Add priv_flag to specifically identify bonding-involved devices. Needed because IFF_MASTER is an unreliable identifier (vlan interfaces above bonding will inherit IFF_MASTER). Misidentification of devices would cause notifier events for other devices to be erroneously processed by bonding, causing various havoc.
Bug discovered by Martin Papik <[EMAIL PROTECTED]>; this patch is modified from his original. Signed-off-by: Martin Papik <[EMAIL PROTECTED]> Signed-off-by: Jay Vosburgh <[EMAIL PROTECTED]> --- netdev-2.6.git-upstream/include/linux/if.h 2006/08/26 17:23:42 1.1 +++ netdev-2.6.git-upstream/include/linux/if.h 2006/08/26 17:24:23 1.2 @@ -59,6 +59,7 @@ #define IFF_SLAVE_INACTIVE 0x4 /* bonding slave not the curr. active */ #define IFF_MASTER_8023AD 0x8 /* bonding master, 802.3ad. */ #define IFF_MASTER_ALB 0x10 /* bonding master, balance-alb. */ +#define IFF_BONDING 0x20 /* bonding master or slave */ #define IF_GET_IFACE 0x0001 /* for querying only */ #define IF_GET_PROTO 0x0002 --- netdev-2.6.git-upstream/drivers/net/bonding/bond_main.c 2006/08/19 15:47:27 1.4 +++ netdev-2.6.git-upstream/drivers/net/bonding/bond_main.c 2006/08/26 17:24:41 1.5 @@ -1371,6 +1371,7 @@ int bond_enslave(struct net_device *bond } new_slave->dev = slave_dev; + slave_dev->priv_flags |= IFF_BONDING; if ((bond->params.mode == BOND_MODE_TLB) || (bond->params.mode == BOND_MODE_ALB)) { @@ -1784,7 +1785,7 @@ int bond_release(struct net_device *bond dev_set_mac_address(slave_dev, &addr); slave_dev->priv_flags &= ~(IFF_MASTER_8023AD | IFF_MASTER_ALB | - IFF_SLAVE_INACTIVE); + IFF_SLAVE_INACTIVE | IFF_BONDING); kfree(slave); @@ -3216,6 +3217,9 @@ static int bond_netdev_event(struct noti (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); @@ -4185,6 +4189,7 @@ static int bond_init(struct net_device * /* 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 -- VGER BF report: U 0.499999 - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html