Skip to contentSkip to conversation

bonding: One fix two features

PATCH In review Jay Vosburgh 4 participants 9 messages Nov 5, 2008 – Apr 14, 2009

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

cover bonding: One fix two features

Three patches for bonding:

Patch 1 adds IPv6 gratuitous neighbor advertisements during
failover.

Patch 2 fixes ALB mode to correctly balance traffic on VLANs
configured above bonding.

Patch 3 adds alternate aggregator selection policies to 802.3ad
mode. The new policies permit gang failover within 802.3ad, providing a means to always have the "best" aggregator be the active aggregator (instead of waiting until all members of the active aggregator fail, as is the case currently).

	Please apply for net-next-2.6.

	-J

---
-Jay Vosburgh, IBM Linux Technology Center, [email protected]

#1 lore raw

PATCH 1/3 bonding: send IPv6 neighbor advertisement on failover

From: Brian Haley <[email protected]>

This patch adds better IPv6 failover support for bonding devices, especially when in active-backup mode and there are only IPv6 addresses configured, as reported by Alex Sidorenko.

- Creates a new file, net/drivers/bonding/bond_ipv6.c, for the IPv6-specific routines. Both regular bonds and VLANs over bonds are supported.

- Adds a new tunable, num_unsol_na, to limit the number of unsolicited IPv6 Neighbor Advertisements that are sent on a failover event. Default is 1.

- Creates two new IPv6 neighbor discovery functions:

ndisc_build_skb()
ndisc_send_skb()

These were required to support VLANs since we have to be able to add the VLAN id to the skb since ndisc_send_na() and friends shouldn't be asked to do this. These two routines are basically __ndisc_send() split into two pieces, in a slightly different order.

- Updates Documentation/networking/bonding.txt and bumps the rev of bond support to 3.4.0.

On failover, this new code will generate one packet:

- An unsolicited IPv6 Neighbor Advertisement, which helps the switch learn that the address has moved to the new slave.

Testing has shown that sending just the NA results in pretty good behavior when in active-back mode, I saw no lost ping packets for example.

---
 Documentation/networking/bonding.txt |   10 ++
 drivers/net/Kconfig                  |    1 +
 drivers/net/bonding/Makefile         |    3 +
 drivers/net/bonding/bond_ipv6.c      |  218 ++++++++++++++++++++++++++++++++++
 drivers/net/bonding/bond_main.c      |   33 +++++-
 drivers/net/bonding/bond_sysfs.c     |   42 +++++++
 drivers/net/bonding/bonding.h        |   34 +++++-
 include/net/ndisc.h                  |   14 ++
 net/ipv6/ndisc.c                     |   92 ++++++++++----
 9 files changed, 416 insertions(+), 31 deletions(-)
 create mode 100644 drivers/net/bonding/bond_ipv6.c
Documentation/networking/bonding.txt +10 -0
index d733a42..3f4d0fa 100644
@@ -551,6 +551,16 @@ num_grat_arp
 	affects only the active-backup mode.  This option was added for
 	bonding version 3.3.0.
 
+num_unsol_na
+
+	Specifies the number of unsolicited IPv6 Neighbor Advertisements
+	to be issued after a failover event.  One unsolicited NA is issued
+	immediately after the failover.
+
+	The valid range is 0 - 255; the default value is 1.  This option
+	affects only the active-backup mode.  This option was added for
+	bonding version 3.4.0.
+
 primary
 
 	A string (eth0, eth2, etc) specifying which slave is the
drivers/net/Kconfig +1 -0
index 0f3e6b2..f1d0a13 100644
@@ -61,6 +61,7 @@ config DUMMY
 config BONDING
 	tristate "Bonding driver support"
 	depends on INET
+	depends on IPV6 || IPV6=n
 	---help---
 	  Say 'Y' or 'M' if you wish to be able to 'bond' multiple Ethernet
 	  Channels together. This is called 'Etherchannel' by Cisco,
drivers/net/bonding/Makefile +3 -0
index 5cdae2b..6f9c6fa 100644
@@ -6,3 +6,6 @@ obj-$(CONFIG_BONDING) += bonding.o
 
 bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o
 
+ipv6-$(subst m,y,$(CONFIG_IPV6)) += bond_ipv6.o
+bonding-objs += $(ipv6-y)
+
drivers/net/bonding/bond_ipv6.c new file +218 -0

218 lines not on this page · raw message

drivers/net/bonding/bond_main.c +31 -2
index 39575d7..798d98c 100644
@@ -89,6 +89,7 @@
 
 static int max_bonds	= BOND_DEFAULT_MAX_BONDS;
 static int num_grat_arp = 1;
+static int num_unsol_na = 1;
 static int miimon	= BOND_LINK_MON_INTERV;
 static int updelay	= 0;
 static int downdelay	= 0;
@@ -107,6 +108,8 @@ module_param(max_bonds, int, 0);
 MODULE_PARM_DESC(max_bonds, "Max number of bonded devices");
 module_param(num_grat_arp, int, 0644);
 MODULE_PARM_DESC(num_grat_arp, "Number of gratuitous ARP packets to send on failover event");
+module_param(num_unsol_na, int, 0644);
+MODULE_PARM_DESC(num_unsol_na, "Number of unsolicited IPv6 Neighbor Advertisements packets to send on failover event");
 module_param(miimon, int, 0);
 MODULE_PARM_DESC(miimon, "Link check interval in milliseconds");
 module_param(updelay, int, 0);
@@ -242,14 +245,13 @@ static int bond_add_vlan(struct bonding *bond, unsigned short vlan_id)
 	dprintk("bond: %s, vlan id %d\n",
 		(bond ? bond->dev->name: "None"), vlan_id);
 
-	vlan = kmalloc(sizeof(struct vlan_entry), GFP_KERNEL);
+	vlan = kzalloc(sizeof(struct vlan_entry), GFP_KERNEL);
 	if (!vlan) {
 		return -ENOMEM;
 	}
 
 	INIT_LIST_HEAD(&vlan->vlan_list);
 	vlan->vlan_id = vlan_id;
-	vlan->vlan_ip = 0;
 
 	write_lock_bh(&bond->lock);
 
@@ -1208,6 +1210,9 @@ void bond_change_active_slave(struct bonding *bond, struct slave *new_active)
 			bond->send_grat_arp = bond->params.num_grat_arp;
 			bond_send_gratuitous_arp(bond);
 
+			bond->send_unsol_na = bond->params.num_unsol_na;
+			bond_send_unsolicited_na(bond);
+
 			write_unlock_bh(&bond->curr_slave_lock);
 			read_unlock(&bond->lock);
 
@@ -2463,6 +2468,12 @@ void bond_mii_monitor(struct work_struct *work)
 		read_unlock(&bond->curr_slave_lock);
 	}
 
+	if (bond->send_unsol_na) {
+		read_lock(&bond->curr_slave_lock);
+		bond_send_unsolicited_na(bond);
+		read_unlock(&bond->curr_slave_lock);
+	}
+
 	if (bond_miimon_inspect(bond)) {
 		read_unlock(&bond->lock);
 		rtnl_lock();
@@ -3158,6 +3169,12 @@ void bond_activebackup_arp_mon(struct work_struct *work)
 		read_unlock(&bond->curr_slave_lock);
 	}
 
+	if (bond->send_unsol_na) {
+		read_lock(&bond->curr_slave_lock);
+		bond_send_unsolicited_na(bond);
+		read_unlock(&bond->curr_slave_lock);
+	}
+
 	if (bond_ab_arp_inspect(bond, delta_in_ticks)) {
 		read_unlock(&bond->lock);
 		rtnl_lock();
@@ -3827,6 +3844,7 @@ static int bond_close(struct net_device *bond_dev)
 	write_lock_bh(&bond->lock);
 
 	bond->send_grat_arp = 0;
+	bond->send_unsol_na = 0;
 
 	/* signal timers not to re-arm */
 	bond->kill_timers = 1;
@@ -4542,6 +4560,7 @@ static int bond_init(struct net_device *bond_dev, struct bond_params *params)
 	bond->primary_slave = NULL;
 	bond->dev = bond_dev;
 	bond->send_grat_arp = 0;
+	bond->send_unsol_na = 0;
 	bond->setup_by_slave = 0;
 	INIT_LIST_HEAD(&bond->vlan_list);
 
@@ -4791,6 +4810,13 @@ static int bond_check_params(struct bond_params *params)
 		num_grat_arp = 1;
 	}
 
+	if (num_unsol_na < 0 || num_unsol_na > 255) {
+		printk(KERN_WARNING DRV_NAME
+		       ": Warning: num_unsol_na (%d) not in range 0-255 so it "
+		       "was reset to 1 \n", num_unsol_na);
+		num_unsol_na = 1;
+	}
+
 	/* reset values for 802.3ad */
 	if (bond_mode == BOND_MODE_8023AD) {
 		if (!miimon) {
@@ -4992,6 +5018,7 @@ static int bond_check_params(struct bond_params *params)
 	params->xmit_policy = xmit_hashtype;
 	params->miimon = miimon;
 	params->num_grat_arp = num_grat_arp;
+	params->num_unsol_na = num_unsol_na;
 	params->arp_interval = arp_interval;
 	params->arp_validate = arp_validate_value;
 	params->updelay = updelay;
@@ -5144,6 +5171,7 @@ static int __init bonding_init(void)
 
 	register_netdevice_notifier(&bond_netdev_notifier);
 	register_inetaddr_notifier(&bond_inetaddr_notifier);
+	bond_register_ipv6_notifier();
 
 	goto out;
 err:
@@ -5166,6 +5194,7 @@ static void __exit bonding_exit(void)
 {
 	unregister_netdevice_notifier(&bond_netdev_notifier);
 	unregister_inetaddr_notifier(&bond_inetaddr_notifier);
+	bond_unregister_ipv6_notifier();
 
 	bond_destroy_sysfs();
 
drivers/net/bonding/bond_sysfs.c +42 -0

54 lines not on this page · raw message

drivers/net/bonding/bonding.h +32 -2
index ffb668d..0491c7c 100644
@@ -19,16 +19,19 @@
 #include <linux/proc_fs.h>
 #include <linux/if_bonding.h>
 #include <linux/kobject.h>
+#include <linux/in6.h>
 #include "bond_3ad.h"
 #include "bond_alb.h"
 
-#define DRV_VERSION	"3.3.0"
-#define DRV_RELDATE	"June 10, 2008"
+#define DRV_VERSION	"3.4.0"
+#define DRV_RELDATE	"October 7, 2008"
 #define DRV_NAME	"bonding"
 #define DRV_DESCRIPTION	"Ethernet Channel Bonding Driver"
 
 #define BOND_MAX_ARP_TARGETS	16
 
+extern struct list_head bond_dev_list;
+
 #ifdef BONDING_DEBUG
 #define dprintk(fmt, args...) \
 	printk(KERN_DEBUG     \
@@ -126,6 +129,7 @@ struct bond_params {
 	int xmit_policy;
 	int miimon;
 	int num_grat_arp;
+	int num_unsol_na;
 	int arp_interval;
 	int arp_validate;
 	int use_carrier;
@@ -148,6 +152,9 @@ struct vlan_entry {
 	struct list_head vlan_list;
 	__be32 vlan_ip;
 	unsigned short vlan_id;
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+	struct in6_addr vlan_ipv6;
+#endif
 };
 
 struct slave {
@@ -195,6 +202,7 @@ struct bonding {
 	rwlock_t curr_slave_lock;
 	s8       kill_timers;
 	s8	 send_grat_arp;
+	s8	 send_unsol_na;
 	s8	 setup_by_slave;
 	struct   net_device_stats stats;
 #ifdef CONFIG_PROC_FS
@@ -218,6 +226,9 @@ struct bonding {
 	struct   delayed_work arp_work;
 	struct   delayed_work alb_work;
 	struct   delayed_work ad_work;
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+	struct   in6_addr master_ipv6;
+#endif
 };
 
 /**
@@ -341,5 +352,24 @@ extern struct bond_parm_tbl xmit_hashtype_tbl[];
 extern struct bond_parm_tbl arp_validate_tbl[];
 extern struct bond_parm_tbl fail_over_mac_tbl[];
 
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+void bond_send_unsolicited_na(struct bonding *bond);
+void bond_register_ipv6_notifier(void);
+void bond_unregister_ipv6_notifier(void);
+#else
+static inline void bond_send_unsolicited_na(struct bonding *bond)
+{
+	return;
+}
+static inline void bond_register_ipv6_notifier(void)
+{
+	return;
+}
+static inline void bond_unregister_ipv6_notifier(void)
+{
+	return;
+}
+#endif
+
 #endif /* _LINUX_BONDING_H */
 
include/net/ndisc.h +14 -0
index 11dd013..ce532f2 100644
@@ -108,6 +108,20 @@ extern void			ndisc_send_redirect(struct sk_buff *skb,
 
 extern int			ndisc_mc_map(struct in6_addr *addr, char *buf, struct net_device *dev, int dir);
 
+extern struct sk_buff		*ndisc_build_skb(struct net_device *dev,
+						 const struct in6_addr *daddr,
+						 const struct in6_addr *saddr,
+						 struct icmp6hdr *icmp6h,
+						 const struct in6_addr *target,
+						 int llinfo);
+
+extern void			ndisc_send_skb(struct sk_buff *skb,
+					       struct net_device *dev,
+					       struct neighbour *neigh,
+					       const struct in6_addr *daddr,
+					       const struct in6_addr *saddr,
+					       struct icmp6hdr *icmp6h);
+
 
 
 /*
net/ipv6/ndisc.c +65 -27

123 lines not on this page · raw message

Signature · 1 line
1.6.0.2
#2 lore raw

PATCH 2/3 bonding: Fix ALB mode to balance traffic on VLANs

The current ALB function that processes incoming ARPs
does not handle traffic for VLANs configured above bonding. This causes traffic on those VLANs to all be assigned the same slave. This patch corrects that misbehavior by locating the bonding interface nested below the VLAN interface.

Bug reported by Sven Anders <[email protected]>, who also
tested an earlier version of this patch and confirmed that it resolved the problem.

---
 drivers/net/bonding/bond_alb.c |   13 ++++++++++---
 1 files changed, 10 insertions(+), 3 deletions(-)
drivers/net/bonding/bond_alb.c +10 -3
index 87437c7..e170fa2 100644
@@ -346,14 +346,18 @@ static void rlb_update_entry_from_arp(struct bonding *bond, struct arp_pkt *arp)
 
 static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct packet_type *ptype, struct net_device *orig_dev)
 {
-	struct bonding *bond = bond_dev->priv;
+	struct bonding *bond;
 	struct arp_pkt *arp = (struct arp_pkt *)skb->data;
 	int res = NET_RX_DROP;
 
 	if (dev_net(bond_dev) != &init_net)
 		goto out;
 
-	if (!(bond_dev->flags & IFF_MASTER))
+	while (bond_dev->priv_flags & IFF_802_1Q_VLAN)
+		bond_dev = vlan_dev_real_dev(bond_dev);
+
+	if (!(bond_dev->priv_flags & IFF_BONDING) ||
+	    !(bond_dev->flags & IFF_MASTER))
 		goto out;
 
 	if (!arp) {
@@ -368,6 +372,9 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct
 
 	if (arp->op_code == htons(ARPOP_REPLY)) {
 		/* update rx hash table for this ARP */
+		printk("rar: update orig %s bond_dev %s\n", orig_dev->name,
+		       bond_dev->name);
+		bond = bond_dev->priv;
 		rlb_update_entry_from_arp(bond, arp);
 		dprintk("Server received an ARP Reply from client\n");
 	}
@@ -818,7 +825,7 @@ static int rlb_initialize(struct bonding *bond)
 
 	/*initialize packet type*/
 	pk_type->type = __constant_htons(ETH_P_ARP);
-	pk_type->dev = bond->dev;
+	pk_type->dev = NULL;
 	pk_type->func = rlb_arp_recv;
 
 	/* register to receive ARPs */
Signature · 1 line
1.6.0.2
#3 lore raw

PATCH 3/3 bonding: alternate agg selection policies for 802.3ad

This patch implements alternative aggregator selection policies
for 802.3ad. The existing policy, now termed "stable," selects the active aggregator by greatest bandwidth, and only reselects a new aggregator if the active aggregator is entirely disabled (no more ports or all ports down).

This patch adds two new policies: bandwidth and count, selecting
the active aggregator by total bandwidth (like the stable policy) or by the number of ports in the aggregator, respectively. These two policies also differ from the stable policy in that they will reselect the active aggregator when availability-related changes occur in the bond (e.g., link state change).

This permits "gang failover" within 802.3ad, allowing redundant
aggregators along parallel paths to always maintain the "best" aggregator as the active aggregator (rather than having to wait for the active to entirely fail).

This patch also updates the driver version to 3.5.0.

---
 Documentation/networking/bonding.txt |   42 +++++
 drivers/net/bonding/bond_3ad.c       |  326 +++++++++++++++++++++-------------
 drivers/net/bonding/bond_3ad.h       |   10 +-
 drivers/net/bonding/bond_main.c      |   30 +++
 drivers/net/bonding/bond_sysfs.c     |   49 +++++
 drivers/net/bonding/bonding.h        |    5 +-
 6 files changed, 333 insertions(+), 129 deletions(-)
Documentation/networking/bonding.txt +42 -0
index 3f4d0fa..5ede747 100644
@@ -194,6 +194,48 @@ or, for backwards compatibility, the option value.  E.g.,
 
 	The parameters are as follows:
 
+ad_select
+
+	Specifies the 802.3ad aggregation selection logic to use.  The
+	possible values and their effects are:
+
+	stable or 0
+
+		The active aggregator is chosen by largest aggregate
+		bandwidth.
+
+		Reselection of the active aggregator occurs only when all
+		slaves of the active aggregator are down or the active
+		aggregator has no slaves.
+
+		This is the default value.
+
+	bandwidth or 1
+
+		The active aggregator is chosen by largest aggregate
+		bandwidth.  Reselection occurs if:
+
+		- A slave is added to or removed from the bond
+
+		- Any slave's link state changes
+
+		- Any slave's 802.3ad association state changes
+
+		- The bond's adminstrative state changes to up
+
+	count or 2
+
+		The active aggregator is chosen by the largest number of
+		ports (slaves).  Reselection occurs as described under the
+		"bandwidth" setting, above.
+
+	The bandwidth and count selection policies permit failover of
+	802.3ad aggregations when partial failure of the active aggregator
+	occurs.  This keeps the aggregator with the highest availability
+	(either in bandwidth or in number of ports) active at all times.
+
+	This option was added in bonding version 3.4.0.
+
 arp_interval
 
 	Specifies the ARP link monitoring frequency in milliseconds.
drivers/net/bonding/bond_3ad.c +203 -123

409 lines not on this page · raw message

drivers/net/bonding/bond_3ad.h +6 -4
index b5ee45f..a803fe0 100644
@@ -42,10 +42,11 @@ typedef struct mac_addr {
 	u8 mac_addr_value[ETH_ALEN];
 } mac_addr_t;
 
-typedef enum {
-	AD_BANDWIDTH = 0,
-	AD_COUNT
-} agg_selection_t;
+enum {
+	BOND_AD_STABLE = 0,
+	BOND_AD_BANDWIDTH = 1,
+	BOND_AD_COUNT = 2,
+};
 
 // rx machine states(43.4.11 in the 802.3ad standard)
 typedef enum {
@@ -277,6 +278,7 @@ void bond_3ad_initialize(struct bonding *bond, u16 tick_resolution, int lacp_fas
 int  bond_3ad_bind_slave(struct slave *slave);
 void bond_3ad_unbind_slave(struct slave *slave);
 void bond_3ad_state_machine_handler(struct work_struct *);
+void bond_3ad_initiate_agg_selection(struct bonding *bond, int timeout);
 void bond_3ad_adapter_speed_changed(struct slave *slave);
 void bond_3ad_adapter_duplex_changed(struct slave *slave);
 void bond_3ad_handle_link_change(struct slave *slave, char link);
drivers/net/bonding/bond_main.c +30 -0
index 798d98c..02de3e0 100644
@@ -97,6 +97,7 @@ static int use_carrier	= 1;
 static char *mode	= NULL;
 static char *primary	= NULL;
 static char *lacp_rate	= NULL;
+static char *ad_select  = 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, };
@@ -130,6 +131,8 @@ MODULE_PARM_DESC(primary, "Primary network device to use");
 module_param(lacp_rate, charp, 0);
 MODULE_PARM_DESC(lacp_rate, "LACPDU tx rate to request from 802.3ad partner "
 			    "(slow/fast)");
+module_param(ad_select, charp, 0);
+MODULE_PARM_DESC(ad_select, "803.ad aggregation selection logic: stable (0, default), bandwidth (1), count (2)");
 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");
@@ -200,6 +203,13 @@ struct bond_parm_tbl fail_over_mac_tbl[] = {
 {	NULL,			-1},
 };
 
+struct bond_parm_tbl ad_select_tbl[] = {
+{	"stable",	BOND_AD_STABLE},
+{	"bandwidth",	BOND_AD_BANDWIDTH},
+{	"count",	BOND_AD_COUNT},
+{	NULL,		-1},
+};
+
 /*-------------------------- Forward declarations ---------------------------*/
 
 static void bond_send_gratuitous_arp(struct bonding *bond);
@@ -3318,6 +3328,8 @@ static void bond_info_show_master(struct seq_file *seq)
 		seq_puts(seq, "\n802.3ad info\n");
 		seq_printf(seq, "LACP rate: %s\n",
 			   (bond->params.lacp_fast) ? "fast" : "slow");
+		seq_printf(seq, "Aggregator selection policy (ad_select): %s\n",
+			   ad_select_tbl[bond->params.ad_select].modename);
 
 		if (bond_3ad_get_active_agg_info(bond, &ad_info)) {
 			seq_printf(seq, "bond %s has no active aggregator\n",
@@ -3824,6 +3836,7 @@ static int bond_open(struct net_device *bond_dev)
 		queue_delayed_work(bond->wq, &bond->ad_work, 0);
 		/* register to receive LACPDUs */
 		bond_register_lacpdu(bond);
+		bond_3ad_initiate_agg_selection(bond, 1);
 	}
 
 	return 0;
@@ -4763,6 +4776,23 @@ static int bond_check_params(struct bond_params *params)
 		}
 	}
 
+	if (ad_select) {
+		params->ad_select = bond_parse_parm(ad_select, ad_select_tbl);
+		if (params->ad_select == -1) {
+			printk(KERN_ERR DRV_NAME
+			       ": Error: Invalid ad_select \"%s\"\n",
+			       ad_select == NULL ? "NULL" : ad_select);
+			return -EINVAL;
+		}
+
+		if (bond_mode != BOND_MODE_8023AD) {
+			printk(KERN_WARNING DRV_NAME
+			       ": ad_select param only affects 802.3ad mode\n");
+		}
+	} else {
+		params->ad_select = BOND_AD_STABLE;
+	}
+
 	if (max_bonds < 0 || max_bonds > INT_MAX) {
 		printk(KERN_WARNING DRV_NAME
 		       ": Warning: max_bonds (%d) not in range %d-%d, so it "
drivers/net/bonding/bond_sysfs.c +49 -0

67 lines not on this page · raw message

drivers/net/bonding/bonding.h +3 -2
index 0491c7c..b5eb8e6 100644
@@ -23,8 +23,8 @@
 #include "bond_3ad.h"
 #include "bond_alb.h"
 
-#define DRV_VERSION	"3.4.0"
-#define DRV_RELDATE	"October 7, 2008"
+#define DRV_VERSION	"3.5.0"
+#define DRV_RELDATE	"November 4, 2008"
 #define DRV_NAME	"bonding"
 #define DRV_DESCRIPTION	"Ethernet Channel Bonding Driver"
 
@@ -137,6 +137,7 @@ struct bond_params {
 	int updelay;
 	int downdelay;
 	int lacp_fast;
+	int ad_select;
 	char primary[IFNAMSIZ];
 	__be32 arp_targets[BOND_MAX_ARP_TARGETS];
 };
Signature · 1 line
1.6.0.2
#4 lore raw

2.6.29 ALB bonding printk()s every arp (was Re: [PATCH 2/3] bonding: Fix ALB mode to balance traffic on VLANs)

In reply to Jay Vosburgh

On 05 Nov 2008 at 01:51, Jay Vosburgh wrote:

JV> The current ALB function that processes incoming ARPs does JV> not handle traffic for VLANs configured above bonding. This JV> causes traffic on those VLANs to all be assigned the same slave. JV> This patch corrects that misbehavior by locating the bonding JV> interface nested below the VLAN interface.

Quoted 6 lines
drivers/net/bonding/bond_alb.c +0 -0
index 87437c7..e170fa2 100644

drivers/net/bonding/bond_alb.c static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct · line 380

Quoted 9 lines
drivers/net/bonding/bond_alb.c +3 -0
@@ -368,6 +372,9 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct
 
 	if (arp->op_code == htons(ARPOP_REPLY)) {
 		/* update rx hash table for this ARP */
+		printk("rar: update orig %s bond_dev %s\n", orig_dev->name,
+		       bond_dev->name);
+		bond = bond_dev->priv;
 		rlb_update_entry_from_arp(bond, arp);
 		dprintk("Server received an ARP Reply from client\n");
 	}

Is the printk() in this patch necessary?

We recently put a 2.6.29 kernel on a router with multiple ALB bonds each of several e1000 devices. It now generates a log entry every time an arp is-at packet arrives at a bonded interface. I'm not sure that's a feature...

Cheers

Duncan

Signature · 3 lines
Duncan Gibb - Technical Director
Sirius Corporation plc - control through freedom
http://www.siriusit.co.uk/ || t: +44 870 608 0063
maintainer
#5 lore raw

Re: 2.6.29 ALB bonding printk()s every arp

From: Duncan Gibb <[email protected]>
Date: Wed, 15 Apr 2009 00:15:04 +0100

Quoted 12 lines · Jay Vosburgh

drivers/net/bonding/bond_alb.c static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct · line 380

Quoted 8 lines
drivers/net/bonding/bond_alb.c +3 -0
 	if (arp->op_code == htons(ARPOP_REPLY)) {
 		/* update rx hash table for this ARP */
+		printk("rar: update orig %s bond_dev %s\n", orig_dev->name,
+		       bond_dev->name);
+		bond = bond_dev->priv;
 		rlb_update_entry_from_arp(bond, arp);
 		dprintk("Server received an ARP Reply from client\n");
 	}

Is the printk() in this patch necessary?

Quoted 4 lines

We recently put a 2.6.29 kernel on a router with multiple ALB bonds each of several e1000 devices. It now generates a log entry every time an arp is-at packet arrives at a bonded interface. I'm not sure that's a feature...

That's pretty annoying. Jay can we just remove this?

#6 lore raw

PATCH 1/1 bonding: Remove debug printk

In reply to Duncan Gibb

Remove debug printk I accidently left in as part of commit:

commit 6146b1a4da98377e4abddc91ba5856bef8f23f1e

Author: Jay Vosburgh <[email protected]>
Date:   Tue Nov 4 17:51:15 2008 -0800

bonding: Fix ALB mode to balance traffic on VLANs

Reported by Duncan Gibb <[email protected]>

drivers/net/bonding/bond_alb.c +0 -2
index 8dc6fbb..553a899 100644
@@ -370,8 +370,6 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct
 
 	if (arp->op_code == htons(ARPOP_REPLY)) {
 		/* update rx hash table for this ARP */
-		printk("rar: update orig %s bond_dev %s\n", orig_dev->name,
-		       bond_dev->name);
 		bond = netdev_priv(bond_dev);
 		rlb_update_entry_from_arp(bond, arp);
 		pr_debug("Server received an ARP Reply from client\n");
maintainer
#7 lore raw

Re: [PATCH] bonding: Remove debug printk

From: Jay Vosburgh <[email protected]>
Date: Tue, 14 Apr 2009 16:51:49 -0700

Quoted 4 lines

Remove debug printk I accidently left in as part of commit:

commit 6146b1a4da98377e4abddc91ba5856bef8f23f1e

Quoted 8 lines
Author: Jay Vosburgh <[email protected]>
Date:   Tue Nov 4 17:51:15 2008 -0800

bonding: Fix ALB mode to balance traffic on VLANs

Reported by Duncan Gibb <[email protected]>

Applied and queued up for 2.6.29-stable

#8 lore raw

Re: [PATCH 1/3] bonding: send IPv6 neighbor advertisement on failover

In reply to Jay Vosburgh
Quoted 736 lines · Jay Vosburgh

From: Brian Haley <[email protected]>

This patch adds better IPv6 failover support for bonding devices, especially when in active-backup mode and there are only IPv6 addresses configured, as reported by Alex Sidorenko.

- Creates a new file, net/drivers/bonding/bond_ipv6.c, for the IPv6-specific routines. Both regular bonds and VLANs over bonds are supported.

- Adds a new tunable, num_unsol_na, to limit the number of unsolicited IPv6 Neighbor Advertisements that are sent on a failover event. Default is 1.

- Creates two new IPv6 neighbor discovery functions:

ndisc_build_skb()
ndisc_send_skb()

These were required to support VLANs since we have to be able to add the VLAN id to the skb since ndisc_send_na() and friends shouldn't be asked to do this. These two routines are basically __ndisc_send() split into two pieces, in a slightly different order.

- Updates Documentation/networking/bonding.txt and bumps the rev of bond support to 3.4.0.

On failover, this new code will generate one packet:

- An unsolicited IPv6 Neighbor Advertisement, which helps the switch learn that the address has moved to the new slave.

Testing has shown that sending just the NA results in pretty good behavior when in active-back mode, I saw no lost ping packets for example.

---
 Documentation/networking/bonding.txt |   10 ++
 drivers/net/Kconfig                  |    1 +
 drivers/net/bonding/Makefile         |    3 +
 drivers/net/bonding/bond_ipv6.c      |  218 ++++++++++++++++++++++++++++++++++
 drivers/net/bonding/bond_main.c      |   33 +++++-
 drivers/net/bonding/bond_sysfs.c     |   42 +++++++
 drivers/net/bonding/bonding.h        |   34 +++++-
 include/net/ndisc.h                  |   14 ++
 net/ipv6/ndisc.c                     |   92 ++++++++++----
 9 files changed, 416 insertions(+), 31 deletions(-)
 create mode 100644 drivers/net/bonding/bond_ipv6.c
Documentation/networking/bonding.txt +10 -0
index d733a42..3f4d0fa 100644
@@ -551,6 +551,16 @@ num_grat_arp
 	affects only the active-backup mode.  This option was added for
 	bonding version 3.3.0.
 
+num_unsol_na
+
+	Specifies the number of unsolicited IPv6 Neighbor Advertisements
+	to be issued after a failover event.  One unsolicited NA is issued
+	immediately after the failover.
+
+	The valid range is 0 - 255; the default value is 1.  This option
+	affects only the active-backup mode.  This option was added for
+	bonding version 3.4.0.
+
 primary
 
 	A string (eth0, eth2, etc) specifying which slave is the
drivers/net/Kconfig +1 -0
index 0f3e6b2..f1d0a13 100644
@@ -61,6 +61,7 @@ config DUMMY
 config BONDING
 	tristate "Bonding driver support"
 	depends on INET
+	depends on IPV6 || IPV6=n
 	---help---
 	  Say 'Y' or 'M' if you wish to be able to 'bond' multiple Ethernet
 	  Channels together. This is called 'Etherchannel' by Cisco,
drivers/net/bonding/Makefile +3 -0
index 5cdae2b..6f9c6fa 100644
@@ -6,3 +6,6 @@ obj-$(CONFIG_BONDING) += bonding.o
 
 bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o
 
+ipv6-$(subst m,y,$(CONFIG_IPV6)) += bond_ipv6.o
+bonding-objs += $(ipv6-y)
+
drivers/net/bonding/bond_ipv6.c new file +218 -0
new file mode 100644
index 0000000..7c78b7b
@@ -0,0 +1,218 @@
+/*
+ * Copyright(c) 2008 Hewlett-Packard Development Company, L.P.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License as published by the
+ * Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+ * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
+ * for more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program; if not, write to the Free Software Foundation, Inc.,
+ * 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+ *
+ * The full GNU General Public License is included in this distribution in the
+ * file called LICENSE.
+ *
+ */
+
+//#define BONDING_DEBUG 1
+
+#include <linux/types.h>
+#include <linux/if_vlan.h>
+#include <net/ipv6.h>
+#include <net/ndisc.h>
+#include <net/addrconf.h>
+#include "bonding.h"
+
+/*
+ * Assign bond->master_ipv6 to the next IPv6 address in the list, or
+ * zero it out if there are none.
+ */
+static void bond_glean_dev_ipv6(struct net_device *dev, struct in6_addr *addr)
+{
+	struct inet6_dev *idev;
+	struct inet6_ifaddr *ifa;
+
+	if (!dev)
+		return;
+
+	idev = in6_dev_get(dev);
+	if (!idev)
+		return;
+
+	read_lock_bh(&idev->lock);
+	ifa = idev->addr_list;
+	if (ifa)
+		ipv6_addr_copy(addr, &ifa->addr);
+	else
+		ipv6_addr_set(addr, 0, 0, 0, 0);
+
+	read_unlock_bh(&idev->lock);
+
+	in6_dev_put(idev);
+}
+
+static void bond_na_send(struct net_device *slave_dev,
+			 struct in6_addr *daddr,
+			 int router,
+			 unsigned short vlan_id)
+{
+	struct in6_addr mcaddr;
+	struct icmp6hdr icmp6h = {
+		.icmp6_type = NDISC_NEIGHBOUR_ADVERTISEMENT,
+	};
+	struct sk_buff *skb;
+
+	icmp6h.icmp6_router = router;
+	icmp6h.icmp6_solicited = 0;
+	icmp6h.icmp6_override = 1;
+
+	addrconf_addr_solict_mult(daddr, &mcaddr);
+
+	dprintk("ipv6 na on slave %s: dest %pI6, src %pI6\n",
+	       slave->name, &mcaddr, daddr);
+
+	skb = ndisc_build_skb(slave_dev, &mcaddr, daddr, &icmp6h, daddr,
+			      ND_OPT_TARGET_LL_ADDR);
+
+	if (!skb) {
+		printk(KERN_ERR DRV_NAME ": NA 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;
+		}
+	}
+
+	ndisc_send_skb(skb, slave_dev, NULL, &mcaddr, daddr, &icmp6h);
+}
+
+/*
+ * Kick out an unsolicited Neighbor Advertisement for an IPv6 address on
+ * the bonding master.  This will help the switch learn our address
+ * if in active-backup mode.
+ *
+ * Caller must hold curr_slave_lock for read or better
+ */
+void bond_send_unsolicited_na(struct bonding *bond)
+{
+	struct slave *slave = bond->curr_active_slave;
+	struct vlan_entry *vlan;
+	struct inet6_dev *idev;
+	int is_router;
+
+	dprintk("bond_send_unsol_na: bond %s slave %s\n", bond->dev->name,
+				slave ? slave->dev->name : "NULL");
+
+	if (!slave || !bond->send_unsol_na ||
+	    test_bit(__LINK_STATE_LINKWATCH_PENDING, &slave->dev->state))
+		return;
+
+	bond->send_unsol_na--;
+
+	idev = in6_dev_get(bond->dev);
+	if (!idev)
+		return;
+
+	is_router = !!idev->cnf.forwarding;
+
+	in6_dev_put(idev);
+
+	if (!ipv6_addr_any(&bond->master_ipv6))
+		bond_na_send(slave->dev, &bond->master_ipv6, is_router, 0);
+
+	list_for_each_entry(vlan, &bond->vlan_list, vlan_list) {
+		if (!ipv6_addr_any(&vlan->vlan_ipv6)) {
+			bond_na_send(slave->dev, &vlan->vlan_ipv6, is_router,
+				     vlan->vlan_id);
+		}
+	}
+}
+
+/*
+ * bond_inet6addr_event: handle inet6addr notifier chain events.
+ *
+ * We keep track of device IPv6 addresses primarily to use as source
+ * addresses in NS probes.
+ *
+ * We track one IPv6 for the main device (if it has one).
+ */
+static int bond_inet6addr_event(struct notifier_block *this,
+				unsigned long event,
+				void *ptr)
+{
+	struct inet6_ifaddr *ifa = ptr;
+	struct net_device *vlan_dev, *event_dev = ifa->idev->dev;
+	struct bonding *bond;
+	struct vlan_entry *vlan;
+
+	if (dev_net(event_dev) != &init_net)
+		return NOTIFY_DONE;
+
+	list_for_each_entry(bond, &bond_dev_list, bond_list) {
+		if (bond->dev == event_dev) {
+			switch (event) {
+			case NETDEV_UP:
+				if (ipv6_addr_any(&bond->master_ipv6))
+					ipv6_addr_copy(&bond->master_ipv6,
+						       &ifa->addr);
+				return NOTIFY_OK;
+			case NETDEV_DOWN:
+				if (ipv6_addr_equal(&bond->master_ipv6,
+						    &ifa->addr))
+					bond_glean_dev_ipv6(bond->dev,
+							    &bond->master_ipv6);
+				return NOTIFY_OK;
+			default:
+				return NOTIFY_DONE;
+			}
+		}
+
+		list_for_each_entry(vlan, &bond->vlan_list, vlan_list) {
+			vlan_dev = vlan_group_get_device(bond->vlgrp,
+							 vlan->vlan_id);
+			if (vlan_dev == event_dev) {
+				switch (event) {
+				case NETDEV_UP:
+					if (ipv6_addr_any(&vlan->vlan_ipv6))
+						ipv6_addr_copy(&vlan->vlan_ipv6,
+							       &ifa->addr);
+					return NOTIFY_OK;
+				case NETDEV_DOWN:
+					if (ipv6_addr_equal(&vlan->vlan_ipv6,
+							    &ifa->addr))
+						bond_glean_dev_ipv6(vlan_dev,
+								    &vlan->vlan_ipv6);
+					return NOTIFY_OK;
+				default:
+					return NOTIFY_DONE;
+				}
+			}
+		}
+	}
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block bond_inet6addr_notifier = {
+	.notifier_call = bond_inet6addr_event,
+};
+
+void bond_register_ipv6_notifier(void)
+{
+	register_inet6addr_notifier(&bond_inet6addr_notifier);
+}
+
+void bond_unregister_ipv6_notifier(void)
+{
+	unregister_inet6addr_notifier(&bond_inet6addr_notifier);
+}
+
drivers/net/bonding/bond_main.c +31 -2
index 39575d7..798d98c 100644
@@ -89,6 +89,7 @@
 
 static int max_bonds	= BOND_DEFAULT_MAX_BONDS;
 static int num_grat_arp = 1;
+static int num_unsol_na = 1;
 static int miimon	= BOND_LINK_MON_INTERV;
 static int updelay	= 0;
 static int downdelay	= 0;
@@ -107,6 +108,8 @@ module_param(max_bonds, int, 0);
 MODULE_PARM_DESC(max_bonds, "Max number of bonded devices");
 module_param(num_grat_arp, int, 0644);
 MODULE_PARM_DESC(num_grat_arp, "Number of gratuitous ARP packets to send on failover event");
+module_param(num_unsol_na, int, 0644);
+MODULE_PARM_DESC(num_unsol_na, "Number of unsolicited IPv6 Neighbor Advertisements packets to send on failover event");
 module_param(miimon, int, 0);
 MODULE_PARM_DESC(miimon, "Link check interval in milliseconds");
 module_param(updelay, int, 0);
@@ -242,14 +245,13 @@ static int bond_add_vlan(struct bonding *bond, unsigned short vlan_id)
 	dprintk("bond: %s, vlan id %d\n",
 		(bond ? bond->dev->name: "None"), vlan_id);
 
-	vlan = kmalloc(sizeof(struct vlan_entry), GFP_KERNEL);
+	vlan = kzalloc(sizeof(struct vlan_entry), GFP_KERNEL);
 	if (!vlan) {
 		return -ENOMEM;
 	}
 
 	INIT_LIST_HEAD(&vlan->vlan_list);
 	vlan->vlan_id = vlan_id;
-	vlan->vlan_ip = 0;
 
 	write_lock_bh(&bond->lock);
 
@@ -1208,6 +1210,9 @@ void bond_change_active_slave(struct bonding *bond, struct slave *new_active)
 			bond->send_grat_arp = bond->params.num_grat_arp;
 			bond_send_gratuitous_arp(bond);
 
+			bond->send_unsol_na = bond->params.num_unsol_na;
+			bond_send_unsolicited_na(bond);
+
 			write_unlock_bh(&bond->curr_slave_lock);
 			read_unlock(&bond->lock);
 
@@ -2463,6 +2468,12 @@ void bond_mii_monitor(struct work_struct *work)
 		read_unlock(&bond->curr_slave_lock);
 	}
 
+	if (bond->send_unsol_na) {
+		read_lock(&bond->curr_slave_lock);
+		bond_send_unsolicited_na(bond);
+		read_unlock(&bond->curr_slave_lock);
+	}
+
 	if (bond_miimon_inspect(bond)) {
 		read_unlock(&bond->lock);
 		rtnl_lock();
@@ -3158,6 +3169,12 @@ void bond_activebackup_arp_mon(struct work_struct *work)
 		read_unlock(&bond->curr_slave_lock);
 	}
 
+	if (bond->send_unsol_na) {
+		read_lock(&bond->curr_slave_lock);
+		bond_send_unsolicited_na(bond);
+		read_unlock(&bond->curr_slave_lock);
+	}
+
 	if (bond_ab_arp_inspect(bond, delta_in_ticks)) {
 		read_unlock(&bond->lock);
 		rtnl_lock();
@@ -3827,6 +3844,7 @@ static int bond_close(struct net_device *bond_dev)
 	write_lock_bh(&bond->lock);
 
 	bond->send_grat_arp = 0;
+	bond->send_unsol_na = 0;
 
 	/* signal timers not to re-arm */
 	bond->kill_timers = 1;
@@ -4542,6 +4560,7 @@ static int bond_init(struct net_device *bond_dev, struct bond_params *params)
 	bond->primary_slave = NULL;
 	bond->dev = bond_dev;
 	bond->send_grat_arp = 0;
+	bond->send_unsol_na = 0;
 	bond->setup_by_slave = 0;
 	INIT_LIST_HEAD(&bond->vlan_list);
 
@@ -4791,6 +4810,13 @@ static int bond_check_params(struct bond_params *params)
 		num_grat_arp = 1;
 	}
 
+	if (num_unsol_na < 0 || num_unsol_na > 255) {
+		printk(KERN_WARNING DRV_NAME
+		       ": Warning: num_unsol_na (%d) not in range 0-255 so it "
+		       "was reset to 1 \n", num_unsol_na);
+		num_unsol_na = 1;
+	}
+
 	/* reset values for 802.3ad */
 	if (bond_mode == BOND_MODE_8023AD) {
 		if (!miimon) {
@@ -4992,6 +5018,7 @@ static int bond_check_params(struct bond_params *params)
 	params->xmit_policy = xmit_hashtype;
 	params->miimon = miimon;
 	params->num_grat_arp = num_grat_arp;
+	params->num_unsol_na = num_unsol_na;
 	params->arp_interval = arp_interval;
 	params->arp_validate = arp_validate_value;
 	params->updelay = updelay;
@@ -5144,6 +5171,7 @@ static int __init bonding_init(void)
 
 	register_netdevice_notifier(&bond_netdev_notifier);
 	register_inetaddr_notifier(&bond_inetaddr_notifier);
+	bond_register_ipv6_notifier();
 
 	goto out;
 err:
@@ -5166,6 +5194,7 @@ static void __exit bonding_exit(void)
 {
 	unregister_netdevice_notifier(&bond_netdev_notifier);
 	unregister_inetaddr_notifier(&bond_inetaddr_notifier);
+	bond_unregister_ipv6_notifier();
 
 	bond_destroy_sysfs();
 
drivers/net/bonding/bond_sysfs.c +42 -0
index e400d7d..8788e3e 100644
@@ -983,6 +983,47 @@ out:
 	return ret;
 }
 static DEVICE_ATTR(num_grat_arp, S_IRUGO | S_IWUSR, bonding_show_n_grat_arp, bonding_store_n_grat_arp);
+
+/*
+ * Show and set the number of unsolicted NA's to send after a failover event.
+ */
+static ssize_t bonding_show_n_unsol_na(struct device *d,
+				       struct device_attribute *attr,
+				       char *buf)
+{
+	struct bonding *bond = to_bond(d);
+
+	return sprintf(buf, "%d\n", bond->params.num_unsol_na);
+}
+
+static ssize_t bonding_store_n_unsol_na(struct device *d,
+					struct device_attribute *attr,
+					const char *buf, size_t count)
+{
+	int new_value, ret = count;
+	struct bonding *bond = to_bond(d);
+
+	if (sscanf(buf, "%d", &new_value) != 1) {
+		printk(KERN_ERR DRV_NAME
+		       ": %s: no num_unsol_na value specified.\n",
+		       bond->dev->name);
+		ret = -EINVAL;
+		goto out;
+	}
+	if (new_value < 0 || new_value > 255) {
+		printk(KERN_ERR DRV_NAME
+		       ": %s: Invalid num_unsol_na value %d not in range 0-255; rejected.\n",
+		       bond->dev->name, new_value);
+		ret = -EINVAL;
+		goto out;
+	} else {
+		bond->params.num_unsol_na = new_value;
+	}
+out:
+	return ret;
+}
+static DEVICE_ATTR(num_unsol_na, S_IRUGO | S_IWUSR, bonding_show_n_unsol_na, bonding_store_n_unsol_na);
+
 /*
  * Show and set the MII monitor interval.  There are two tricky bits
  * here.  First, if MII monitoring is activated, then we must disable
@@ -1420,6 +1461,7 @@ static struct attribute *per_bond_attrs[] = {
 	&dev_attr_lacp_rate.attr,
 	&dev_attr_xmit_hash_policy.attr,
 	&dev_attr_num_grat_arp.attr,
+	&dev_attr_num_unsol_na.attr,
 	&dev_attr_miimon.attr,
 	&dev_attr_primary.attr,
 	&dev_attr_use_carrier.attr,
drivers/net/bonding/bonding.h +32 -2
index ffb668d..0491c7c 100644
@@ -19,16 +19,19 @@
 #include <linux/proc_fs.h>
 #include <linux/if_bonding.h>
 #include <linux/kobject.h>
+#include <linux/in6.h>
 #include "bond_3ad.h"
 #include "bond_alb.h"
 
-#define DRV_VERSION	"3.3.0"
-#define DRV_RELDATE	"June 10, 2008"
+#define DRV_VERSION	"3.4.0"
+#define DRV_RELDATE	"October 7, 2008"
 #define DRV_NAME	"bonding"
 #define DRV_DESCRIPTION	"Ethernet Channel Bonding Driver"
 
 #define BOND_MAX_ARP_TARGETS	16
 
+extern struct list_head bond_dev_list;
+
 #ifdef BONDING_DEBUG
 #define dprintk(fmt, args...) \
 	printk(KERN_DEBUG     \
@@ -126,6 +129,7 @@ struct bond_params {
 	int xmit_policy;
 	int miimon;
 	int num_grat_arp;
+	int num_unsol_na;
 	int arp_interval;
 	int arp_validate;
 	int use_carrier;
@@ -148,6 +152,9 @@ struct vlan_entry {
 	struct list_head vlan_list;
 	__be32 vlan_ip;
 	unsigned short vlan_id;
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+	struct in6_addr vlan_ipv6;
+#endif
 };
 
 struct slave {
@@ -195,6 +202,7 @@ struct bonding {
 	rwlock_t curr_slave_lock;
 	s8       kill_timers;
 	s8	 send_grat_arp;
+	s8	 send_unsol_na;
 	s8	 setup_by_slave;
 	struct   net_device_stats stats;
 #ifdef CONFIG_PROC_FS
@@ -218,6 +226,9 @@ struct bonding {
 	struct   delayed_work arp_work;
 	struct   delayed_work alb_work;
 	struct   delayed_work ad_work;
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+	struct   in6_addr master_ipv6;
+#endif
 };
 
 /**
@@ -341,5 +352,24 @@ extern struct bond_parm_tbl xmit_hashtype_tbl[];
 extern struct bond_parm_tbl arp_validate_tbl[];
 extern struct bond_parm_tbl fail_over_mac_tbl[];
 
+#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
+void bond_send_unsolicited_na(struct bonding *bond);
+void bond_register_ipv6_notifier(void);
+void bond_unregister_ipv6_notifier(void);
+#else
+static inline void bond_send_unsolicited_na(struct bonding *bond)
+{
+	return;
+}
+static inline void bond_register_ipv6_notifier(void)
+{
+	return;
+}
+static inline void bond_unregister_ipv6_notifier(void)
+{
+	return;
+}
+#endif
+
 #endif /* _LINUX_BONDING_H */
 
include/net/ndisc.h +14 -0
index 11dd013..ce532f2 100644
@@ -108,6 +108,20 @@ extern void			ndisc_send_redirect(struct sk_buff *skb,
 
 extern int			ndisc_mc_map(struct in6_addr *addr, char *buf, struct net_device *dev, int dir);
 
+extern struct sk_buff		*ndisc_build_skb(struct net_device *dev,
+						 const struct in6_addr *daddr,
+						 const struct in6_addr *saddr,
+						 struct icmp6hdr *icmp6h,
+						 const struct in6_addr *target,
+						 int llinfo);
+
+extern void			ndisc_send_skb(struct sk_buff *skb,
+					       struct net_device *dev,
+					       struct neighbour *neigh,
+					       const struct in6_addr *daddr,
+					       const struct in6_addr *saddr,
+					       struct icmp6hdr *icmp6h);
+
 
 
 /*
net/ipv6/ndisc.c +56 -27
index 2a6752d..fbf451c 100644
@@ -437,38 +437,20 @@ static void pndisc_destructor(struct pneigh_entry *n)
 	ipv6_dev_mc_dec(dev, &maddr);
 }
 
-/*
- *	Send a Neighbour Advertisement
- */
-static void __ndisc_send(struct net_device *dev,
-			 struct neighbour *neigh,
-			 const struct in6_addr *daddr,
-			 const struct in6_addr *saddr,
-			 struct icmp6hdr *icmp6h, const struct in6_addr *target,
-			 int llinfo)
+struct sk_buff *ndisc_build_skb(struct net_device *dev,
+				const struct in6_addr *daddr,
+				const struct in6_addr *saddr,
+				struct icmp6hdr *icmp6h,
+				const struct in6_addr *target,
+				int llinfo)
 {
-	struct flowi fl;
-	struct dst_entry *dst;
 	struct net *net = dev_net(dev);
 	struct sock *sk = net->ipv6.ndisc_sk;
 	struct sk_buff *skb;
 	struct icmp6hdr *hdr;
-	struct inet6_dev *idev;
 	int len;
 	int err;
-	u8 *opt, type;
-
-	type = icmp6h->icmp6_type;
-
-	icmpv6_flow_init(sk, &fl, type, saddr, daddr, dev->ifindex);
-
-	dst = icmp6_dst_alloc(dev, neigh, daddr);
-	if (!dst)
-		return;
-
-	err = xfrm_lookup(&dst, &fl, NULL, 0);
-	if (err < 0)
-		return;
+	u8 *opt;
 
 	if (!dev->addr_len)
 		llinfo = 0;
@@ -485,8 +467,7 @@ static void __ndisc_send(struct net_device *dev,
 		ND_PRINTK0(KERN_ERR
 			   "ICMPv6 ND: %s() failed to allocate an skb.\n",
 			   __func__);
-		dst_release(dst);
-		return;
+		return NULL;
 	}
 
 	skb_reserve(skb, LL_RESERVED_SPACE(dev));
@@ -513,6 +494,42 @@ static void __ndisc_send(struct net_device *dev,
 					   csum_partial((__u8 *) hdr,
 							len, 0));
 
+	return skb;
+}
+
+EXPORT_SYMBOL(ndisc_build_skb);
+
+void ndisc_send_skb(struct sk_buff *skb,
+		    struct net_device *dev,
+		    struct neighbour *neigh,
+		    const struct in6_addr *daddr,
+		    const struct in6_addr *saddr,
+		    struct icmp6hdr *icmp6h)
+{
+	struct flowi fl;
+	struct dst_entry *dst;
+	struct net *net = dev_net(dev);
+	struct sock *sk = net->ipv6.ndisc_sk;
+	struct inet6_dev *idev;
+	int err;
+	u8 type;
+
+	type = icmp6h->icmp6_type;
+
+	icmpv6_flow_init(sk, &fl, type, saddr, daddr, dev->ifindex);
+
+	dst = icmp6_dst_alloc(dev, neigh, daddr);
+	if (!dst) {
+		kfree_skb(skb);
+		return;
+	}
+
+	err = xfrm_lookup(&dst, &fl, NULL, 0);
+	if (err < 0) {
+		kfree_skb(skb);
+		return;
+	}
+
 	skb->dst = dst;
 
 	idev = in6_dev_get(dst->dev);
@@ -529,6 +546,27 @@ static void __ndisc_send(struct net_device *dev,
 		in6_dev_put(idev);
 }
 
+EXPORT_SYMBOL(ndisc_send_skb);
+
+/*
+ *	Send a Neighbour Discover packet
+ */
+static void __ndisc_send(struct net_device *dev,
+			 struct neighbour *neigh,
+			 const struct in6_addr *daddr,
+			 const struct in6_addr *saddr,
+			 struct icmp6hdr *icmp6h, const struct in6_addr *target,
+			 int llinfo)
+{

net/ipv6/ndisc.c static void __ndisc_send(struct net_device *dev, · line 570

Quoted 10 lines
net/ipv6/ndisc.c +9 -0
+	struct sk_buff *skb;
+
+	skb = ndisc_build_skb(dev, daddr, saddr, icmp6h, target, llinfo);
+	if (!skb)
+		return;
+
+	ndisc_send_skb(skb, dev, neigh, daddr, saddr, icmp6h);
+}
+
 static void ndisc_send_na(struct net_device *dev, struct neighbour *neigh,

applied 1-3