related to this aggregator is the one we want to remove + select_new_active_agg = aggregator->is_active; + // clear the aggregator + ad_clear_agg(aggregator); + if (select_new_active_agg) { + printk(KERN_INFO "Removing an active aggregator\n"); + // select new active aggregator + ad_agg_selection_logic(__get_first_agg(port)); + } + } + } + + dprintk("Unbinding port %d\n", port->actor_port_number); + // find the aggregator that this port is connected to + temp_aggregator = __get_first_agg(port); + for (; temp_aggregator; temp_aggregator = __get_next_agg(temp_aggregator)) { + prev_port = NULL; + // search the port in the aggregator's related ports + for (temp_port=temp_aggregator->lag_ports; temp_port; prev_port=temp_port, temp_port=temp_port->next_port_in_aggregator) { + if (temp_port == port) { // the aggregator found - detach the port from this aggregator + if (prev_port) { + prev_port->next_port_in_aggregator = temp_port->next_port_in_aggregator; + } else { + temp_aggregator->lag_ports = temp_port->next_port_in_aggregator; + } + temp_aggregator->num_of_ports--; + if (temp_aggregator->num_of_ports==0) { + select_new_active_agg = temp_aggregator->is_active; + // clear the aggregator + ad_clear_agg(temp_aggregator); + if (select_new_active_agg) { + printk(KERN_INFO "Removing an active aggregator\n"); + // select new active aggregator + ad_agg_selection_logic(__get_first_agg(port)); + } + } + break; + } + } + } + port->slave=NULL; +} + +/** + * bond_3ad_state_machine_handler - handle state machines timeout + * @bond: bonding struct to work on + * + * The state machine handling concept in this module is to check every tick + * which state machine should operate any function. The execution order is + * round robin, so when we have an interaction between state machines, the + * reply of one to each other might be delayed until next tick. + * + * This function also complete the initialization when the agg_select_timer + * times out, and it selects an aggregator for the ports that are yet not + * related to any aggregator, and selects the active aggregator for a bond. + */ +void bond_3ad_state_machine_handler(struct bonding *bond) +{ + struct port *port; + struct aggregator *aggregator; + + read_lock(&bond->lock); + + if (bond->kill_timers) { + goto out; + } + + //check if there are any slaves + if (bond->slave_cnt == 0) { + goto re_arm; + } + + // check if agg_select_timer timer after initialize is timed out + if (BOND_AD_INFO(bond).agg_select_timer && !(--BOND_AD_INFO(bond).agg_select_timer)) { + // select the active aggregator for the bond + if ((port = __get_first_port(bond))) { + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Warning: bond's first port is uninitialized\n"); + goto re_arm; + } + + aggregator = __get_first_agg(port); + ad_agg_selection_logic(aggregator); + } + } + + // for each port run the state machines + for (port = __get_first_port(bond); port; port = __get_next_port(port)) { + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Warning: Found an uninitialized port\n"); + goto re_arm; + } + + ad_rx_machine(NULL, port); + ad_periodic_machine(port); + ad_port_selection_logic(port); + ad_mux_machine(port); + ad_tx_machine(port); + + // turn off the BEGIN bit, since we already handled it + if (port->sm_vars & AD_PORT_BEGIN) { + port->sm_vars &= ~AD_PORT_BEGIN; + } + } + +re_arm: + mod_timer(&(BOND_AD_INFO(bond).ad_timer), jiffies + ad_delta_in_ticks); +out: + read_unlock(&bond->lock); +} + +/** + * bond_3ad_rx_indication - handle a received frame + * @lacpdu: received lacpdu + * @slave: slave struct to work on + * @length: length of the data received + * + * It is assumed that frames that were sent on this NIC don't returned as new + * received frames (loopback). Since only the payload is given to this + * function, it check for loopback. + */ +static void bond_3ad_rx_indication(struct lacpdu *lacpdu, struct slave *slave, u16 length) +{ + struct port *port; + + if (length >= sizeof(struct lacpdu)) { + + port = &(SLAVE_AD_INFO(slave).port); + + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Warning: port of slave %s is uninitialized\n", slave->dev->name); + return; + } + + switch (lacpdu->subtype) { + case AD_TYPE_LACPDU: + __ntohs_lacpdu(lacpdu); + dprintk("Received LACPDU on port %d\n", port->actor_port_number); + ad_rx_machine(lacpdu, port); + break; + + case AD_TYPE_MARKER: + // No need to convert fields to Little Endian since we don't use the marker's fields. + + switch (((struct marker *)lacpdu)->tlv_type) { + case AD_MARKER_INFORMATION_SUBTYPE: + dprintk("Received Marker Information on port %d\n", port->actor_port_number); + ad_marker_info_received((struct marker *)lacpdu, port); + break; + + case AD_MARKER_RESPONSE_SUBTYPE: + dprintk("Received Marker Response on port %d\n", port->actor_port_number); + ad_marker_response_received((struct marker *)lacpdu, port); + break; + + default: + dprintk("Received an unknown Marker subtype on slot %d\n", port->actor_port_number); + } + } + } +} + +/** + * bond_3ad_adapter_speed_changed - handle a slave's speed change indication + * @slave: slave struct to work on + * + * Handle reselection of aggregator (if needed) for this port. + */ +void bond_3ad_adapter_speed_changed(struct slave *slave) +{ + struct port *port; + + port = &(SLAVE_AD_INFO(slave).port); + + // if slave is null, the whole port is not initialized + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Warning: speed changed for uninitialized port on %s\n", + slave->dev->name); + return; + } + + port->actor_admin_port_key &= ~AD_SPEED_KEY_BITS; + port->actor_oper_port_key=port->actor_admin_port_key |= (__get_link_speed(port) << 1); + dprintk("Port %d changed speed\n", port->actor_port_number); + // there is no need to reselect a new aggregator, just signal the + // state machines to reinitialize + port->sm_vars |= AD_PORT_BEGIN; +} + +/** + * bond_3ad_adapter_duplex_changed - handle a slave's duplex change indication + * @slave: slave struct to work on + * + * Handle reselection of aggregator (if needed) for this port. + */ +void bond_3ad_adapter_duplex_changed(struct slave *slave) +{ + struct port *port; + + port=&(SLAVE_AD_INFO(slave).port); + + // if slave is null, the whole port is not initialized + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Warning: duplex changed for uninitialized port on %s\n", + slave->dev->name); + return; + } + + port->actor_admin_port_key &= ~AD_DUPLEX_KEY_BITS; + port->actor_oper_port_key=port->actor_admin_port_key |= __get_duplex(port); + dprintk("Port %d changed duplex\n", port->actor_port_number); + // there is no need to reselect a new aggregator, just signal the + // state machines to reinitialize + port->sm_vars |= AD_PORT_BEGIN; +} + +/** + * bond_3ad_handle_link_change - handle a slave's link status change indication + * @slave: slave struct to work on + * @status: whether the link is now up or down + * + * Handle reselection of aggregator (if needed) for this port. + */ +void bond_3ad_handle_link_change(struct slave *slave, char link) +{ + struct port *port; + + port = &(SLAVE_AD_INFO(slave).port); + + // if slave is null, the whole port is not initialized + if (!port->slave) { + printk(KERN_WARNING DRV_NAME ": Warning: link status changed for uninitialized port on %s\n", + slave->dev->name); + return; + } + + // on link down we are zeroing duplex and speed since some of the adaptors(ce1000.lan) report full duplex/speed instead of N/A(duplex) / 0(speed) + // on link up we are forcing recheck on the duplex and speed since some of he adaptors(ce1000.lan) report + if (link == BOND_LINK_UP) { + port->is_enabled = 1; + port->actor_admin_port_key &= ~AD_DUPLEX_KEY_BITS; + port->actor_oper_port_key=port->actor_admin_port_key |= __get_duplex(port); + port->actor_admin_port_key &= ~AD_SPEED_KEY_BITS; + port->actor_oper_port_key=port->actor_admin_port_key |= (__get_link_speed(port) << 1); + } else { + /* link has failed */ + port->is_enabled = 0; + port->actor_admin_port_key &= ~AD_DUPLEX_KEY_BITS; + port->actor_oper_port_key= (port->actor_admin_port_key &= ~AD_SPEED_KEY_BITS); + } + //BOND_PRINT_DBG(("Port %d changed link status to %s", port->actor_port_number, ((link == BOND_LINK_UP)?"UP":"DOWN"))); + // there is no need to reselect a new aggregator, just signal the + // state machines to reinitialize + port->sm_vars |= AD_PORT_BEGIN; +} + +/** + * bond_3ad_get_active_agg_info - get information of the active aggregator + * @bond: bonding struct to work on + * @ad_info: ad_info struct to fill with the bond's info + * + * Returns: 0 on success + * < 0 on error + */ +int bond_3ad_get_active_agg_info(struct bonding *bond, struct ad_info *ad_info) +{ + struct aggregator *aggregator = NULL; + struct port *port; + + for (port = __get_first_port(bond); port; port = __get_next_port(port)) { + if (port->aggregator && port->aggregator->is_active) { + aggregator = port->aggregator; + break; + } + } + + if (aggregator) { + ad_info->aggregator_id = aggregator->aggregator_identifier; + ad_info->ports = aggregator->num_of_ports; + ad_info->actor_key = aggregator->actor_oper_aggregator_key; + ad_info->partner_key = aggregator->partner_oper_aggregator_key; + memcpy(ad_info->partner_system, aggregator->partner_system.mac_addr_value, ETH_ALEN); + return 0; + } + + return -1; +} + +int bond_3ad_xmit_xor(struct sk_buff *skb, struct net_device *dev) +{ + struct slave *slave, *start_at; + struct bonding *bond = dev->priv; + struct ethhdr *data = (struct ethhdr *)skb->data; + int slave_agg_no; + int slaves_in_agg; + int agg_id; + int i; + struct ad_info ad_info; + int res = 1; + + /* make sure that the slaves list will + * not change during tx + */ + read_lock(&bond->lock); + + if (!BOND_IS_OK(bond)) { + goto out; + } + + if (bond_3ad_get_active_agg_info(bond, &ad_info)) { + printk(KERN_DEBUG "ERROR: bond_3ad_get_active_agg_info failed\n"); + goto out; + } + + slaves_in_agg = ad_info.ports; + agg_id = ad_info.aggregator_id; + + if (slaves_in_agg == 0) { + /*the aggregator is empty*/ + printk(KERN_DEBUG "ERROR: active aggregator is empty\n"); + goto out; + } + + slave_agg_no = (data->h_dest[5]^bond->dev->dev_addr[5]) % slaves_in_agg; + + bond_for_each_slave(bond, slave, i) { + struct aggregator *agg = SLAVE_AD_INFO(slave).port.aggregator; + + if (agg && (agg->aggregator_identifier == agg_id)) { + slave_agg_no--; + if (slave_agg_no < 0) { + break; + } + } + } + + if (slave_agg_no >= 0) { + printk(KERN_ERR DRV_NAME ": Error: Couldn't find a slave to tx on for aggregator ID %d\n", agg_id); + goto out; + } + + start_at = slave; + + bond_for_each_slave_from(bond, slave, i, start_at) { + int slave_agg_id = 0; + struct aggregator *agg = SLAVE_AD_INFO(slave).port.aggregator; + + if (agg) { + slave_agg_id = agg->aggregator_identifier; + } + + if (SLAVE_IS_OK(slave) && agg && (slave_agg_id == agg_id)) { + res = bond_dev_queue_xmit(bond, skb, slave->dev); + break; + } + } + +out: + if (res) { + /* no suitable interface, frame not sent */ + dev_kfree_skb(skb); + } + read_unlock(&bond->lock); + return 0; +} + +int bond_3ad_lacpdu_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type* ptype) +{ + struct bonding *bond = dev->priv; + struct slave *slave = NULL; + int ret = NET_RX_DROP; + + if (!(dev->flags & IFF_MASTER)) { + goto out; + } + + read_lock(&bond->lock); + slave = bond_get_slave_by_dev((struct bonding *)dev->priv, + skb->real_dev); + if (slave == NULL) { + goto out_unlock; + } + + bond_3ad_rx_indication((struct lacpdu *) skb->data, slave, skb->len); + + ret = NET_RX_SUCCESS; + +out_unlock: + read_unlock(&bond->lock); +out: + dev_kfree_skb(skb); + + return ret; +} + diff --git a/drivers/net/bonding/bond_3ad.h b/drivers/net/bonding/bond_3ad.h new file mode 100644 index 00000000000..f4682389418 --- /dev/null +++ b/drivers/net/bonding/bond_3ad.h @@ -0,0 +1,300 @@ +/* + * 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. + * + * + * Changes: + * + * 2003/05/01 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Amir Noam <amir.noam at intel dot com> + * - Added support for lacp_rate module param. + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Renamed bond_3ad_link_status_changed() to + * bond_3ad_handle_link_change() for compatibility with TLB. + * + * 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Code cleanup and style changes + */ + +#ifndef __BOND_3AD_H__ +#define __BOND_3AD_H__ + +#include <asm/byteorder.h> +#include <linux/skbuff.h> +#include <linux/netdevice.h> + +// General definitions +#define BOND_ETH_P_LACPDU 0x8809 +#define PKT_TYPE_LACPDU __constant_htons(BOND_ETH_P_LACPDU) +#define AD_TIMER_INTERVAL 100 /*msec*/ + +#define MULTICAST_LACPDU_ADDR {0x01, 0x80, 0xC2, 0x00, 0x00, 0x02} +#define AD_MULTICAST_LACPDU_ADDR {MULTICAST_LACPDU_ADDR} + +#define AD_LACP_SLOW 0 +#define AD_LACP_FAST 1 + +typedef struct mac_addr { + u8 mac_addr_value[ETH_ALEN]; +} mac_addr_t; + +typedef enum { + AD_BANDWIDTH = 0, + AD_COUNT +} agg_selection_t; + +// rx machine states(43.4.11 in the 802.3ad standard) +typedef enum { + AD_RX_DUMMY, + AD_RX_INITIALIZE, // rx Machine + AD_RX_PORT_DISABLED, // rx Machine + AD_RX_LACP_DISABLED, // rx Machine + AD_RX_EXPIRED, // rx Machine + AD_RX_DEFAULTED, // rx Machine + AD_RX_CURRENT // rx Machine +} rx_states_t; + +// periodic machine states(43.4.12 in the 802.3ad standard) +typedef enum { + AD_PERIODIC_DUMMY, + AD_NO_PERIODIC, // periodic machine + AD_FAST_PERIODIC, // periodic machine + AD_SLOW_PERIODIC, // periodic machine + AD_PERIODIC_TX // periodic machine +} periodic_states_t; + +// mux machine states(43.4.13 in the 802.3ad standard) +typedef enum { + AD_MUX_DUMMY, + AD_MUX_DETACHED, // mux machine + AD_MUX_WAITING, // mux machine + AD_MUX_ATTACHED, // mux machine + AD_MUX_COLLECTING_DISTRIBUTING // mux machine +} mux_states_t; + +// tx machine states(43.4.15 in the 802.3ad standard) +typedef enum { + AD_TX_DUMMY, + AD_TRANSMIT // tx Machine +} tx_states_t; + +// rx indication types +typedef enum { + AD_TYPE_LACPDU = 1, // type lacpdu + AD_TYPE_MARKER // type marker +} pdu_type_t; + +// rx marker indication types +typedef enum { + AD_MARKER_INFORMATION_SUBTYPE = 1, // marker imformation subtype + AD_MARKER_RESPONSE_SUBTYPE // marker response subtype +} marker_subtype_t; + +// timers types(43.4.9 in the 802.3ad standard) +typedef enum { + AD_CURRENT_WHILE_TIMER, + AD_ACTOR_CHURN_TIMER, + AD_PERIODIC_TIMER, + AD_PARTNER_CHURN_TIMER, + AD_WAIT_WHILE_TIMER +} ad_timers_t; + +#pragma pack(1) + +typedef struct ad_header { + struct mac_addr destination_address; + struct mac_addr source_address; + u16 length_type; +} ad_header_t; + +// Link Aggregation Control Protocol(LACP) data unit structure(43.4.2.2 in the 802.3ad standard) +typedef struct lacpdu { + u8 subtype; // = LACP(= 0x01) + u8 version_number; + u8 tlv_type_actor_info; // = actor information(type/length/value) + u8 actor_information_length; // = 20 + u16 actor_system_priority; + struct mac_addr actor_system; + u16 actor_key; + u16 actor_port_priority; + u16 actor_port; + u8 actor_state; + u8 reserved_3_1[3]; // = 0 + u8 tlv_type_partner_info; // = partner information + u8 partner_information_length; // = 20 + u16 partner_system_priority; + struct mac_addr partner_system; + u16 partner_key; + u16 partner_port_priority; + u16 partner_port; + u8 partner_state; + u8 reserved_3_2[3]; // = 0 + u8 tlv_type_collector_info; // = collector information + u8 collector_information_length; // = 16 + u16 collector_max_delay; + u8 reserved_12[12]; + u8 tlv_type_terminator; // = terminator + u8 terminator_length; // = 0 + u8 reserved_50[50]; // = 0 +} lacpdu_t; + +typedef struct lacpdu_header { + struct ad_header ad_header; + struct lacpdu lacpdu; +} lacpdu_header_t; + +// Marker Protocol Data Unit(PDU) structure(43.5.3.2 in the 802.3ad standard) +typedef struct marker { + u8 subtype; // = 0x02 (marker PDU) + u8 version_number; // = 0x01 + u8 tlv_type; // = 0x01 (marker information) + // = 0x02 (marker response information) + u8 marker_length; // = 0x16 + u16 requester_port; // The number assigned to the port by the requester + struct mac_addr requester_system; // The requester's system id + u32 requester_transaction_id; // The transaction id allocated by the requester, + u16 pad; // = 0 + u8 tlv_type_terminator; // = 0x00 + u8 terminator_length; // = 0x00 + u8 reserved_90[90]; // = 0 +} marker_t; + +typedef struct marker_header { + struct ad_header ad_header; + struct marker marker; +} marker_header_t; + +#pragma pack() + +struct slave; +struct bonding; +struct ad_info; +struct port; + +#ifdef __ia64__ +#pragma pack(8) +#endif + +// aggregator structure(43.4.5 in the 802.3ad standard) +typedef struct aggregator { + struct mac_addr aggregator_mac_address; + u16 aggregator_identifier; + u16 is_individual; // BOOLEAN + u16 actor_admin_aggregator_key; + u16 actor_oper_aggregator_key; + struct mac_addr partner_system; + u16 partner_system_priority; + u16 partner_oper_aggregator_key; + u16 receive_state; // BOOLEAN + u16 transmit_state; // BOOLEAN + struct port *lag_ports; + // ****** PRIVATE PARAMETERS ****** + struct slave *slave; // pointer to the bond slave that this aggregator belongs to + u16 is_active; // BOOLEAN. Indicates if this aggregator is active + u16 num_of_ports; +} aggregator_t; + +// port structure(43.4.6 in the 802.3ad standard) +typedef struct port { + u16 actor_port_number; + u16 actor_port_priority; + struct mac_addr actor_system; // This parameter is added here although it is not specified in the standard, just for simplification + u16 actor_system_priority; // This parameter is added here although it is not specified in the standard, just for simplification + u16 actor_port_aggregator_identifier; + u16 ntt; // BOOLEAN + u16 actor_admin_port_key; + u16 actor_oper_port_key; + u8 actor_admin_port_state; + u8 actor_oper_port_state; + struct mac_addr partner_admin_system; + struct mac_addr partner_oper_system; + u16 partner_admin_system_priority; + u16 partner_oper_system_priority; + u16 partner_admin_key; + u16 partner_oper_key; + u16 partner_admin_port_number; + u16 partner_oper_port_number; + u16 partner_admin_port_priority; + u16 partner_oper_port_priority; + u8 partner_admin_port_state; + u8 partner_oper_port_state; + u16 is_enabled; // BOOLEAN + // ****** PRIVATE PARAMETERS ****** + u16 sm_vars; // all state machines variables for this port + rx_states_t sm_rx_state; // state machine rx state + u16 sm_rx_timer_counter; // state machine rx timer counter + periodic_states_t sm_periodic_state;// state machine periodic state + u16 sm_periodic_timer_counter; // state machine periodic timer counter + mux_states_t sm_mux_state; // state machine mux state + u16 sm_mux_timer_counter; // state machine mux timer counter + tx_states_t sm_tx_state; // state machine tx state + u16 sm_tx_timer_counter; // state machine tx timer counter(allways on - enter to transmit state 3 time per second) + struct slave *slave; // pointer to the bond slave that this port belongs to + struct aggregator *aggregator; // pointer to an aggregator that this port related to + struct port *next_port_in_aggregator; // Next port on the linked list of the parent aggregator + u32 transaction_id; // continuous number for identification of Marker PDU's; + struct lacpdu lacpdu; // the lacpdu that will be sent for this port +} port_t; + +// system structure +typedef struct ad_system { + u16 sys_priority; + struct mac_addr sys_mac_addr; +} ad_system_t; + +#ifdef __ia64__ +#pragma pack() +#endif + +// ================= AD Exported structures to the main bonding code ================== +#define BOND_AD_INFO(bond) ((bond)->ad_info) +#define SLAVE_AD_INFO(slave) ((slave)->ad_info) + +struct ad_bond_info { + ad_system_t system; // 802.3ad system structure + u32 agg_select_timer; // Timer to select aggregator after all adapter's hand shakes + u32 agg_select_mode; // Mode of selection of active aggregator(bandwidth/count) + int lacp_fast; /* whether fast periodic tx should be + * requested + */ + struct timer_list ad_timer; + struct packet_type ad_pkt_type; +}; + +struct ad_slave_info { + struct aggregator aggregator; // 802.3ad aggregator structure + struct port port; // 802.3ad port structure + spinlock_t rx_machine_lock; // To avoid race condition between callback and receive interrupt + u16 id; +}; + +// ================= AD Exported functions to the main bonding code ================== +void bond_3ad_initialize(struct bonding *bond, u16 tick_resolution, int lacp_fast); +int bond_3ad_bind_slave(struct slave *slave); +void bond_3ad_unbind_slave(struct slave *slave); +void bond_3ad_state_machine_handler(struct bonding *bond); +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); +int bond_3ad_get_active_agg_info(struct bonding *bond, struct ad_info *ad_info); +int bond_3ad_xmit_xor(struct sk_buff *skb, struct net_device *dev); +int bond_3ad_lacpdu_recv(struct sk_buff *skb, struct net_device *dev, struct packet_type* ptype); +#endif //__BOND_3AD_H__ + diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c new file mode 100644 index 00000000000..5ce606d9dc0 --- /dev/null +++ b/drivers/net/bonding/bond_alb.c @@ -0,0 +1,1696 @@ +/* + * 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. + * + * + * Changes: + * + * 2003/06/25 - Shmulik Hen <shmulik.hen at intel dot com> + * - Fixed signed/unsigned calculation errors that caused load sharing + * to collapse to one slave under very heavy UDP Tx stress. + * + * 2003/08/06 - Amir Noam <amir.noam at intel dot com> + * - Add support for setting bond's MAC address with special + * handling required for ALB/TLB. + * + * 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Code cleanup and style changes + * + * 2003/12/30 - Amir Noam <amir.noam at intel dot com> + * - Fixed: Cannot remove and re-enslave the original active slave. + * + * 2004/01/14 - Shmulik Hen <shmulik.hen at intel dot com> + * - Add capability to tag self generated packets in ALB/TLB modes. + */ + +//#define BONDING_DEBUG 1 + +#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> +#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}; +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]; + u16 type; + u8 padding[ETH_ZLEN - ETH_HLEN]; +}; + +struct arp_pkt { + u16 hw_addr_space; + u16 prot_addr_space; + u8 hw_addr_len; + u8 prot_addr_len; + u16 op_code; + u8 mac_src[ETH_ALEN]; /* sender hardware address */ + u32 ip_src; /* sender IP address */ + u8 mac_dst[ETH_ALEN]; /* target hardware address */ + u32 ip_dst; /* target IP address */ +}; +#pragma pack() + +/* Forward declaration */ +static void alb_send_learning_packets(struct slave *slave, u8 mac_addr[]); + +static inline u8 _simple_hash(u8 *hash_start, int hash_size) +{ + 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) +{ + spin_lock(&(BOND_ALB_INFO(bond).tx_hashtbl_lock)); +} + +static inline void _unlock_tx_hashtbl(struct bonding *bond) +{ + spin_unlock(&(BOND_ALB_INFO(bond).tx_hashtbl_lock)); +} + +/* 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; + + 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; + } + + _unlock_tx_hashtbl(bond); + + tlb_init_slave(slave); +} + +/* 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); + int i; + + spin_lock_init(&(bond_info->tx_hashtbl_lock)); + + _lock_tx_hashtbl(bond); + + bond_info->tx_hashtbl = kmalloc(size, GFP_KERNEL); + if (!bond_info->tx_hashtbl) { + printk(KERN_ERR DRV_NAME + ": Error: %s: Failed to allocate TLB hash table\n", + bond->dev->name); + _unlock_tx_hashtbl(bond); + return -1; + } + + memset(bond_info->tx_hashtbl, 0, size); + + 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) +{ + spin_lock(&(BOND_ALB_INFO(bond).rx_hashtbl_lock)); +} + +static inline void _unlock_rx_hashtbl(struct bonding *bond) +{ + spin_unlock(&(BOND_ALB_INFO(bond).rx_hashtbl_lock)); +} + +/* 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); +} + +static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct packet_type *ptype) +{ + struct bonding *bond = bond_dev->priv; + struct arp_pkt *arp = (struct arp_pkt *)skb->data; + int res = NET_RX_DROP; + + if (!(bond_dev->flags & IFF_MASTER)) { + goto out; + } + + if (!arp) { + dprintk("Packet has no ARP data\n"); + goto out; + } + + if (skb->len < sizeof(struct arp_pkt)) { + dprintk("Packet is too small to be an ARP\n"); + goto out; + } + + if (arp->op_code == htons(ARPOP_REPLY)) { + /* update rx hash table for this ARP */ + rlb_update_entry_from_arp(bond, arp); + dprintk("Server received an ARP Reply from client\n"); + } + + 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) { + bond->alb_info.primary_is_promisc = 1; + dev_set_promiscuity(bond->curr_active_slave->dev, 1); + } + + 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); + + write_lock(&bond->curr_slave_lock); + + if (slave != bond->curr_active_slave) { + rlb_teach_disabled_mac_on_primary(bond, slave->dev->dev_addr); + } + + write_unlock(&bond->curr_slave_lock); +} + +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 + ": Error: failed to create an ARP packet\n"); + 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 + ": Error: failed to insert VLAN tag\n"); + 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 */ +static void rlb_req_update_subnet_clients(struct bonding *bond, u32 src_ip) +{ + 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 + ": Error: found a client with no channel in " + "the client's hash table\n"); + 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)); + struct arp_pkt *arp = (struct arp_pkt *)skb->nh.raw; + 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)) { + unsigned short vlan_id; + int res = vlan_get_tag(skb, &vlan_id); + if (!res) { + client_info->tag = 1; + client_info->vlan_id = vlan_id; + } + } + + 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) +{ + struct arp_pkt *arp = (struct arp_pkt *)skb->nh.raw; + struct slave *tx_slave = NULL; + + if (arp->op_code == __constant_htons(ARPOP_REPLY)) { + /* 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); + } + dprintk("Server sent ARP Reply packet\n"); + } else if (arp->op_code == __constant_htons(ARPOP_REQUEST)) { + /* 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); + dprintk("Server sent ARP Request packet\n"); + } + + 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); + int size = RLB_HASH_TABLE_SIZE * sizeof(struct rlb_client_info); + int i; + + spin_lock_init(&(bond_info->rx_hashtbl_lock)); + + _lock_rx_hashtbl(bond); + + bond_info->rx_hashtbl = kmalloc(size, GFP_KERNEL); + if (!bond_info->rx_hashtbl) { + printk(KERN_ERR DRV_NAME + ": Error: %s: Failed to allocate RLB hash table\n", + bond->dev->name); + _unlock_rx_hashtbl(bond); + return -1; + } + + 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*/ + pk_type->type = __constant_htons(ETH_P_ARP); + pk_type->dev = bond->dev; + 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); + pkt.type = __constant_htons(ETH_P_LOOP); + + 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); + + skb->mac.raw = data; + skb->nh.raw = data + ETH_HLEN; + 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 + ": Error: failed to insert VLAN tag\n"); + 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 + ": Error: dev_set_mac_address of dev %s failed! ALB " + "mode requires that the base driver support setting " + "the hw address also when the network device's " + "interface is open\n", + dev->name); + return -EOPNOTSUPP; + } + return 0; +} + +/* Caller must hold bond lock for write or curr_slave_lock for write*/ +static void alb_swap_mac_addr(struct bonding *bond, struct slave *slave1, struct slave *slave2) +{ + struct slave *disabled_slave = NULL; + u8 tmp_mac_addr[ETH_ALEN]; + int slaves_state_differ; + + slaves_state_differ = (SLAVE_IS_OK(slave1) != SLAVE_IS_OK(slave2)); + + 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); + + /* 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. + * + * Caller must hold bond lock + */ +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) { + alb_swap_mac_addr(bond, slave, tmp_slave); + } + } +} + +/** + * 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; + } + } + + if (found) { + /* a slave was found that is using the mac address + * of the new slave + */ + printk(KERN_ERR DRV_NAME + ": Error: the hw address of slave %s is not " + "unique - cannot enslave it!", + slave->dev->name); + return -EINVAL; + } + + return 0; + } + + /* 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 + ": 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); + } +} + diff --git a/drivers/net/bonding/bond_alb.h b/drivers/net/bonding/bond_alb.h new file mode 100644 index 00000000000..e4091cd8d65 --- /dev/null +++ b/drivers/net/bonding/bond_alb.h @@ -0,0 +1,141 @@ +/* + * 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. + * + * + * Changes: + * + * 2003/08/06 - Amir Noam <amir.noam at intel dot com> + * - Add support for setting bond's MAC address with special + * handling required for ALB/TLB. + * + * 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Code cleanup and style changes + */ + +#ifndef __BOND_ALB_H__ +#define __BOND_ALB_H__ + +#include <linux/if_ether.h> + +struct bonding; +struct slave; + +#define BOND_ALB_INFO(bond) ((bond)->alb_info) +#define SLAVE_TLB_INFO(slave) ((slave)->tlb_info) + +struct tlb_client_info { + struct slave *tx_slave; /* A pointer to slave used for transmiting + * packets to a Client that the Hash function + * gave this entry index. + */ + u32 tx_bytes; /* Each Client acumulates the BytesTx that + * were tranmitted to it, and after each + * CallBack the LoadHistory is devided + * by the balance interval + */ + u32 load_history; /* This field contains the amount of Bytes + * that were transmitted to this client by + * the server on the previous balance + * interval in Bps. + */ + u32 next; /* The next Hash table entry index, assigned + * to use the same adapter for transmit. + */ + u32 prev; /* The previous Hash table entry index, + * assigned to use the same + */ +}; + +/* ------------------------------------------------------------------------- + * struct rlb_client_info contains all info related to a specific rx client + * connection. This is the Clients Hash Table entry struct + * ------------------------------------------------------------------------- + */ +struct rlb_client_info { + u32 ip_src; /* the server IP address */ + u32 ip_dst; /* the client IP address */ + u8 mac_dst[ETH_ALEN]; /* the client MAC address */ + u32 next; /* The next Hash table entry index */ + u32 prev; /* The previous Hash table entry index */ + u8 assigned; /* checking whether this entry is assigned */ + u8 ntt; /* flag - need to transmit client info */ + struct slave *slave; /* the slave assigned to this client */ + u8 tag; /* flag - need to tag skb */ + unsigned short vlan_id; /* VLAN tag associated with IP address */ +}; + +struct tlb_slave_info { + u32 head; /* Index to the head of the bi-directional clients + * hash table entries list. The entries in the list + * are the entries that were assigned to use this + * slave for transmit. + */ + u32 load; /* Each slave sums the loadHistory of all clients + * assigned to it + */ +}; + +struct alb_bond_info { + struct timer_list alb_timer; + struct tlb_client_info *tx_hashtbl; /* Dynamically allocated */ + spinlock_t tx_hashtbl_lock; + u32 unbalanced_load; + int tx_rebalance_counter; + int lp_counter; + /* -------- rlb parameters -------- */ + int rlb_enabled; + struct packet_type rlb_pkt_type; + struct rlb_client_info *rx_hashtbl; /* Receive hash table */ + spinlock_t rx_hashtbl_lock; + u32 rx_hashtbl_head; + u8 rx_ntt; /* flag - need to transmit + * to all rx clients + */ + struct slave *next_rx_slave;/* next slave to be assigned + * to a new rx client for + */ + u32 rlb_interval_counter; + u8 primary_is_promisc; /* boolean */ + u32 rlb_promisc_timeout_counter;/* counts primary + * promiscuity time + */ + u32 rlb_update_delay_counter; + u32 rlb_update_retry_counter;/* counter of retries + * of client update + */ + u8 rlb_rebalance; /* flag - indicates that the + * rx traffic should be + * rebalanced + */ + struct vlan_entry *current_alb_vlan; +}; + +int bond_alb_initialize(struct bonding *bond, int rlb_enabled); +void bond_alb_deinitialize(struct bonding *bond); +int bond_alb_init_slave(struct bonding *bond, struct slave *slave); +void bond_alb_deinit_slave(struct bonding *bond, struct slave *slave); +void bond_alb_handle_link_change(struct bonding *bond, struct slave *slave, char link); +void bond_alb_handle_active_change(struct bonding *bond, struct slave *new_slave); +int bond_alb_xmit(struct sk_buff *skb, struct net_device *bond_dev); +void bond_alb_monitor(struct bonding *bond); +int bond_alb_set_mac_address(struct net_device *bond_dev, void *addr); +void bond_alb_clear_vlan(struct bonding *bond, unsigned short vlan_id); +#endif /* __BOND_ALB_H__ */ + diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c new file mode 100644 index 00000000000..770e28f98fd --- /dev/null +++ b/drivers/net/bonding/bond_main.c @@ -0,0 +1,4708 @@ +/* + * originally based on the dummy device. + * + * Copyright 1999, Thomas Davis, tadavis@lbl.gov. + * Licensed under the GPL. Based on dummy.c, and eql.c devices. + * + * bonding.c: an Ethernet Bonding driver + * + * This is useful to talk to a Cisco EtherChannel compatible equipment: + * Cisco 5500 + * Sun Trunking (Solaris) + * Alteon AceDirector Trunks + * Linux Bonding + * and probably many L2 switches ... + * + * How it works: + * ifconfig bond0 ipaddress netmask up + * will setup a network device, with an ip address. No mac address + * will be assigned at this time. The hw mac address will come from + * the first slave bonded to the channel. All slaves will then use + * this hw mac address. + * + * ifconfig bond0 down + * will release all slaves, marking them as down. + * + * ifenslave bond0 eth0 + * will attach eth0 to bond0 as a slave. eth0 hw mac address will either + * a: be used as initial mac address + * b: if a hw mac address already is there, eth0's hw mac address + * will then be set from bond0. + * + * v0.1 - first working version. + * v0.2 - changed stats to be calculated by summing slaves stats. + * + * Changes: + * Arnaldo Carvalho de Melo <acme@conectiva.com.br> + * - fix leaks on failure at bond_init + * + * 2000/09/30 - Willy Tarreau <willy at meta-x.org> + * - added trivial code to release a slave device. + * - fixed security bug (CAP_NET_ADMIN not checked) + * - implemented MII link monitoring to disable dead links : + * All MII capable slaves are checked every <miimon> milliseconds + * (100 ms seems good). This value can be changed by passing it to + * insmod. A value of zero disables the monitoring (default). + * - fixed an infinite loop in bond_xmit_roundrobin() when there's no + * good slave. + * - made the code hopefully SMP safe + * + * 2000/10/03 - Willy Tarreau <willy at meta-x.org> + * - optimized slave lists based on relevant suggestions from Thomas Davis + * - implemented active-backup method to obtain HA with two switches: + * stay as long as possible on the same active interface, while we + * also monitor the backup one (MII link status) because we want to know + * if we are able to switch at any time. ( pass "mode=1" to insmod ) + * - lots of stress testings because we need it to be more robust than the + * wires ! :-> + * + * 2000/10/09 - Willy Tarreau <willy at meta-x.org> + * - added up and down delays after link state change. + * - optimized the slaves chaining so that when we run forward, we never + * repass through the bond itself, but we can find it by searching + * backwards. Renders the deletion more difficult, but accelerates the + * scan. + * - smarter enslaving and releasing. + * - finer and more robust SMP locking + * + * 2000/10/17 - Willy Tarreau <willy at meta-x.org> + * - fixed two potential SMP race conditions + * + * 2000/10/18 - Willy Tarreau <willy at meta-x.org> + * - small fixes to the monitoring FSM in case of zero delays + * 2000/11/01 - Willy Tarreau <willy at meta-x.org> + * - fixed first slave not automatically used in trunk mode. + * 2000/11/10 : spelling of "EtherChannel" corrected. + * 2000/11/13 : fixed a race condition in case of concurrent accesses to ioctl(). + * 2000/12/16 : fixed improper usage of rtnl_exlock_nowait(). + * + * 2001/1/3 - Chad N. Tindel <ctindel at ieee dot org> + * - The bonding driver now simulates MII status monitoring, just like + * a normal network device. It will show that the link is down iff + * every slave in the bond shows that their links are down. If at least + * one slave is up, the bond's MII status will appear as up. + * + * 2001/2/7 - Chad N. Tindel <ctindel at ieee dot org> + * - Applications can now query the bond from user space to get + * information which may be useful. They do this by calling + * the BOND_INFO_QUERY ioctl. Once the app knows how many slaves + * are in the bond, it can call the BOND_SLAVE_INFO_QUERY ioctl to + * get slave specific information (# link failures, etc). See + * <linux/if_bonding.h> for more details. The structs of interest + * are ifbond and ifslave. + * + * 2001/4/5 - Chad N. Tindel <ctindel at ieee dot org> + * - Ported to 2.4 Kernel + * + * 2001/5/2 - Jeffrey E. Mast <jeff at mastfamily dot com> + * - When a device is detached from a bond, the slave device is no longer + * left thinking that is has a master. + * + * 2001/5/16 - Jeffrey E. Mast <jeff at mastfamily dot com> + * - memset did not appropriately initialized the bond rw_locks. Used + * rwlock_init to initialize to unlocked state to prevent deadlock when + * first attempting a lock + * - Called SET_MODULE_OWNER for bond device + * + * 2001/5/17 - Tim Anderson <tsa at mvista.com> + * - 2 paths for releasing for slave release; 1 through ioctl + * and 2) through close. Both paths need to release the same way. + * - the free slave in bond release is changing slave status before + * the free. The netdev_set_master() is intended to change slave state + * so it should not be done as part of the release process. + * - Simple rule for slave state at release: only the active in A/B and + * only one in the trunked case. + * + * 2001/6/01 - Tim Anderson <tsa at mvista.com> + * - Now call dev_close when releasing a slave so it doesn't screw up + * out routing table. + * + * 2001/6/01 - Chad N. Tindel <ctindel at ieee dot org> + * - Added /proc support for getting bond and slave information. + * Information is in /proc/net/<bond device>/info. + * - Changed the locking when calling bond_close to prevent deadlock. + * + * 2001/8/05 - Janice Girouard <girouard at us.ibm.com> + * - correct problem where refcnt of slave is not incremented in bond_ioctl + * so the system hangs when halting. + * - correct locking problem when unable to malloc in bond_enslave. + * - adding bond_xmit_xor logic. + * - adding multiple bond device support. + * + * 2001/8/13 - Erik Habbinga <erik_habbinga at hp dot com> + * - correct locking problem with rtnl_exlock_nowait + * + * 2001/8/23 - Janice Girouard <girouard at us.ibm.com> + * - bzero initial dev_bonds, to correct oops + * - convert SIOCDEVPRIVATE to new MII ioctl calls + * + * 2001/9/13 - Takao Indoh <indou dot takao at jp dot fujitsu dot com> + * - Add the BOND_CHANGE_ACTIVE ioctl implementation + * + * 2001/9/14 - Mark Huth <mhuth at mvista dot com> + * - Change MII_LINK_READY to not check for end of auto-negotiation, + * but only for an up link. + * + * 2001/9/20 - Chad N. Tindel <ctindel at ieee dot org> + * - Add the device field to bonding_t. Previously the net_device + * corresponding to a bond wasn't available from the bonding_t + * structure. + * + * 2001/9/25 - Janice Girouard <girouard at us.ibm.com> + * - add arp_monitor for active backup mode + * + * 2001/10/23 - Takao Indoh <indou dot takao at jp dot fujitsu dot com> + * - Various memory leak fixes + * + * 2001/11/5 - Mark Huth <mark dot huth at mvista dot com> + * - Don't take rtnl lock in bond_mii_monitor as it deadlocks under + * certain hotswap conditions. + * Note: this same change may be required in bond_arp_monitor ??? + * - Remove possibility of calling bond_sethwaddr with NULL slave_dev ptr + * - Handle hot swap ethernet interface deregistration events to remove + * kernel oops following hot swap of enslaved interface + * + * 2002/1/2 - Chad N. Tindel <ctindel at ieee dot org> + * - Restore original slave flags at release time. + * + * 2002/02/18 - Erik Habbinga <erik_habbinga at hp dot com> + * - bond_release(): calling kfree on our_slave after call to + * bond_restore_slave_flags, not before + * - bond_enslave(): saving slave flags into original_flags before + * call to netdev_set_master, so the IFF_SLAVE flag doesn't end + * up in original_flags + * + * 2002/04/05 - Mark Smith <mark.smith at comdev dot cc> and + * Steve Mead <steve.mead at comdev dot cc> + * - Port Gleb Natapov's multicast support patchs from 2.4.12 + * to 2.4.18 adding support for multicast. + * + * 2002/06/10 - Tony Cureington <tony.cureington * hp_com> + * - corrected uninitialized pointer (ifr.ifr_data) in bond_check_dev_link; + * actually changed function to use MIIPHY, then MIIREG, and finally + * ETHTOOL to determine the link status + * - fixed bad ifr_data pointer assignments in bond_ioctl + * - corrected mode 1 being reported as active-backup in bond_get_info; + * also added text to distinguish type of load balancing (rr or xor) + * - change arp_ip_target module param from "1-12s" (array of 12 ptrs) + * to "s" (a single ptr) + * + * 2002/08/30 - Jay Vosburgh <fubar at us dot ibm dot com> + * - Removed acquisition of xmit_lock in set_multicast_list; caused + * deadlock on SMP (lock is held by caller). + * - Revamped SIOCGMIIPHY, SIOCGMIIREG portion of bond_check_dev_link(). + * + * 2002/09/18 - Jay Vosburgh <fubar at us dot ibm dot com> + * - Fixed up bond_check_dev_link() (and callers): removed some magic + * numbers, banished local MII_ defines, wrapped ioctl calls to + * prevent EFAULT errors + * + * 2002/9/30 - Jay Vosburgh <fubar at us dot ibm dot com> + * - make sure the ip target matches the arp_target before saving the + * hw address. + * + * 2002/9/30 - Dan Eisner <eisner at 2robots dot com> + * - make sure my_ip is set before taking down the link, since + * not all switches respond if the source ip is not set. + * + * 2002/10/8 - Janice Girouard <girouard at us dot ibm dot com> + * - read in the local ip address when enslaving a device + * - add primary support + * - make sure 2*arp_interval has passed when a new device + * is brought on-line before taking it down. + * + * 2002/09/11 - Philippe De Muyter <phdm at macqel dot be> + * - Added bond_xmit_broadcast logic. + * - Added bond_mode() support function. + * + * 2002/10/26 - Laurent Deniel <laurent.deniel at free.fr> + * - allow to register multicast addresses only on active slave + * (useful in active-backup mode) + * - add multicast module parameter + * - fix deletion of multicast groups after unloading module + * + * 2002/11/06 - Kameshwara Rayaprolu <kameshwara.rao * wipro_com> + * - Changes to prevent panic from closing the device twice; if we close + * the device in bond_release, we must set the original_flags to down + * so it won't be closed again by the network layer. + * + * 2002/11/07 - Tony Cureington <tony.cureington * hp_com> + * - Fix arp_target_hw_addr memory leak + * - Created activebackup_arp_monitor function to handle arp monitoring + * in active backup mode - the bond_arp_monitor had several problems... + * such as allowing slaves to tx arps sequentially without any delay + * for a response + * - Renamed bond_arp_monitor to loadbalance_arp_monitor and re-wrote + * this function to just handle arp monitoring in load-balancing mode; + * it is a lot more compact now + * - Changes to ensure one and only one slave transmits in active-backup + * mode + * - Robustesize parameters; warn users about bad combinations of + * parameters; also if miimon is specified and a network driver does + * not support MII or ETHTOOL, inform the user of this + * - Changes to support link_failure_count when in arp monitoring mode + * - Fix up/down delay reported in /proc + * - Added version; log version; make version available from "modinfo -d" + * - Fixed problem in bond_check_dev_link - if the first IOCTL (SIOCGMIIPH) + * failed, the ETHTOOL ioctl never got a chance + * + * 2002/11/16 - Laurent Deniel <laurent.deniel at free.fr> + * - fix multicast handling in activebackup_arp_monitor + * - remove one unnecessary and confusing curr_active_slave == slave test + * in activebackup_arp_monitor + * + * 2002/11/17 - Laurent Deniel <laurent.deniel at free.fr> + * - fix bond_slave_info_query when slave_id = num_slaves + * + * 2002/11/19 - Janice Girouard <girouard at us dot ibm dot com> + * - correct ifr_data reference. Update ifr_data reference + * to mii_ioctl_data struct values to avoid confusion. + * + * 2002/11/22 - Bert Barbe <bert.barbe at oracle dot com> + * - Add support for multiple arp_ip_target + * + * 2002/12/13 - Jay Vosburgh <fubar at us dot ibm dot com> + * - Changed to allow text strings for mode and multicast, e.g., + * insmod bonding mode=active-backup. The numbers still work. + * One change: an invalid choice will cause module load failure, + * rather than the previous behavior of just picking one. + * - Minor cleanups; got rid of dup ctype stuff, atoi function + * + * 2003/02/07 - Jay Vosburgh <fubar at us dot ibm dot com> + * - Added use_carrier module parameter that causes miimon to + * use netif_carrier_ok() test instead of MII/ETHTOOL ioctls. + * - Minor cleanups; consolidated ioctl calls to one function. + * + * 2003/02/07 - Tony Cureington <tony.cureington * hp_com> + * - Fix bond_mii_monitor() logic error that could result in + * bonding round-robin mode ignoring links after failover/recovery + * + * 2003/03/17 - Jay Vosburgh <fubar at us dot ibm dot com> + * - kmalloc fix (GFP_KERNEL to GFP_ATOMIC) reported by + * Shmulik dot Hen at intel.com. + * - Based on discussion on mailing list, changed use of + * update_slave_cnt(), created wrapper functions for adding/removing + * slaves, changed bond_xmit_xor() to check slave_cnt instead of + * checking slave and slave->dev (which only worked by accident). + * - Misc code cleanup: get arp_send() prototype from header file, + * add max_bonds to bonding.txt. + * + * 2003/03/18 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Shmulik Hen <shmulik.hen at intel dot com> + * - Make sure only bond_attach_slave() and bond_detach_slave() can + * manipulate the slave list, including slave_cnt, even when in + * bond_release_all(). + * - Fixed hang in bond_release() with traffic running: + * netdev_set_master() must not be called from within the bond lock. + * + * 2003/03/18 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Shmulik Hen <shmulik.hen at intel dot com> + * - Fixed hang in bond_enslave() with traffic running: + * netdev_set_master() must not be called from within the bond lock. + * + * 2003/03/18 - Amir Noam <amir.noam at intel dot com> + * - Added support for getting slave's speed and duplex via ethtool. + * Needed for 802.3ad and other future modes. + * + * 2003/03/18 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Shmulik Hen <shmulik.hen at intel dot com> + * - Enable support of modes that need to use the unique mac address of + * each slave. + * * bond_enslave(): Moved setting the slave's mac address, and + * openning it, from the application to the driver. This breaks + * backward comaptibility with old versions of ifenslave that open + * the slave before enalsving it !!!. + * * bond_release(): The driver also takes care of closing the slave + * and restoring its original mac address. + * - Removed the code that restores all base driver's flags. + * Flags are automatically restored once all undo stages are done + * properly. + * - Block possibility of enslaving before the master is up. This + * prevents putting the system in an unstable state. + * + * 2003/03/18 - Amir Noam <amir.noam at intel dot com>, + * Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Shmulik Hen <shmulik.hen at intel dot com> + * - Added support for IEEE 802.3ad Dynamic link aggregation mode. + * + * 2003/05/01 - Amir Noam <amir.noam at intel dot com> + * - Added ABI version control to restore compatibility between + * new/old ifenslave and new/old bonding. + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Fixed bug in bond_release_all(): save old value of curr_active_slave + * before setting it to NULL. + * - Changed driver versioning scheme to include version number instead + * of release date (that is already in another field). There are 3 + * fields X.Y.Z where: + * X - Major version - big behavior changes + * Y - Minor version - addition of features + * Z - Extra version - minor changes and bug fixes + * The current version is 1.0.0 as a base line. + * + * 2003/05/01 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Amir Noam <amir.noam at intel dot com> + * - Added support for lacp_rate module param. + * - Code beautification and style changes (mainly in comments). + * new version - 1.0.1 + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Based on discussion on mailing list, changed locking scheme + * to use lock/unlock or lock_bh/unlock_bh appropriately instead + * of lock_irqsave/unlock_irqrestore. The new scheme helps exposing + * hidden bugs and solves system hangs that occurred due to the fact + * that holding lock_irqsave doesn't prevent softirqs from running. + * This also increases total throughput since interrupts are not + * blocked on each transmitted packets or monitor timeout. + * new version - 2.0.0 + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Added support for Transmit load balancing mode. + * - Concentrate all assignments of curr_active_slave to a single point + * so specific modes can take actions when the primary adapter is + * changed. + * - Take the updelay parameter into consideration during bond_enslave + * since some adapters loose their link during setting the device. + * - Renamed bond_3ad_link_status_changed() to + * bond_3ad_handle_link_change() for compatibility with TLB. + * new version - 2.1.0 + * + * 2003/05/01 - Tsippy Mendelson <tsippy.mendelson at intel dot com> + * - Added support for Adaptive load balancing mode which is + * equivalent to Transmit load balancing + Receive load balancing. + * new version - 2.2.0 + * + * 2003/05/15 - Jay Vosburgh <fubar at us dot ibm dot com> + * - Applied fix to activebackup_arp_monitor posted to bonding-devel + * by Tony Cureington <tony.cureington * hp_com>. Fixes ARP + * monitor endless failover bug. Version to 2.2.10 + * + * 2003/05/20 - Amir Noam <amir.noam at intel dot com> + * - Fixed bug in ABI version control - Don't commit to a specific + * ABI version if receiving unsupported ioctl commands. + * + * 2003/05/22 - Jay Vosburgh <fubar at us dot ibm dot com> + * - Fix ifenslave -c causing bond to loose existing routes; + * added bond_set_mac_address() that doesn't require the + * bond to be down. + * - In conjunction with fix for ifenslave -c, in + * bond_change_active(), changing to the already active slave + * is no longer an error (it successfully does nothing). + * + * 2003/06/30 - Amir Noam <amir.noam at intel dot com> + * - Fixed bond_change_active() for ALB/TLB modes. + * Version to 2.2.14. + * + * 2003/07/29 - Amir Noam <amir.noam at intel dot com> + * - Fixed ARP monitoring bug. + * Version to 2.2.15. + * + * 2003/07/31 - Willy Tarreau <willy at ods dot org> + * - Fixed kernel panic when using ARP monitoring without + * setting bond's IP address. + * Version to 2.2.16. + * + * 2003/08/06 - Amir Noam <amir.noam at intel dot com> + * - Back port from 2.6: use alloc_netdev(); fix /proc handling; + * made stats a part of bond struct so no need to allocate + * and free it separately; use standard list operations instead + * of pre-allocated array of bonds. + * Version to 2.3.0. + * + * 2003/08/07 - Jay Vosburgh <fubar at us dot ibm dot com>, + * Amir Noam <amir.noam at intel dot com> and + * Shmulik Hen <shmulik.hen at intel dot com> + * - Propagating master's settings: Distinguish between modes that + * use a primary slave from those that don't, and propagate settings + * accordingly; Consolidate change_active opeartions and add + * reselect_active and find_best opeartions; Decouple promiscuous + * handling from the multicast mode setting; Add support for changing + * HW address and MTU with proper unwind; Consolidate procfs code, + * add CHANGENAME handler; Enhance netdev notification handling. + * Version to 2.4.0. + * + * 2003/09/15 - Stephen Hemminger <shemminger at osdl dot org>, + * Amir Noam <amir.noam at intel dot com> + * - Convert /proc to seq_file interface. + * Change /proc/net/bondX/info to /proc/net/bonding/bondX. + * Set version to 2.4.1. + * + * 2003/11/20 - Amir Noam <amir.noam at intel dot com> + * - Fix /proc creation/destruction. + * + * 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Massive cleanup - Set version to 2.5.0 + * Code changes: + * o Consolidate format of prints and debug prints. + * o Remove bonding_t/slave_t typedefs and consolidate all casts. + * o Remove dead code and unnecessary checks. + * o Consolidate starting/stopping timers. + * o Consolidate handling of primary module param throughout the code. + * o Removed multicast module param support - all settings are done + * according to mode. + * o Slave list iteration - bond is no longer part of the list, + * added cyclic list iteration macros. + * o Consolidate error handling in all xmit functions. + * Style changes: + * o Consolidate function naming and declarations. + * o Consolidate function params and local variables names. + * o Consolidate return values. + * o Consolidate curly braces. + * o Consolidate conditionals format. + * o Change struct member names and types. + * o Chomp trailing spaces, remove empty lines, fix indentations. + * o Re-organize code according to context. + * + * 2003/12/30 - Amir Noam <amir.noam at intel dot com> + * - Fixed: Cannot remove and re-enslave the original active slave. + * - Fixed: Releasing the original active slave causes mac address + * duplication. + * - Add support for slaves that use ethtool_ops. + * Set version to 2.5.3. + * + * 2004/01/05 - Amir Noam <amir.noam at intel dot com> + * - Save bonding parameters per bond instead of using the global values. + * Set version to 2.5.4. + * + * 2004/01/14 - Shmulik Hen <shmulik.hen at intel dot com> + * - Enhance VLAN support: + * * Add support for VLAN hardware acceleration capable slaves. + * * Add capability to tag self generated packets in ALB/TLB modes. + * Set version to 2.6.0. + * 2004/10/29 - Mitch Williams <mitch.a.williams at intel dot com> + * - Fixed bug when unloading module while using 802.3ad. If + * spinlock debugging is turned on, this causes a stack dump. + * Solution is to move call to dev_remove_pack outside of the + * spinlock. + * Set version to 2.6.1. + * + */ + +//#define BONDING_DEBUG 1 + +#include <linux/config.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/sched.h> +#include <linux/types.h> +#include <linux/fcntl.h> +#include <linux/interrupt.h> +#include <linux/ptrace.h> +#include <linux/ioport.h> +#include <linux/in.h> +#include <linux/ip.h> +#include <linux/slab.h> +#include <linux/string.h> +#include <linux/init.h> +#include <linux/timer.h> +#include <linux/socket.h> +#include <linux/ctype.h> +#include <linux/inet.h> +#include <linux/bitops.h> +#include <asm/system.h> +#include <asm/io.h> +#include <asm/dma.h> +#include <asm/uaccess.h> +#include <linux/errno.h> +#include <linux/netdevice.h> +#include <linux/inetdevice.h> +#include <linux/etherdevice.h> +#include <linux/skbuff.h> +#include <net/sock.h> +#include <linux/rtnetlink.h> +#include <linux/proc_fs.h> +#include <linux/seq_file.h> +#include <linux/smp.h> +#include <linux/if_ether.h> +#include <net/arp.h> +#include <linux/mii.h> +#include <linux/ethtool.h> +#include <linux/if_vlan.h> +#include <linux/if_bonding.h> +#include "bonding.h" +#include "bond_3ad.h" +#include "bond_alb.h" + +/*---------------------------- Module parameters ----------------------------*/ + +/* monitor all links that often (in milliseconds). <=0 disables monitoring */ +#define BOND_LINK_MON_INTERV 0 +#define BOND_LINK_ARP_INTERV 0 + +static int max_bonds = BOND_DEFAULT_MAX_BONDS; +static int miimon = BOND_LINK_MON_INTERV; +static int updelay = 0; +static int downdelay = 0; +static int use_carrier = 1; +static char *mode = NULL; +static char *primary = NULL; +static char *lacp_rate = NULL; +static int arp_interval = BOND_LINK_ARP_INTERV; +static char *arp_ip_target[BOND_MAX_ARP_TARGETS] = { NULL, }; + +module_param(max_bonds, int, 0); +MODULE_PARM_DESC(max_bonds, "Max number of bonded devices"); +module_param(miimon, int, 0); +MODULE_PARM_DESC(miimon, "Link check interval in milliseconds"); +module_param(updelay, int, 0); +MODULE_PARM_DESC(updelay, "Delay before considering link up, in milliseconds"); +module_param(downdelay, int, 0); +MODULE_PARM_DESC(downdelay, "Delay before considering link down, in milliseconds"); +module_param(use_carrier, int, 0); +MODULE_PARM_DESC(use_carrier, "Use netif_carrier_ok (vs MII ioctls) in miimon; 0 for off, 1 for on (default)"); +module_param(mode, charp, 0); +MODULE_PARM_DESC(mode, "Mode of operation : 0 for round robin, 1 for active-backup, 2 for xor"); +module_param(primary, charp, 0); +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(arp_interval, int, 0); +MODULE_PARM_DESC(arp_interval, "arp interval in milliseconds"); +module_param_array(arp_ip_target, charp, NULL, 0); +MODULE_PARM_DESC(arp_ip_target, "arp targets in n.n.n.n form"); + +/*----------------------------- Global variables ----------------------------*/ + +static const char *version = + DRV_DESCRIPTION ": v" DRV_VERSION " (" DRV_RELDATE ")\n"; + +static LIST_HEAD(bond_dev_list); + +#ifdef CONFIG_PROC_FS +static struct proc_dir_entry *bond_proc_dir = NULL; +#endif + +static u32 arp_target[BOND_MAX_ARP_TARGETS] = { 0, } ; +static int arp_ip_count = 0; +static u32 my_ip = 0; +static int bond_mode = BOND_MODE_ROUNDROBIN; +static int lacp_fast = 0; +static int app_abi_ver = 0; +static int orig_app_abi_ver = -1; /* This is used to save the first ABI version + * we receive from the application. Once set, + * it won't be changed, and the module will + * refuse to enslave/release interfaces if the + * command comes from an application using + * another ABI version. + */ + +struct bond_parm_tbl { + char *modename; + int mode; +}; + +static struct bond_parm_tbl bond_lacp_tbl[] = { +{ "slow", AD_LACP_SLOW}, +{ "fast", AD_LACP_FAST}, +{ NULL, -1}, +}; + +static struct bond_parm_tbl bond_mode_tbl[] = { +{ "balance-rr", BOND_MODE_ROUNDROBIN}, +{ "active-backup", BOND_MODE_ACTIVEBACKUP}, +{ "balance-xor", BOND_MODE_XOR}, +{ "broadcast", BOND_MODE_BROADCAST}, +{ "802.3ad", BOND_MODE_8023AD}, +{ "balance-tlb", BOND_MODE_TLB}, +{ "balance-alb", BOND_MODE_ALB}, +{ NULL, -1}, +}; + +/*-------------------------- Forward declarations ---------------------------*/ + +static inline void bond_set_mode_ops(struct net_device *bond_dev, int mode); + +/*---------------------------- General routines -----------------------------*/ + +static const char *bond_mode_name(int mode) +{ + switch (mode) { + case BOND_MODE_ROUNDROBIN : + return "load balancing (round-robin)"; + case BOND_MODE_ACTIVEBACKUP : + return "fault-tolerance (active-backup)"; + case BOND_MODE_XOR : + return "load balancing (xor)"; + case BOND_MODE_BROADCAST : + return "fault-tolerance (broadcast)"; + case BOND_MODE_8023AD: + return "IEEE 802.3ad Dynamic link aggregation"; + case BOND_MODE_TLB: + return "transmit load balancing"; + case BOND_MODE_ALB: + return "adaptive load balancing"; + default: + return "unknown"; + } +} + +/*---------------------------------- VLAN -----------------------------------*/ + +/** + * bond_add_vlan - add a new vlan id on bond + * @bond: bond that got the notification + * @vlan_id: the vlan id to add + * + * Returns -ENOMEM if allocation failed. + */ +static int bond_add_vlan(struct bonding *bond, unsigned short vlan_id) +{ + struct vlan_entry *vlan; + + dprintk("bond: %s, vlan id %d\n", + (bond ? bond->dev->name: "None"), vlan_id); + + vlan = kmalloc(sizeof(struct vlan_entry), GFP_KERNEL); + if (!vlan) { + return -ENOMEM; + } + + INIT_LIST_HEAD(&vlan->vlan_list); + vlan->vlan_id = vlan_id; + + write_lock_bh(&bond->lock); + + list_add_tail(&vlan->vlan_list, &bond->vlan_list); + + write_unlock_bh(&bond->lock); + + dprintk("added VLAN ID %d on bond %s\n", vlan_id, bond->dev->name); + + return 0; +} + +/** + * bond_del_vlan - delete a vlan id from bond + * @bond: bond that got the notification + * @vlan_id: the vlan id to delete + * + * returns -ENODEV if @vlan_id was not found in @bond. + */ +static int bond_del_vlan(struct bonding *bond, unsigned short vlan_id) +{ + struct vlan_entry *vlan, *next; + int res = -ENODEV; + + dprintk("bond: %s, vlan id %d\n", bond->dev->name, vlan_id); + + write_lock_bh(&bond->lock); + + list_for_each_entry_safe(vlan, next, &bond->vlan_list, vlan_list) { + if (vlan->vlan_id == vlan_id) { + list_del(&vlan->vlan_list); + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + bond_alb_clear_vlan(bond, vlan_id); + } + + dprintk("removed VLAN ID %d from bond %s\n", vlan_id, + bond->dev->name); + + kfree(vlan); + + if (list_empty(&bond->vlan_list) && + (bond->slave_cnt == 0)) { + /* Last VLAN removed and no slaves, so + * restore block on adding VLANs. This will + * be removed once new slaves that are not + * VLAN challenged will be added. + */ + bond->dev->features |= NETIF_F_VLAN_CHALLENGED; + } + + res = 0; + goto out; + } + } + + dprintk("couldn't find VLAN ID %d in bond %s\n", vlan_id, + bond->dev->name); + +out: + write_unlock_bh(&bond->lock); + return res; +} + +/** + * bond_has_challenged_slaves + * @bond: the bond we're working on + * + * Searches the slave list. Returns 1 if a vlan challenged slave + * was found, 0 otherwise. + * + * Assumes bond->lock is held. + */ +static int bond_has_challenged_slaves(struct bonding *bond) +{ + struct slave *slave; + int i; + + bond_for_each_slave(bond, slave, i) { + if (slave->dev->features & NETIF_F_VLAN_CHALLENGED) { + dprintk("found VLAN challenged slave - %s\n", + slave->dev->name); + return 1; + } + } + + dprintk("no VLAN challenged slaves found\n"); + return 0; +} + +/** + * bond_next_vlan - safely skip to the next item in the vlans list. + * @bond: the bond we're working on + * @curr: item we're advancing from + * + * Returns %NULL if list is empty, bond->next_vlan if @curr is %NULL, + * or @curr->next otherwise (even if it is @curr itself again). + * + * Caller must hold bond->lock + */ +struct vlan_entry *bond_next_vlan(struct bonding *bond, struct vlan_entry *curr) +{ + struct vlan_entry *next, *last; + + if (list_empty(&bond->vlan_list)) { + return NULL; + } + + if (!curr) { + next = list_entry(bond->vlan_list.next, + struct vlan_entry, vlan_list); + } else { + last = list_entry(bond->vlan_list.prev, + struct vlan_entry, vlan_list); + if (last == curr) { + next = list_entry(bond->vlan_list.next, + struct vlan_entry, vlan_list); + } else { + next = list_entry(curr->vlan_list.next, + struct vlan_entry, vlan_list); + } + } + + return next; +} + +/** + * bond_dev_queue_xmit - Prepare skb for xmit. + * + * @bond: bond device that got this skb for tx. + * @skb: hw accel VLAN tagged skb to transmit + * @slave_dev: slave that is supposed to xmit this skbuff + * + * When the bond gets an skb to transmit that is + * already hardware accelerated VLAN tagged, and it + * needs to relay this skb to a slave that is not + * hw accel capable, the skb needs to be "unaccelerated", + * i.e. strip the hwaccel tag and re-insert it as part + * of the payload. + */ +int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb, struct net_device *slave_dev) +{ + unsigned short vlan_id; + + if (!list_empty(&bond->vlan_list) && + !(slave_dev->features & NETIF_F_HW_VLAN_TX) && + vlan_get_tag(skb, &vlan_id) == 0) { + skb->dev = slave_dev; + skb = vlan_put_tag(skb, vlan_id); + if (!skb) { + /* vlan_put_tag() frees the skb in case of error, + * so return success here so the calling functions + * won't attempt to free is again. + */ + return 0; + } + } else { + skb->dev = slave_dev; + } + + skb->priority = 1; + dev_queue_xmit(skb); + + return 0; +} + +/* + * In the following 3 functions, bond_vlan_rx_register(), bond_vlan_rx_add_vid + * and bond_vlan_rx_kill_vid, We don't protect the slave list iteration with a + * lock because: + * a. This operation is performed in IOCTL context, + * b. The operation is protected by the RTNL semaphore in the 8021q code, + * c. Holding a lock with BH disabled while directly calling a base driver + * entry point is generally a BAD idea. + * + * The design of synchronization/protection for this operation in the 8021q + * module is good for one or more VLAN devices over a single physical device + * and cannot be extended for a teaming solution like bonding, so there is a + * potential race condition here where a net device from the vlan group might + * be referenced (either by a base driver or the 8021q code) while it is being + * removed from the system. However, it turns out we're not making matters + * worse, and if it works for regular VLAN usage it will work here too. +*/ + +/** + * bond_vlan_rx_register - Propagates registration to slaves + * @bond_dev: bonding net device that got called + * @grp: vlan group being registered + */ +static void bond_vlan_rx_register(struct net_device *bond_dev, struct vlan_group *grp) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave; + int i; + + bond->vlgrp = grp; + + bond_for_each_slave(bond, slave, i) { + struct net_device *slave_dev = slave->dev; + + if ((slave_dev->features & NETIF_F_HW_VLAN_RX) && + slave_dev->vlan_rx_register) { + slave_dev->vlan_rx_register(slave_dev, grp); + } + } +} + +/** + * bond_vlan_rx_add_vid - Propagates adding an id to slaves + * @bond_dev: bonding net device that got called + * @vid: vlan id being added + */ +static void bond_vlan_rx_add_vid(struct net_device *bond_dev, uint16_t vid) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave; + int i, res; + + bond_for_each_slave(bond, slave, i) { + struct net_device *slave_dev = slave->dev; + + if ((slave_dev->features & NETIF_F_HW_VLAN_FILTER) && + slave_dev->vlan_rx_add_vid) { + slave_dev->vlan_rx_add_vid(slave_dev, vid); + } + } + + res = bond_add_vlan(bond, vid); + if (res) { + printk(KERN_ERR DRV_NAME + ": %s: Failed to add vlan id %d\n", + bond_dev->name, vid); + } +} + +/** + * bond_vlan_rx_kill_vid - Propagates deleting an id to slaves + * @bond_dev: bonding net device that got called + * @vid: vlan id being removed + */ +static void bond_vlan_rx_kill_vid(struct net_device *bond_dev, uint16_t vid) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave; + struct net_device *vlan_dev; + int i, res; + + bond_for_each_slave(bond, slave, i) { + struct net_device *slave_dev = slave->dev; + + if ((slave_dev->features & NETIF_F_HW_VLAN_FILTER) && + slave_dev->vlan_rx_kill_vid) { + /* Save and then restore vlan_dev in the grp array, + * since the slave's driver might clear it. + */ + vlan_dev = bond->vlgrp->vlan_devices[vid]; + slave_dev->vlan_rx_kill_vid(slave_dev, vid); + bond->vlgrp->vlan_devices[vid] = vlan_dev; + } + } + + res = bond_del_vlan(bond, vid); + if (res) { + printk(KERN_ERR DRV_NAME + ": %s: Failed to remove vlan id %d\n", + bond_dev->name, vid); + } +} + +static void bond_add_vlans_on_slave(struct bonding *bond, struct net_device *slave_dev) +{ + struct vlan_entry *vlan; + + write_lock_bh(&bond->lock); + + if (list_empty(&bond->vlan_list)) { + goto out; + } + + if ((slave_dev->features & NETIF_F_HW_VLAN_RX) && + slave_dev->vlan_rx_register) { + slave_dev->vlan_rx_register(slave_dev, bond->vlgrp); + } + + if (!(slave_dev->features & NETIF_F_HW_VLAN_FILTER) || + !(slave_dev->vlan_rx_add_vid)) { + goto out; + } + + list_for_each_entry(vlan, &bond->vlan_list, vlan_list) { + slave_dev->vlan_rx_add_vid(slave_dev, vlan->vlan_id); + } + +out: + write_unlock_bh(&bond->lock); +} + +static void bond_del_vlans_from_slave(struct bonding *bond, struct net_device *slave_dev) +{ + struct vlan_entry *vlan; + struct net_device *vlan_dev; + + write_lock_bh(&bond->lock); + + if (list_empty(&bond->vlan_list)) { + goto out; + } + + if (!(slave_dev->features & NETIF_F_HW_VLAN_FILTER) || + !(slave_dev->vlan_rx_kill_vid)) { + goto unreg; + } + + list_for_each_entry(vlan, &bond->vlan_list, vlan_list) { + /* Save and then restore vlan_dev in the grp array, + * since the slave's driver might clear it. + */ + vlan_dev = bond->vlgrp->vlan_devices[vlan->vlan_id]; + slave_dev->vlan_rx_kill_vid(slave_dev, vlan->vlan_id); + bond->vlgrp->vlan_devices[vlan->vlan_id] = vlan_dev; + } + +unreg: + if ((slave_dev->features & NETIF_F_HW_VLAN_RX) && + slave_dev->vlan_rx_register) { + slave_dev->vlan_rx_register(slave_dev, NULL); + } + +out: + write_unlock_bh(&bond->lock); +} + +/*------------------------------- Link status -------------------------------*/ + +/* + * Get link speed and duplex from the slave's base driver + * using ethtool. If for some reason the call fails or the + * values are invalid, fake speed and duplex to 100/Full + * and return error. + */ +static int bond_update_speed_duplex(struct slave *slave) +{ + struct net_device *slave_dev = slave->dev; + static int (* ioctl)(struct net_device *, struct ifreq *, int); + struct ifreq ifr; + struct ethtool_cmd etool; + + /* Fake speed and duplex */ + slave->speed = SPEED_100; + slave->duplex = DUPLEX_FULL; + + if (slave_dev->ethtool_ops) { + u32 res; + + if (!slave_dev->ethtool_ops->get_settings) { + return -1; + } + + res = slave_dev->ethtool_ops->get_settings(slave_dev, &etool); + if (res < 0) { + return -1; + } + + goto verify; + } + + ioctl = slave_dev->do_ioctl; + strncpy(ifr.ifr_name, slave_dev->name, IFNAMSIZ); + etool.cmd = ETHTOOL_GSET; + ifr.ifr_data = (char*)&etool; + if (!ioctl || (IOCTL(slave_dev, &ifr, SIOCETHTOOL) < 0)) { + return -1; + } + +verify: + switch (etool.speed) { + case SPEED_10: + case SPEED_100: + case SPEED_1000: + break; + default: + return -1; + } + + switch (etool.duplex) { + case DUPLEX_FULL: + case DUPLEX_HALF: + break; + default: + return -1; + } + + slave->speed = etool.speed; + slave->duplex = etool.duplex; + + return 0; +} + +/* + * if <dev> supports MII link status reporting, check its link status. + * + * We either do MII/ETHTOOL ioctls, or check netif_carrier_ok(), + * depening upon the setting of the use_carrier parameter. + * + * Return either BMSR_LSTATUS, meaning that the link is up (or we + * can't tell and just pretend it is), or 0, meaning that the link is + * down. + * + * If reporting is non-zero, instead of faking link up, return -1 if + * both ETHTOOL and MII ioctls fail (meaning the device does not + * support them). If use_carrier is set, return whatever it says. + * It'd be nice if there was a good way to tell if a driver supports + * netif_carrier, but there really isn't. + */ +static int bond_check_dev_link(struct bonding *bond, struct net_device *slave_dev, int reporting) +{ + static int (* ioctl)(struct net_device *, struct ifreq *, int); + struct ifreq ifr; + struct mii_ioctl_data *mii; + struct ethtool_value etool; + + if (bond->params.use_carrier) { + return netif_carrier_ok(slave_dev) ? BMSR_LSTATUS : 0; + } + + ioctl = slave_dev->do_ioctl; + if (ioctl) { + /* TODO: set pointer to correct ioctl on a per team member */ + /* bases to make this more efficient. that is, once */ + /* we determine the correct ioctl, we will always */ + /* call it and not the others for that team */ + /* member. */ + + /* + * We cannot assume that SIOCGMIIPHY will also read a + * register; not all network drivers (e.g., e100) + * support that. + */ + + /* Yes, the mii is overlaid on the ifreq.ifr_ifru */ + strncpy(ifr.ifr_name, slave_dev->name, IFNAMSIZ); + mii = if_mii(&ifr); + if (IOCTL(slave_dev, &ifr, SIOCGMIIPHY) == 0) { + mii->reg_num = MII_BMSR; + if (IOCTL(slave_dev, &ifr, SIOCGMIIREG) == 0) { + return (mii->val_out & BMSR_LSTATUS); + } + } + } + + /* try SIOCETHTOOL ioctl, some drivers cache ETHTOOL_GLINK */ + /* for a period of time so we attempt to get link status */ + /* from it last if the above MII ioctls fail... */ + if (slave_dev->ethtool_ops) { + if (slave_dev->ethtool_ops->get_link) { + u32 link; + + link = slave_dev->ethtool_ops->get_link(slave_dev); + + return link ? BMSR_LSTATUS : 0; + } + } + + if (ioctl) { + strncpy(ifr.ifr_name, slave_dev->name, IFNAMSIZ); + etool.cmd = ETHTOOL_GLINK; + ifr.ifr_data = (char*)&etool; + if (IOCTL(slave_dev, &ifr, SIOCETHTOOL) == 0) { + if (etool.data == 1) { + return BMSR_LSTATUS; + } else { + dprintk("SIOCETHTOOL shows link down\n"); + return 0; + } + } + } + + /* + * If reporting, report that either there's no dev->do_ioctl, + * or both SIOCGMIIREG and SIOCETHTOOL failed (meaning that we + * cannot report link status). If not reporting, pretend + * we're ok. + */ + return (reporting ? -1 : BMSR_LSTATUS); +} + +/*----------------------------- Multicast list ------------------------------*/ + +/* + * Returns 0 if dmi1 and dmi2 are the same, non-0 otherwise + */ +static inline int bond_is_dmi_same(struct dev_mc_list *dmi1, struct dev_mc_list *dmi2) +{ + return memcmp(dmi1->dmi_addr, dmi2->dmi_addr, dmi1->dmi_addrlen) == 0 && + dmi1->dmi_addrlen == dmi2->dmi_addrlen; +} + +/* + * returns dmi entry if found, NULL otherwise + */ +static struct dev_mc_list *bond_mc_list_find_dmi(struct dev_mc_list *dmi, struct dev_mc_list *mc_list) +{ + struct dev_mc_list *idmi; + + for (idmi = mc_list; idmi; idmi = idmi->next) { + if (bond_is_dmi_same(dmi, idmi)) { + return idmi; + } + } + + return NULL; +} + +/* + * Push the promiscuity flag down to appropriate slaves + */ +static void bond_set_promiscuity(struct bonding *bond, int inc) +{ + if (USES_PRIMARY(bond->params.mode)) { + /* write lock already acquired */ + if (bond->curr_active_slave) { + dev_set_promiscuity(bond->curr_active_slave->dev, inc); + } + } else { + struct slave *slave; + int i; + bond_for_each_slave(bond, slave, i) { + dev_set_promiscuity(slave->dev, inc); + } + } +} + +/* + * Push the allmulti flag down to all slaves + */ +static void bond_set_allmulti(struct bonding *bond, int inc) +{ + if (USES_PRIMARY(bond->params.mode)) { + /* write lock already acquired */ + if (bond->curr_active_slave) { + dev_set_allmulti(bond->curr_active_slave->dev, inc); + } + } else { + struct slave *slave; + int i; + bond_for_each_slave(bond, slave, i) { + dev_set_allmulti(slave->dev, inc); + } + } +} + +/* + * Add a Multicast address to slaves + * according to mode + */ +static void bond_mc_add(struct bonding *bond, void *addr, int alen) +{ + if (USES_PRIMARY(bond->params.mode)) { + /* write lock already acquired */ + if (bond->curr_active_slave) { + dev_mc_add(bond->curr_active_slave->dev, addr, alen, 0); + } + } else { + struct slave *slave; + int i; + bond_for_each_slave(bond, slave, i) { + dev_mc_add(slave->dev, addr, alen, 0); + } + } +} + +/* + * Remove a multicast address from slave + * according to mode + */ +static void bond_mc_delete(struct bonding *bond, void *addr, int alen) +{ + if (USES_PRIMARY(bond->params.mode)) { + /* write lock already acquired */ + if (bond->curr_active_slave) { + dev_mc_delete(bond->curr_active_slave->dev, addr, alen, 0); + } + } else { + struct slave *slave; + int i; + bond_for_each_slave(bond, slave, i) { + dev_mc_delete(slave->dev, addr, alen, 0); + } + } +} + +/* + * Totally destroys the mc_list in bond + */ +static void bond_mc_list_destroy(struct bonding *bond) +{ + struct dev_mc_list *dmi; + + dmi = bond->mc_list; + while (dmi) { + bond->mc_list = dmi->next; + kfree(dmi); + dmi = bond->mc_list; + } +} + +/* + * Copy all the Multicast addresses from src to the bonding device dst + */ +static int bond_mc_list_copy(struct dev_mc_list *mc_list, struct bonding *bond, int gpf_flag) +{ + struct dev_mc_list *dmi, *new_dmi; + + for (dmi = mc_list; dmi; dmi = dmi->next) { + new_dmi = kmalloc(sizeof(struct dev_mc_list), gpf_flag); + + if (!new_dmi) { + /* FIXME: Potential memory leak !!! */ + return -ENOMEM; + } + + new_dmi->next = bond->mc_list; + bond->mc_list = new_dmi; + new_dmi->dmi_addrlen = dmi->dmi_addrlen; + memcpy(new_dmi->dmi_addr, dmi->dmi_addr, dmi->dmi_addrlen); + new_dmi->dmi_users = dmi->dmi_users; + new_dmi->dmi_gusers = dmi->dmi_gusers; + } + + return 0; +} + +/* + * flush all members of flush->mc_list from device dev->mc_list + */ +static void bond_mc_list_flush(struct net_device *bond_dev, struct net_device *slave_dev) +{ + struct bonding *bond = bond_dev->priv; + struct dev_mc_list *dmi; + + for (dmi = bond_dev->mc_list; dmi; dmi = dmi->next) { + dev_mc_delete(slave_dev, dmi->dmi_addr, dmi->dmi_addrlen, 0); + } + + if (bond->params.mode == BOND_MODE_8023AD) { + /* del lacpdu mc addr from mc list */ + u8 lacpdu_multicast[ETH_ALEN] = MULTICAST_LACPDU_ADDR; + + dev_mc_delete(slave_dev, lacpdu_multicast, ETH_ALEN, 0); + } +} + +/*--------------------------- Active slave change ---------------------------*/ + +/* + * Update the mc list and multicast-related flags for the new and + * old active slaves (if any) according to the multicast mode, and + * promiscuous flags unconditionally. + */ +static void bond_mc_swap(struct bonding *bond, struct slave *new_active, struct slave *old_active) +{ + struct dev_mc_list *dmi; + + if (!USES_PRIMARY(bond->params.mode)) { + /* nothing to do - mc list is already up-to-date on + * all slaves + */ + return; + } + + if (old_active) { + if (bond->dev->flags & IFF_PROMISC) { + dev_set_promiscuity(old_active->dev, -1); + } + + if (bond->dev->flags & IFF_ALLMULTI) { + dev_set_allmulti(old_active->dev, -1); + } + + for (dmi = bond->dev->mc_list; dmi; dmi = dmi->next) { + dev_mc_delete(old_active->dev, dmi->dmi_addr, dmi->dmi_addrlen, 0); + } + } + + if (new_active) { + if (bond->dev->flags & IFF_PROMISC) { + dev_set_promiscuity(new_active->dev, 1); + } + + if (bond->dev->flags & IFF_ALLMULTI) { + dev_set_allmulti(new_active->dev, 1); + } + + for (dmi = bond->dev->mc_list; dmi; dmi = dmi->next) { + dev_mc_add(new_active->dev, dmi->dmi_addr, dmi->dmi_addrlen, 0); + } + } +} + +/** + * find_best_interface - select the best available slave to be the active one + * @bond: our bonding struct + * + * Warning: Caller must hold curr_slave_lock for writing. + */ +static struct slave *bond_find_best_slave(struct bonding *bond) +{ + struct slave *new_active, *old_active; + struct slave *bestslave = NULL; + int mintime = bond->params.updelay; + int i; + + new_active = old_active = bond->curr_active_slave; + + if (!new_active) { /* there were no active slaves left */ + if (bond->slave_cnt > 0) { /* found one slave */ + new_active = bond->first_slave; + } else { + return NULL; /* still no slave, return NULL */ + } + } + + /* first try the primary link; if arping, a link must tx/rx traffic + * before it can be considered the curr_active_slave - also, we would skip + * slaves between the curr_active_slave and primary_slave that may be up + * and able to arp + */ + if ((bond->primary_slave) && + (!bond->params.arp_interval) && + (IS_UP(bond->primary_slave->dev))) { + new_active = bond->primary_slave; + } + + /* remember where to stop iterating over the slaves */ + old_active = new_active; + + bond_for_each_slave_from(bond, new_active, i, old_active) { + if (IS_UP(new_active->dev)) { + if (new_active->link == BOND_LINK_UP) { + return new_active; + } else if (new_active->link == BOND_LINK_BACK) { + /* link up, but waiting for stabilization */ + if (new_active->delay < mintime) { + mintime = new_active->delay; + bestslave = new_active; + } + } + } + } + + return bestslave; +} + +/** + * change_active_interface - change the active slave into the specified one + * @bond: our bonding struct + * @new: the new slave to make the active one + * + * Set the new slave to the bond's settings and unset them on the old + * curr_active_slave. + * Setting include flags, mc-list, promiscuity, allmulti, etc. + * + * If @new's link state is %BOND_LINK_BACK we'll set it to %BOND_LINK_UP, + * because it is apparently the best available slave we have, even though its + * updelay hasn't timed out yet. + * + * Warning: Caller must hold curr_slave_lock for writing. + */ +static void bond_change_active_slave(struct bonding *bond, struct slave *new_active) +{ + struct slave *old_active = bond->curr_active_slave; + + if (old_active == new_active) { + return; + } + + if (new_active) { + if (new_active->link == BOND_LINK_BACK) { + if (USES_PRIMARY(bond->params.mode)) { + printk(KERN_INFO DRV_NAME + ": %s: making interface %s the new " + "active one %d ms earlier.\n", + bond->dev->name, new_active->dev->name, + (bond->params.updelay - new_active->delay) * bond->params.miimon); + } + + new_active->delay = 0; + new_active->link = BOND_LINK_UP; + new_active->jiffies = jiffies; + + if (bond->params.mode == BOND_MODE_8023AD) { + bond_3ad_handle_link_change(new_active, BOND_LINK_UP); + } + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + bond_alb_handle_link_change(bond, new_active, BOND_LINK_UP); + } + } else { + if (USES_PRIMARY(bond->params.mode)) { + printk(KERN_INFO DRV_NAME + ": %s: making interface %s the new " + "active one.\n", + bond->dev->name, new_active->dev->name); + } + } + } + + if (bond->params.mode == BOND_MODE_ACTIVEBACKUP) { + if (old_active) { + bond_set_slave_inactive_flags(old_active); + } + + if (new_active) { + bond_set_slave_active_flags(new_active); + } + } + + if (USES_PRIMARY(bond->params.mode)) { + bond_mc_swap(bond, new_active, old_active); + } + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + bond_alb_handle_active_change(bond, new_active); + } else { + bond->curr_active_slave = new_active; + } +} + +/** + * bond_select_active_slave - select a new active slave, if needed + * @bond: our bonding struct + * + * This functions shoud be called when one of the following occurs: + * - The old curr_active_slave has been released or lost its link. + * - The primary_slave has got its link back. + * - A slave has got its link back and there's no old curr_active_slave. + * + * Warning: Caller must hold curr_slave_lock for writing. + */ +static void bond_select_active_slave(struct bonding *bond) +{ + struct slave *best_slave; + + best_slave = bond_find_best_slave(bond); + if (best_slave != bond->curr_active_slave) { + bond_change_active_slave(bond, best_slave); + } +} + +/*--------------------------- slave list handling ---------------------------*/ + +/* + * This function attaches the slave to the end of list. + * + * bond->lock held for writing by caller. + */ +static void bond_attach_slave(struct bonding *bond, struct slave *new_slave) +{ + if (bond->first_slave == NULL) { /* attaching the first slave */ + new_slave->next = new_slave; + new_slave->prev = new_slave; + bond->first_slave = new_slave; + } else { + new_slave->next = bond->first_slave; + new_slave->prev = bond->first_slave->prev; + new_slave->next->prev = new_slave; + new_slave->prev->next = new_slave; + } + + bond->slave_cnt++; +} + +/* + * This function detaches the slave from the list. + * WARNING: no check is made to verify if the slave effectively + * belongs to <bond>. + * Nothing is freed on return, structures are just unchained. + * If any slave pointer in bond was pointing to <slave>, + * it should be changed by the calling function. + * + * bond->lock held for writing by caller. + */ +static void bond_detach_slave(struct bonding *bond, struct slave *slave) +{ + if (slave->next) { + slave->next->prev = slave->prev; + } + + if (slave->prev) { + slave->prev->next = slave->next; + } + + if (bond->first_slave == slave) { /* slave is the first slave */ + if (bond->slave_cnt > 1) { /* there are more slave */ + bond->first_slave = slave->next; + } else { + bond->first_slave = NULL; /* slave was the last one */ + } + } + + slave->next = NULL; + slave->prev = NULL; + bond->slave_cnt--; +} + +/*---------------------------------- IOCTL ----------------------------------*/ + +static int bond_sethwaddr(struct net_device *bond_dev, struct net_device *slave_dev) +{ + dprintk("bond_dev=%p\n", bond_dev); + dprintk("slave_dev=%p\n", slave_dev); + dprintk("slave_dev->addr_len=%d\n", slave_dev->addr_len); + memcpy(bond_dev->dev_addr, slave_dev->dev_addr, slave_dev->addr_len); + return 0; +} + +/* enslave device <slave> to bond device <master> */ +static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *new_slave = NULL; + struct dev_mc_list *dmi; + struct sockaddr addr; + int link_reporting; + int old_features = bond_dev->features; + int res = 0; + + if (slave_dev->do_ioctl == NULL) { + printk(KERN_WARNING DRV_NAME + ": Warning : no link monitoring support for %s\n", + slave_dev->name); + } + + /* bond must be initialized by bond_open() before enslaving */ + if (!(bond_dev->flags & IFF_UP)) { + dprintk("Error, master_dev is not up\n"); + return -EPERM; + } + + /* already enslaved */ + if (slave_dev->flags & IFF_SLAVE) { + dprintk("Error, Device was already enslaved\n"); + return -EBUSY; + } + + /* vlan challenged mutual exclusion */ + /* no need to lock since we're protected by rtnl_lock */ + if (slave_dev->features & NETIF_F_VLAN_CHALLENGED) { + dprintk("%s: NETIF_F_VLAN_CHALLENGED\n", slave_dev->name); + if (!list_empty(&bond->vlan_list)) { + printk(KERN_ERR DRV_NAME + ": Error: cannot enslave VLAN " + "challenged slave %s on VLAN enabled " + "bond %s\n", slave_dev->name, + bond_dev->name); + return -EPERM; + } else { + printk(KERN_WARNING DRV_NAME + ": Warning: enslaved VLAN challenged " + "slave %s. Adding VLANs will be blocked as " + "long as %s is part of bond %s\n", + slave_dev->name, slave_dev->name, + bond_dev->name); + bond_dev->features |= NETIF_F_VLAN_CHALLENGED; + } + } else { + dprintk("%s: ! NETIF_F_VLAN_CHALLENGED\n", slave_dev->name); + if (bond->slave_cnt == 0) { + /* First slave, and it is not VLAN challenged, + * so remove the block of adding VLANs over the bond. + */ + bond_dev->features &= ~NETIF_F_VLAN_CHALLENGED; + } + } + + if (app_abi_ver >= 1) { + /* The application is using an ABI, which requires the + * slave interface to be closed. + */ + if ((slave_dev->flags & IFF_UP)) { + printk(KERN_ERR DRV_NAME + ": Error: %s is up\n", + slave_dev->name); + res = -EPERM; + goto err_undo_flags; + } + + if (slave_dev->set_mac_address == NULL) { + printk(KERN_ERR DRV_NAME + ": Error: The slave device you specified does " + "not support setting the MAC address.\n"); + printk(KERN_ERR + "Your kernel likely does not support slave " + "devices.\n"); + + res = -EOPNOTSUPP; + goto err_undo_flags; + } + } else { + /* The application is not using an ABI, which requires the + * slave interface to be open. + */ + if (!(slave_dev->flags & IFF_UP)) { + printk(KERN_ERR DRV_NAME + ": Error: %s is not running\n", + slave_dev->name); + res = -EINVAL; + goto err_undo_flags; + } + + if ((bond->params.mode == BOND_MODE_8023AD) || + (bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + printk(KERN_ERR DRV_NAME + ": Error: to use %s mode, you must upgrade " + "ifenslave.\n", + bond_mode_name(bond->params.mode)); + res = -EOPNOTSUPP; + goto err_undo_flags; + } + } + + new_slave = kmalloc(sizeof(struct slave), GFP_KERNEL); + if (!new_slave) { + res = -ENOMEM; + goto err_undo_flags; + } + + memset(new_slave, 0, sizeof(struct slave)); + + /* save slave's original flags before calling + * netdev_set_master and dev_open + */ + new_slave->original_flags = slave_dev->flags; + + if (app_abi_ver >= 1) { + /* save slave's original ("permanent") mac address for + * modes that needs it, and for restoring it upon release, + * and then set it to the master's address + */ + memcpy(new_slave->perm_hwaddr, slave_dev->dev_addr, ETH_ALEN); + + /* set slave to master's mac address + * The application already set the master's + * mac address to that of the first slave + */ + memcpy(addr.sa_data, bond_dev->dev_addr, bond_dev->addr_len); + addr.sa_family = slave_dev->type; + res = dev_set_mac_address(slave_dev, &addr); + if (res) { + dprintk("Error %d calling set_mac_address\n", res); + goto err_free; + } + + /* open the slave since the application closed it */ + res = dev_open(slave_dev); + if (res) { + dprintk("Openning slave %s failed\n", slave_dev->name); + goto err_restore_mac; + } + } + + res = netdev_set_master(slave_dev, bond_dev); + if (res) { + dprintk("Error %d calling netdev_set_master\n", res); + if (app_abi_ver < 1) { + goto err_free; + } else { + goto err_close; + } + } + + new_slave->dev = slave_dev; + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + /* bond_alb_init_slave() must be called before all other stages since + * it might fail and we do not want to have to undo everything + */ + res = bond_alb_init_slave(bond, new_slave); + if (res) { + goto err_unset_master; + } + } + + /* If the mode USES_PRIMARY, then the new slave gets the + * master's promisc (and mc) settings only if it becomes the + * curr_active_slave, and that is taken care of later when calling + * bond_change_active() + */ + if (!USES_PRIMARY(bond->params.mode)) { + /* set promiscuity level to new slave */ + if (bond_dev->flags & IFF_PROMISC) { + dev_set_promiscuity(slave_dev, 1); + } + + /* set allmulti level to new slave */ + if (bond_dev->flags & IFF_ALLMULTI) { + dev_set_allmulti(slave_dev, 1); + } + + /* upload master's mc_list to new slave */ + for (dmi = bond_dev->mc_list; dmi; dmi = dmi->next) { + dev_mc_add (slave_dev, dmi->dmi_addr, dmi->dmi_addrlen, 0); + } + } + + if (bond->params.mode == BOND_MODE_8023AD) { + /* add lacpdu mc addr to mc list */ + u8 lacpdu_multicast[ETH_ALEN] = MULTICAST_LACPDU_ADDR; + + dev_mc_add(slave_dev, lacpdu_multicast, ETH_ALEN, 0); + } + + bond_add_vlans_on_slave(bond, slave_dev); + + write_lock_bh(&bond->lock); + + bond_attach_slave(bond, new_slave); + + new_slave->delay = 0; + new_slave->link_failure_count = 0; + + if (bond->params.miimon && !bond->params.use_carrier) { + link_reporting = bond_check_dev_link(bond, slave_dev, 1); + + if ((link_reporting == -1) && !bond->params.arp_interval) { + /* + * miimon is set but a bonded network driver + * does not support ETHTOOL/MII and + * arp_interval is not set. Note: if + * use_carrier is enabled, we will never go + * here (because netif_carrier is always + * supported); thus, we don't need to change + * the messages for netif_carrier. + */ + printk(KERN_WARNING DRV_NAME + ": Warning: MII and ETHTOOL support not " + "available for interface %s, and " + "arp_interval/arp_ip_target module parameters " + "not specified, thus bonding will not detect " + "link failures! see bonding.txt for details.\n", + slave_dev->name); + } else if (link_reporting == -1) { + /* unable get link status using mii/ethtool */ + printk(KERN_WARNING DRV_NAME + ": Warning: can't get link status from " + "interface %s; the network driver associated " + "with this interface does not support MII or " + "ETHTOOL link status reporting, thus miimon " + "has no effect on this interface.\n", + slave_dev->name); + } + } + + /* check for initial state */ + if (!bond->params.miimon || + (bond_check_dev_link(bond, slave_dev, 0) == BMSR_LSTATUS)) { + if (bond->params.updelay) { + dprintk("Initial state of slave_dev is " + "BOND_LINK_BACK\n"); + new_slave->link = BOND_LINK_BACK; + new_slave->delay = bond->params.updelay; + } else { + dprintk("Initial state of slave_dev is " + "BOND_LINK_UP\n"); + new_slave->link = BOND_LINK_UP; + } + new_slave->jiffies = jiffies; + } else { + dprintk("Initial state of slave_dev is " + "BOND_LINK_DOWN\n"); + new_slave->link = BOND_LINK_DOWN; + } + + if (bond_update_speed_duplex(new_slave) && + (new_slave->link != BOND_LINK_DOWN)) { + printk(KERN_WARNING DRV_NAME + ": Warning: failed to get speed and duplex from %s, " + "assumed to be 100Mb/sec and Full.\n", + new_slave->dev->name); + + if (bond->params.mode == BOND_MODE_8023AD) { + printk(KERN_WARNING + "Operation of 802.3ad mode requires ETHTOOL " + "support in base driver for proper aggregator " + "selection.\n"); + } + } + + if (USES_PRIMARY(bond->params.mode) && bond->params.primary[0]) { + /* if there is a primary slave, remember it */ + if (strcmp(bond->params.primary, new_slave->dev->name) == 0) { + bond->primary_slave = new_slave; + } + } + + switch (bond->params.mode) { + case BOND_MODE_ACTIVEBACKUP: + /* if we're in active-backup mode, we need one and only one active + * interface. The backup interfaces will have their NOARP flag set + * because we need them to be completely deaf and not to respond to + * any ARP request on the network to avoid fooling a switch. Thus, + * since we guarantee that curr_active_slave always point to the last + * usable interface, we just have to verify this interface's flag. + */ + if (((!bond->curr_active_slave) || + (bond->curr_active_slave->dev->flags & IFF_NOARP)) && + (new_slave->link != BOND_LINK_DOWN)) { + dprintk("This is the first active slave\n"); + /* first slave or no active slave yet, and this link + is OK, so make this interface the active one */ + bond_change_active_slave(bond, new_slave); + } else { + dprintk("This is just a backup slave\n"); + bond_set_slave_inactive_flags(new_slave); + } + break; + case BOND_MODE_8023AD: + /* in 802.3ad mode, the internal mechanism + * will activate the slaves in the selected + * aggregator + */ + bond_set_slave_inactive_flags(new_slave); + /* if this is the first slave */ + if (bond->slave_cnt == 1) { + SLAVE_AD_INFO(new_slave).id = 1; + /* Initialize AD with the number of times that the AD timer is called in 1 second + * can be called only after the mac address of the bond is set + */ + bond_3ad_initialize(bond, 1000/AD_TIMER_INTERVAL, + bond->params.lacp_fast); + } else { + SLAVE_AD_INFO(new_slave).id = + SLAVE_AD_INFO(new_slave->prev).id + 1; + } + + bond_3ad_bind_slave(new_slave); + break; + case BOND_MODE_TLB: + case BOND_MODE_ALB: + new_slave->state = BOND_STATE_ACTIVE; + if ((!bond->curr_active_slave) && + (new_slave->link != BOND_LINK_DOWN)) { + /* first slave or no active slave yet, and this link + * is OK, so make this interface the active one + */ + bond_change_active_slave(bond, new_slave); + } + break; + default: + dprintk("This slave is always active in trunk mode\n"); + + /* always active in trunk mode */ + new_slave->state = BOND_STATE_ACTIVE; + + /* In trunking mode there is little meaning to curr_active_slave + * anyway (it holds no special properties of the bond device), + * so we can change it without calling change_active_interface() + */ + if (!bond->curr_active_slave) { + bond->curr_active_slave = new_slave; + } + break; + } /* switch(bond_mode) */ + + write_unlock_bh(&bond->lock); + + if (app_abi_ver < 1) { + /* + * !!! This is to support old versions of ifenslave. + * We can remove this in 2.5 because our ifenslave takes + * care of this for us. + * We check to see if the master has a mac address yet. + * If not, we'll give it the mac address of our slave device. + */ + int ndx = 0; + + for (ndx = 0; ndx < bond_dev->addr_len; ndx++) { + dprintk("Checking ndx=%d of bond_dev->dev_addr\n", + ndx); + if (bond_dev->dev_addr[ndx] != 0) { + dprintk("Found non-zero byte at ndx=%d\n", + ndx); + break; + } + } + + if (ndx == bond_dev->addr_len) { + /* + * We got all the way through the address and it was + * all 0's. + */ + dprintk("%s doesn't have a MAC address yet. \n", + bond_dev->name); + dprintk("Going to give assign it from %s.\n", + slave_dev->name); + bond_sethwaddr(bond_dev, slave_dev); + } + } + + printk(KERN_INFO DRV_NAME + ": %s: enslaving %s as a%s interface with a%s link.\n", + bond_dev->name, slave_dev->name, + new_slave->state == BOND_STATE_ACTIVE ? "n active" : " backup", + new_slave->link != BOND_LINK_DOWN ? "n up" : " down"); + + /* enslave is successful */ + return 0; + +/* Undo stages on error */ +err_unset_master: + netdev_set_master(slave_dev, NULL); + +err_close: + dev_close(slave_dev); + +err_restore_mac: + memcpy(addr.sa_data, new_slave->perm_hwaddr, ETH_ALEN); + addr.sa_family = slave_dev->type; + dev_set_mac_address(slave_dev, &addr); + +err_free: + kfree(new_slave); + +err_undo_flags: + bond_dev->features = old_features; + + return res; +} + +/* + * Try to release the slave device <slave> from the bond device <master> + * It is legal to access curr_active_slave without a lock because all the function + * is write-locked. + * + * The rules for slave state should be: + * for Active/Backup: + * Active stays on all backups go down + * for Bonded connections: + * The first up interface should be left on and all others downed. + */ +static int bond_release(struct net_device *bond_dev, struct net_device *slave_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave, *oldcurrent; + struct sockaddr addr; + int mac_addr_differ; + + /* slave is not a slave or master is not master of this slave */ + if (!(slave_dev->flags & IFF_SLAVE) || + (slave_dev->master != bond_dev)) { + printk(KERN_ERR DRV_NAME + ": Error: %s: cannot release %s.\n", + bond_dev->name, slave_dev->name); + return -EINVAL; + } + + write_lock_bh(&bond->lock); + + slave = bond_get_slave_by_dev(bond, slave_dev); + if (!slave) { + /* not a slave of this bond */ + printk(KERN_INFO DRV_NAME + ": %s: %s not enslaved\n", + bond_dev->name, slave_dev->name); + return -EINVAL; + } + + mac_addr_differ = memcmp(bond_dev->dev_addr, + slave->perm_hwaddr, + ETH_ALEN); + if (!mac_addr_differ && (bond->slave_cnt > 1)) { + printk(KERN_WARNING DRV_NAME + ": Warning: the permanent HWaddr of %s " + "- %02X:%02X:%02X:%02X:%02X:%02X - is " + "still in use by %s. Set the HWaddr of " + "%s to a different address to avoid " + "conflicts.\n", + slave_dev->name, + slave->perm_hwaddr[0], + slave->perm_hwaddr[1], + slave->perm_hwaddr[2], + slave->perm_hwaddr[3], + slave->perm_hwaddr[4], + slave->perm_hwaddr[5], + bond_dev->name, + slave_dev->name); + } + + /* Inform AD package of unbinding of slave. */ + if (bond->params.mode == BOND_MODE_8023AD) { + /* must be called before the slave is + * detached from the list + */ + bond_3ad_unbind_slave(slave); + } + + printk(KERN_INFO DRV_NAME + ": %s: releasing %s interface %s\n", + bond_dev->name, + (slave->state == BOND_STATE_ACTIVE) + ? "active" : "backup", + slave_dev->name); + + oldcurrent = bond->curr_active_slave; + + bond->current_arp_slave = NULL; + + /* release the slave from its bond */ + bond_detach_slave(bond, slave); + + if (bond->primary_slave == slave) { + bond->primary_slave = NULL; + } + + if (oldcurrent == slave) { + bond_change_active_slave(bond, NULL); + } + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + /* Must be called only after the slave has been + * detached from the list and the curr_active_slave + * has been cleared (if our_slave == old_current), + * but before a new active slave is selected. + */ + bond_alb_deinit_slave(bond, slave); + } + + if (oldcurrent == slave) { + bond_select_active_slave(bond); + + if (!bond->curr_active_slave) { + printk(KERN_INFO DRV_NAME + ": %s: now running without any active " + "interface !\n", + bond_dev->name); + } + } + + if (bond->slave_cnt == 0) { + /* if the last slave was removed, zero the mac address + * of the master so it will be set by the application + * to the mac address of the first slave + */ + memset(bond_dev->dev_addr, 0, bond_dev->addr_len); + + if (list_empty(&bond->vlan_list)) { + bond_dev->features |= NETIF_F_VLAN_CHALLENGED; + } else { + printk(KERN_WARNING DRV_NAME + ": Warning: clearing HW address of %s while it " + "still has VLANs.\n", + bond_dev->name); + printk(KERN_WARNING DRV_NAME + ": When re-adding slaves, make sure the bond's " + "HW address matches its VLANs'.\n"); + } + } else if ((bond_dev->features & NETIF_F_VLAN_CHALLENGED) && + !bond_has_challenged_slaves(bond)) { + printk(KERN_INFO DRV_NAME + ": last VLAN challenged slave %s " + "left bond %s. VLAN blocking is removed\n", + slave_dev->name, bond_dev->name); + bond_dev->features &= ~NETIF_F_VLAN_CHALLENGED; + } + + write_unlock_bh(&bond->lock); + + bond_del_vlans_from_slave(bond, slave_dev); + + /* If the mode USES_PRIMARY, then we should only remove its + * promisc and mc settings if it was the curr_active_slave, but that was + * already taken care of above when we detached the slave + */ + if (!USES_PRIMARY(bond->params.mode)) { + /* unset promiscuity level from slave */ + if (bond_dev->flags & IFF_PROMISC) { + dev_set_promiscuity(slave_dev, -1); + } + + /* unset allmulti level from slave */ + if (bond_dev->flags & IFF_ALLMULTI) { + dev_set_allmulti(slave_dev, -1); + } + + /* flush master's mc_list from slave */ + bond_mc_list_flush(bond_dev, slave_dev); + } + + netdev_set_master(slave_dev, NULL); + + /* close slave before restoring its mac address */ + dev_close(slave_dev); + + if (app_abi_ver >= 1) { + /* restore original ("permanent") mac address */ + memcpy(addr.sa_data, slave->perm_hwaddr, ETH_ALEN); + addr.sa_family = slave_dev->type; + dev_set_mac_address(slave_dev, &addr); + } + + /* restore the original state of the + * IFF_NOARP flag that might have been + * set by bond_set_slave_inactive_flags() + */ + if ((slave->original_flags & IFF_NOARP) == 0) { + slave_dev->flags &= ~IFF_NOARP; + } + + kfree(slave); + + return 0; /* deletion OK */ +} + +/* + * This function releases all slaves. + */ +static int bond_release_all(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave; + struct net_device *slave_dev; + struct sockaddr addr; + + write_lock_bh(&bond->lock); + + if (bond->slave_cnt == 0) { + goto out; + } + + bond->current_arp_slave = NULL; + bond->primary_slave = NULL; + bond_change_active_slave(bond, NULL); + + while ((slave = bond->first_slave) != NULL) { + /* Inform AD package of unbinding of slave + * before slave is detached from the list. + */ + if (bond->params.mode == BOND_MODE_8023AD) { + bond_3ad_unbind_slave(slave); + } + + slave_dev = slave->dev; + bond_detach_slave(bond, slave); + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + /* must be called only after the slave + * has been detached from the list + */ + bond_alb_deinit_slave(bond, slave); + } + + /* now that the slave is detached, unlock and perform + * all the undo steps that should not be called from + * within a lock. + */ + write_unlock_bh(&bond->lock); + + bond_del_vlans_from_slave(bond, slave_dev); + + /* If the mode USES_PRIMARY, then we should only remove its + * promisc and mc settings if it was the curr_active_slave, but that was + * already taken care of above when we detached the slave + */ + if (!USES_PRIMARY(bond->params.mode)) { + /* unset promiscuity level from slave */ + if (bond_dev->flags & IFF_PROMISC) { + dev_set_promiscuity(slave_dev, -1); + } + + /* unset allmulti level from slave */ + if (bond_dev->flags & IFF_ALLMULTI) { + dev_set_allmulti(slave_dev, -1); + } + + /* flush master's mc_list from slave */ + bond_mc_list_flush(bond_dev, slave_dev); + } + + netdev_set_master(slave_dev, NULL); + + /* close slave before restoring its mac address */ + dev_close(slave_dev); + + if (app_abi_ver >= 1) { + /* restore original ("permanent") mac address*/ + memcpy(addr.sa_data, slave->perm_hwaddr, ETH_ALEN); + addr.sa_family = slave_dev->type; + dev_set_mac_address(slave_dev, &addr); + } + + /* restore the original state of the IFF_NOARP flag that might have + * been set by bond_set_slave_inactive_flags() + */ + if ((slave->original_flags & IFF_NOARP) == 0) { + slave_dev->flags &= ~IFF_NOARP; + } + + kfree(slave); + + /* re-acquire the lock before getting the next slave */ + write_lock_bh(&bond->lock); + } + + /* zero the mac address of the master so it will be + * set by the application to the mac address of the + * first slave + */ + memset(bond_dev->dev_addr, 0, bond_dev->addr_len); + + if (list_empty(&bond->vlan_list)) { + bond_dev->features |= NETIF_F_VLAN_CHALLENGED; + } else { + printk(KERN_WARNING DRV_NAME + ": Warning: clearing HW address of %s while it " + "still has VLANs.\n", + bond_dev->name); + printk(KERN_WARNING DRV_NAME + ": When re-adding slaves, make sure the bond's " + "HW address matches its VLANs'.\n"); + } + + printk(KERN_INFO DRV_NAME + ": %s: released all slaves\n", + bond_dev->name); + +out: + write_unlock_bh(&bond->lock); + + return 0; +} + +/* + * This function changes the active slave to slave <slave_dev>. + * It returns -EINVAL in the following cases. + * - <slave_dev> is not found in the list. + * - There is not active slave now. + * - <slave_dev> is already active. + * - The link state of <slave_dev> is not BOND_LINK_UP. + * - <slave_dev> is not running. + * In these cases, this fuction does nothing. + * In the other cases, currnt_slave pointer is changed and 0 is returned. + */ +static int bond_ioctl_change_active(struct net_device *bond_dev, struct net_device *slave_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *old_active = NULL; + struct slave *new_active = NULL; + int res = 0; + + if (!USES_PRIMARY(bond->params.mode)) { + return -EINVAL; + } + + /* Verify that master_dev is indeed the master of slave_dev */ + if (!(slave_dev->flags & IFF_SLAVE) || + (slave_dev->master != bond_dev)) { + return -EINVAL; + } + + write_lock_bh(&bond->lock); + + old_active = bond->curr_active_slave; + new_active = bond_get_slave_by_dev(bond, slave_dev); + + /* + * Changing to the current active: do nothing; return success. + */ + if (new_active && (new_active == old_active)) { + write_unlock_bh(&bond->lock); + return 0; + } + + if ((new_active) && + (old_active) && + (new_active->link == BOND_LINK_UP) && + IS_UP(new_active->dev)) { + bond_change_active_slave(bond, new_active); + } else { + res = -EINVAL; + } + + write_unlock_bh(&bond->lock); + + return res; +} + +static int bond_ethtool_ioctl(struct net_device *bond_dev, struct ifreq *ifr) +{ + struct ethtool_drvinfo info; + void __user *addr = ifr->ifr_data; + uint32_t cmd; + + if (get_user(cmd, (uint32_t __user *)addr)) { + return -EFAULT; + } + + switch (cmd) { + case ETHTOOL_GDRVINFO: + if (copy_from_user(&info, addr, sizeof(info))) { + return -EFAULT; + } + + if (strcmp(info.driver, "ifenslave") == 0) { + int new_abi_ver; + char *endptr; + + new_abi_ver = simple_strtoul(info.fw_version, + &endptr, 0); + if (*endptr) { + printk(KERN_ERR DRV_NAME + ": Error: got invalid ABI " + "version from application\n"); + + return -EINVAL; + } + + if (orig_app_abi_ver == -1) { + orig_app_abi_ver = new_abi_ver; + } + + app_abi_ver = new_abi_ver; + } + + strncpy(info.driver, DRV_NAME, 32); + strncpy(info.version, DRV_VERSION, 32); + snprintf(info.fw_version, 32, "%d", BOND_ABI_VERSION); + + if (copy_to_user(addr, &info, sizeof(info))) { + return -EFAULT; + } + + return 0; + default: + return -EOPNOTSUPP; + } +} + +static int bond_info_query(struct net_device *bond_dev, struct ifbond *info) +{ + struct bonding *bond = bond_dev->priv; + + info->bond_mode = bond->params.mode; + info->miimon = bond->params.miimon; + + read_lock_bh(&bond->lock); + info->num_slaves = bond->slave_cnt; + read_unlock_bh(&bond->lock); + + return 0; +} + +static int bond_slave_info_query(struct net_device *bond_dev, struct ifslave *info) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave; + int i, found = 0; + + if (info->slave_id < 0) { + return -ENODEV; + } + + read_lock_bh(&bond->lock); + + bond_for_each_slave(bond, slave, i) { + if (i == (int)info->slave_id) { + found = 1; + break; + } + } + + read_unlock_bh(&bond->lock); + + if (found) { + strcpy(info->slave_name, slave->dev->name); + info->link = slave->link; + info->state = slave->state; + info->link_failure_count = slave->link_failure_count; + } else { + return -ENODEV; + } + + return 0; +} + +/*-------------------------------- Monitoring -------------------------------*/ + +/* this function is called regularly to monitor each slave's link. */ +static void bond_mii_monitor(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave, *oldcurrent; + int do_failover = 0; + int delta_in_ticks; + int i; + + read_lock(&bond->lock); + + delta_in_ticks = (bond->params.miimon * HZ) / 1000; + + if (bond->kill_timers) { + goto out; + } + + if (bond->slave_cnt == 0) { + goto re_arm; + } + + /* we will try to read the link status of each of our slaves, and + * set their IFF_RUNNING flag appropriately. For each slave not + * supporting MII status, we won't do anything so that a user-space + * program could monitor the link itself if needed. + */ + + read_lock(&bond->curr_slave_lock); + oldcurrent = bond->curr_active_slave; + read_unlock(&bond->curr_slave_lock); + + bond_for_each_slave(bond, slave, i) { + struct net_device *slave_dev = slave->dev; + int link_state; + u16 old_speed = slave->speed; + u8 old_duplex = slave->duplex; + + link_state = bond_check_dev_link(bond, slave_dev, 0); + + switch (slave->link) { + case BOND_LINK_UP: /* the link was up */ + if (link_state == BMSR_LSTATUS) { + /* link stays up, nothing more to do */ + break; + } else { /* link going down */ + slave->link = BOND_LINK_FAIL; + slave->delay = bond->params.downdelay; + + if (slave->link_failure_count < UINT_MAX) { + slave->link_failure_count++; + } + + if (bond->params.downdelay) { + printk(KERN_INFO DRV_NAME + ": %s: link status down for %s " + "interface %s, disabling it in " + "%d ms.\n", + bond_dev->name, + IS_UP(slave_dev) + ? ((bond->params.mode == BOND_MODE_ACTIVEBACKUP) + ? ((slave == oldcurrent) + ? "active " : "backup ") + : "") + : "idle ", + slave_dev->name, + bond->params.downdelay * bond->params.miimon); + } + } + /* no break ! fall through the BOND_LINK_FAIL test to + ensure proper action to be taken + */ + case BOND_LINK_FAIL: /* the link has just gone down */ + if (link_state != BMSR_LSTATUS) { + /* link stays down */ + if (slave->delay <= 0) { + /* link down for too long time */ + slave->link = BOND_LINK_DOWN; + + /* in active/backup mode, we must + * completely disable this interface + */ + if ((bond->params.mode == BOND_MODE_ACTIVEBACKUP) || + (bond->params.mode == BOND_MODE_8023AD)) { + bond_set_slave_inactive_flags(slave); + } + + printk(KERN_INFO DRV_NAME + ": %s: link status definitely " + "down for interface %s, " + "disabling it\n", + bond_dev->name, + slave_dev->name); + + /* notify ad that the link status has changed */ + if (bond->params.mode == BOND_MODE_8023AD) { + bond_3ad_handle_link_change(slave, BOND_LINK_DOWN); + } + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + bond_alb_handle_link_change(bond, slave, BOND_LINK_DOWN); + } + + if (slave == oldcurrent) { + do_failover = 1; + } + } else { + slave->delay--; + } + } else { + /* link up again */ + slave->link = BOND_LINK_UP; + slave->jiffies = jiffies; + printk(KERN_INFO DRV_NAME + ": %s: link status up again after %d " + "ms for interface %s.\n", + bond_dev->name, + (bond->params.downdelay - slave->delay) * bond->params.miimon, + slave_dev->name); + } + break; + case BOND_LINK_DOWN: /* the link was down */ + if (link_state != BMSR_LSTATUS) { + /* the link stays down, nothing more to do */ + break; + } else { /* link going up */ + slave->link = BOND_LINK_BACK; + slave->delay = bond->params.updelay; + + if (bond->params.updelay) { + /* if updelay == 0, no need to + advertise about a 0 ms delay */ + printk(KERN_INFO DRV_NAME + ": %s: link status up for " + "interface %s, enabling it " + "in %d ms.\n", + bond_dev->name, + slave_dev->name, + bond->params.updelay * bond->params.miimon); + } + } + /* no break ! fall through the BOND_LINK_BACK state in + case there's something to do. + */ + case BOND_LINK_BACK: /* the link has just come back */ + if (link_state != BMSR_LSTATUS) { + /* link down again */ + slave->link = BOND_LINK_DOWN; + + printk(KERN_INFO DRV_NAME + ": %s: link status down again after %d " + "ms for interface %s.\n", + bond_dev->name, + (bond->params.updelay - slave->delay) * bond->params.miimon, + slave_dev->name); + } else { + /* link stays up */ + if (slave->delay == 0) { + /* now the link has been up for long time enough */ + slave->link = BOND_LINK_UP; + slave->jiffies = jiffies; + + if (bond->params.mode == BOND_MODE_8023AD) { + /* prevent it from being the active one */ + slave->state = BOND_STATE_BACKUP; + } else if (bond->params.mode != BOND_MODE_ACTIVEBACKUP) { + /* make it immediately active */ + slave->state = BOND_STATE_ACTIVE; + } else if (slave != bond->primary_slave) { + /* prevent it from being the active one */ + slave->state = BOND_STATE_BACKUP; + } + + printk(KERN_INFO DRV_NAME + ": %s: link status definitely " + "up for interface %s.\n", + bond_dev->name, + slave_dev->name); + + /* notify ad that the link status has changed */ + if (bond->params.mode == BOND_MODE_8023AD) { + bond_3ad_handle_link_change(slave, BOND_LINK_UP); + } + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + bond_alb_handle_link_change(bond, slave, BOND_LINK_UP); + } + + if ((!oldcurrent) || + (slave == bond->primary_slave)) { + do_failover = 1; + } + } else { + slave->delay--; + } + } + break; + default: + /* Should not happen */ + printk(KERN_ERR "bonding: Error: %s Illegal value (link=%d)\n", + slave->dev->name, slave->link); + goto out; + } /* end of switch (slave->link) */ + + bond_update_speed_duplex(slave); + + if (bond->params.mode == BOND_MODE_8023AD) { + if (old_speed != slave->speed) { + bond_3ad_adapter_speed_changed(slave); + } + + if (old_duplex != slave->duplex) { + bond_3ad_adapter_duplex_changed(slave); + } + } + + } /* end of for */ + + if (do_failover) { + write_lock(&bond->curr_slave_lock); + + bond_select_active_slave(bond); + + if (oldcurrent && !bond->curr_active_slave) { + printk(KERN_INFO DRV_NAME + ": %s: now running without any active " + "interface !\n", + bond_dev->name); + } + + write_unlock(&bond->curr_slave_lock); + } + +re_arm: + if (bond->params.miimon) { + mod_timer(&bond->mii_timer, jiffies + delta_in_ticks); + } +out: + read_unlock(&bond->lock); +} + +static void bond_arp_send_all(struct bonding *bond, struct slave *slave) +{ + int i; + u32 *targets = bond->params.arp_targets; + + for (i = 0; (i < BOND_MAX_ARP_TARGETS) && targets[i]; i++) { + arp_send(ARPOP_REQUEST, ETH_P_ARP, targets[i], slave->dev, + my_ip, NULL, slave->dev->dev_addr, + NULL); + } +} + +/* + * this function is called regularly to monitor each slave's link + * ensuring that traffic is being sent and received when arp monitoring + * is used in load-balancing mode. if the adapter has been dormant, then an + * arp is transmitted to generate traffic. see activebackup_arp_monitor for + * arp monitoring in active backup mode. + */ +static void bond_loadbalance_arp_mon(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave, *oldcurrent; + int do_failover = 0; + int delta_in_ticks; + int i; + + read_lock(&bond->lock); + + delta_in_ticks = (bond->params.arp_interval * HZ) / 1000; + + if (bond->kill_timers) { + goto out; + } + + if (bond->slave_cnt == 0) { + goto re_arm; + } + + read_lock(&bond->curr_slave_lock); + oldcurrent = bond->curr_active_slave; + read_unlock(&bond->curr_slave_lock); + + /* see if any of the previous devices are up now (i.e. they have + * xmt and rcv traffic). the curr_active_slave does not come into + * the picture unless it is null. also, slave->jiffies is not needed + * here because we send an arp on each slave and give a slave as + * long as it needs to get the tx/rx within the delta. + * TODO: what about up/down delay in arp mode? it wasn't here before + * so it can wait + */ + bond_for_each_slave(bond, slave, i) { + if (slave->link != BOND_LINK_UP) { + if (((jiffies - slave->dev->trans_start) <= delta_in_ticks) && + ((jiffies - slave->dev->last_rx) <= delta_in_ticks)) { + + slave->link = BOND_LINK_UP; + slave->state = BOND_STATE_ACTIVE; + + /* primary_slave has no meaning in round-robin + * mode. the window of a slave being up and + * curr_active_slave being null after enslaving + * is closed. + */ + if (!oldcurrent) { + printk(KERN_INFO DRV_NAME + ": %s: link status definitely " + "up for interface %s, ", + bond_dev->name, + slave->dev->name); + do_failover = 1; + } else { + printk(KERN_INFO DRV_NAME + ": %s: interface %s is now up\n", + bond_dev->name, + slave->dev->name); + } + } + } else { + /* slave->link == BOND_LINK_UP */ + + /* not all switches will respond to an arp request + * when the source ip is 0, so don't take the link down + * if we don't know our ip yet + */ + if (((jiffies - slave->dev->trans_start) >= (2*delta_in_ticks)) || + (((jiffies - slave->dev->last_rx) >= (2*delta_in_ticks)) && + my_ip)) { + + slave->link = BOND_LINK_DOWN; + slave->state = BOND_STATE_BACKUP; + + if (slave->link_failure_count < UINT_MAX) { + slave->link_failure_count++; + } + + printk(KERN_INFO DRV_NAME + ": %s: interface %s is now down.\n", + bond_dev->name, + slave->dev->name); + + if (slave == oldcurrent) { + do_failover = 1; + } + } + } + + /* note: if switch is in round-robin mode, all links + * must tx arp to ensure all links rx an arp - otherwise + * links may oscillate or not come up at all; if switch is + * in something like xor mode, there is nothing we can + * do - all replies will be rx'ed on same link causing slaves + * to be unstable during low/no traffic periods + */ + if (IS_UP(slave->dev)) { + bond_arp_send_all(bond, slave); + } + } + + if (do_failover) { + write_lock(&bond->curr_slave_lock); + + bond_select_active_slave(bond); + + if (oldcurrent && !bond->curr_active_slave) { + printk(KERN_INFO DRV_NAME + ": %s: now running without any active " + "interface !\n", + bond_dev->name); + } + + write_unlock(&bond->curr_slave_lock); + } + +re_arm: + if (bond->params.arp_interval) { + mod_timer(&bond->arp_timer, jiffies + delta_in_ticks); + } +out: + read_unlock(&bond->lock); +} + +/* + * When using arp monitoring in active-backup mode, this function is + * called to determine if any backup slaves have went down or a new + * current slave needs to be found. + * The backup slaves never generate traffic, they are considered up by merely + * receiving traffic. If the current slave goes down, each backup slave will + * be given the opportunity to tx/rx an arp before being taken down - this + * prevents all slaves from being taken down due to the current slave not + * sending any traffic for the backups to receive. The arps are not necessarily + * necessary, any tx and rx traffic will keep the current slave up. While any + * rx traffic will keep the backup slaves up, the current slave is responsible + * for generating traffic to keep them up regardless of any other traffic they + * may have received. + * see loadbalance_arp_monitor for arp monitoring in load balancing mode + */ +static void bond_activebackup_arp_mon(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave; + int delta_in_ticks; + int i; + + read_lock(&bond->lock); + + delta_in_ticks = (bond->params.arp_interval * HZ) / 1000; + + if (bond->kill_timers) { + goto out; + } + + if (bond->slave_cnt == 0) { + goto re_arm; + } + + /* determine if any slave has come up or any backup slave has + * gone down + * TODO: what about up/down delay in arp mode? it wasn't here before + * so it can wait + */ + bond_for_each_slave(bond, slave, i) { + if (slave->link != BOND_LINK_UP) { + if ((jiffies - slave->dev->last_rx) <= delta_in_ticks) { + + slave->link = BOND_LINK_UP; + + write_lock(&bond->curr_slave_lock); + + if ((!bond->curr_active_slave) && + ((jiffies - slave->dev->trans_start) <= delta_in_ticks)) { + bond_change_active_slave(bond, slave); + bond->current_arp_slave = NULL; + } else if (bond->curr_active_slave != slave) { + /* this slave has just come up but we + * already have a current slave; this + * can also happen if bond_enslave adds + * a new slave that is up while we are + * searching for a new slave + */ + bond_set_slave_inactive_flags(slave); + bond->current_arp_slave = NULL; + } + + if (slave == bond->curr_active_slave) { + printk(KERN_INFO DRV_NAME + ": %s: %s is up and now the " + "active interface\n", + bond_dev->name, + slave->dev->name); + } else { + printk(KERN_INFO DRV_NAME + ": %s: backup interface %s is " + "now up\n", + bond_dev->name, + slave->dev->name); + } + + write_unlock(&bond->curr_slave_lock); + } + } else { + read_lock(&bond->curr_slave_lock); + + if ((slave != bond->curr_active_slave) && + (!bond->current_arp_slave) && + (((jiffies - slave->dev->last_rx) >= 3*delta_in_ticks) && + my_ip)) { + /* a backup slave has gone down; three times + * the delta allows the current slave to be + * taken out before the backup slave. + * note: a non-null current_arp_slave indicates + * the curr_active_slave went down and we are + * searching for a new one; under this + * condition we only take the curr_active_slave + * down - this gives each slave a chance to + * tx/rx traffic before being taken out + */ + + read_unlock(&bond->curr_slave_lock); + + slave->link = BOND_LINK_DOWN; + + if (slave->link_failure_count < UINT_MAX) { + slave->link_failure_count++; + } + + bond_set_slave_inactive_flags(slave); + + printk(KERN_INFO DRV_NAME + ": %s: backup interface %s is now down\n", + bond_dev->name, + slave->dev->name); + } else { + read_unlock(&bond->curr_slave_lock); + } + } + } + + read_lock(&bond->curr_slave_lock); + slave = bond->curr_active_slave; + read_unlock(&bond->curr_slave_lock); + + if (slave) { + /* if we have sent traffic in the past 2*arp_intervals but + * haven't xmit and rx traffic in that time interval, select + * a different slave. slave->jiffies is only updated when + * a slave first becomes the curr_active_slave - not necessarily + * after every arp; this ensures the slave has a full 2*delta + * before being taken out. if a primary is being used, check + * if it is up and needs to take over as the curr_active_slave + */ + if ((((jiffies - slave->dev->trans_start) >= (2*delta_in_ticks)) || + (((jiffies - slave->dev->last_rx) >= (2*delta_in_ticks)) && + my_ip)) && + ((jiffies - slave->jiffies) >= 2*delta_in_ticks)) { + + slave->link = BOND_LINK_DOWN; + + if (slave->link_failure_count < UINT_MAX) { + slave->link_failure_count++; + } + + printk(KERN_INFO DRV_NAME + ": %s: link status down for active interface " + "%s, disabling it\n", + bond_dev->name, + slave->dev->name); + + write_lock(&bond->curr_slave_lock); + + bond_select_active_slave(bond); + slave = bond->curr_active_slave; + + write_unlock(&bond->curr_slave_lock); + + bond->current_arp_slave = slave; + + if (slave) { + slave->jiffies = jiffies; + } + } else if ((bond->primary_slave) && + (bond->primary_slave != slave) && + (bond->primary_slave->link == BOND_LINK_UP)) { + /* at this point, slave is the curr_active_slave */ + printk(KERN_INFO DRV_NAME + ": %s: changing from interface %s to primary " + "interface %s\n", + bond_dev->name, + slave->dev->name, + bond->primary_slave->dev->name); + + /* primary is up so switch to it */ + write_lock(&bond->curr_slave_lock); + bond_change_active_slave(bond, bond->primary_slave); + write_unlock(&bond->curr_slave_lock); + + slave = bond->primary_slave; + slave->jiffies = jiffies; + } else { + bond->current_arp_slave = NULL; + } + + /* the current slave must tx an arp to ensure backup slaves + * rx traffic + */ + if (slave && my_ip) { + bond_arp_send_all(bond, slave); + } + } + + /* if we don't have a curr_active_slave, search for the next available + * backup slave from the current_arp_slave and make it the candidate + * for becoming the curr_active_slave + */ + if (!slave) { + if (!bond->current_arp_slave) { + bond->current_arp_slave = bond->first_slave; + } + + if (bond->current_arp_slave) { + bond_set_slave_inactive_flags(bond->current_arp_slave); + + /* search for next candidate */ + bond_for_each_slave_from(bond, slave, i, bond->current_arp_slave) { + if (IS_UP(slave->dev)) { + slave->link = BOND_LINK_BACK; + bond_set_slave_active_flags(slave); + bond_arp_send_all(bond, slave); + slave->jiffies = jiffies; + bond->current_arp_slave = slave; + break; + } + + /* if the link state is up at this point, we + * mark it down - this can happen if we have + * simultaneous link failures and + * reselect_active_interface doesn't make this + * one the current slave so it is still marked + * up when it is actually down + */ + if (slave->link == BOND_LINK_UP) { + slave->link = BOND_LINK_DOWN; + if (slave->link_failure_count < UINT_MAX) { + slave->link_failure_count++; + } + + bond_set_slave_inactive_flags(slave); + + printk(KERN_INFO DRV_NAME + ": %s: backup interface %s is " + "now down.\n", + bond_dev->name, + slave->dev->name); + } + } + } + } + +re_arm: + if (bond->params.arp_interval) { + mod_timer(&bond->arp_timer, jiffies + delta_in_ticks); + } +out: + read_unlock(&bond->lock); +} + +/*------------------------------ proc/seq_file-------------------------------*/ + +#ifdef CONFIG_PROC_FS + +#define SEQ_START_TOKEN ((void *)1) + +static void *bond_info_seq_start(struct seq_file *seq, loff_t *pos) +{ + struct bonding *bond = seq->private; + loff_t off = 0; + struct slave *slave; + int i; + + /* make sure the bond won't be taken away */ + read_lock(&dev_base_lock); + read_lock_bh(&bond->lock); + + if (*pos == 0) { + return SEQ_START_TOKEN; + } + + bond_for_each_slave(bond, slave, i) { + if (++off == *pos) { + return slave; + } + } + + return NULL; +} + +static void *bond_info_seq_next(struct seq_file *seq, void *v, loff_t *pos) +{ + struct bonding *bond = seq->private; + struct slave *slave = v; + + ++*pos; + if (v == SEQ_START_TOKEN) { + return bond->first_slave; + } + + slave = slave->next; + + return (slave == bond->first_slave) ? NULL : slave; +} + +static void bond_info_seq_stop(struct seq_file *seq, void *v) +{ + struct bonding *bond = seq->private; + + read_unlock_bh(&bond->lock); + read_unlock(&dev_base_lock); +} + +static void bond_info_show_master(struct seq_file *seq) +{ + struct bonding *bond = seq->private; + struct slave *curr; + + read_lock(&bond->curr_slave_lock); + curr = bond->curr_active_slave; + read_unlock(&bond->curr_slave_lock); + + seq_printf(seq, "Bonding Mode: %s\n", + bond_mode_name(bond->params.mode)); + + if (USES_PRIMARY(bond->params.mode)) { + seq_printf(seq, "Primary Slave: %s\n", + (bond->params.primary[0]) ? + bond->params.primary : "None"); + + seq_printf(seq, "Currently Active Slave: %s\n", + (curr) ? curr->dev->name : "None"); + } + + seq_printf(seq, "MII Status: %s\n", (curr) ? "up" : "down"); + seq_printf(seq, "MII Polling Interval (ms): %d\n", bond->params.miimon); + seq_printf(seq, "Up Delay (ms): %d\n", + bond->params.updelay * bond->params.miimon); + seq_printf(seq, "Down Delay (ms): %d\n", + bond->params.downdelay * bond->params.miimon); + + if (bond->params.mode == BOND_MODE_8023AD) { + struct ad_info ad_info; + + seq_puts(seq, "\n802.3ad info\n"); + seq_printf(seq, "LACP rate: %s\n", + (bond->params.lacp_fast) ? "fast" : "slow"); + + if (bond_3ad_get_active_agg_info(bond, &ad_info)) { + seq_printf(seq, "bond %s has no active aggregator\n", + bond->dev->name); + } else { + seq_printf(seq, "Active Aggregator Info:\n"); + + seq_printf(seq, "\tAggregator ID: %d\n", + ad_info.aggregator_id); + seq_printf(seq, "\tNumber of ports: %d\n", + ad_info.ports); + seq_printf(seq, "\tActor Key: %d\n", + ad_info.actor_key); + seq_printf(seq, "\tPartner Key: %d\n", + ad_info.partner_key); + seq_printf(seq, "\tPartner Mac Address: %02x:%02x:%02x:%02x:%02x:%02x\n", + ad_info.partner_system[0], + ad_info.partner_system[1], + ad_info.partner_system[2], + ad_info.partner_system[3], + ad_info.partner_system[4], + ad_info.partner_system[5]); + } + } +} + +static void bond_info_show_slave(struct seq_file *seq, const struct slave *slave) +{ + struct bonding *bond = seq->private; + + seq_printf(seq, "\nSlave Interface: %s\n", slave->dev->name); + seq_printf(seq, "MII Status: %s\n", + (slave->link == BOND_LINK_UP) ? "up" : "down"); + seq_printf(seq, "Link Failure Count: %d\n", + slave->link_failure_count); + + if (app_abi_ver >= 1) { + seq_printf(seq, + "Permanent HW addr: %02x:%02x:%02x:%02x:%02x:%02x\n", + slave->perm_hwaddr[0], + slave->perm_hwaddr[1], + slave->perm_hwaddr[2], + slave->perm_hwaddr[3], + slave->perm_hwaddr[4], + slave->perm_hwaddr[5]); + } + + if (bond->params.mode == BOND_MODE_8023AD) { + const struct aggregator *agg + = SLAVE_AD_INFO(slave).port.aggregator; + + if (agg) { + seq_printf(seq, "Aggregator ID: %d\n", + agg->aggregator_identifier); + } else { + seq_puts(seq, "Aggregator ID: N/A\n"); + } + } +} + +static int bond_info_seq_show(struct seq_file *seq, void *v) +{ + if (v == SEQ_START_TOKEN) { + seq_printf(seq, "%s\n", version); + bond_info_show_master(seq); + } else { + bond_info_show_slave(seq, v); + } + + return 0; +} + +static struct seq_operations bond_info_seq_ops = { + .start = bond_info_seq_start, + .next = bond_info_seq_next, + .stop = bond_info_seq_stop, + .show = bond_info_seq_show, +}; + +static int bond_info_open(struct inode *inode, struct file *file) +{ + struct seq_file *seq; + struct proc_dir_entry *proc; + int res; + + res = seq_open(file, &bond_info_seq_ops); + if (!res) { + /* recover the pointer buried in proc_dir_entry data */ + seq = file->private_data; + proc = PDE(inode); + seq->private = proc->data; + } + + return res; +} + +static struct file_operations bond_info_fops = { + .owner = THIS_MODULE, + .open = bond_info_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int bond_create_proc_entry(struct bonding *bond) +{ + struct net_device *bond_dev = bond->dev; + + if (bond_proc_dir) { + bond->proc_entry = create_proc_entry(bond_dev->name, + S_IRUGO, + bond_proc_dir); + if (bond->proc_entry == NULL) { + printk(KERN_WARNING DRV_NAME + ": Warning: Cannot create /proc/net/%s/%s\n", + DRV_NAME, bond_dev->name); + } else { + bond->proc_entry->data = bond; + bond->proc_entry->proc_fops = &bond_info_fops; + bond->proc_entry->owner = THIS_MODULE; + memcpy(bond->proc_file_name, bond_dev->name, IFNAMSIZ); + } + } + + return 0; +} + +static void bond_remove_proc_entry(struct bonding *bond) +{ + if (bond_proc_dir && bond->proc_entry) { + remove_proc_entry(bond->proc_file_name, bond_proc_dir); + memset(bond->proc_file_name, 0, IFNAMSIZ); + bond->proc_entry = NULL; + } +} + +/* Create the bonding directory under /proc/net, if doesn't exist yet. + * Caller must hold rtnl_lock. + */ +static void bond_create_proc_dir(void) +{ + int len = strlen(DRV_NAME); + + for (bond_proc_dir = proc_net->subdir; bond_proc_dir; + bond_proc_dir = bond_proc_dir->next) { + if ((bond_proc_dir->namelen == len) && + !memcmp(bond_proc_dir->name, DRV_NAME, len)) { + break; + } + } + + if (!bond_proc_dir) { + bond_proc_dir = proc_mkdir(DRV_NAME, proc_net); + if (bond_proc_dir) { + bond_proc_dir->owner = THIS_MODULE; + } else { + printk(KERN_WARNING DRV_NAME + ": Warning: cannot create /proc/net/%s\n", + DRV_NAME); + } + } +} + +/* Destroy the bonding directory under /proc/net, if empty. + * Caller must hold rtnl_lock. + */ +static void bond_destroy_proc_dir(void) +{ + struct proc_dir_entry *de; + + if (!bond_proc_dir) { + return; + } + + /* verify that the /proc dir is empty */ + for (de = bond_proc_dir->subdir; de; de = de->next) { + /* ignore . and .. */ + if (*(de->name) != '.') { + break; + } + } + + if (de) { + if (bond_proc_dir->owner == THIS_MODULE) { + bond_proc_dir->owner = NULL; + } + } else { + remove_proc_entry(DRV_NAME, proc_net); + bond_proc_dir = NULL; + } +} +#endif /* CONFIG_PROC_FS */ + +/*-------------------------- netdev event handling --------------------------*/ + +/* + * Change device name + */ +static int bond_event_changename(struct bonding *bond) +{ +#ifdef CONFIG_PROC_FS + bond_remove_proc_entry(bond); + bond_create_proc_entry(bond); +#endif + + return NOTIFY_DONE; +} + +static int bond_master_netdev_event(unsigned long event, struct net_device *bond_dev) +{ + struct bonding *event_bond = bond_dev->priv; + + switch (event) { + case NETDEV_CHANGENAME: + return bond_event_changename(event_bond); + case NETDEV_UNREGISTER: + /* + * TODO: remove a bond from the list? + */ + break; + default: + break; + } + + return NOTIFY_DONE; +} + +static int bond_slave_netdev_event(unsigned long event, struct net_device *slave_dev) +{ + struct net_device *bond_dev = slave_dev->master; + + switch (event) { + case NETDEV_UNREGISTER: + if (bond_dev) { + bond_release(bond_dev, slave_dev); + } + break; + case NETDEV_CHANGE: + /* + * TODO: is this what we get if somebody + * sets up a hierarchical bond, then rmmod's + * one of the slave bonding devices? + */ + break; + case NETDEV_DOWN: + /* + * ... Or is it this? + */ + break; + case NETDEV_CHANGEMTU: + /* + * TODO: Should slaves be allowed to + * independently alter their MTU? For + * an active-backup bond, slaves need + * not be the same type of device, so + * MTUs may vary. For other modes, + * slaves arguably should have the + * same MTUs. To do this, we'd need to + * take over the slave's change_mtu + * function for the duration of their + * servitude. + */ + break; + case NETDEV_CHANGENAME: + /* + * TODO: handle changing the primary's name + */ + break; + default: + break; + } + + return NOTIFY_DONE; +} + +/* + * bond_netdev_event: handle netdev notifier chain events. + * + * This function receives events for the netdev chain. The caller (an + * ioctl handler calling notifier_call_chain) holds the necessary + * locks for us to safely manipulate the slave devices (RTNL lock, + * dev_probe_lock). + */ +static int bond_netdev_event(struct notifier_block *this, unsigned long event, void *ptr) +{ + struct net_device *event_dev = (struct net_device *)ptr; + + dprintk("event_dev: %s, event: %lx\n", + (event_dev ? event_dev->name : "None"), + event); + + if (event_dev->flags & IFF_MASTER) { + dprintk("IFF_MASTER\n"); + return bond_master_netdev_event(event, event_dev); + } + + if (event_dev->flags & IFF_SLAVE) { + dprintk("IFF_SLAVE\n"); + return bond_slave_netdev_event(event, event_dev); + } + + return NOTIFY_DONE; +} + +static struct notifier_block bond_netdev_notifier = { + .notifier_call = bond_netdev_event, +}; + +/*-------------------------- Packet type handling ---------------------------*/ + +/* register to receive lacpdus on a bond */ +static void bond_register_lacpdu(struct bonding *bond) +{ + struct packet_type *pk_type = &(BOND_AD_INFO(bond).ad_pkt_type); + + /* initialize packet type */ + pk_type->type = PKT_TYPE_LACPDU; + pk_type->dev = bond->dev; + pk_type->func = bond_3ad_lacpdu_recv; + + dev_add_pack(pk_type); +} + +/* unregister to receive lacpdus on a bond */ +static void bond_unregister_lacpdu(struct bonding *bond) +{ + dev_remove_pack(&(BOND_AD_INFO(bond).ad_pkt_type)); +} + +/*-------------------------- Device entry points ----------------------------*/ + +static int bond_open(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct timer_list *mii_timer = &bond->mii_timer; + struct timer_list *arp_timer = &bond->arp_timer; + + bond->kill_timers = 0; + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + struct timer_list *alb_timer = &(BOND_ALB_INFO(bond).alb_timer); + + /* bond_alb_initialize must be called before the timer + * is started. + */ + if (bond_alb_initialize(bond, (bond->params.mode == BOND_MODE_ALB))) { + /* something went wrong - fail the open operation */ + return -1; + } + + init_timer(alb_timer); + alb_timer->expires = jiffies + 1; + alb_timer->data = (unsigned long)bond; + alb_timer->function = (void *)&bond_alb_monitor; + add_timer(alb_timer); + } + + if (bond->params.miimon) { /* link check interval, in milliseconds. */ + init_timer(mii_timer); + mii_timer->expires = jiffies + 1; + mii_timer->data = (unsigned long)bond_dev; + mii_timer->function = (void *)&bond_mii_monitor; + add_timer(mii_timer); + } + + if (bond->params.arp_interval) { /* arp interval, in milliseconds. */ + init_timer(arp_timer); + arp_timer->expires = jiffies + 1; + arp_timer->data = (unsigned long)bond_dev; + if (bond->params.mode == BOND_MODE_ACTIVEBACKUP) { + arp_timer->function = (void *)&bond_activebackup_arp_mon; + } else { + arp_timer->function = (void *)&bond_loadbalance_arp_mon; + } + add_timer(arp_timer); + } + + if (bond->params.mode == BOND_MODE_8023AD) { + struct timer_list *ad_timer = &(BOND_AD_INFO(bond).ad_timer); + init_timer(ad_timer); + ad_timer->expires = jiffies + 1; + ad_timer->data = (unsigned long)bond; + ad_timer->function = (void *)&bond_3ad_state_machine_handler; + add_timer(ad_timer); + + /* register to receive LACPDUs */ + bond_register_lacpdu(bond); + } + + return 0; +} + +static int bond_close(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + + if (bond->params.mode == BOND_MODE_8023AD) { + /* Unregister the receive of LACPDUs */ + bond_unregister_lacpdu(bond); + } + + write_lock_bh(&bond->lock); + + bond_mc_list_destroy(bond); + + /* signal timers not to re-arm */ + bond->kill_timers = 1; + + write_unlock_bh(&bond->lock); + + /* del_timer_sync must run without holding the bond->lock + * because a running timer might be trying to hold it too + */ + + if (bond->params.miimon) { /* link check interval, in milliseconds. */ + del_timer_sync(&bond->mii_timer); + } + + if (bond->params.arp_interval) { /* arp interval, in milliseconds. */ + del_timer_sync(&bond->arp_timer); + } + + switch (bond->params.mode) { + case BOND_MODE_8023AD: + del_timer_sync(&(BOND_AD_INFO(bond).ad_timer)); + break; + case BOND_MODE_TLB: + case BOND_MODE_ALB: + del_timer_sync(&(BOND_ALB_INFO(bond).alb_timer)); + break; + default: + break; + } + + /* Release the bonded slaves */ + bond_release_all(bond_dev); + + if ((bond->params.mode == BOND_MODE_TLB) || + (bond->params.mode == BOND_MODE_ALB)) { + /* Must be called only after all + * slaves have been released + */ + bond_alb_deinitialize(bond); + } + + return 0; +} + +static struct net_device_stats *bond_get_stats(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct net_device_stats *stats = &(bond->stats), *sstats; + struct slave *slave; + int i; + + memset(stats, 0, sizeof(struct net_device_stats)); + + read_lock_bh(&bond->lock); + + bond_for_each_slave(bond, slave, i) { + sstats = slave->dev->get_stats(slave->dev); + + stats->rx_packets += sstats->rx_packets; + stats->rx_bytes += sstats->rx_bytes; + stats->rx_errors += sstats->rx_errors; + stats->rx_dropped += sstats->rx_dropped; + + stats->tx_packets += sstats->tx_packets; + stats->tx_bytes += sstats->tx_bytes; + stats->tx_errors += sstats->tx_errors; + stats->tx_dropped += sstats->tx_dropped; + + stats->multicast += sstats->multicast; + stats->collisions += sstats->collisions; + + stats->rx_length_errors += sstats->rx_length_errors; + stats->rx_over_errors += sstats->rx_over_errors; + stats->rx_crc_errors += sstats->rx_crc_errors; + stats->rx_frame_errors += sstats->rx_frame_errors; + stats->rx_fifo_errors += sstats->rx_fifo_errors; + stats->rx_missed_errors += sstats->rx_missed_errors; + + stats->tx_aborted_errors += sstats->tx_aborted_errors; + stats->tx_carrier_errors += sstats->tx_carrier_errors; + stats->tx_fifo_errors += sstats->tx_fifo_errors; + stats->tx_heartbeat_errors += sstats->tx_heartbeat_errors; + stats->tx_window_errors += sstats->tx_window_errors; + } + + read_unlock_bh(&bond->lock); + + return stats; +} + +static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd) +{ + struct net_device *slave_dev = NULL; + struct ifbond k_binfo; + struct ifbond __user *u_binfo = NULL; + struct ifslave k_sinfo; + struct ifslave __user *u_sinfo = NULL; + struct mii_ioctl_data *mii = NULL; + int prev_abi_ver = orig_app_abi_ver; + int res = 0; + + dprintk("bond_ioctl: master=%s, cmd=%d\n", + bond_dev->name, cmd); + + switch (cmd) { + case SIOCETHTOOL: + return bond_ethtool_ioctl(bond_dev, ifr); + case SIOCGMIIPHY: + mii = if_mii(ifr); + if (!mii) { + return -EINVAL; + } + mii->phy_id = 0; + /* Fall Through */ + case SIOCGMIIREG: + /* + * We do this again just in case we were called by SIOCGMIIREG + * instead of SIOCGMIIPHY. + */ + mii = if_mii(ifr); + if (!mii) { + return -EINVAL; + } + + if (mii->reg_num == 1) { + struct bonding *bond = bond_dev->priv; + mii->val_out = 0; + read_lock_bh(&bond->lock); + read_lock(&bond->curr_slave_lock); + if (bond->curr_active_slave) { + mii->val_out = BMSR_LSTATUS; + } + read_unlock(&bond->curr_slave_lock); + read_unlock_bh(&bond->lock); + } + + return 0; + case BOND_INFO_QUERY_OLD: + case SIOCBONDINFOQUERY: + u_binfo = (struct ifbond __user *)ifr->ifr_data; + + if (copy_from_user(&k_binfo, u_binfo, sizeof(ifbond))) { + return -EFAULT; + } + + res = bond_info_query(bond_dev, &k_binfo); + if (res == 0) { + if (copy_to_user(u_binfo, &k_binfo, sizeof(ifbond))) { + return -EFAULT; + } + } + + return res; + case BOND_SLAVE_INFO_QUERY_OLD: + case SIOCBONDSLAVEINFOQUERY: + u_sinfo = (struct ifslave __user *)ifr->ifr_data; + + if (copy_from_user(&k_sinfo, u_sinfo, sizeof(ifslave))) { + return -EFAULT; + } + + res = bond_slave_info_query(bond_dev, &k_sinfo); + if (res == 0) { + if (copy_to_user(u_sinfo, &k_sinfo, sizeof(ifslave))) { + return -EFAULT; + } + } + + return res; + default: + /* Go on */ + break; + } + + if (!capable(CAP_NET_ADMIN)) { + return -EPERM; + } + + if (orig_app_abi_ver == -1) { + /* no orig_app_abi_ver was provided yet, so we'll use the + * current one from now on, even if it's 0 + */ + orig_app_abi_ver = app_abi_ver; + + } else if (orig_app_abi_ver != app_abi_ver) { + printk(KERN_ERR DRV_NAME + ": Error: already using ifenslave ABI version %d; to " + "upgrade ifenslave to version %d, you must first " + "reload bonding.\n", + orig_app_abi_ver, app_abi_ver); + return -EINVAL; + } + + slave_dev = dev_get_by_name(ifr->ifr_slave); + + dprintk("slave_dev=%p: \n", slave_dev); + + if (!slave_dev) { + res = -ENODEV; + } else { + dprintk("slave_dev->name=%s: \n", slave_dev->name); + switch (cmd) { + case BOND_ENSLAVE_OLD: + case SIOCBONDENSLAVE: + res = bond_enslave(bond_dev, slave_dev); + break; + case BOND_RELEASE_OLD: + case SIOCBONDRELEASE: + res = bond_release(bond_dev, slave_dev); + break; + case BOND_SETHWADDR_OLD: + case SIOCBONDSETHWADDR: + res = bond_sethwaddr(bond_dev, slave_dev); + break; + case BOND_CHANGE_ACTIVE_OLD: + case SIOCBONDCHANGEACTIVE: + res = bond_ioctl_change_active(bond_dev, slave_dev); + break; + default: + res = -EOPNOTSUPP; + } + + dev_put(slave_dev); + } + + if (res < 0) { + /* The ioctl failed, so there's no point in changing the + * orig_app_abi_ver. We'll restore it's value just in case + * we've changed it earlier in this function. + */ + orig_app_abi_ver = prev_abi_ver; + } + + return res; +} + +static void bond_set_multicast_list(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct dev_mc_list *dmi; + + write_lock_bh(&bond->lock); + + /* + * Do promisc before checking multicast_mode + */ + if ((bond_dev->flags & IFF_PROMISC) && !(bond->flags & IFF_PROMISC)) { + bond_set_promiscuity(bond, 1); + } + + if (!(bond_dev->flags & IFF_PROMISC) && (bond->flags & IFF_PROMISC)) { + bond_set_promiscuity(bond, -1); + } + + /* set allmulti flag to slaves */ + if ((bond_dev->flags & IFF_ALLMULTI) && !(bond->flags & IFF_ALLMULTI)) { + bond_set_allmulti(bond, 1); + } + + if (!(bond_dev->flags & IFF_ALLMULTI) && (bond->flags & IFF_ALLMULTI)) { + bond_set_allmulti(bond, -1); + } + + bond->flags = bond_dev->flags; + + /* looking for addresses to add to slaves' mc list */ + for (dmi = bond_dev->mc_list; dmi; dmi = dmi->next) { + if (!bond_mc_list_find_dmi(dmi, bond->mc_list)) { + bond_mc_add(bond, dmi->dmi_addr, dmi->dmi_addrlen); + } + } + + /* looking for addresses to delete from slaves' list */ + for (dmi = bond->mc_list; dmi; dmi = dmi->next) { + if (!bond_mc_list_find_dmi(dmi, bond_dev->mc_list)) { + bond_mc_delete(bond, dmi->dmi_addr, dmi->dmi_addrlen); + } + } + + /* save master's multicast list */ + bond_mc_list_destroy(bond); + bond_mc_list_copy(bond_dev->mc_list, bond, GFP_ATOMIC); + + write_unlock_bh(&bond->lock); +} + +/* + * Change the MTU of all of a master's slaves to match the master + */ +static int bond_change_mtu(struct net_device *bond_dev, int new_mtu) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave, *stop_at; + int res = 0; + int i; + + dprintk("bond=%p, name=%s, new_mtu=%d\n", bond, + (bond_dev ? bond_dev->name : "None"), new_mtu); + + /* Can't hold bond->lock with bh disabled here since + * some base drivers panic. On the other hand we can't + * hold bond->lock without bh disabled because we'll + * deadlock. The only solution is to rely on the fact + * that we're under rtnl_lock here, and the slaves + * list won't change. This doesn't solve the problem + * of setting the slave's MTU while it is + * transmitting, but the assumption is that the base + * driver can handle that. + * + * TODO: figure out a way to safely iterate the slaves + * list, but without holding a lock around the actual + * call to the base driver. + */ + + bond_for_each_slave(bond, slave, i) { + dprintk("s %p s->p %p c_m %p\n", slave, + slave->prev, slave->dev->change_mtu); + res = dev_set_mtu(slave->dev, new_mtu); + + if (res) { + /* If we failed to set the slave's mtu to the new value + * we must abort the operation even in ACTIVE_BACKUP + * mode, because if we allow the backup slaves to have + * different mtu values than the active slave we'll + * need to change their mtu when doing a failover. That + * means changing their mtu from timer context, which + * is probably not a good idea. + */ + dprintk("err %d %s\n", res, slave->dev->name); + goto unwind; + } + } + + bond_dev->mtu = new_mtu; + + return 0; + +unwind: + /* 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) { + int tmp_res; + + tmp_res = dev_set_mtu(slave->dev, bond_dev->mtu); + if (tmp_res) { + dprintk("unwind err %d dev %s\n", tmp_res, + slave->dev->name); + } + } + + return res; +} + +/* + * Change HW address + * + * Note that many devices must be down to change the HW address, and + * downing the master releases all slaves. We can make bonds full of + * bonding devices to test this, however. + */ +static int bond_set_mac_address(struct net_device *bond_dev, void *addr) +{ + struct bonding *bond = bond_dev->priv; + struct sockaddr *sa = addr, tmp_sa; + struct slave *slave, *stop_at; + int res = 0; + int i; + + dprintk("bond=%p, name=%s\n", bond, (bond_dev ? bond_dev->name : "None")); + + if (!is_valid_ether_addr(sa->sa_data)) { + return -EADDRNOTAVAIL; + } + + /* Can't hold bond->lock with bh disabled here since + * some base drivers panic. On the other hand we can't + * hold bond->lock without bh disabled because we'll + * deadlock. The only solution is to rely on the fact + * that we're under rtnl_lock here, and the slaves + * list won't change. This doesn't solve the problem + * of setting the slave's hw address while it is + * transmitting, but the assumption is that the base + * driver can handle that. + * + * TODO: figure out a way to safely iterate the slaves + * list, but without holding a lock around the actual + * call to the base driver. + */ + + bond_for_each_slave(bond, slave, i) { + dprintk("slave %p %s\n", slave, slave->dev->name); + + if (slave->dev->set_mac_address == NULL) { + res = -EOPNOTSUPP; + dprintk("EOPNOTSUPP %s\n", slave->dev->name); + goto unwind; + } + + res = dev_set_mac_address(slave->dev, addr); + if (res) { + /* TODO: consider downing the slave + * and retry ? + * User should expect communications + * breakage anyway until ARP finish + * updating, so... + */ + dprintk("err %d %s\n", res, slave->dev->name); + goto unwind; + } + } + + /* success */ + memcpy(bond_dev->dev_addr, sa->sa_data, bond_dev->addr_len); + return 0; + +unwind: + memcpy(tmp_sa.sa_data, bond_dev->dev_addr, bond_dev->addr_len); + tmp_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) { + int tmp_res; + + tmp_res = dev_set_mac_address(slave->dev, &tmp_sa); + if (tmp_res) { + dprintk("unwind err %d dev %s\n", tmp_res, + slave->dev->name); + } + } + + return res; +} + +static int bond_xmit_roundrobin(struct sk_buff *skb, struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave, *start_at; + int i; + int res = 1; + + read_lock(&bond->lock); + + if (!BOND_IS_OK(bond)) { + goto out; + } + + read_lock(&bond->curr_slave_lock); + slave = start_at = bond->curr_active_slave; + read_unlock(&bond->curr_slave_lock); + + if (!slave) { + goto out; + } + + bond_for_each_slave_from(bond, slave, i, start_at) { + if (IS_UP(slave->dev) && + (slave->link == BOND_LINK_UP) && + (slave->state == BOND_STATE_ACTIVE)) { + res = bond_dev_queue_xmit(bond, skb, slave->dev); + + write_lock(&bond->curr_slave_lock); + bond->curr_active_slave = slave->next; + write_unlock(&bond->curr_slave_lock); + + break; + } + } + + +out: + if (res) { + /* no suitable interface, frame not sent */ + dev_kfree_skb(skb); + } + read_unlock(&bond->lock); + return 0; +} + +/* + * in active-backup mode, we know that bond->curr_active_slave is always valid if + * the bond has a usable interface. + */ +static int bond_xmit_activebackup(struct sk_buff *skb, struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + int res = 1; + + /* if we are sending arp packets, try to at least + identify our own ip address */ + if (bond->params.arp_interval && !my_ip && + (skb->protocol == __constant_htons(ETH_P_ARP))) { + char *the_ip = (char *)skb->data + + sizeof(struct ethhdr) + + sizeof(struct arphdr) + + ETH_ALEN; + memcpy(&my_ip, the_ip, 4); + } + + read_lock(&bond->lock); + read_lock(&bond->curr_slave_lock); + + if (!BOND_IS_OK(bond)) { + goto out; + } + + if (bond->curr_active_slave) { /* one usable interface */ + res = bond_dev_queue_xmit(bond, skb, bond->curr_active_slave->dev); + } + +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; +} + +/* + * in XOR mode, we determine the output device by performing xor on + * the source and destination hw adresses. If this device is not + * enabled, find the next slave following this xor slave. + */ +static int bond_xmit_xor(struct sk_buff *skb, struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct ethhdr *data = (struct ethhdr *)skb->data; + struct slave *slave, *start_at; + int slave_no; + int i; + int res = 1; + + read_lock(&bond->lock); + + if (!BOND_IS_OK(bond)) { + goto out; + } + + slave_no = (data->h_dest[5]^bond_dev->dev_addr[5]) % bond->slave_cnt; + + bond_for_each_slave(bond, slave, i) { + slave_no--; + if (slave_no < 0) { + break; + } + } + + start_at = slave; + + bond_for_each_slave_from(bond, slave, i, start_at) { + if (IS_UP(slave->dev) && + (slave->link == BOND_LINK_UP) && + (slave->state == BOND_STATE_ACTIVE)) { + res = bond_dev_queue_xmit(bond, skb, slave->dev); + break; + } + } + +out: + if (res) { + /* no suitable interface, frame not sent */ + dev_kfree_skb(skb); + } + read_unlock(&bond->lock); + return 0; +} + +/* + * in broadcast mode, we send everything to all usable interfaces. + */ +static int bond_xmit_broadcast(struct sk_buff *skb, struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + struct slave *slave, *start_at; + struct net_device *tx_dev = NULL; + int i; + int res = 1; + + read_lock(&bond->lock); + + if (!BOND_IS_OK(bond)) { + goto out; + } + + read_lock(&bond->curr_slave_lock); + start_at = bond->curr_active_slave; + read_unlock(&bond->curr_slave_lock); + + if (!start_at) { + goto out; + } + + bond_for_each_slave_from(bond, slave, i, start_at) { + if (IS_UP(slave->dev) && + (slave->link == BOND_LINK_UP) && + (slave->state == BOND_STATE_ACTIVE)) { + if (tx_dev) { + struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC); + if (!skb2) { + printk(KERN_ERR DRV_NAME + ": Error: bond_xmit_broadcast(): " + "skb_clone() failed\n"); + continue; + } + + res = bond_dev_queue_xmit(bond, skb2, tx_dev); + if (res) { + dev_kfree_skb(skb2); + continue; + } + } + tx_dev = slave->dev; + } + } + + if (tx_dev) { + res = bond_dev_queue_xmit(bond, skb, tx_dev); + } + +out: + if (res) { + /* no suitable interface, frame not sent */ + dev_kfree_skb(skb); + } + /* frame sent to all suitable interfaces */ + read_unlock(&bond->lock); + return 0; +} + +/*------------------------- Device initialization ---------------------------*/ + +/* + * set bond mode specific net device operations + */ +static inline void bond_set_mode_ops(struct net_device *bond_dev, int mode) +{ + switch (mode) { + case BOND_MODE_ROUNDROBIN: + bond_dev->hard_start_xmit = bond_xmit_roundrobin; + break; + case BOND_MODE_ACTIVEBACKUP: + bond_dev->hard_start_xmit = bond_xmit_activebackup; + break; + case BOND_MODE_XOR: + bond_dev->hard_start_xmit = bond_xmit_xor; + break; + case BOND_MODE_BROADCAST: + bond_dev->hard_start_xmit = bond_xmit_broadcast; + break; + case BOND_MODE_8023AD: + bond_dev->hard_start_xmit = bond_3ad_xmit_xor; + break; + case BOND_MODE_TLB: + case BOND_MODE_ALB: + bond_dev->hard_start_xmit = bond_alb_xmit; + bond_dev->set_mac_address = bond_alb_set_mac_address; + break; + default: + /* Should never happen, mode already checked */ + printk(KERN_ERR DRV_NAME + ": Error: Unknown bonding mode %d\n", + mode); + break; + } +} + +/* + * Does not allocate but creates a /proc entry. + * Allowed to fail. + */ +static int __init bond_init(struct net_device *bond_dev, struct bond_params *params) +{ + struct bonding *bond = bond_dev->priv; + + dprintk("Begin bond_init for %s\n", bond_dev->name); + + /* initialize rwlocks */ + rwlock_init(&bond->lock); + rwlock_init(&bond->curr_slave_lock); + + bond->params = *params; /* copy params struct */ + + /* Initialize pointers */ + bond->first_slave = NULL; + bond->curr_active_slave = NULL; + bond->current_arp_slave = NULL; + bond->primary_slave = NULL; + bond->dev = bond_dev; + INIT_LIST_HEAD(&bond->vlan_list); + + /* Initialize the device entry points */ + bond_dev->open = bond_open; + bond_dev->stop = bond_close; + bond_dev->get_stats = bond_get_stats; + bond_dev->do_ioctl = bond_do_ioctl; + bond_dev->set_multicast_list = bond_set_multicast_list; + bond_dev->change_mtu = bond_change_mtu; + bond_dev->set_mac_address = bond_set_mac_address; + + bond_set_mode_ops(bond_dev, bond->params.mode); + + bond_dev->destructor = free_netdev; + + /* Initialize the device options */ + bond_dev->tx_queue_len = 0; + bond_dev->flags |= IFF_MASTER|IFF_MULTICAST; + + /* At first, we block adding VLANs. That's the only way to + * prevent problems that occur when adding VLANs over an + * empty bond. The block will be removed once non-challenged + * slaves are enslaved. + */ + bond_dev->features |= NETIF_F_VLAN_CHALLENGED; + + /* don't acquire bond device's xmit_lock when + * transmitting */ + bond_dev->features |= NETIF_F_LLTX; + + /* By default, we declare the bond to be fully + * VLAN hardware accelerated capable. Special + * care is taken in the various xmit functions + * when there are slaves that are not hw accel + * capable + */ + bond_dev->vlan_rx_register = bond_vlan_rx_register; + bond_dev->vlan_rx_add_vid = bond_vlan_rx_add_vid; + bond_dev->vlan_rx_kill_vid = bond_vlan_rx_kill_vid; + bond_dev->features |= (NETIF_F_HW_VLAN_TX | + NETIF_F_HW_VLAN_RX | + NETIF_F_HW_VLAN_FILTER); + +#ifdef CONFIG_PROC_FS + bond_create_proc_entry(bond); +#endif + + list_add_tail(&bond->bond_list, &bond_dev_list); + + return 0; +} + +/* De-initialize device specific data. + * Caller must hold rtnl_lock. + */ +static inline void bond_deinit(struct net_device *bond_dev) +{ + struct bonding *bond = bond_dev->priv; + + list_del(&bond->bond_list); + +#ifdef CONFIG_PROC_FS + bond_remove_proc_entry(bond); +#endif +} + +/* Unregister and free all bond devices. + * Caller must hold rtnl_lock. + */ +static void bond_free_all(void) +{ + struct bonding *bond, *nxt; + + list_for_each_entry_safe(bond, nxt, &bond_dev_list, bond_list) { + struct net_device *bond_dev = bond->dev; + + unregister_netdevice(bond_dev); + bond_deinit(bond_dev); + } + +#ifdef CONFIG_PROC_FS + bond_destroy_proc_dir(); +#endif +} + +/*------------------------- Module initialization ---------------------------*/ + +/* + * Convert string input module parms. Accept either the + * number of the mode or its string name. + */ +static inline int bond_parse_parm(char *mode_arg, struct bond_parm_tbl *tbl) +{ + int i; + + for (i = 0; tbl[i].modename; i++) { + if ((isdigit(*mode_arg) && + tbl[i].mode == simple_strtol(mode_arg, NULL, 0)) || + (strncmp(mode_arg, tbl[i].modename, + strlen(tbl[i].modename)) == 0)) { + return tbl[i].mode; + } + } + + return -1; +} + +static int bond_check_params(struct bond_params *params) +{ + /* + * Convert string parameters. + */ + if (mode) { + bond_mode = bond_parse_parm(mode, bond_mode_tbl); + if (bond_mode == -1) { + printk(KERN_ERR DRV_NAME + ": Error: Invalid bonding mode \"%s\"\n", + mode == NULL ? "NULL" : mode); + return -EINVAL; + } + } + + if (lacp_rate) { + if (bond_mode != BOND_MODE_8023AD) { + printk(KERN_INFO DRV_NAME + ": lacp_rate param is irrelevant in mode %s\n", + bond_mode_name(bond_mode)); + } else { + lacp_fast = bond_parse_parm(lacp_rate, bond_lacp_tbl); + if (lacp_fast == -1) { + printk(KERN_ERR DRV_NAME + ": Error: Invalid lacp rate \"%s\"\n", + lacp_rate == NULL ? "NULL" : lacp_rate); + return -EINVAL; + } + } + } + + if (max_bonds < 1 || max_bonds > INT_MAX) { + printk(KERN_WARNING DRV_NAME + ": Warning: max_bonds (%d) not in range %d-%d, so it " + "was reset to BOND_DEFAULT_MAX_BONDS (%d)", + max_bonds, 1, INT_MAX, BOND_DEFAULT_MAX_BONDS); + max_bonds = BOND_DEFAULT_MAX_BONDS; + } + + if (miimon < 0) { + printk(KERN_WARNING DRV_NAME + ": Warning: miimon module parameter (%d), " + "not in range 0-%d, so it was reset to %d\n", + miimon, INT_MAX, BOND_LINK_MON_INTERV); + miimon = BOND_LINK_MON_INTERV; + } + + if (updelay < 0) { + printk(KERN_WARNING DRV_NAME + ": Warning: updelay module parameter (%d), " + "not in range 0-%d, so it was reset to 0\n", + updelay, INT_MAX); + updelay = 0; + } + + if (downdelay < 0) { + printk(KERN_WARNING DRV_NAME + ": Warning: downdelay module parameter (%d), " + "not in range 0-%d, so it was reset to 0\n", + downdelay, INT_MAX); + downdelay = 0; + } + + if ((use_carrier != 0) && (use_carrier != 1)) { + printk(KERN_WARNING DRV_NAME + ": Warning: use_carrier module parameter (%d), " + "not of valid value (0/1), so it was set to 1\n", + use_carrier); + use_carrier = 1; + } + + /* reset values for 802.3ad */ + if (bond_mode == BOND_MODE_8023AD) { + if (!miimon) { + printk(KERN_WARNING DRV_NAME + ": Warning: miimon must be specified, " + "otherwise bonding will not detect link " + "failure, speed and duplex which are " + "essential for 802.3ad operation\n"); + printk(KERN_WARNING "Forcing miimon to 100msec\n"); + miimon = 100; + } + } + + /* reset values for TLB/ALB */ + if ((bond_mode == BOND_MODE_TLB) || + (bond_mode == BOND_MODE_ALB)) { + if (!miimon) { + printk(KERN_WARNING DRV_NAME + ": Warning: miimon must be specified, " + "otherwise bonding will not detect link " + "failure and link speed which are essential " + "for TLB/ALB load balancing\n"); + printk(KERN_WARNING "Forcing miimon to 100msec\n"); + miimon = 100; + } + } + + if (bond_mode == BOND_MODE_ALB) { + printk(KERN_NOTICE DRV_NAME + ": In ALB mode you might experience client " + "disconnections upon reconnection of a link if the " + "bonding module updelay parameter (%d msec) is " + "incompatible with the forwarding delay time of the " + "switch\n", + updelay); + } + + if (!miimon) { + if (updelay || downdelay) { + /* just warn the user the up/down delay will have + * no effect since miimon is zero... + */ + printk(KERN_WARNING DRV_NAME + ": Warning: miimon module parameter not set " + "and updelay (%d) or downdelay (%d) module " + "parameter is set; updelay and downdelay have " + "no effect unless miimon is set\n", + updelay, downdelay); + } + } else { + /* don't allow arp monitoring */ + if (arp_interval) { + printk(KERN_WARNING DRV_NAME + ": Warning: miimon (%d) and arp_interval (%d) " + "can't be used simultaneously, disabling ARP " + "monitoring\n", + miimon, arp_interval); + arp_interval = 0; + } + + if ((updelay % miimon) != 0) { + printk(KERN_WARNING DRV_NAME + ": Warning: updelay (%d) is not a multiple " + "of miimon (%d), updelay rounded to %d ms\n", + updelay, miimon, (updelay / miimon) * miimon); + } + + updelay /= miimon; + + if ((downdelay % miimon) != 0) { + printk(KERN_WARNING DRV_NAME + ": Warning: downdelay (%d) is not a multiple " + "of miimon (%d), downdelay rounded to %d ms\n", + downdelay, miimon, + (downdelay / miimon) * miimon); + } + + downdelay /= miimon; + } + + if (arp_interval < 0) { + printk(KERN_WARNING DRV_NAME + ": Warning: arp_interval module parameter (%d) " + ", not in range 0-%d, so it was reset to %d\n", + arp_interval, INT_MAX, BOND_LINK_ARP_INTERV); + arp_interval = BOND_LINK_ARP_INTERV; + } + + for (arp_ip_count = 0; + (arp_ip_count < BOND_MAX_ARP_TARGETS) && arp_ip_target[arp_ip_count]; + arp_ip_count++) { + /* not complete check, but should be good enough to + catch mistakes */ + if (!isdigit(arp_ip_target[arp_ip_count][0])) { + printk(KERN_WARNING DRV_NAME + ": Warning: bad arp_ip_target module parameter " + "(%s), ARP monitoring will not be performed\n", + arp_ip_target[arp_ip_count]); + arp_interval = 0; + } else { + u32 ip = in_aton(arp_ip_target[arp_ip_count]); + arp_target[arp_ip_count] = ip; + } + } + + if (arp_interval && !arp_ip_count) { + /* don't allow arping if no arp_ip_target given... */ + printk(KERN_WARNING DRV_NAME + ": Warning: arp_interval module parameter (%d) " + "specified without providing an arp_ip_target " + "parameter, arp_interval was reset to 0\n", + arp_interval); + arp_interval = 0; + } + + if (miimon) { + printk(KERN_INFO DRV_NAME + ": MII link monitoring set to %d ms\n", + miimon); + } else if (arp_interval) { + int i; + + printk(KERN_INFO DRV_NAME + ": ARP monitoring set to %d ms with %d target(s):", + arp_interval, arp_ip_count); + + for (i = 0; i < arp_ip_count; i++) + printk (" %s", arp_ip_target[i]); + + printk("\n"); + + } else { + /* miimon and arp_interval not set, we need one so things + * work as expected, see bonding.txt for details + */ + printk(KERN_WARNING DRV_NAME + ": Warning: either miimon or arp_interval and " + "arp_ip_target module parameters must be specified, " + "otherwise bonding will not detect link failures! see " + "bonding.txt for details.\n"); + } + + if (primary && !USES_PRIMARY(bond_mode)) { + /* currently, using a primary only makes sense + * in active backup, TLB or ALB modes + */ + printk(KERN_WARNING DRV_NAME + ": Warning: %s primary device specified but has no " + "effect in %s mode\n", + primary, bond_mode_name(bond_mode)); + primary = NULL; + } + + /* fill params struct with the proper values */ + params->mode = bond_mode; + params->miimon = miimon; + params->arp_interval = arp_interval; + params->updelay = updelay; + params->downdelay = downdelay; + params->use_carrier = use_carrier; + params->lacp_fast = lacp_fast; + params->primary[0] = 0; + + if (primary) { + strncpy(params->primary, primary, IFNAMSIZ); + params->primary[IFNAMSIZ - 1] = 0; + } + + memcpy(params->arp_targets, arp_target, sizeof(arp_target)); + + return 0; +} + +static int __init bonding_init(void) +{ + struct bond_params params; + int i; + int res; + + printk(KERN_INFO "%s", version); + + res = bond_check_params(¶ms); + if (res) { + return res; + } + + rtnl_lock(); + +#ifdef CONFIG_PROC_FS + bond_create_proc_dir(); +#endif + + for (i = 0; i < max_bonds; i++) { + struct net_device *bond_dev; + + bond_dev = alloc_netdev(sizeof(struct bonding), "", ether_setup); + if (!bond_dev) { + res = -ENOMEM; + goto out_err; + } + + res = dev_alloc_name(bond_dev, "bond%d"); + if (res < 0) { + free_netdev(bond_dev); + goto out_err; + } + + /* bond_init() must be called after dev_alloc_name() (for the + * /proc files), but before register_netdevice(), because we + * need to set function pointers. + */ + res = bond_init(bond_dev, ¶ms); + if (res < 0) { + free_netdev(bond_dev); + goto out_err; + } + + SET_MODULE_OWNER(bond_dev); + + res = register_netdevice(bond_dev); + if (res < 0) { + bond_deinit(bond_dev); + free_netdev(bond_dev); + goto out_err; + } + } + + rtnl_unlock(); + register_netdevice_notifier(&bond_netdev_notifier); + + return 0; + +out_err: + /* free and unregister all bonds that were successfully added */ + bond_free_all(); + + rtnl_unlock(); + + return res; +} + +static void __exit bonding_exit(void) +{ + unregister_netdevice_notifier(&bond_netdev_notifier); + + rtnl_lock(); + bond_free_all(); + rtnl_unlock(); +} + +module_init(bonding_init); +module_exit(bonding_exit); +MODULE_LICENSE("GPL"); +MODULE_VERSION(DRV_VERSION); +MODULE_DESCRIPTION(DRV_DESCRIPTION ", v" DRV_VERSION); +MODULE_AUTHOR("Thomas Davis, tadavis@lbl.gov and many others"); +MODULE_SUPPORTED_DEVICE("most ethernet devices"); + +/* + * Local variables: + * c-indent-level: 8 + * c-basic-offset: 8 + * tab-width: 8 + * End: + */ + diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h new file mode 100644 index 00000000000..8c325308489 --- /dev/null +++ b/drivers/net/bonding/bonding.h @@ -0,0 +1,252 @@ +/* + * Bond several ethernet interfaces into a Cisco, running 'Etherchannel'. + * + * Portions are (c) Copyright 1995 Simon "Guru Aleph-Null" Janes + * NCM: Network and Communications Management, Inc. + * + * BUT, I'm the one who modified it for ethernet, so: + * (c) Copyright 1999, Thomas Davis, tadavis@lbl.gov + * + * This software may be used and distributed according to the terms + * of the GNU Public License, incorporated herein by reference. + * + * + * 2003/03/18 - Amir Noam <amir.noam at intel dot com>, + * Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Shmulik Hen <shmulik.hen at intel dot com> + * - Added support for IEEE 802.3ad Dynamic link aggregation mode. + * + * 2003/05/01 - Tsippy Mendelson <tsippy.mendelson at intel dot com> and + * Amir Noam <amir.noam at intel dot com> + * - Code beautification and style changes (mainly in comments). + * + * 2003/05/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Added support for Transmit load balancing mode. + * + * 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com> + * - Code cleanup and style changes + */ + +#ifndef _LINUX_BONDING_H +#define _LINUX_BONDING_H + +#include <linux/timer.h> +#include <linux/proc_fs.h> +#include <linux/if_bonding.h> +#include "bond_3ad.h" +#include "bond_alb.h" + +#define DRV_VERSION "2.6.1" +#define DRV_RELDATE "October 29, 2004" +#define DRV_NAME "bonding" +#define DRV_DESCRIPTION "Ethernet Channel Bonding Driver" + +#define BOND_MAX_ARP_TARGETS 16 + +#ifdef BONDING_DEBUG +#define dprintk(fmt, args...) \ + printk(KERN_DEBUG \ + DRV_NAME ": %s() %d: " fmt, __FUNCTION__, __LINE__ , ## args ) +#else +#define dprintk(fmt, args...) +#endif /* BONDING_DEBUG */ + +#define IS_UP(dev) \ + ((((dev)->flags & IFF_UP) == IFF_UP) && \ + netif_running(dev) && \ + netif_carrier_ok(dev)) + +/* + * Checks whether bond is ready for transmit. + * + * Caller must hold bond->lock + */ +#define BOND_IS_OK(bond) \ + (((bond)->dev->flags & IFF_UP) && \ + netif_running((bond)->dev) && \ + ((bond)->slave_cnt > 0)) + +/* + * Checks whether slave is ready for transmit. + */ +#define SLAVE_IS_OK(slave) \ + (((slave)->dev->flags & IFF_UP) && \ + netif_running((slave)->dev) && \ + ((slave)->link == BOND_LINK_UP) && \ + ((slave)->state == BOND_STATE_ACTIVE)) + + +#define USES_PRIMARY(mode) \ + (((mode) == BOND_MODE_ACTIVEBACKUP) || \ + ((mode) == BOND_MODE_TLB) || \ + ((mode) == BOND_MODE_ALB)) + +/* + * Less bad way to call ioctl from within the kernel; this needs to be + * done some other way to get the call out of interrupt context. + * Needs "ioctl" variable to be supplied by calling context. + */ +#define IOCTL(dev, arg, cmd) ({ \ + int res = 0; \ + mm_segment_t fs = get_fs(); \ + set_fs(get_ds()); \ + res = ioctl(dev, arg, cmd); \ + set_fs(fs); \ + res; }) + +/** + * bond_for_each_slave_from - iterate the slaves list from a starting point + * @bond: the bond holding this list. + * @pos: current slave. + * @cnt: counter for max number of moves + * @start: starting point. + * + * Caller must hold bond->lock + */ +#define bond_for_each_slave_from(bond, pos, cnt, start) \ + for (cnt = 0, pos = start; \ + cnt < (bond)->slave_cnt; \ + cnt++, pos = (pos)->next) + +/** + * bond_for_each_slave_from_to - iterate the slaves list from start point to stop point + * @bond: the bond holding this list. + * @pos: current slave. + * @cnt: counter for number max of moves + * @start: start point. + * @stop: stop point. + * + * Caller must hold bond->lock + */ +#define bond_for_each_slave_from_to(bond, pos, cnt, start, stop) \ + for (cnt = 0, pos = start; \ + ((cnt < (bond)->slave_cnt) && (pos != (stop)->next)); \ + cnt++, pos = (pos)->next) + +/** + * bond_for_each_slave - iterate the slaves list from head + * @bond: the bond holding this list. + * @pos: current slave. + * @cnt: counter for max number of moves + * + * Caller must hold bond->lock + */ +#define bond_for_each_slave(bond, pos, cnt) \ + bond_for_each_slave_from(bond, pos, cnt, (bond)->first_slave) + + +struct bond_params { + int mode; + int miimon; + int arp_interval; + int use_carrier; + int updelay; + int downdelay; + int lacp_fast; + char primary[IFNAMSIZ]; + u32 arp_targets[BOND_MAX_ARP_TARGETS]; +}; + +struct vlan_entry { + struct list_head vlan_list; + unsigned short vlan_id; +}; + +struct slave { + struct net_device *dev; /* first - usefull for panic debug */ + struct slave *next; + struct slave *prev; + s16 delay; + u32 jiffies; + s8 link; /* one of BOND_LINK_XXXX */ + s8 state; /* one of BOND_STATE_XXXX */ + u32 original_flags; + u32 link_failure_count; + u16 speed; + u8 duplex; + u8 perm_hwaddr[ETH_ALEN]; + struct ad_slave_info ad_info; /* HUGE - better to dynamically alloc */ + struct tlb_slave_info tlb_info; +}; + +/* + * Here are the locking policies for the two bonding locks: + * + * 1) Get bond->lock when reading/writing slave list. + * 2) Get bond->curr_slave_lock when reading/writing bond->curr_active_slave. + * (It is unnecessary when the write-lock is put with bond->lock.) + * 3) When we lock with bond->curr_slave_lock, we must lock with bond->lock + * beforehand. + */ +struct bonding { + struct net_device *dev; /* first - usefull for panic debug */ + struct slave *first_slave; + struct slave *curr_active_slave; + struct slave *current_arp_slave; + struct slave *primary_slave; + s32 slave_cnt; /* never change this value outside the attach/detach wrappers */ + rwlock_t lock; + rwlock_t curr_slave_lock; + struct timer_list mii_timer; + struct timer_list arp_timer; + s8 kill_timers; + struct net_device_stats stats; +#ifdef CONFIG_PROC_FS + struct proc_dir_entry *proc_entry; + char proc_file_name[IFNAMSIZ]; +#endif /* CONFIG_PROC_FS */ + struct list_head bond_list; + struct dev_mc_list *mc_list; + u16 flags; + struct ad_bond_info ad_info; + struct alb_bond_info alb_info; + struct bond_params params; + struct list_head vlan_list; + struct vlan_group *vlgrp; +}; + +/** + * Returns NULL if the net_device does not belong to any of the bond's slaves + * + * Caller must hold bond lock for read + */ +extern inline struct slave *bond_get_slave_by_dev(struct bonding *bond, struct net_device *slave_dev) +{ + struct slave *slave = NULL; + int i; + + bond_for_each_slave(bond, slave, i) { + if (slave->dev == slave_dev) { + break; + } + } + + return slave; +} + +extern inline struct bonding *bond_get_bond_by_slave(struct slave *slave) +{ + if (!slave || !slave->dev->master) { + return NULL; + } + + return (struct bonding *)slave->dev->master->priv; +} + +extern inline void bond_set_slave_inactive_flags(struct slave *slave) +{ + slave->state = BOND_STATE_BACKUP; + slave->dev->flags |= IFF_NOARP; +} + +extern inline void bond_set_slave_active_flags(struct slave *slave) +{ + slave->state = BOND_STATE_ACTIVE; + slave->dev->flags &= ~IFF_NOARP; +} + +struct vlan_entry *bond_next_vlan(struct bonding *bond, struct vlan_entry *curr); +int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb, struct net_device *slave_dev); + +#endif /* _LINUX_BONDING_H */ + |