[PATCH] bonding: suppress duplicate packets

Originally submitted by Kenzo Iwami; his original description is:

The current bonding driver receives duplicate packets when broadcast/
multicast packets are sent by other devices or packets are flooded by the
switch. In this patch, new flags are added in priv_flags of net_device
structure to let the bonding driver discard duplicate packets in
dev.c:skb_bond().

	Modified by Jay Vosburgh to change a define name, update some
comments, rearrange the new skb_bond() for clarity, clear all bonding
priv_flags on slave release, and update the driver version.

Signed-off-by: Kenzo Iwami <k-iwami@cj.jp.nec.com>
Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>

drivers/net/bonding/bond_main.c

@@ -1040,6 +1040,10 @@ void bond_change_active_slave(struct bonding *bond, struct slave *new_active)
 	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;
 	}
@@ -1443,15 +1447,16 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev)
 
 	switch (bond->params.mode) {
 	case BOND_MODE_ACTIVEBACKUP:
-		/* if we're in active-backup mode, we need one and only one active
+		/* if we're in active-backup mode, we need one and
-		 * interface. The backup interfaces will have their NOARP flag set
+		 * only one active interface. The backup interfaces
-		 * because we need them to be completely deaf and not to respond to
+		 * will have their SLAVE_INACTIVE flag set because we
-		 * any ARP request on the network to avoid fooling a switch. Thus,
+		 * need them to be drop all packets. Thus, since we
-		 * since we guarantee that curr_active_slave always point to the last
+		 * guarantee that curr_active_slave always point to
-		 * usable interface, we just have to verify this interface's flag.
+		 * 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)) &&
+		     (bond->curr_active_slave->dev->priv_flags & IFF_SLAVE_INACTIVE)) &&
 		    (new_slave->link != BOND_LINK_DOWN)) {
 			dprintk("This is the first active slave\n");
 			/* first slave or no active slave yet, and this link
@@ -1492,6 +1497,8 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev)
 			 * 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:
@@ -1724,13 +1731,8 @@ int bond_release(struct net_device *bond_dev, struct net_device *slave_dev)
 	addr.sa_family = slave_dev->type;
 	dev_set_mac_address(slave_dev, &addr);
 
-	/* restore the original state of the
+	slave_dev->priv_flags &= ~(IFF_MASTER_8023AD | IFF_MASTER_ALB |
-	 * IFF_NOARP flag that might have been
+				   IFF_SLAVE_INACTIVE);
-	 * set by bond_set_slave_inactive_flags()
-	 */
-	if ((slave->original_flags & IFF_NOARP) == 0) {
-		slave_dev->flags &= ~IFF_NOARP;
-	}
 
 	kfree(slave);
 
@@ -1816,12 +1818,8 @@ static int bond_release_all(struct net_device *bond_dev)
 		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
+		slave_dev->priv_flags &= ~(IFF_MASTER_8023AD | IFF_MASTER_ALB |
-		 * been set by bond_set_slave_inactive_flags()
+					   IFF_SLAVE_INACTIVE);
-		 */
-		if ((slave->original_flags & IFF_NOARP) == 0) {
-			slave_dev->flags &= ~IFF_NOARP;
-		}
 
 		kfree(slave);
 
@@ -4061,14 +4059,17 @@ void bond_set_mode_ops(struct bonding *bond, int mode)
 		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;

drivers/net/bonding/bond_sysfs.c

@@ -424,6 +424,12 @@ static ssize_t bonding_store_mode(struct class_device *cd, const char *buf, size
 		ret = -EINVAL;
 		goto out;
 	} else {
+		if (bond->params.mode == BOND_MODE_8023AD)
+			bond_unset_master_3ad_flags(bond);
+
+		if (bond->params.mode == BOND_MODE_ALB)
+			bond_unset_master_alb_flags(bond);
+
 		bond->params.mode = new_value;
 		bond_set_mode_ops(bond, bond->params.mode);
 		printk(KERN_INFO DRV_NAME ": %s: setting mode to %s (%d).\n",

drivers/net/bonding/bonding.h

@@ -22,8 +22,8 @@
 #include "bond_3ad.h"
 #include "bond_alb.h"
 
-#define DRV_VERSION	"3.0.1"
+#define DRV_VERSION	"3.0.2"
-#define DRV_RELDATE	"January 9, 2006"
+#define DRV_RELDATE	"February 21, 2006"
 #define DRV_NAME	"bonding"
 #define DRV_DESCRIPTION	"Ethernet Channel Bonding Driver"
 
@@ -230,14 +230,37 @@ static inline struct bonding *bond_get_bond_by_slave(struct slave *slave)
 
 static inline void bond_set_slave_inactive_flags(struct slave *slave)
 {
-	slave->state = BOND_STATE_BACKUP;
+	struct bonding *bond = slave->dev->master->priv;
-	slave->dev->flags |= IFF_NOARP;
+	if (bond->params.mode != BOND_MODE_TLB &&
+	    bond->params.mode != BOND_MODE_ALB)
+		slave->state = BOND_STATE_BACKUP;
+	slave->dev->priv_flags |= IFF_SLAVE_INACTIVE;
 }
 
 static inline void bond_set_slave_active_flags(struct slave *slave)
 {
 	slave->state = BOND_STATE_ACTIVE;
-	slave->dev->flags &= ~IFF_NOARP;
+	slave->dev->priv_flags &= ~IFF_SLAVE_INACTIVE;
+}
+
+static inline void bond_set_master_3ad_flags(struct bonding *bond)
+{
+	bond->dev->priv_flags |= IFF_MASTER_8023AD;
+}
+
+static inline void bond_unset_master_3ad_flags(struct bonding *bond)
+{
+	bond->dev->priv_flags &= ~IFF_MASTER_8023AD;
+}
+
+static inline void bond_set_master_alb_flags(struct bonding *bond)
+{
+	bond->dev->priv_flags |= IFF_MASTER_ALB;
+}
+
+static inline void bond_unset_master_alb_flags(struct bonding *bond)
+{
+	bond->dev->priv_flags &= ~IFF_MASTER_ALB;
 }
 
 struct vlan_entry *bond_next_vlan(struct bonding *bond, struct vlan_entry *curr);

include/linux/if.h

@@ -52,6 +52,9 @@
 /* Private (from user) interface flags (netdevice->priv_flags). */
 #define IFF_802_1Q_VLAN 0x1             /* 802.1Q VLAN device.          */
 #define IFF_EBRIDGE	0x2		/* Ethernet bridging device.	*/
+#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 IF_GET_IFACE	0x0001		/* for querying only */
 #define IF_GET_PROTO	0x0002

include/linux/if_ether.h

@@ -61,6 +61,7 @@
 #define ETH_P_8021Q	0x8100          /* 802.1Q VLAN Extended Header  */
 #define ETH_P_IPX	0x8137		/* IPX over DIX			*/
 #define ETH_P_IPV6	0x86DD		/* IPv6 over bluebook		*/
+#define ETH_P_SLOW	0x8809		/* Slow Protocol. See 802.3ad 43B */
 #define ETH_P_WCCP	0x883E		/* Web-cache coordination protocol
 					 * defined in draft-wilson-wrec-wccp-v2-00.txt */
 #define ETH_P_PPP_DISC	0x8863		/* PPPoE discovery messages     */

net/core/dev.c

@@ -1446,8 +1446,29 @@ static inline struct net_device *skb_bond(struct sk_buff *skb)
 {
 	struct net_device *dev = skb->dev;
 
-	if (dev->master)
+	if (dev->master) {
+		/*
+		 * On bonding slaves other than the currently active
+		 * slave, suppress duplicates except for 802.3ad
+		 * ETH_P_SLOW and alb non-mcast/bcast.
+		 */
+		if (dev->priv_flags & IFF_SLAVE_INACTIVE) {
+			if (dev->master->priv_flags & IFF_MASTER_ALB) {
+				if (skb->pkt_type != PACKET_BROADCAST &&
+				    skb->pkt_type != PACKET_MULTICAST)
+					goto keep;
+			}
+
+			if (dev->master->priv_flags & IFF_MASTER_8023AD &&
+			    skb->protocol == __constant_htons(ETH_P_SLOW))
+				goto keep;
+		
+			kfree_skb(skb);
+			return NULL;
+		}
+keep:
 		skb->dev = dev->master;
+	}
 
 	return dev;
 }
@@ -1591,6 +1612,9 @@ int netif_receive_skb(struct sk_buff *skb)
 
 	orig_dev = skb_bond(skb);
 
+	if (!orig_dev)
+		return NET_RX_DROP;
+
 	__get_cpu_var(netdev_rx_stat).total++;
 
 	skb->h.raw = skb->nh.raw = skb->data;