[RFC PATCH 0/2] bonding: generic netlink, multi-link mode

[RFC PATCH 0/2] bonding: generic netlink, multi-link mode

[Jay Vosburgh]

	These patches add support to bonding for generic netlink and a new
multi-link mode.  At the moment, I'm looking primarily for discussion
about the generic netlink and implementation of multi-link.

	First, in patch 1, is a generic netlink infrastructure for
bonding.  This patch provides a "get mode" command and a "slave link state
change" asychnronous notification via a netlink multicast group.  One long
term goal is to have bonding be controlled via netlink, both for
administrative purposes (add / remove slaves, etc) and policy (slave A is
better than slave B).  I'd appreciate feedback from netlink savvy folks as
to whether this is the appropriate starting point.

	Second, in patch 2, is the multi-link kernel code itself, which is
at present a work in progress.  Here, I'm primarily looking for comments
regarding the control interface for this mode.

	As implemented, this is a new mode to bonding, controlled via
generic netlink commands from a user space daemon.  Slave assignment for
outgoing traffic is handled directly by bonding (the mapping table used by
multi-link is within bonding itself, and the usual transmit hash policy is
applied to the set of slaves allowable for a given destination).

	In some private discussion with Andy, he suggested that this would
be better if it utilized the recently added queue mapping facility within
bonding, and then having the queue (and thus slave) assignments performed
at the qdisc level (via a tc filter) instead of within bonding itself.
This, I believe, would require a new tc filter that implements the ability
to set a skb queue_mapping in a hash (of protocol data in the packet) or
round robin fashion.  In this case, the tc filter would also incorporate
all of the netlink functionality for communicating with the user space
daemon (to permit the mappings to be updated).

	Thoughts?

	Lastly, a description of the multi-link system itself.  This is a
reimplementation of a load balancing scheme that has been available on AIX
for some time.  It operates essentially as a load balancer by subnet, with
a UDP-based protocol to exchange multi-link topology information between
participating systems.  Hosts participating in multi-link have IP
addresses in a separate subnet.  Interfaces enslaved to multi-link do not
lose their assigned IP address information, and may also operate
separately from multi-link.

	One notable feature is that multi-link provides load balancing
facilities for network devices that cannot change their MAC address, such
as Infiniband.

	For example, given two systems as follows:

host A:
bond0		10.88.0.1/16
slave eth0	10.0.0.1/16
slave eth1	10.1.0.1/16
slave eth2	10.2.0.1/16

host B:
bond0		10.88.0.2/16
slave eth0	10.0.0.2/16
slave eth1	10.1.0.2/16
slave eth2	10.2.0.2/16

	in this case, host A's bond0 running multi-link would load balance
traffic from 10.88.0.1 to 10.88.0.2 across eth0, eth1 and eth2.  The user
space daemon negotiates the link set to use with other participating
hosts, and communicates that to the multi-link implementation.

	-J

---
	-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com

[RFC PATCH 1/2] bonding: generic netlink infrastructure

[Jay Vosburgh]

	Generic netlink infrastructure for bonding.  Includes two
netlink operations: notification for slave link state change, and a
"get mode" netlink command.
---
 drivers/net/bonding/Makefile    |    2 +-
 drivers/net/bonding/bond_main.c |   39 +++++++++++++++++----------------------
 drivers/net/bonding/bonding.h   |    1 +
 include/linux/if_bonding.h      |   23 +++++++++++++++++++++++
 4 files changed, 42 insertions(+), 23 deletions(-)

diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile
index 6f9c6fa..26848a2 100644
--- a/drivers/net/bonding/Makefile
+++ b/drivers/net/bonding/Makefile
@@ -4,7 +4,7 @@
 
 obj-$(CONFIG_BONDING) += bonding.o
 
-bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o
+bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_netlink.o
 
 ipv6-$(subst m,y,$(CONFIG_IPV6)) += bond_ipv6.o
 bonding-objs += $(ipv6-y)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index bb33b3b..4d3a2c8 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -83,6 +83,7 @@
 #include "bonding.h"
 #include "bond_3ad.h"
 #include "bond_alb.h"
+#include "bond_netlink.h"
 
 /*---------------------------- Module parameters ----------------------------*/
 
@@ -2417,6 +2418,8 @@ static void bond_miimon_commit(struct bonding *bond)
 				bond_alb_handle_link_change(bond, slave,
 							    BOND_LINK_UP);
 
+			bond_nl_link_change(bond, slave, BOND_LINK_UP);
+
 			if (!bond->curr_active_slave ||
 			    (slave == bond->primary_slave))
 				goto do_failover;
@@ -2444,6 +2447,8 @@ static void bond_miimon_commit(struct bonding *bond)
 				bond_alb_handle_link_change(bond, slave,
 							    BOND_LINK_DOWN);
 
+			bond_nl_link_change(bond, slave, BOND_LINK_DOWN);
+
 			if (slave == bond->curr_active_slave)
 				goto do_failover;
 
@@ -2865,6 +2870,7 @@ void bond_loadbalance_arp_mon(struct work_struct *work)
 						bond->dev->name,
 						slave->dev->name);
 				}
+				bond_nl_link_change(bond, slave, BOND_LINK_UP);
 			}
 		} else {
 			/* slave->link == BOND_LINK_UP */
@@ -2892,6 +2898,8 @@ void bond_loadbalance_arp_mon(struct work_struct *work)
 
 				if (slave == oldcurrent)
 					do_failover = 1;
+
+				bond_nl_link_change(bond, slave, BOND_LINK_DOWN);
 			}
 		}
 
@@ -3038,6 +3046,8 @@ static void bond_ab_arp_commit(struct bonding *bond, int delta_in_ticks)
 				pr_info("%s: link status definitely up for interface %s.\n",
 					bond->dev->name, slave->dev->name);
 
+				bond_nl_link_change(bond, slave, BOND_LINK_UP);
+
 				if (!bond->curr_active_slave ||
 				    (slave == bond->primary_slave))
 					goto do_failover;
@@ -3056,6 +3066,8 @@ static void bond_ab_arp_commit(struct bonding *bond, int delta_in_ticks)
 			pr_info("%s: link status definitely down for interface %s, disabling it\n",
 				bond->dev->name, slave->dev->name);
 
+			bond_nl_link_change(bond, slave, BOND_LINK_DOWN);
+
 			if (slave == bond->curr_active_slave) {
 				bond->current_arp_slave = NULL;
 				goto do_failover;
@@ -4683,7 +4695,7 @@ static void bond_destructor(struct net_device *bond_dev)
 	free_netdev(bond_dev);
 }
 
-static void bond_setup(struct net_device *bond_dev)
+void bond_setup(struct net_device *bond_dev)
 {
 	struct bonding *bond = netdev_priv(bond_dev);
 
@@ -5191,24 +5203,6 @@ static int bond_init(struct net_device *bond_dev)
 	return 0;
 }
 
-static int bond_validate(struct nlattr *tb[], struct nlattr *data[])
-{
-	if (tb[IFLA_ADDRESS]) {
-		if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
-			return -EINVAL;
-		if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
-			return -EADDRNOTAVAIL;
-	}
-	return 0;
-}
-
-static struct rtnl_link_ops bond_link_ops __read_mostly = {
-	.kind		= "bond",
-	.priv_size	= sizeof(struct bonding),
-	.setup		= bond_setup,
-	.validate	= bond_validate,
-};
-
 /* Create a new bond based on the specified name and bonding parameters.
  * If name is NULL, obtain a suitable "bond%d" name for us.
  * Caller must NOT hold rtnl_lock; we need to release it here before we
@@ -5216,6 +5210,7 @@ static struct rtnl_link_ops bond_link_ops __read_mostly = {
  */
 int bond_create(struct net *net, const char *name)
 {
+	extern struct rtnl_link_ops bond_link_ops;
 	struct net_device *bond_dev;
 	int res;
 
@@ -5304,7 +5299,7 @@ static int __init bonding_init(void)
 	if (res)
 		goto out;
 
-	res = rtnl_link_register(&bond_link_ops);
+	res = bond_netlink_init();
 	if (res)
 		goto err_link;
 
@@ -5325,7 +5320,7 @@ static int __init bonding_init(void)
 out:
 	return res;
 err:
-	rtnl_link_unregister(&bond_link_ops);
+	bond_netlink_fini();
 err_link:
 	unregister_pernet_subsys(&bond_net_ops);
 #ifdef CONFIG_NET_POLL_CONTROLLER
@@ -5343,7 +5338,7 @@ static void __exit bonding_exit(void)
 
 	bond_destroy_sysfs();
 
-	rtnl_link_unregister(&bond_link_ops);
+	bond_netlink_fini();
 	unregister_pernet_subsys(&bond_net_ops);
 
 #ifdef CONFIG_NET_POLL_CONTROLLER
diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h
index ad3ae46..db7bb06 100644
--- a/drivers/net/bonding/bonding.h
+++ b/drivers/net/bonding/bonding.h
@@ -380,6 +380,7 @@ void bond_select_active_slave(struct bonding *bond);
 void bond_change_active_slave(struct bonding *bond, struct slave *new_active);
 void bond_register_arp(struct bonding *);
 void bond_unregister_arp(struct bonding *);
+extern void bond_setup(struct net_device *bond_dev);
 
 struct bond_net {
 	struct net *		net;	/* Associated network namespace */
diff --git a/include/linux/if_bonding.h b/include/linux/if_bonding.h
index a17edda..b03d832 100644
--- a/include/linux/if_bonding.h
+++ b/include/linux/if_bonding.h
@@ -114,6 +114,29 @@ struct ad_info {
 	__u8 partner_system[ETH_ALEN];
 };
 
+enum {
+	BOND_GENL_ATTR_UNSPEC = 0,
+	BOND_GENL_ATTR_MASTER_INDEX,
+	BOND_GENL_ATTR_SLAVE_INDEX,
+	BOND_GENL_ATTR_MODE,
+	BOND_GENL_ATTR_SLAVE_LINK,
+	__BOND_GENL_ATTR_MAX,
+};
+
+#define BOND_GENL_ATTR_MAX (__BOND_GENL_ATTR_MAX - 1)
+
+enum {
+	BOND_GENL_CMD_UNSPEC = 0,
+	BOND_GENL_CMD_GET_MODE,
+	BOND_GENL_SLAVE_LINK,
+	__BOND_GENL_MAX,
+};
+
+#define BOND_GENL_MAX (__BOND_GENL_MAX - 1)
+
+#define BOND_GENL_VERSION	1
+#define BOND_GENL_MC_GROUP	"bond_mc_group"
+
 #endif /* _LINUX_IF_BONDING_H */
 
 /*
-- 
1.6.0.2

[RFC PATCH 2/2] bonding: add multi-link mode

[Jay Vosburgh]

	Adds multi-link mode for bonding.

	This mode performs per-subnet balancing, wherein each slave is
typically a member of a discrete IP subnet, and the multi-link (ML)
addresses exist in a subnet of their own.  A user space daemon runs the
ML discovery protocol, which locates other ML hosts and exchanges link
information.  The daemon then informs bonding of the appropriate set of
slaves to reach a particular ML destination.  The ML daemon also monitors
the links to insure continued availabilty.

	Note that ML slaves maintain their assigned IP addresses, and
may operate outside the scope of the bond.
---
 drivers/net/bonding/Makefile    |    3 ++-
 drivers/net/bonding/bond_main.c |   34 +++++++++++++++++++++++++++++++---
 drivers/net/bonding/bonding.h   |   13 +++++++++++++
 include/linux/if.h              |    1 +
 include/linux/if_bonding.h      |   15 +++++++++++++++
 net/core/dev.c                  |   37 +++++++++++++++++++++++++++++--------
 6 files changed, 91 insertions(+), 12 deletions(-)

diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile
index 26848a2..0b6ed50 100644
--- a/drivers/net/bonding/Makefile
+++ b/drivers/net/bonding/Makefile
@@ -4,7 +4,8 @@
 
 obj-$(CONFIG_BONDING) += bonding.o
 
-bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_netlink.o
+bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_netlink.o \
+	bond_ml.o
 
 ipv6-$(subst m,y,$(CONFIG_IPV6)) += bond_ipv6.o
 bonding-objs += $(ipv6-y)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 4d3a2c8..1399949 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -200,6 +200,7 @@ const struct bond_parm_tbl bond_mode_tbl[] = {
 {	"802.3ad",		BOND_MODE_8023AD},
 {	"balance-tlb",		BOND_MODE_TLB},
 {	"balance-alb",		BOND_MODE_ALB},
+{	"multi-link",		BOND_MODE_ML},
 {	NULL,			-1},
 };
 
@@ -257,9 +258,10 @@ static const char *bond_mode_name(int mode)
 		[BOND_MODE_8023AD] = "IEEE 802.3ad Dynamic link aggregation",
 		[BOND_MODE_TLB] = "transmit load balancing",
 		[BOND_MODE_ALB] = "adaptive load balancing",
+		[BOND_MODE_ML] = "multi-link",
 	};
 
-	if (mode < 0 || mode > BOND_MODE_ALB)
+	if (mode < 0 || mode > BOND_MODE_ML)
 		return "unknown";
 
 	return names[mode];
@@ -1603,7 +1605,7 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev)
 	 */
 	memcpy(new_slave->perm_hwaddr, slave_dev->dev_addr, ETH_ALEN);
 
-	if (!bond->params.fail_over_mac) {
+	if (!bond->params.fail_over_mac && bond->params.mode != BOND_MODE_ML) {
 		/*
 		 * Set slave to master's mac address.  The application already
 		 * set the master's mac address to that of the first slave
@@ -2097,6 +2099,9 @@ static int bond_release_all(struct net_device *bond_dev)
 		if (bond->params.mode == BOND_MODE_8023AD)
 			bond_3ad_unbind_slave(slave);
 
+		if (bond->params.mode == BOND_MODE_ML)
+			bond_ml_unbind_slave(bond, slave);
+
 		slave_dev = slave->dev;
 		bond_detach_slave(bond, slave);
 
@@ -3357,6 +3362,8 @@ static void bond_info_show_master(struct seq_file *seq)
 			seq_printf(seq, "\tPartner Mac Address: %pM\n",
 				   ad_info.partner_system);
 		}
+	} else if (bond->params.mode == BOND_MODE_ML) {
+		bond_ml_show_proc(seq, bond);
 	}
 }
 
@@ -3843,6 +3850,11 @@ static int bond_open(struct net_device *bond_dev)
 		bond_3ad_initiate_agg_selection(bond, 1);
 	}
 
+	if (bond->params.mode == BOND_MODE_ML) {
+		INIT_DELAYED_WORK(&bond->ml_work, bond_ml_monitor);
+		queue_delayed_work(bond->wq, &bond->ml_work, 0);
+	}
+
 	return 0;
 }
 
@@ -3884,6 +3896,9 @@ static int bond_close(struct net_device *bond_dev)
 	case BOND_MODE_ALB:
 		cancel_delayed_work(&bond->alb_work);
 		break;
+	case BOND_MODE_ML:
+		cancel_delayed_work(&bond->ml_work);
+		break;
 	default:
 		break;
 	}
@@ -4602,6 +4617,8 @@ static netdev_tx_t bond_start_xmit(struct sk_buff *skb, struct net_device *dev)
 	case BOND_MODE_ALB:
 	case BOND_MODE_TLB:
 		return bond_alb_xmit(skb, dev);
+	case BOND_MODE_ML:
+		return bond_xmit_ml(skb, dev);
 	default:
 		/* Should never happen, mode already checked */
 		pr_err("%s: Error: Unknown bonding mode %d\n",
@@ -4639,6 +4656,11 @@ void bond_set_mode_ops(struct bonding *bond, int mode)
 		/* FALLTHRU */
 	case BOND_MODE_TLB:
 		break;
+	case BOND_MODE_ML:
+		bond_set_xmit_hash_policy(bond);
+		bond_set_master_ml_flags(bond);
+		bond_ml_init(bond);
+		break;
 	default:
 		/* Should never happen, mode already checked */
 		pr_err("%s: Error: Unknown bonding mode %d\n",
@@ -4713,7 +4735,6 @@ void bond_setup(struct net_device *bond_dev)
 	ether_setup(bond_dev);
 	bond_dev->netdev_ops = &bond_netdev_ops;
 	bond_dev->ethtool_ops = &bond_ethtool_ops;
-	bond_set_mode_ops(bond, bond->params.mode);
 
 	bond_dev->destructor = bond_destructor;
 
@@ -4726,6 +4747,8 @@ void bond_setup(struct net_device *bond_dev)
 	if (bond->params.arp_interval)
 		bond_dev->priv_flags |= IFF_MASTER_ARPMON;
 
+	bond_set_mode_ops(bond, bond->params.mode);
+
 	/* At first, we block adding VLANs. That's the only way to
 	 * prevent problems that occur when adding VLANs over an
 	 * empty bond. The block will be removed once non-challenged
@@ -4773,6 +4796,10 @@ static void bond_work_cancel_all(struct bonding *bond)
 	    delayed_work_pending(&bond->ad_work))
 		cancel_delayed_work(&bond->ad_work);
 
+	if (bond->params.mode == BOND_MODE_ML &&
+	    delayed_work_pending(&bond->ml_work))
+		cancel_delayed_work(&bond->ml_work);
+
 	if (delayed_work_pending(&bond->mcast_work))
 		cancel_delayed_work(&bond->mcast_work);
 }
@@ -4858,6 +4885,7 @@ static int bond_check_params(struct bond_params *params)
 
 	if (xmit_hash_policy) {
 		if ((bond_mode != BOND_MODE_XOR) &&
+		    (bond_mode != BOND_MODE_ML) &&
 		    (bond_mode != BOND_MODE_8023AD)) {
 			pr_info("xmit_hash_policy param is irrelevant in mode %s\n",
 			       bond_mode_name(bond_mode));
diff --git a/drivers/net/bonding/bonding.h b/drivers/net/bonding/bonding.h
index db7bb06..13b9dd5 100644
--- a/drivers/net/bonding/bonding.h
+++ b/drivers/net/bonding/bonding.h
@@ -23,6 +23,7 @@
 #include <linux/in6.h>
 #include "bond_3ad.h"
 #include "bond_alb.h"
+#include "bond_ml.h"
 
 #define DRV_VERSION	"3.7.0"
 #define DRV_RELDATE	"June 2, 2010"
@@ -246,6 +247,7 @@ struct bonding {
 	u16      rr_tx_counter;
 	struct   ad_bond_info ad_info;
 	struct   alb_bond_info alb_info;
+	struct   ml_bond_info ml_info;
 	struct   bond_params params;
 	struct   list_head vlan_list;
 	struct   vlan_group *vlgrp;
@@ -255,6 +257,7 @@ struct bonding {
 	struct   delayed_work arp_work;
 	struct   delayed_work alb_work;
 	struct   delayed_work ad_work;
+	struct   delayed_work ml_work;
 	struct   delayed_work mcast_work;
 #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
 	struct   in6_addr master_ipv6;
@@ -361,6 +364,16 @@ static inline void bond_unset_master_alb_flags(struct bonding *bond)
 	bond->dev->priv_flags &= ~IFF_MASTER_ALB;
 }
 
+static inline void bond_set_master_ml_flags(struct bonding *bond)
+{
+	bond->dev->priv_flags |= IFF_MASTER_ML;
+}
+
+static inline void bond_unset_master_ml_flags(struct bonding *bond)
+{
+	bond->dev->priv_flags &= ~IFF_MASTER_ML;
+}
+
 struct vlan_entry *bond_next_vlan(struct bonding *bond, struct vlan_entry *curr);
 int bond_dev_queue_xmit(struct bonding *bond, struct sk_buff *skb, struct net_device *slave_dev);
 int bond_create(struct net *net, const char *name);
diff --git a/include/linux/if.h b/include/linux/if.h
index 1239599..826b06f 100644
--- a/include/linux/if.h
+++ b/include/linux/if.h
@@ -77,6 +77,7 @@
 #define IFF_BRIDGE_PORT	0x8000		/* device used as bridge port */
 #define IFF_OVS_DATAPATH	0x10000	/* device used as Open vSwitch
 					 * datapath port */
+#define IFF_MASTER_ML	0x20000		/* bonding master, multi-link */
 
 #define IF_GET_IFACE	0x0001		/* for querying only */
 #define IF_GET_PROTO	0x0002
diff --git a/include/linux/if_bonding.h b/include/linux/if_bonding.h
index b03d832..15c8773 100644
--- a/include/linux/if_bonding.h
+++ b/include/linux/if_bonding.h
@@ -70,6 +70,7 @@
 #define BOND_MODE_8023AD        4
 #define BOND_MODE_TLB           5
 #define BOND_MODE_ALB		6 /* TLB + RLB (receive load balancing) */
+#define BOND_MODE_ML		7
 
 /* each slave's link has 4 states */
 #define BOND_LINK_UP    0           /* link is up and running */
@@ -114,12 +115,22 @@ struct ad_info {
 	__u8 partner_system[ETH_ALEN];
 };
 
+struct bond_ml_route {
+	__u16 lif_index;
+	struct in_addr laddr;
+	struct in_addr raddr;
+};
+
 enum {
 	BOND_GENL_ATTR_UNSPEC = 0,
 	BOND_GENL_ATTR_MASTER_INDEX,
 	BOND_GENL_ATTR_SLAVE_INDEX,
 	BOND_GENL_ATTR_MODE,
 	BOND_GENL_ATTR_SLAVE_LINK,
+	BOND_GENL_ATTR_ML_LADDR,
+	BOND_GENL_ATTR_ML_RADDR,
+	BOND_GENL_ATTR_ML_MLADDR,
+	BOND_GENL_ATTR_ML_INDEX,
 	__BOND_GENL_ATTR_MAX,
 };
 
@@ -129,6 +140,10 @@ enum {
 	BOND_GENL_CMD_UNSPEC = 0,
 	BOND_GENL_CMD_GET_MODE,
 	BOND_GENL_SLAVE_LINK,
+	BOND_GENL_ML_CMD_RT_ADD,
+	BOND_GENL_ML_CMD_RT_DEL,
+	BOND_GENL_ML_CMD_RT_FLUSH,
+	BOND_GENL_ML_CMD_DISCOVERY,
 	__BOND_GENL_MAX,
 };
 
diff --git a/net/core/dev.c b/net/core/dev.c
index d28b3a0..02b653b 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2921,10 +2921,28 @@ static inline void skb_bond_set_mac_by_master(struct sk_buff *skb,
 /* On bonding slaves other than the currently active slave, suppress
  * duplicates except for 802.3ad ETH_P_SLOW, alb non-mcast/bcast, and
  * ARP on active-backup slaves with arp_validate enabled.
+ * Additionally, set skb->dev appropriately for the mode / action.
  */
 int __skb_bond_should_drop(struct sk_buff *skb, struct net_device *master)
 {
 	struct net_device *dev = skb->dev;
+	struct iphdr *iph;
+
+	if (master->priv_flags & IFF_MASTER_ML) {
+		if (skb->protocol == htons(ETH_P_IP)) {
+			iph = ip_hdr(skb);
+			if (!iph)
+				goto out;
+
+			/* For ML, assign to master only if traffic is for
+			 * master, as slaves keep their assigned IP addresses
+			 */
+			if (!ip_route_input(skb, iph->daddr, iph->saddr, 0,
+					    master))
+				skb->dev = master;
+		}
+		return 0;
+	}
 
 	if (master->priv_flags & IFF_MASTER_ARPMON)
 		dev->last_rx = jiffies;
@@ -2941,19 +2959,22 @@ int __skb_bond_should_drop(struct sk_buff *skb, struct net_device *master)
 	if (dev->priv_flags & IFF_SLAVE_INACTIVE) {
 		if ((dev->priv_flags & IFF_SLAVE_NEEDARP) &&
 		    skb->protocol == __cpu_to_be16(ETH_P_ARP))
-			return 0;
+			goto out;
 
 		if (master->priv_flags & IFF_MASTER_ALB) {
 			if (skb->pkt_type != PACKET_BROADCAST &&
 			    skb->pkt_type != PACKET_MULTICAST)
-				return 0;
+				goto out;
 		}
 		if (master->priv_flags & IFF_MASTER_8023AD &&
 		    skb->protocol == __cpu_to_be16(ETH_P_SLOW))
-			return 0;
+			goto out;
 
 		return 1;
 	}
+
+out:
+	skb->dev = master;
 	return 0;
 }
 EXPORT_SYMBOL(__skb_bond_should_drop);
@@ -2981,6 +3002,10 @@ static int __netif_receive_skb(struct sk_buff *skb)
 	if (!skb->skb_iif)
 		skb->skb_iif = skb->dev->ifindex;
 
+	skb_reset_network_header(skb);
+	skb_reset_transport_header(skb);
+	skb->mac_len = skb->network_header - skb->mac_header;
+
 	/*
 	 * bonding note: skbs received on inactive slaves should only
 	 * be delivered to pkt handlers that are exact matches.  Also
@@ -2997,14 +3022,10 @@ static int __netif_receive_skb(struct sk_buff *skb)
 		if (skb_bond_should_drop(skb, master)) {
 			skb->deliver_no_wcard = 1;
 			null_or_orig = orig_dev; /* deliver only exact match */
-		} else
-			skb->dev = master;
+		}
 	}
 
 	__this_cpu_inc(softnet_data.processed);
-	skb_reset_network_header(skb);
-	skb_reset_transport_header(skb);
-	skb->mac_len = skb->network_header - skb->mac_header;
 
 	pt_prev = NULL;
 
-- 
1.6.0.2

Re: [RFC PATCH 1/2] bonding: generic netlink infrastructure

[David Miller]

From: Jay Vosburgh <fubar@us.ibm.com>
Date: Mon, 13 Dec 2010 14:07:34 -0800

> -bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o
> +bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_netlink.o

The new bond_netlink.c file is missing from the patch.

Forgotten "git add" most likely. :-)