X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=drivers%2Fnet%2Fbonding%2Fbond_main.c;h=6482aed4bb7cddec1c6879e47a3a26a495332359;hb=97bf2856c6014879bd04983a3e9dfcdac1e7fe85;hp=778fd075451300b48b2fd7fb3a961214b09931ff;hpb=5273a3df6485dc2ad6aa7ddd441b9a21970f003b;p=linux-2.6.git diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 778fd0754..6482aed4b 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -29,451 +29,10 @@ * 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 - * - fix leaks on failure at bond_init - * - * 2000/09/30 - Willy Tarreau - * - 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 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 - * - 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 - * - 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 - * - fixed two potential SMP race conditions - * - * 2000/10/18 - Willy Tarreau - * - small fixes to the monitoring FSM in case of zero delays - * 2000/11/01 - Willy Tarreau - * - 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 - * - 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 - * - 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 - * for more details. The structs of interest - * are ifbond and ifslave. - * - * 2001/4/5 - Chad N. Tindel - * - Ported to 2.4 Kernel - * - * 2001/5/2 - Jeffrey E. Mast - * - 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 - * - 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 - * - 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 - * - Now call dev_close when releasing a slave so it doesn't screw up - * out routing table. - * - * 2001/6/01 - Chad N. Tindel - * - Added /proc support for getting bond and slave information. - * Information is in /proc/net//info. - * - Changed the locking when calling bond_close to prevent deadlock. - * - * 2001/8/05 - Janice Girouard - * - 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 - * - correct locking problem with rtnl_exlock_nowait - * - * 2001/8/23 - Janice Girouard - * - bzero initial dev_bonds, to correct oops - * - convert SIOCDEVPRIVATE to new MII ioctl calls - * - * 2001/9/13 - Takao Indoh - * - Add the BOND_CHANGE_ACTIVE ioctl implementation - * - * 2001/9/14 - Mark Huth - * - Change MII_LINK_READY to not check for end of auto-negotiation, - * but only for an up link. - * - * 2001/9/20 - Chad N. Tindel - * - 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 - * - add arp_monitor for active backup mode - * - * 2001/10/23 - Takao Indoh - * - Various memory leak fixes - * - * 2001/11/5 - Mark Huth - * - 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 - * - Restore original slave flags at release time. - * - * 2002/02/18 - Erik Habbinga - * - 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 and - * Steve Mead - * - Port Gleb Natapov's multicast support patchs from 2.4.12 - * to 2.4.18 adding support for multicast. - * - * 2002/06/10 - Tony Cureington - * - 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 - * - 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 - * - 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 - * - make sure the ip target matches the arp_target before saving the - * hw address. - * - * 2002/9/30 - Dan Eisner - * - 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 - * - 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 - * - Added bond_xmit_broadcast logic. - * - Added bond_mode() support function. - * - * 2002/10/26 - Laurent Deniel - * - 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 - * - 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 - * - 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 - * - 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 - * - fix bond_slave_info_query when slave_id = num_slaves - * - * 2002/11/19 - Janice Girouard - * - correct ifr_data reference. Update ifr_data reference - * to mii_ioctl_data struct values to avoid confusion. - * - * 2002/11/22 - Bert Barbe - * - Add support for multiple arp_ip_target - * - * 2002/12/13 - Jay Vosburgh - * - 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 - * - 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 - * - Fix bond_mii_monitor() logic error that could result in - * bonding round-robin mode ignoring links after failover/recovery - * - * 2003/03/17 - Jay Vosburgh - * - 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 and - * Shmulik Hen - * - 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 and - * Shmulik Hen - * - 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 - * - Added support for getting slave's speed and duplex via ethtool. - * Needed for 802.3ad and other future modes. - * - * 2003/03/18 - Tsippy Mendelson and - * Shmulik Hen - * - 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 , - * Tsippy Mendelson and - * Shmulik Hen - * - Added support for IEEE 802.3ad Dynamic link aggregation mode. - * - * 2003/05/01 - Amir Noam - * - Added ABI version control to restore compatibility between - * new/old ifenslave and new/old bonding. - * - * 2003/05/01 - Shmulik Hen - * - 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 and - * Amir Noam - * - 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 - * - 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 - * - 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 - * - 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 - * - Applied fix to activebackup_arp_monitor posted to bonding-devel - * by Tony Cureington . Fixes ARP - * monitor endless failover bug. Version to 2.2.10 - * - * 2003/05/20 - Amir Noam - * - Fixed bug in ABI version control - Don't commit to a specific - * ABI version if receiving unsupported ioctl commands. - * - * 2003/05/22 - Jay Vosburgh - * - 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 - * - Fixed bond_change_active() for ALB/TLB modes. - * Version to 2.2.14. - * - * 2003/07/29 - Amir Noam - * - Fixed ARP monitoring bug. - * Version to 2.2.15. - * - * 2003/07/31 - Willy Tarreau - * - Fixed kernel panic when using ARP monitoring without - * setting bond's IP address. - * Version to 2.2.16. - * - * 2003/08/06 - Amir Noam - * - 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 , - * Amir Noam and - * Shmulik Hen - * - 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 , - * Amir Noam - * - 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 - * - Fix /proc creation/destruction. - * - * 2003/12/01 - Shmulik Hen - * - 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 - * - 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 - * - Save bonding parameters per bond instead of using the global values. - * Set version to 2.5.4. - * - * 2004/01/14 - Shmulik Hen - * - 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. */ //#define BONDING_DEBUG 1 -#include #include #include #include @@ -483,7 +42,10 @@ #include #include #include +#include #include +#include +#include #include #include #include @@ -491,8 +53,8 @@ #include #include #include +#include #include -#include #include #include #include @@ -512,6 +74,7 @@ #include #include #include +#include #include "bonding.h" #include "bond_3ad.h" #include "bond_alb.h" @@ -530,67 +93,70 @@ static int use_carrier = 1; static char *mode = NULL; static char *primary = NULL; static char *lacp_rate = NULL; +static char *xmit_hash_policy = NULL; static int arp_interval = BOND_LINK_ARP_INTERV; static char *arp_ip_target[BOND_MAX_ARP_TARGETS] = { NULL, }; +static char *arp_validate = NULL; +struct bond_params bonding_defaults; -MODULE_PARM(max_bonds, "i"); +module_param(max_bonds, int, 0); MODULE_PARM_DESC(max_bonds, "Max number of bonded devices"); -MODULE_PARM(miimon, "i"); +module_param(miimon, int, 0); MODULE_PARM_DESC(miimon, "Link check interval in milliseconds"); -MODULE_PARM(updelay, "i"); +module_param(updelay, int, 0); MODULE_PARM_DESC(updelay, "Delay before considering link up, in milliseconds"); -MODULE_PARM(downdelay, "i"); -MODULE_PARM_DESC(downdelay, "Delay before considering link down, in milliseconds"); -MODULE_PARM(use_carrier, "i"); -MODULE_PARM_DESC(use_carrier, "Use netif_carrier_ok (vs MII ioctls) in miimon; 0 for off, 1 for on (default)"); -MODULE_PARM(mode, "s"); -MODULE_PARM_DESC(mode, "Mode of operation : 0 for round robin, 1 for active-backup, 2 for xor"); -MODULE_PARM(primary, "s"); +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 balance-rr, " + "1 for active-backup, 2 for balance-xor, " + "3 for broadcast, 4 for 802.3ad, 5 for balance-tlb, " + "6 for balance-alb"); +module_param(primary, charp, 0); MODULE_PARM_DESC(primary, "Primary network device to use"); -MODULE_PARM(lacp_rate, "s"); -MODULE_PARM_DESC(lacp_rate, "LACPDU tx rate to request from 802.3ad partner (slow/fast)"); -MODULE_PARM(arp_interval, "i"); +module_param(lacp_rate, charp, 0); +MODULE_PARM_DESC(lacp_rate, "LACPDU tx rate to request from 802.3ad partner " + "(slow/fast)"); +module_param(xmit_hash_policy, charp, 0); +MODULE_PARM_DESC(xmit_hash_policy, "XOR hashing method: 0 for layer 2 (default)" + ", 1 for layer 3+4"); +module_param(arp_interval, int, 0); MODULE_PARM_DESC(arp_interval, "arp interval in milliseconds"); -MODULE_PARM(arp_ip_target, "1-" __MODULE_STRING(BOND_MAX_ARP_TARGETS) "s"); +module_param_array(arp_ip_target, charp, NULL, 0); MODULE_PARM_DESC(arp_ip_target, "arp targets in n.n.n.n form"); +module_param(arp_validate, charp, 0); +MODULE_PARM_DESC(arp_validate, "validate src/dst of ARP probes: none (default), active, backup or all"); /*----------------------------- Global variables ----------------------------*/ -static const char *version = +static const char * const version = DRV_DESCRIPTION ": v" DRV_VERSION " (" DRV_RELDATE ")\n"; -static LIST_HEAD(bond_dev_list); +LIST_HEAD(bond_dev_list); #ifdef CONFIG_PROC_FS static struct proc_dir_entry *bond_proc_dir = NULL; #endif +extern struct rw_semaphore bonding_rwsem; 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 xmit_hashtype= BOND_XMIT_POLICY_LAYER2; 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[] = { + +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[] = { +struct bond_parm_tbl bond_mode_tbl[] = { { "balance-rr", BOND_MODE_ROUNDROBIN}, { "active-backup", BOND_MODE_ACTIVEBACKUP}, { "balance-xor", BOND_MODE_XOR}, @@ -601,13 +167,27 @@ static struct bond_parm_tbl bond_mode_tbl[] = { { NULL, -1}, }; +struct bond_parm_tbl xmit_hashtype_tbl[] = { +{ "layer2", BOND_XMIT_POLICY_LAYER2}, +{ "layer3+4", BOND_XMIT_POLICY_LAYER34}, +{ NULL, -1}, +}; + +struct bond_parm_tbl arp_validate_tbl[] = { +{ "none", BOND_ARP_VALIDATE_NONE}, +{ "active", BOND_ARP_VALIDATE_ACTIVE}, +{ "backup", BOND_ARP_VALIDATE_BACKUP}, +{ "all", BOND_ARP_VALIDATE_ALL}, +{ NULL, -1}, +}; + /*-------------------------- Forward declarations ---------------------------*/ -static inline void bond_set_mode_ops(struct net_device *bond_dev, int mode); +static void bond_send_gratuitous_arp(struct bonding *bond); /*---------------------------- General routines -----------------------------*/ -static const char *bond_mode_name(int mode) +const char *bond_mode_name(int mode) { switch (mode) { case BOND_MODE_ROUNDROBIN : @@ -652,6 +232,7 @@ static int bond_add_vlan(struct bonding *bond, unsigned short vlan_id) INIT_LIST_HEAD(&vlan->vlan_list); vlan->vlan_id = vlan_id; + vlan->vlan_ip = 0; write_lock_bh(&bond->lock); @@ -786,29 +367,20 @@ struct vlan_entry *bond_next_vlan(struct bonding *bond, struct vlan_entry *curr) * @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 tarnsmit that is + * 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. - * - * Assumption - once a VLAN device is created over the bond device, all - * packets are going to be hardware accelerated VLAN tagged since the IP - * binding is done over the VLAN device */ int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb, struct net_device *slave_dev) { unsigned short vlan_id; - int res; if (!list_empty(&bond->vlan_list) && - !(slave_dev->features & NETIF_F_HW_VLAN_TX)) { - res = vlan_get_tag(skb, &vlan_id); - if (res) { - return -EINVAL; - } - + !(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) { @@ -892,7 +464,7 @@ static void bond_vlan_rx_add_vid(struct net_device *bond_dev, uint16_t vid) res = bond_add_vlan(bond, vid); if (res) { printk(KERN_ERR DRV_NAME - ": %s: Failed to add vlan id %d\n", + ": %s: Error: Failed to add vlan id %d\n", bond_dev->name, vid); } } @@ -926,7 +498,7 @@ static void bond_vlan_rx_kill_vid(struct net_device *bond_dev, uint16_t vid) res = bond_del_vlan(bond, vid); if (res) { printk(KERN_ERR DRV_NAME - ": %s: Failed to remove vlan id %d\n", + ": %s: Error: Failed to remove vlan id %d\n", bond_dev->name, vid); } } @@ -996,6 +568,42 @@ out: /*------------------------------- Link status -------------------------------*/ +/* + * Set the carrier state for the master according to the state of its + * slaves. If any slaves are up, the master is up. In 802.3ad mode, + * do special 802.3ad magic. + * + * Returns zero if carrier state does not change, nonzero if it does. + */ +static int bond_set_carrier(struct bonding *bond) +{ + struct slave *slave; + int i; + + if (bond->slave_cnt == 0) + goto down; + + if (bond->params.mode == BOND_MODE_8023AD) + return bond_3ad_set_carrier(bond); + + bond_for_each_slave(bond, slave, i) { + if (slave->link == BOND_LINK_UP) { + if (!netif_carrier_ok(bond->dev)) { + netif_carrier_on(bond->dev); + return 1; + } + return 0; + } + } + +down: + if (netif_carrier_ok(bond->dev)) { + netif_carrier_off(bond->dev); + return 1; + } + return 0; +} + /* * Get link speed and duplex from the slave's base driver * using ethtool. If for some reason the call fails or the @@ -1014,7 +622,7 @@ static int bond_update_speed_duplex(struct slave *slave) slave->duplex = DUPLEX_FULL; if (slave_dev->ethtool_ops) { - u32 res; + int res; if (!slave_dev->ethtool_ops->get_settings) { return -1; @@ -1041,6 +649,7 @@ verify: case SPEED_10: case SPEED_100: case SPEED_1000: + case SPEED_10000: break; default: return -1; @@ -1103,7 +712,7 @@ static int bond_check_dev_link(struct bonding *bond, struct net_device *slave_de /* Yes, the mii is overlaid on the ifreq.ifr_ifru */ strncpy(ifr.ifr_name, slave_dev->name, IFNAMSIZ); - mii = (struct mii_ioctl_data *)&ifr.ifr_data; + mii = if_mii(&ifr); if (IOCTL(slave_dev, &ifr, SIOCGMIIPHY) == 0) { mii->reg_num = MII_BMSR; if (IOCTL(slave_dev, &ifr, SIOCGMIIREG) == 0) { @@ -1271,12 +880,13 @@ static void bond_mc_list_destroy(struct bonding *bond) /* * 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) +static int bond_mc_list_copy(struct dev_mc_list *mc_list, struct bonding *bond, + gfp_t gfp_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); + new_dmi = kmalloc(sizeof(struct dev_mc_list), gfp_flag); if (!new_dmi) { /* FIXME: Potential memory leak !!! */ @@ -1430,7 +1040,7 @@ static struct slave *bond_find_best_slave(struct bonding *bond) * * Warning: Caller must hold curr_slave_lock for writing. */ -static void bond_change_active_slave(struct bonding *bond, struct slave *new_active) +void bond_change_active_slave(struct bonding *bond, struct slave *new_active) { struct slave *old_active = bond->curr_active_slave; @@ -1470,16 +1080,6 @@ static void bond_change_active_slave(struct bonding *bond, struct slave *new_act } } - 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); } @@ -1487,9 +1087,24 @@ static void bond_change_active_slave(struct bonding *bond, struct slave *new_act if ((bond->params.mode == BOND_MODE_TLB) || (bond->params.mode == BOND_MODE_ALB)) { bond_alb_handle_active_change(bond, new_active); + if (old_active) + bond_set_slave_inactive_flags(old_active); + if (new_active) + bond_set_slave_active_flags(new_active); } else { bond->curr_active_slave = new_active; } + + 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); + } + bond_send_gratuitous_arp(bond); + } } /** @@ -1503,13 +1118,27 @@ static void bond_change_active_slave(struct bonding *bond, struct slave *new_act * * Warning: Caller must hold curr_slave_lock for writing. */ -static void bond_select_active_slave(struct bonding *bond) +void bond_select_active_slave(struct bonding *bond) { struct slave *best_slave; + int rv; best_slave = bond_find_best_slave(bond); if (best_slave != bond->curr_active_slave) { bond_change_active_slave(bond, best_slave); + rv = bond_set_carrier(bond); + if (!rv) + return; + + if (netif_carrier_ok(bond->dev)) { + printk(KERN_INFO DRV_NAME + ": %s: first active interface up!\n", + bond->dev->name); + } else { + printk(KERN_INFO DRV_NAME ": %s: " + "now running without any active interface !\n", + bond->dev->name); + } } } @@ -1571,7 +1200,7 @@ static void bond_detach_slave(struct bonding *bond, struct slave *slave) /*---------------------------------- IOCTL ----------------------------------*/ -static int bond_sethwaddr(struct net_device *bond_dev, struct net_device *slave_dev) +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); @@ -1580,8 +1209,51 @@ static int bond_sethwaddr(struct net_device *bond_dev, struct net_device *slave_ return 0; } +#define BOND_INTERSECT_FEATURES \ + (NETIF_F_SG | NETIF_F_ALL_CSUM | NETIF_F_TSO | NETIF_F_UFO) + +/* + * Compute the common dev->feature set available to all slaves. Some + * feature bits are managed elsewhere, so preserve feature bits set on + * master device that are not part of the examined set. + */ +static int bond_compute_features(struct bonding *bond) +{ + unsigned long features = BOND_INTERSECT_FEATURES; + struct slave *slave; + struct net_device *bond_dev = bond->dev; + unsigned short max_hard_header_len = ETH_HLEN; + int i; + + bond_for_each_slave(bond, slave, i) { + features &= (slave->dev->features & BOND_INTERSECT_FEATURES); + if (slave->dev->hard_header_len > max_hard_header_len) + max_hard_header_len = slave->dev->hard_header_len; + } + + if ((features & NETIF_F_SG) && + !(features & NETIF_F_ALL_CSUM)) + features &= ~NETIF_F_SG; + + /* + * features will include NETIF_F_TSO (NETIF_F_UFO) iff all + * slave devices support NETIF_F_TSO (NETIF_F_UFO), which + * implies that all slaves also support scatter-gather + * (NETIF_F_SG), which implies that features also includes + * NETIF_F_SG. So no need to check whether we have an + * illegal combination of NETIF_F_{TSO,UFO} and + * !NETIF_F_SG + */ + + features |= (bond_dev->features & ~BOND_INTERSECT_FEATURES); + bond_dev->features = features; + bond_dev->hard_header_len = max_hard_header_len; + + return 0; +} + /* enslave device to bond device */ -static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev) +int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev) { struct bonding *bond = bond_dev->priv; struct slave *new_slave = NULL; @@ -1591,10 +1263,11 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de int old_features = bond_dev->features; int res = 0; - if (slave_dev->do_ioctl == NULL) { + if (!bond->params.use_carrier && slave_dev->ethtool_ops == NULL && + slave_dev->do_ioctl == NULL) { printk(KERN_WARNING DRV_NAME - ": Warning : no link monitoring support for %s\n", - slave_dev->name); + ": %s: Warning: no link monitoring support for %s\n", + bond_dev->name, slave_dev->name); } /* bond must be initialized by bond_open() before enslaving */ @@ -1615,17 +1288,17 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de 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 " + ": %s: Error: cannot enslave VLAN " "challenged slave %s on VLAN enabled " - "bond %s\n", slave_dev->name, + "bond %s\n", bond_dev->name, slave_dev->name, bond_dev->name); return -EPERM; } else { printk(KERN_WARNING DRV_NAME - ": Warning: enslaved VLAN challenged " + ": %s: 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, slave_dev->name, slave_dev->name, bond_dev->name); bond_dev->features |= NETIF_F_VLAN_CHALLENGED; } @@ -1639,51 +1312,35 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de } } - 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"); + /* + * Old ifenslave binaries are no longer supported. These can + * be identified with moderate accurary by the state of the slave: + * the current ifenslave will set the interface down prior to + * enslaving it; the old ifenslave will not. + */ + if ((slave_dev->flags & IFF_UP)) { + printk(KERN_ERR DRV_NAME ": %s is up. " + "This may be due to an out of date ifenslave.\n", + slave_dev->name); + res = -EPERM; + goto err_undo_flags; + } - 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 (slave_dev->set_mac_address == NULL) { + printk(KERN_ERR DRV_NAME + ": %s: Error: The slave device you specified does " + "not support setting the MAC address. " + "Your kernel likely does not support slave " + "devices.\n", bond_dev->name); + res = -EOPNOTSUPP; + 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; - } + if (slave_dev->get_stats == NULL) { + printk(KERN_NOTICE DRV_NAME + ": %s: the driver for slave device %s does not provide " + "get_stats function, network statistics will be " + "inaccurate.\n", bond_dev->name, slave_dev->name); } new_slave = kmalloc(sizeof(struct slave), GFP_KERNEL); @@ -1699,44 +1356,40 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de */ 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); + /* + * Save slave's original ("permanent") mac address for modes + * that need 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 = slave_dev->set_mac_address(slave_dev, &addr); - if (res) { - dprintk("Error %d calling set_mac_address\n", res); - goto err_free; - } + /* + * 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; - } + /* 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; - } + goto err_close; } new_slave->dev = slave_dev; + slave_dev->priv_flags |= IFF_BONDING; if ((bond->params.mode == BOND_MODE_TLB) || (bond->params.mode == BOND_MODE_ALB)) { @@ -1787,6 +1440,10 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de new_slave->delay = 0; new_slave->link_failure_count = 0; + bond_compute_features(bond); + + new_slave->last_arp_rx = jiffies; + if (bond->params.miimon && !bond->params.use_carrier) { link_reporting = bond_check_dev_link(bond, slave_dev, 1); @@ -1801,21 +1458,21 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de * the messages for netif_carrier. */ printk(KERN_WARNING DRV_NAME - ": Warning: MII and ETHTOOL support not " + ": %s: 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); + bond_dev->name, 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 " + ": %s: 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); + bond_dev->name, slave_dev->name); } } @@ -1842,15 +1499,15 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de if (bond_update_speed_duplex(new_slave) && (new_slave->link != BOND_LINK_DOWN)) { printk(KERN_WARNING DRV_NAME - ": Warning: failed to get speed/duplex from %s, speed " - "forced to 100Mbps, duplex forced to Full.\n", - new_slave->dev->name); + ": %s: Warning: failed to get speed and duplex from %s, " + "assumed to be 100Mb/sec and Full.\n", + bond_dev->name, new_slave->dev->name); if (bond->params.mode == BOND_MODE_8023AD) { - printk(KERN_WARNING - "Operation of 802.3ad mode requires ETHTOOL " + printk(KERN_WARNING DRV_NAME + ": %s: Warning: Operation of 802.3ad mode requires ETHTOOL " "support in base driver for proper aggregator " - "selection.\n"); + "selection.\n", bond_dev->name); } } @@ -1863,24 +1520,8 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de 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); - } + bond_set_slave_inactive_flags(new_slave); + bond_select_active_slave(bond); break; case BOND_MODE_8023AD: /* in 802.3ad mode, the internal mechanism @@ -1912,6 +1553,8 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de * is OK, so make this interface the active one */ bond_change_active_slave(bond, new_slave); + } else { + bond_set_slave_inactive_flags(new_slave); } break; default: @@ -1930,40 +1573,13 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de break; } /* switch(bond_mode) */ - write_unlock_bh(&bond->lock); + bond_set_carrier(bond); - 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; - } - } + write_unlock_bh(&bond->lock); - 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); - } - } + res = bond_create_slave_symlinks(bond_dev, slave_dev); + if (res) + goto err_unset_master; printk(KERN_INFO DRV_NAME ": %s: enslaving %s as a%s interface with a%s link.\n", @@ -1984,14 +1600,14 @@ err_close: err_restore_mac: memcpy(addr.sa_data, new_slave->perm_hwaddr, ETH_ALEN); addr.sa_family = slave_dev->type; - slave_dev->set_mac_address(slave_dev, &addr); + dev_set_mac_address(slave_dev, &addr); err_free: kfree(new_slave); err_undo_flags: bond_dev->features = old_features; - + return res; } @@ -2006,7 +1622,7 @@ err_undo_flags: * 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) +int bond_release(struct net_device *bond_dev, struct net_device *slave_dev) { struct bonding *bond = bond_dev->priv; struct slave *slave, *oldcurrent; @@ -2017,7 +1633,7 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de if (!(slave_dev->flags & IFF_SLAVE) || (slave_dev->master != bond_dev)) { printk(KERN_ERR DRV_NAME - ": Error: %s: cannot release %s.\n", + ": %s: Error: cannot release %s.\n", bond_dev->name, slave_dev->name); return -EINVAL; } @@ -2030,6 +1646,7 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de printk(KERN_INFO DRV_NAME ": %s: %s not enslaved\n", bond_dev->name, slave_dev->name); + write_unlock_bh(&bond->lock); return -EINVAL; } @@ -2038,11 +1655,12 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de ETH_ALEN); if (!mac_addr_differ && (bond->slave_cnt > 1)) { printk(KERN_WARNING DRV_NAME - ": Warning: the permanent HWaddr of %s " + ": %s: 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", + bond_dev->name, slave_dev->name, slave->perm_hwaddr[0], slave->perm_hwaddr[1], @@ -2076,6 +1694,8 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de /* release the slave from its bond */ bond_detach_slave(bond, slave); + bond_compute_features(bond); + if (bond->primary_slave == slave) { bond->primary_slave = NULL; } @@ -2094,18 +1714,12 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de bond_alb_deinit_slave(bond, slave); } - if (oldcurrent == 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) { + bond_set_carrier(bond); + /* 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 @@ -2116,24 +1730,28 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de bond_dev->features |= NETIF_F_VLAN_CHALLENGED; } else { printk(KERN_WARNING DRV_NAME - ": Warning: clearing HW address of %s while it " + ": %s: Warning: clearing HW address of %s while it " "still has VLANs.\n", - bond_dev->name); + bond_dev->name, bond_dev->name); printk(KERN_WARNING DRV_NAME - ": When re-adding slaves, make sure the bond's " - "HW address matches its VLANs'.\n"); + ": %s: When re-adding slaves, make sure the bond's " + "HW address matches its VLANs'.\n", + bond_dev->name); } } else if ((bond_dev->features & NETIF_F_VLAN_CHALLENGED) && !bond_has_challenged_slaves(bond)) { printk(KERN_INFO DRV_NAME - ": last VLAN challenged slave %s " + ": %s: last VLAN challenged slave %s " "left bond %s. VLAN blocking is removed\n", - slave_dev->name, bond_dev->name); + bond_dev->name, slave_dev->name, bond_dev->name); bond_dev->features &= ~NETIF_F_VLAN_CHALLENGED; } write_unlock_bh(&bond->lock); + /* must do this from outside any spinlocks */ + bond_destroy_slave_symlinks(bond_dev, slave_dev); + bond_del_vlans_from_slave(bond, slave_dev); /* If the mode USES_PRIMARY, then we should only remove its @@ -2160,20 +1778,14 @@ static int bond_release(struct net_device *bond_dev, struct net_device *slave_de /* 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; - slave_dev->set_mac_address(slave_dev, &addr); - } + /* 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; - } + slave_dev->priv_flags &= ~(IFF_MASTER_8023AD | IFF_MASTER_ALB | + IFF_SLAVE_INACTIVE | IFF_BONDING | + IFF_SLAVE_NEEDARP); kfree(slave); @@ -2192,6 +1804,8 @@ static int bond_release_all(struct net_device *bond_dev) write_lock_bh(&bond->lock); + netif_carrier_off(bond_dev); + if (bond->slave_cnt == 0) { goto out; } @@ -2219,12 +1833,15 @@ static int bond_release_all(struct net_device *bond_dev) bond_alb_deinit_slave(bond, slave); } + bond_compute_features(bond); + /* 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_destroy_slave_symlinks(bond_dev, slave_dev); bond_del_vlans_from_slave(bond, slave_dev); /* If the mode USES_PRIMARY, then we should only remove its @@ -2251,19 +1868,13 @@ static int bond_release_all(struct net_device *bond_dev) /* 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; - slave_dev->set_mac_address(slave_dev, &addr); - } + /* 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; - } + slave_dev->priv_flags &= ~(IFF_MASTER_8023AD | IFF_MASTER_ALB | + IFF_SLAVE_INACTIVE); kfree(slave); @@ -2281,12 +1892,13 @@ static int bond_release_all(struct net_device *bond_dev) bond_dev->features |= NETIF_F_VLAN_CHALLENGED; } else { printk(KERN_WARNING DRV_NAME - ": Warning: clearing HW address of %s while it " + ": %s: Warning: clearing HW address of %s while it " "still has VLANs.\n", - bond_dev->name); + bond_dev->name, bond_dev->name); printk(KERN_WARNING DRV_NAME - ": When re-adding slaves, make sure the bond's " - "HW address matches its VLANs'.\n"); + ": %s: When re-adding slaves, make sure the bond's " + "HW address matches its VLANs'.\n", + bond_dev->name); } printk(KERN_INFO DRV_NAME @@ -2354,57 +1966,6 @@ static int bond_ioctl_change_active(struct net_device *bond_dev, struct net_devi return res; } -static int bond_ethtool_ioctl(struct net_device *bond_dev, struct ifreq *ifr) -{ - struct ethtool_drvinfo info; - void *addr = ifr->ifr_data; - uint32_t cmd; - - if (get_user(cmd, (uint32_t *)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; @@ -2455,7 +2016,7 @@ static int bond_slave_info_query(struct net_device *bond_dev, struct ifslave *in /*-------------------------------- Monitoring -------------------------------*/ /* this function is called regularly to monitor each slave's link. */ -static void bond_mii_monitor(struct net_device *bond_dev) +void bond_mii_monitor(struct net_device *bond_dev) { struct bonding *bond = bond_dev->priv; struct slave *slave, *oldcurrent; @@ -2654,8 +2215,11 @@ static void bond_mii_monitor(struct net_device *bond_dev) break; default: /* Should not happen */ - printk(KERN_ERR "bonding: Error: %s Illegal value (link=%d)\n", - slave->dev->name, slave->link); + printk(KERN_ERR DRV_NAME + ": %s: Error: %s Illegal value (link=%d)\n", + bond_dev->name, + slave->dev->name, + slave->link); goto out; } /* end of switch (slave->link) */ @@ -2666,48 +2230,318 @@ static void bond_mii_monitor(struct net_device *bond_dev) bond_3ad_adapter_speed_changed(slave); } - if (old_duplex != slave->duplex) { - bond_3ad_adapter_duplex_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); + + write_unlock(&bond->curr_slave_lock); + } else + bond_set_carrier(bond); + +re_arm: + if (bond->params.miimon) { + mod_timer(&bond->mii_timer, jiffies + delta_in_ticks); + } +out: + read_unlock(&bond->lock); +} + + +static u32 bond_glean_dev_ip(struct net_device *dev) +{ + struct in_device *idev; + struct in_ifaddr *ifa; + __be32 addr = 0; + + if (!dev) + return 0; + + rcu_read_lock(); + idev = __in_dev_get_rcu(dev); + if (!idev) + goto out; + + ifa = idev->ifa_list; + if (!ifa) + goto out; + + addr = ifa->ifa_local; +out: + rcu_read_unlock(); + return addr; +} + +static int bond_has_ip(struct bonding *bond) +{ + struct vlan_entry *vlan, *vlan_next; + + if (bond->master_ip) + return 1; + + if (list_empty(&bond->vlan_list)) + return 0; + + list_for_each_entry_safe(vlan, vlan_next, &bond->vlan_list, + vlan_list) { + if (vlan->vlan_ip) + return 1; + } + + return 0; +} + +static int bond_has_this_ip(struct bonding *bond, u32 ip) +{ + struct vlan_entry *vlan, *vlan_next; + + if (ip == bond->master_ip) + return 1; + + if (list_empty(&bond->vlan_list)) + return 0; + + list_for_each_entry_safe(vlan, vlan_next, &bond->vlan_list, + vlan_list) { + if (ip == vlan->vlan_ip) + return 1; + } + + return 0; +} + +/* + * We go to the (large) trouble of VLAN tagging ARP frames because + * switches in VLAN mode (especially if ports are configured as + * "native" to a VLAN) might not pass non-tagged frames. + */ +static void bond_arp_send(struct net_device *slave_dev, int arp_op, u32 dest_ip, u32 src_ip, unsigned short vlan_id) +{ + struct sk_buff *skb; + + dprintk("arp %d on slave %s: dst %x src %x vid %d\n", arp_op, + slave_dev->name, dest_ip, src_ip, vlan_id); + + skb = arp_create(arp_op, ETH_P_ARP, dest_ip, slave_dev, src_ip, + NULL, slave_dev->dev_addr, NULL); + + if (!skb) { + printk(KERN_ERR DRV_NAME ": ARP packet allocation failed\n"); + return; + } + if (vlan_id) { + skb = vlan_put_tag(skb, vlan_id); + if (!skb) { + printk(KERN_ERR DRV_NAME ": failed to insert VLAN tag\n"); + return; + } + } + arp_xmit(skb); +} + + +static void bond_arp_send_all(struct bonding *bond, struct slave *slave) +{ + int i, vlan_id, rv; + u32 *targets = bond->params.arp_targets; + struct vlan_entry *vlan, *vlan_next; + struct net_device *vlan_dev; + struct flowi fl; + struct rtable *rt; + + for (i = 0; (i < BOND_MAX_ARP_TARGETS); i++) { + if (!targets[i]) + continue; + dprintk("basa: target %x\n", targets[i]); + if (list_empty(&bond->vlan_list)) { + dprintk("basa: empty vlan: arp_send\n"); + bond_arp_send(slave->dev, ARPOP_REQUEST, targets[i], + bond->master_ip, 0); + continue; + } + + /* + * If VLANs are configured, we do a route lookup to + * determine which VLAN interface would be used, so we + * can tag the ARP with the proper VLAN tag. + */ + memset(&fl, 0, sizeof(fl)); + fl.fl4_dst = targets[i]; + fl.fl4_tos = RTO_ONLINK; + + rv = ip_route_output_key(&rt, &fl); + if (rv) { + if (net_ratelimit()) { + printk(KERN_WARNING DRV_NAME + ": %s: no route to arp_ip_target %u.%u.%u.%u\n", + bond->dev->name, NIPQUAD(fl.fl4_dst)); + } + continue; + } + + /* + * This target is not on a VLAN + */ + if (rt->u.dst.dev == bond->dev) { + ip_rt_put(rt); + dprintk("basa: rtdev == bond->dev: arp_send\n"); + bond_arp_send(slave->dev, ARPOP_REQUEST, targets[i], + bond->master_ip, 0); + continue; + } + + vlan_id = 0; + list_for_each_entry_safe(vlan, vlan_next, &bond->vlan_list, + vlan_list) { + vlan_dev = bond->vlgrp->vlan_devices[vlan->vlan_id]; + if (vlan_dev == rt->u.dst.dev) { + vlan_id = vlan->vlan_id; + dprintk("basa: vlan match on %s %d\n", + vlan_dev->name, vlan_id); + break; } } - } /* end of for */ + if (vlan_id) { + ip_rt_put(rt); + bond_arp_send(slave->dev, ARPOP_REQUEST, targets[i], + vlan->vlan_ip, vlan_id); + continue; + } - if (do_failover) { - write_lock(&bond->curr_slave_lock); + if (net_ratelimit()) { + printk(KERN_WARNING DRV_NAME + ": %s: no path to arp_ip_target %u.%u.%u.%u via rt.dev %s\n", + bond->dev->name, NIPQUAD(fl.fl4_dst), + rt->u.dst.dev ? rt->u.dst.dev->name : "NULL"); + } + ip_rt_put(rt); + } +} - bond_select_active_slave(bond); +/* + * Kick out a gratuitous ARP for an IP on the bonding master plus one + * for each VLAN above us. + */ +static void bond_send_gratuitous_arp(struct bonding *bond) +{ + struct slave *slave = bond->curr_active_slave; + struct vlan_entry *vlan; + struct net_device *vlan_dev; - if (oldcurrent && !bond->curr_active_slave) { - printk(KERN_INFO DRV_NAME - ": %s: now running without any active " - "interface !\n", - bond_dev->name); - } + dprintk("bond_send_grat_arp: bond %s slave %s\n", bond->dev->name, + slave ? slave->dev->name : "NULL"); + if (!slave) + return; - write_unlock(&bond->curr_slave_lock); + if (bond->master_ip) { + bond_arp_send(slave->dev, ARPOP_REPLY, bond->master_ip, + bond->master_ip, 0); } -re_arm: - if (bond->params.miimon) { - mod_timer(&bond->mii_timer, jiffies + delta_in_ticks); + list_for_each_entry(vlan, &bond->vlan_list, vlan_list) { + vlan_dev = bond->vlgrp->vlan_devices[vlan->vlan_id]; + if (vlan->vlan_ip) { + bond_arp_send(slave->dev, ARPOP_REPLY, vlan->vlan_ip, + vlan->vlan_ip, vlan->vlan_id); + } } -out: - read_unlock(&bond->lock); } -static void bond_arp_send_all(struct bonding *bond, struct slave *slave) +static void bond_validate_arp(struct bonding *bond, struct slave *slave, u32 sip, u32 tip) { int i; u32 *targets = bond->params.arp_targets; + 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); + dprintk("bva: sip %u.%u.%u.%u tip %u.%u.%u.%u t[%d] " + "%u.%u.%u.%u bhti(tip) %d\n", + NIPQUAD(sip), NIPQUAD(tip), i, NIPQUAD(targets[i]), + bond_has_this_ip(bond, tip)); + if (sip == targets[i]) { + if (bond_has_this_ip(bond, tip)) + slave->last_arp_rx = jiffies; + return; + } } } +static int bond_arp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_type *pt, struct net_device *orig_dev) +{ + struct arphdr *arp; + struct slave *slave; + struct bonding *bond; + unsigned char *arp_ptr; + u32 sip, tip; + + if (!(dev->priv_flags & IFF_BONDING) || !(dev->flags & IFF_MASTER)) + goto out; + + bond = dev->priv; + read_lock(&bond->lock); + + dprintk("bond_arp_rcv: bond %s skb->dev %s orig_dev %s\n", + bond->dev->name, skb->dev ? skb->dev->name : "NULL", + orig_dev ? orig_dev->name : "NULL"); + + slave = bond_get_slave_by_dev(bond, orig_dev); + if (!slave || !slave_do_arp_validate(bond, slave)) + goto out_unlock; + + /* ARP header, plus 2 device addresses, plus 2 IP addresses. */ + if (!pskb_may_pull(skb, (sizeof(struct arphdr) + + (2 * dev->addr_len) + + (2 * sizeof(u32))))) + goto out_unlock; + + arp = skb->nh.arph; + if (arp->ar_hln != dev->addr_len || + skb->pkt_type == PACKET_OTHERHOST || + skb->pkt_type == PACKET_LOOPBACK || + arp->ar_hrd != htons(ARPHRD_ETHER) || + arp->ar_pro != htons(ETH_P_IP) || + arp->ar_pln != 4) + goto out_unlock; + + arp_ptr = (unsigned char *)(arp + 1); + arp_ptr += dev->addr_len; + memcpy(&sip, arp_ptr, 4); + arp_ptr += 4 + dev->addr_len; + memcpy(&tip, arp_ptr, 4); + + dprintk("bond_arp_rcv: %s %s/%d av %d sv %d sip %u.%u.%u.%u" + " tip %u.%u.%u.%u\n", bond->dev->name, slave->dev->name, + slave->state, bond->params.arp_validate, + slave_do_arp_validate(bond, slave), NIPQUAD(sip), NIPQUAD(tip)); + + /* + * Backup slaves won't see the ARP reply, but do come through + * here for each ARP probe (so we swap the sip/tip to validate + * the probe). In a "redundant switch, common router" type of + * configuration, the ARP probe will (hopefully) travel from + * the active, through one switch, the router, then the other + * switch before reaching the backup. + */ + if (slave->state == BOND_STATE_ACTIVE) + bond_validate_arp(bond, slave, sip, tip); + else + bond_validate_arp(bond, slave, tip, sip); + +out_unlock: + read_unlock(&bond->lock); +out: + dev_kfree_skb(skb); + return NET_RX_SUCCESS; +} + /* * this function is called regularly to monitor each slave's link * ensuring that traffic is being sent and received when arp monitoring @@ -2715,7 +2549,7 @@ static void bond_arp_send_all(struct bonding *bond, struct slave *slave) * 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) +void bond_loadbalance_arp_mon(struct net_device *bond_dev) { struct bonding *bond = bond_dev->priv; struct slave *slave, *oldcurrent; @@ -2783,7 +2617,7 @@ static void bond_loadbalance_arp_mon(struct net_device *bond_dev) */ if (((jiffies - slave->dev->trans_start) >= (2*delta_in_ticks)) || (((jiffies - slave->dev->last_rx) >= (2*delta_in_ticks)) && - my_ip)) { + bond_has_ip(bond))) { slave->link = BOND_LINK_DOWN; slave->state = BOND_STATE_BACKUP; @@ -2820,13 +2654,6 @@ static void bond_loadbalance_arp_mon(struct net_device *bond_dev) 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); } @@ -2853,7 +2680,7 @@ out: * 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) +void bond_activebackup_arp_mon(struct net_device *bond_dev) { struct bonding *bond = bond_dev->priv; struct slave *slave; @@ -2879,7 +2706,8 @@ static void bond_activebackup_arp_mon(struct net_device *bond_dev) */ bond_for_each_slave(bond, slave, i) { if (slave->link != BOND_LINK_UP) { - if ((jiffies - slave->dev->last_rx) <= delta_in_ticks) { + if ((jiffies - slave_last_rx(bond, slave)) <= + delta_in_ticks) { slave->link = BOND_LINK_UP; @@ -2900,12 +2728,15 @@ static void bond_activebackup_arp_mon(struct net_device *bond_dev) bond->current_arp_slave = NULL; } + bond_set_carrier(bond); + 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); + netif_carrier_on(bond->dev); } else { printk(KERN_INFO DRV_NAME ": %s: backup interface %s is " @@ -2921,8 +2752,8 @@ static void bond_activebackup_arp_mon(struct net_device *bond_dev) if ((slave != bond->curr_active_slave) && (!bond->current_arp_slave) && - (((jiffies - slave->dev->last_rx) >= 3*delta_in_ticks) && - my_ip)) { + (((jiffies - slave_last_rx(bond, slave)) >= 3*delta_in_ticks) && + bond_has_ip(bond))) { /* a backup slave has gone down; three times * the delta allows the current slave to be * taken out before the backup slave. @@ -2968,8 +2799,8 @@ static void bond_activebackup_arp_mon(struct net_device *bond_dev) * 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_last_rx(bond, slave)) >= (2*delta_in_ticks)) && + bond_has_ip(bond))) && ((jiffies - slave->jiffies) >= 2*delta_in_ticks)) { slave->link = BOND_LINK_DOWN; @@ -3021,7 +2852,7 @@ static void bond_activebackup_arp_mon(struct net_device *bond_dev) /* the current slave must tx an arp to ensure backup slaves * rx traffic */ - if (slave && my_ip) { + if (slave && bond_has_ip(bond)) { bond_arp_send_all(bond, slave); } } @@ -3039,7 +2870,7 @@ static void bond_activebackup_arp_mon(struct net_device *bond_dev) 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) { + bond_for_each_slave_from(bond, slave, i, bond->current_arp_slave->next) { if (IS_UP(slave->dev)) { slave->link = BOND_LINK_BACK; bond_set_slave_active_flags(slave); @@ -3139,6 +2970,8 @@ static void bond_info_show_master(struct seq_file *seq) { struct bonding *bond = seq->private; struct slave *curr; + int i; + u32 target; read_lock(&bond->curr_slave_lock); curr = bond->curr_active_slave; @@ -3147,22 +2980,51 @@ static void bond_info_show_master(struct seq_file *seq) seq_printf(seq, "Bonding Mode: %s\n", bond_mode_name(bond->params.mode)); + if (bond->params.mode == BOND_MODE_XOR || + bond->params.mode == BOND_MODE_8023AD) { + seq_printf(seq, "Transmit Hash Policy: %s (%d)\n", + xmit_hashtype_tbl[bond->params.xmit_policy].modename, + bond->params.xmit_policy); + } + if (USES_PRIMARY(bond->params.mode)) { seq_printf(seq, "Primary Slave: %s\n", - (bond->params.primary[0]) ? - bond->params.primary : "None"); + (bond->primary_slave) ? + bond->primary_slave->dev->name : "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 Status: %s\n", netif_carrier_ok(bond->dev) ? + "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); + + /* ARP information */ + if(bond->params.arp_interval > 0) { + int printed=0; + seq_printf(seq, "ARP Polling Interval (ms): %d\n", + bond->params.arp_interval); + + seq_printf(seq, "ARP IP target/s (n.n.n.n form):"); + + for(i = 0; (i < BOND_MAX_ARP_TARGETS) ;i++) { + if (!bond->params.arp_targets[i]) + continue; + if (printed) + seq_printf(seq, ","); + target = ntohl(bond->params.arp_targets[i]); + seq_printf(seq, " %d.%d.%d.%d", HIPQUAD(target)); + printed = 1; + } + seq_printf(seq, "\n"); + } + if (bond->params.mode == BOND_MODE_8023AD) { struct ad_info ad_info; @@ -3202,19 +3064,14 @@ static void bond_info_show_slave(struct seq_file *seq, const struct slave *slave 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", + seq_printf(seq, "Link Failure Count: %u\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]); - } + 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 @@ -3373,7 +3230,10 @@ static int bond_event_changename(struct bonding *bond) bond_remove_proc_entry(bond); bond_create_proc_entry(bond); #endif - + down_write(&(bonding_rwsem)); + bond_destroy_sysfs_entry(bond); + bond_create_sysfs_entry(bond); + up_write(&(bonding_rwsem)); return NOTIFY_DONE; } @@ -3399,6 +3259,7 @@ static int bond_master_netdev_event(unsigned long event, struct net_device *bond static int bond_slave_netdev_event(unsigned long event, struct net_device *slave_dev) { struct net_device *bond_dev = slave_dev->master; + struct bonding *bond = bond_dev->priv; switch (event) { case NETDEV_UNREGISTER: @@ -3437,6 +3298,9 @@ static int bond_slave_netdev_event(unsigned long event, struct net_device *slave * TODO: handle changing the primary's name */ break; + case NETDEV_FEAT_CHANGE: + bond_compute_features(bond); + break; default: break; } @@ -3448,7 +3312,7 @@ static int bond_slave_netdev_event(unsigned long event, struct net_device *slave * 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 + * ioctl handler calling blocking_notifier_call_chain) holds the necessary * locks for us to safely manipulate the slave devices (RTNL lock, * dev_probe_lock). */ @@ -3460,6 +3324,9 @@ static int bond_netdev_event(struct notifier_block *this, unsigned long event, v (event_dev ? event_dev->name : "None"), event); + if (!(event_dev->priv_flags & IFF_BONDING)) + return NOTIFY_DONE; + if (event_dev->flags & IFF_MASTER) { dprintk("IFF_MASTER\n"); return bond_master_netdev_event(event, event_dev); @@ -3473,10 +3340,67 @@ static int bond_netdev_event(struct notifier_block *this, unsigned long event, v return NOTIFY_DONE; } +/* + * bond_inetaddr_event: handle inetaddr notifier chain events. + * + * We keep track of device IPs primarily to use as source addresses in + * ARP monitor probes (rather than spewing out broadcasts all the time). + * + * We track one IP for the main device (if it has one), plus one per VLAN. + */ +static int bond_inetaddr_event(struct notifier_block *this, unsigned long event, void *ptr) +{ + struct in_ifaddr *ifa = ptr; + struct net_device *vlan_dev, *event_dev = ifa->ifa_dev->dev; + struct bonding *bond, *bond_next; + struct vlan_entry *vlan, *vlan_next; + + list_for_each_entry_safe(bond, bond_next, &bond_dev_list, bond_list) { + if (bond->dev == event_dev) { + switch (event) { + case NETDEV_UP: + bond->master_ip = ifa->ifa_local; + return NOTIFY_OK; + case NETDEV_DOWN: + bond->master_ip = bond_glean_dev_ip(bond->dev); + return NOTIFY_OK; + default: + return NOTIFY_DONE; + } + } + + if (list_empty(&bond->vlan_list)) + continue; + + list_for_each_entry_safe(vlan, vlan_next, &bond->vlan_list, + vlan_list) { + vlan_dev = bond->vlgrp->vlan_devices[vlan->vlan_id]; + if (vlan_dev == event_dev) { + switch (event) { + case NETDEV_UP: + vlan->vlan_ip = ifa->ifa_local; + return NOTIFY_OK; + case NETDEV_DOWN: + vlan->vlan_ip = + bond_glean_dev_ip(vlan_dev); + return NOTIFY_OK; + default: + return NOTIFY_DONE; + } + } + } + } + return NOTIFY_DONE; +} + static struct notifier_block bond_netdev_notifier = { .notifier_call = bond_netdev_event, }; +static struct notifier_block bond_inetaddr_notifier = { + .notifier_call = bond_inetaddr_event, +}; + /*-------------------------- Packet type handling ---------------------------*/ /* register to receive lacpdus on a bond */ @@ -3498,6 +3422,61 @@ static void bond_unregister_lacpdu(struct bonding *bond) dev_remove_pack(&(BOND_AD_INFO(bond).ad_pkt_type)); } +void bond_register_arp(struct bonding *bond) +{ + struct packet_type *pt = &bond->arp_mon_pt; + + pt->type = htons(ETH_P_ARP); + pt->dev = NULL; /*bond->dev;XXX*/ + pt->func = bond_arp_rcv; + dev_add_pack(pt); +} + +void bond_unregister_arp(struct bonding *bond) +{ + dev_remove_pack(&bond->arp_mon_pt); +} + +/*---------------------------- Hashing Policies -----------------------------*/ + +/* + * Hash for the the output device based upon layer 3 and layer 4 data. If + * the packet is a frag or not TCP or UDP, just use layer 3 data. If it is + * altogether not IP, mimic bond_xmit_hash_policy_l2() + */ +static int bond_xmit_hash_policy_l34(struct sk_buff *skb, + struct net_device *bond_dev, int count) +{ + struct ethhdr *data = (struct ethhdr *)skb->data; + struct iphdr *iph = skb->nh.iph; + u16 *layer4hdr = (u16 *)((u32 *)iph + iph->ihl); + int layer4_xor = 0; + + if (skb->protocol == __constant_htons(ETH_P_IP)) { + if (!(iph->frag_off & __constant_htons(IP_MF|IP_OFFSET)) && + (iph->protocol == IPPROTO_TCP || + iph->protocol == IPPROTO_UDP)) { + layer4_xor = htons((*layer4hdr ^ *(layer4hdr + 1))); + } + return (layer4_xor ^ + ((ntohl(iph->saddr ^ iph->daddr)) & 0xffff)) % count; + + } + + return (data->h_dest[5] ^ bond_dev->dev_addr[5]) % count; +} + +/* + * Hash for the output device based upon layer 2 data + */ +static int bond_xmit_hash_policy_l2(struct sk_buff *skb, + struct net_device *bond_dev, int count) +{ + struct ethhdr *data = (struct ethhdr *)skb->data; + + return (data->h_dest[5] ^ bond_dev->dev_addr[5]) % count; +} + /*-------------------------- Device entry points ----------------------------*/ static int bond_open(struct net_device *bond_dev) @@ -3544,6 +3523,9 @@ static int bond_open(struct net_device *bond_dev) } else { arp_timer->function = (void *)&bond_loadbalance_arp_mon; } + if (bond->params.arp_validate) + bond_register_arp(bond); + add_timer(arp_timer); } @@ -3566,15 +3548,17 @@ static int bond_close(struct net_device *bond_dev) { struct bonding *bond = bond_dev->priv; - write_lock_bh(&bond->lock); - - bond_mc_list_destroy(bond); - if (bond->params.mode == BOND_MODE_8023AD) { /* Unregister the receive of LACPDUs */ bond_unregister_lacpdu(bond); } + if (bond->params.arp_validate) + bond_unregister_arp(bond); + + write_lock_bh(&bond->lock); + + /* signal timers not to re-arm */ bond->kill_timers = 1; @@ -3604,8 +3588,6 @@ static int bond_close(struct net_device *bond_dev) break; } - /* Release the bonded slaves */ - bond_release_all(bond_dev); if ((bond->params.mode == BOND_MODE_TLB) || (bond->params.mode == BOND_MODE_ALB)) { @@ -3630,33 +3612,35 @@ static struct net_device_stats *bond_get_stats(struct net_device *bond_dev) 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; + if (slave->dev->get_stats) { + 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); @@ -3667,20 +3651,19 @@ static struct net_device_stats *bond_get_stats(struct net_device *bond_dev) static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd) { struct net_device *slave_dev = NULL; - struct ifbond *u_binfo = NULL, k_binfo; - struct ifslave *u_sinfo = NULL, k_sinfo; + 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 = (struct mii_ioctl_data *)&ifr->ifr_data; + mii = if_mii(ifr); if (!mii) { return -EINVAL; } @@ -3691,7 +3674,7 @@ static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd * We do this again just in case we were called by SIOCGMIIREG * instead of SIOCGMIIPHY. */ - mii = (struct mii_ioctl_data *)&ifr->ifr_data; + mii = if_mii(ifr); if (!mii) { return -EINVAL; } @@ -3701,7 +3684,7 @@ static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd mii->val_out = 0; read_lock_bh(&bond->lock); read_lock(&bond->curr_slave_lock); - if (bond->curr_active_slave) { + if (netif_carrier_ok(bond->dev)) { mii->val_out = BMSR_LSTATUS; } read_unlock(&bond->curr_slave_lock); @@ -3711,7 +3694,7 @@ static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd return 0; case BOND_INFO_QUERY_OLD: case SIOCBONDINFOQUERY: - u_binfo = (struct ifbond *)ifr->ifr_data; + u_binfo = (struct ifbond __user *)ifr->ifr_data; if (copy_from_user(&k_binfo, u_binfo, sizeof(ifbond))) { return -EFAULT; @@ -3727,7 +3710,7 @@ static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd return res; case BOND_SLAVE_INFO_QUERY_OLD: case SIOCBONDSLAVEINFOQUERY: - u_sinfo = (struct ifslave *)ifr->ifr_data; + u_sinfo = (struct ifslave __user *)ifr->ifr_data; if (copy_from_user(&k_sinfo, u_sinfo, sizeof(ifslave))) { return -EFAULT; @@ -3750,21 +3733,7 @@ static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd 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; - } - + down_write(&(bonding_rwsem)); slave_dev = dev_get_by_name(ifr->ifr_slave); dprintk("slave_dev=%p: \n", slave_dev); @@ -3797,14 +3766,7 @@ static int bond_do_ioctl(struct net_device *bond_dev, struct ifreq *ifr, int cmd 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; - } - + up_write(&(bonding_rwsem)); return res; } @@ -3889,12 +3851,8 @@ static int bond_change_mtu(struct net_device *bond_dev, int new_mtu) bond_for_each_slave(bond, slave, i) { dprintk("s %p s->p %p c_m %p\n", slave, slave->prev, slave->dev->change_mtu); - if (slave->dev->change_mtu) { - res = slave->dev->change_mtu(slave->dev, new_mtu); - } else { - slave->dev->mtu = new_mtu; - res = 0; - } + + res = dev_set_mtu(slave->dev, new_mtu); if (res) { /* If we failed to set the slave's mtu to the new value @@ -3920,14 +3878,10 @@ unwind: bond_for_each_slave_from_to(bond, slave, i, bond->first_slave, stop_at) { int tmp_res; - if (slave->dev->change_mtu) { - tmp_res = slave->dev->change_mtu(slave->dev, bond_dev->mtu); - if (tmp_res) { - dprintk("unwind err %d dev %s\n", tmp_res, - slave->dev->name); - } - } else { - slave->dev->mtu = bond_dev->mtu; + 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); } } @@ -3979,7 +3933,7 @@ static int bond_set_mac_address(struct net_device *bond_dev, void *addr) goto unwind; } - res = slave->dev->set_mac_address(slave->dev, addr); + res = dev_set_mac_address(slave->dev, addr); if (res) { /* TODO: consider downing the slave * and retry ? @@ -4005,7 +3959,7 @@ unwind: bond_for_each_slave_from_to(bond, slave, i, bond->first_slave, stop_at) { int tmp_res; - tmp_res = slave->dev->set_mac_address(slave->dev, &tmp_sa); + 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); @@ -4060,6 +4014,43 @@ out: return 0; } +static void bond_activebackup_xmit_copy(struct sk_buff *skb, + struct bonding *bond, + struct slave *slave) +{ + struct sk_buff *skb2 = skb_copy(skb, GFP_ATOMIC); + struct ethhdr *eth_data; + u8 *hwaddr; + int res; + + if (!skb2) { + printk(KERN_ERR DRV_NAME ": Error: " + "bond_activebackup_xmit_copy(): skb_copy() failed\n"); + return; + } + + skb2->mac.raw = (unsigned char *)skb2->data; + eth_data = eth_hdr(skb2); + + /* Pick an appropriate source MAC address + * -- use slave's perm MAC addr, unless used by bond + * -- otherwise, borrow active slave's perm MAC addr + * since that will not be used + */ + hwaddr = slave->perm_hwaddr; + if (!memcmp(eth_data->h_source, hwaddr, ETH_ALEN)) + hwaddr = bond->curr_active_slave->perm_hwaddr; + + /* Set source MAC address appropriately */ + memcpy(eth_data->h_source, hwaddr, ETH_ALEN); + + res = bond_dev_queue_xmit(bond, skb2, slave->dev); + if (res) + dev_kfree_skb(skb2); + + return; +} + /* * in active-backup mode, we know that bond->curr_active_slave is always valid if * the bond has a usable interface. @@ -4069,17 +4060,6 @@ static int bond_xmit_activebackup(struct sk_buff *skb, struct net_device *bond_d 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); @@ -4087,10 +4067,26 @@ static int bond_xmit_activebackup(struct sk_buff *skb, struct net_device *bond_d goto out; } - if (bond->curr_active_slave) { /* one usable interface */ - res = bond_dev_queue_xmit(bond, skb, bond->curr_active_slave->dev); + if (!bond->curr_active_slave) + goto out; + + /* Xmit IGMP frames on all slaves to ensure rapid fail-over + for multicast traffic on snooping switches */ + if (skb->protocol == __constant_htons(ETH_P_IP) && + skb->nh.iph->protocol == IPPROTO_IGMP) { + struct slave *slave, *active_slave; + int i; + + active_slave = bond->curr_active_slave; + bond_for_each_slave_from_to(bond, slave, i, active_slave->next, + active_slave->prev) + if (IS_UP(slave->dev) && + (slave->link == BOND_LINK_UP)) + bond_activebackup_xmit_copy(skb, bond, slave); } + res = bond_dev_queue_xmit(bond, skb, bond->curr_active_slave->dev); + out: if (res) { /* no suitable interface, frame not sent */ @@ -4102,14 +4098,13 @@ out: } /* - * 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. + * In bond_xmit_xor() , we determine the output device by using a pre- + * determined xmit_hash_policy(), If the selected device is not enabled, + * find the next active 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; @@ -4121,7 +4116,7 @@ static int bond_xmit_xor(struct sk_buff *skb, struct net_device *bond_dev) goto out; } - slave_no = (data->h_dest[5]^bond_dev->dev_addr[5]) % bond->slave_cnt; + slave_no = bond->xmit_hash_policy(skb, bond_dev, bond->slave_cnt); bond_for_each_slave(bond, slave, i) { slave_no--; @@ -4183,8 +4178,9 @@ static int bond_xmit_broadcast(struct sk_buff *skb, struct net_device *bond_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"); + ": %s: Error: bond_xmit_broadcast(): " + "skb_clone() failed\n", + bond_dev->name); continue; } @@ -4212,20 +4208,15 @@ out: return 0; } -#ifdef CONFIG_NET_FASTROUTE -static int bond_accept_fastpath(struct net_device *bond_dev, struct dst_entry *dst) -{ - return -1; -} -#endif - /*------------------------- Device initialization ---------------------------*/ /* * set bond mode specific net device operations */ -static inline void bond_set_mode_ops(struct net_device *bond_dev, int mode) +void bond_set_mode_ops(struct bonding *bond, int mode) { + struct net_device *bond_dev = bond->dev; + switch (mode) { case BOND_MODE_ROUNDROBIN: bond_dev->hard_start_xmit = bond_xmit_roundrobin; @@ -4235,32 +4226,60 @@ static inline void bond_set_mode_ops(struct net_device *bond_dev, int mode) break; case BOND_MODE_XOR: bond_dev->hard_start_xmit = bond_xmit_xor; + if (bond->params.xmit_policy == BOND_XMIT_POLICY_LAYER34) + bond->xmit_hash_policy = bond_xmit_hash_policy_l34; + else + bond->xmit_hash_policy = bond_xmit_hash_policy_l2; break; case BOND_MODE_BROADCAST: bond_dev->hard_start_xmit = bond_xmit_broadcast; break; case BOND_MODE_8023AD: + bond_set_master_3ad_flags(bond); bond_dev->hard_start_xmit = bond_3ad_xmit_xor; + if (bond->params.xmit_policy == BOND_XMIT_POLICY_LAYER34) + bond->xmit_hash_policy = bond_xmit_hash_policy_l34; + else + bond->xmit_hash_policy = bond_xmit_hash_policy_l2; break; - case BOND_MODE_TLB: case BOND_MODE_ALB: + bond_set_master_alb_flags(bond); + /* FALLTHRU */ + case BOND_MODE_TLB: 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", + ": %s: Error: Unknown bonding mode %d\n", + bond_dev->name, mode); break; } } +static void bond_ethtool_get_drvinfo(struct net_device *bond_dev, + struct ethtool_drvinfo *drvinfo) +{ + strncpy(drvinfo->driver, DRV_NAME, 32); + strncpy(drvinfo->version, DRV_VERSION, 32); + snprintf(drvinfo->fw_version, 32, "%d", BOND_ABI_VERSION); +} + +static const struct ethtool_ops bond_ethtool_ops = { + .get_tx_csum = ethtool_op_get_tx_csum, + .get_tso = ethtool_op_get_tso, + .get_ufo = ethtool_op_get_ufo, + .get_sg = ethtool_op_get_sg, + .get_drvinfo = bond_ethtool_get_drvinfo, +}; + /* * 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) +static int bond_init(struct net_device *bond_dev, struct bond_params *params) { struct bonding *bond = bond_dev->priv; @@ -4285,20 +4304,19 @@ static int __init bond_init(struct net_device *bond_dev, struct bond_params *par bond_dev->stop = bond_close; bond_dev->get_stats = bond_get_stats; bond_dev->do_ioctl = bond_do_ioctl; + bond_dev->ethtool_ops = &bond_ethtool_ops; 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_set_mode_ops(bond, bond->params.mode); bond_dev->destructor = free_netdev; -#ifdef CONFIG_NET_FASTROUTE - bond_dev->accept_fastpath = bond_accept_fastpath; -#endif /* Initialize the device options */ bond_dev->tx_queue_len = 0; bond_dev->flags |= IFF_MASTER|IFF_MULTICAST; + bond_dev->priv_flags |= IFF_BONDING; /* At first, we block adding VLANs. That's the only way to * prevent problems that occur when adding VLANs over an @@ -4307,6 +4325,10 @@ static int __init bond_init(struct net_device *bond_dev, struct bond_params *par */ bond_dev->features |= NETIF_F_VLAN_CHALLENGED; + /* don't acquire bond device's netif_tx_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 @@ -4332,7 +4354,7 @@ static int __init bond_init(struct net_device *bond_dev, struct bond_params *par /* De-initialize device specific data. * Caller must hold rtnl_lock. */ -static inline void bond_deinit(struct net_device *bond_dev) +void bond_deinit(struct net_device *bond_dev) { struct bonding *bond = bond_dev->priv; @@ -4353,6 +4375,9 @@ static void bond_free_all(void) list_for_each_entry_safe(bond, nxt, &bond_dev_list, bond_list) { struct net_device *bond_dev = bond->dev; + bond_mc_list_destroy(bond); + /* Release the bonded slaves */ + bond_release_all(bond_dev); unregister_netdevice(bond_dev); bond_deinit(bond_dev); } @@ -4368,7 +4393,7 @@ static void bond_free_all(void) * 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 bond_parse_parm(char *mode_arg, struct bond_parm_tbl *tbl) { int i; @@ -4386,6 +4411,8 @@ static inline int bond_parse_parm(char *mode_arg, struct bond_parm_tbl *tbl) static int bond_check_params(struct bond_params *params) { + int arp_validate_value; + /* * Convert string parameters. */ @@ -4399,6 +4426,25 @@ static int bond_check_params(struct bond_params *params) } } + if (xmit_hash_policy) { + if ((bond_mode != BOND_MODE_XOR) && + (bond_mode != BOND_MODE_8023AD)) { + printk(KERN_INFO DRV_NAME + ": xor_mode param is irrelevant in mode %s\n", + bond_mode_name(bond_mode)); + } else { + xmit_hashtype = bond_parse_parm(xmit_hash_policy, + xmit_hashtype_tbl); + if (xmit_hashtype == -1) { + printk(KERN_ERR DRV_NAME + ": Error: Invalid xmit_hash_policy \"%s\"\n", + xmit_hash_policy == NULL ? "NULL" : + xmit_hash_policy); + return -EINVAL; + } + } + } + if (lacp_rate) { if (bond_mode != BOND_MODE_8023AD) { printk(KERN_INFO DRV_NAME @@ -4418,7 +4464,7 @@ static int bond_check_params(struct bond_params *params) 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)", + "was reset to BOND_DEFAULT_MAX_BONDS (%d)\n", max_bonds, 1, INT_MAX, BOND_DEFAULT_MAX_BONDS); max_bonds = BOND_DEFAULT_MAX_BONDS; } @@ -4570,6 +4616,29 @@ static int bond_check_params(struct bond_params *params) arp_interval = 0; } + if (arp_validate) { + if (bond_mode != BOND_MODE_ACTIVEBACKUP) { + printk(KERN_ERR DRV_NAME + ": arp_validate only supported in active-backup mode\n"); + return -EINVAL; + } + if (!arp_interval) { + printk(KERN_ERR DRV_NAME + ": arp_validate requires arp_interval\n"); + return -EINVAL; + } + + arp_validate_value = bond_parse_parm(arp_validate, + arp_validate_tbl); + if (arp_validate_value == -1) { + printk(KERN_ERR DRV_NAME + ": Error: invalid arp_validate \"%s\"\n", + arp_validate == NULL ? "NULL" : arp_validate); + return -EINVAL; + } + } else + arp_validate_value = 0; + if (miimon) { printk(KERN_INFO DRV_NAME ": MII link monitoring set to %d ms\n", @@ -4578,8 +4647,10 @@ static int bond_check_params(struct bond_params *params) int i; printk(KERN_INFO DRV_NAME - ": ARP monitoring set to %d ms with %d target(s):", - arp_interval, arp_ip_count); + ": ARP monitoring set to %d ms, validate %s, with %d target(s):", + arp_interval, + arp_validate_tbl[arp_validate_value].modename, + arp_ip_count); for (i = 0; i < arp_ip_count; i++) printk (" %s", arp_ip_target[i]); @@ -4610,8 +4681,10 @@ static int bond_check_params(struct bond_params *params) /* fill params struct with the proper values */ params->mode = bond_mode; + params->xmit_policy = xmit_hashtype; params->miimon = miimon; params->arp_interval = arp_interval; + params->arp_validate = arp_validate_value; params->updelay = updelay; params->downdelay = downdelay; params->use_carrier = use_carrier; @@ -4628,86 +4701,120 @@ static int bond_check_params(struct bond_params *params) return 0; } +static struct lock_class_key bonding_netdev_xmit_lock_key; + +/* Create a new bond based on the specified name and bonding parameters. + * Caller must NOT hold rtnl_lock; we need to release it here before we + * set up our sysfs entries. + */ +int bond_create(char *name, struct bond_params *params, struct bonding **newbond) +{ + struct net_device *bond_dev; + int res; + + rtnl_lock(); + bond_dev = alloc_netdev(sizeof(struct bonding), name, ether_setup); + if (!bond_dev) { + printk(KERN_ERR DRV_NAME + ": %s: eek! can't alloc netdev!\n", + name); + res = -ENOMEM; + goto out_rtnl; + } + + /* 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, params); + if (res < 0) { + goto out_netdev; + } + + SET_MODULE_OWNER(bond_dev); + + res = register_netdevice(bond_dev); + if (res < 0) { + goto out_bond; + } + + lockdep_set_class(&bond_dev->_xmit_lock, &bonding_netdev_xmit_lock_key); + + if (newbond) + *newbond = bond_dev->priv; + + netif_carrier_off(bond_dev); + + rtnl_unlock(); /* allows sysfs registration of net device */ + res = bond_create_sysfs_entry(bond_dev->priv); + goto done; +out_bond: + bond_deinit(bond_dev); +out_netdev: + free_netdev(bond_dev); +out_rtnl: + rtnl_unlock(); +done: + return res; +} + static int __init bonding_init(void) { - struct bond_params params; int i; int res; + char new_bond_name[8]; /* Enough room for 999 bonds at init. */ printk(KERN_INFO "%s", version); - res = bond_check_params(¶ms); + res = bond_check_params(&bonding_defaults); if (res) { - return res; + goto out; } - 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; - } + sprintf(new_bond_name, "bond%d",i); + res = bond_create(new_bond_name,&bonding_defaults, NULL); + if (res) + goto err; } - rtnl_unlock(); - register_netdevice_notifier(&bond_netdev_notifier); + res = bond_create_sysfs(); + if (res) + goto err; - return 0; + register_netdevice_notifier(&bond_netdev_notifier); + register_inetaddr_notifier(&bond_inetaddr_notifier); -out_err: - /* free and unregister all bonds that were successfully added */ + goto out; +err: + rtnl_lock(); bond_free_all(); - + bond_destroy_sysfs(); rtnl_unlock(); - +out: return res; + } static void __exit bonding_exit(void) { unregister_netdevice_notifier(&bond_netdev_notifier); + unregister_inetaddr_notifier(&bond_inetaddr_notifier); rtnl_lock(); bond_free_all(); + bond_destroy_sysfs(); 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");