From 6cf3f41e6c08bca6641a695449791c38a25f35ff Mon Sep 17 00:00:00 2001
From: Jay Vosburgh <fubar@us.ibm.com>
Date: Mon, 3 Nov 2008 18:16:50 -0800
Subject: bonding, net: Move last_rx update into bonding recv logic

	The only user of the net_device->last_rx field is bonding.
This patch adds a conditional update of last_rx to the bonding special
logic in skb_bond_should_drop, causing last_rx to only be updated when
the ARP monitor is running.

	This frees network device drivers from the necessity of
updating last_rx, which can have cache line thrash issues.

Signed-off-by: Jay Vosburgh <fubar@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 32 ++++++++++++++++++--------------
 1 file changed, 18 insertions(+), 14 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 9d77b1d7dca..f1b0dbe5846 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1742,22 +1742,26 @@ static inline int skb_bond_should_drop(struct sk_buff *skb)
 	struct net_device *dev = skb->dev;
 	struct net_device *master = dev->master;
 
-	if (master &&
-	    (dev->priv_flags & IFF_SLAVE_INACTIVE)) {
-		if ((dev->priv_flags & IFF_SLAVE_NEEDARP) &&
-		    skb->protocol == __constant_htons(ETH_P_ARP))
-			return 0;
-
-		if (master->priv_flags & IFF_MASTER_ALB) {
-			if (skb->pkt_type != PACKET_BROADCAST &&
-			    skb->pkt_type != PACKET_MULTICAST)
+	if (master) {
+		if (master->priv_flags & IFF_MASTER_ARPMON)
+			dev->last_rx = jiffies;
+
+		if (dev->priv_flags & IFF_SLAVE_INACTIVE) {
+			if ((dev->priv_flags & IFF_SLAVE_NEEDARP) &&
+			    skb->protocol == __constant_htons(ETH_P_ARP))
 				return 0;
-		}
-		if (master->priv_flags & IFF_MASTER_8023AD &&
-		    skb->protocol == __constant_htons(ETH_P_SLOW))
-			return 0;
 
-		return 1;
+			if (master->priv_flags & IFF_MASTER_ALB) {
+				if (skb->pkt_type != PACKET_BROADCAST &&
+				    skb->pkt_type != PACKET_MULTICAST)
+					return 0;
+			}
+			if (master->priv_flags & IFF_MASTER_8023AD &&
+			    skb->protocol == __constant_htons(ETH_P_SLOW))
+				return 0;
+
+			return 1;
+		}
 	}
 	return 0;
 }
-- 
cgit v1.2.3-70-g09d2


From ae33bc40c0d96d02f51a996482ea7e41c5152695 Mon Sep 17 00:00:00 2001
From: "Eric W. Biederman" <ebiederm@xmission.com>
Date: Wed, 5 Nov 2008 16:00:02 -0800
Subject: net: Guaranetee the proper ordering of the loopback device.

I was recently hunting a bug that occurred in network namespace
cleanup.  In looking at the code it became apparrent that we have
and will continue to have cases where if we have anything going
on in a network namespace there will be assumptions that the
loopback device is present.   Things like sending igmp unsubscribe
messages when we bring down network devices invokes the routing
code which assumes that at least the loopback driver is present.

Therefore to avoid magic initcall ordering hackery that is hard
to follow and hard to get right insert a call to register the
loopback device directly from net_dev_init().    This guarantes
that the loopback device is the first device registered and
the last network device to go away.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/loopback.c    | 13 ++-----------
 include/linux/netdevice.h |  1 +
 net/core/dev.c            | 12 ++++++++++++
 3 files changed, 15 insertions(+), 11 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index 91d08585a6d..c4516b580ba 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -204,17 +204,8 @@ static __net_exit void loopback_net_exit(struct net *net)
 	unregister_netdev(dev);
 }
 
-static struct pernet_operations __net_initdata loopback_net_ops = {
+/* Registered in net/core/dev.c */
+struct pernet_operations __net_initdata loopback_net_ops = {
        .init = loopback_net_init,
        .exit = loopback_net_exit,
 };
-
-static int __init loopback_init(void)
-{
-	return register_pernet_device(&loopback_net_ops);
-}
-
-/* Loopback is special. It should be initialized before any other network
- * device and network subsystem.
- */
-fs_initcall(loopback_init);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index f1b0dbe5846..12d7f4469dc 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1766,6 +1766,7 @@ static inline int skb_bond_should_drop(struct sk_buff *skb)
 	return 0;
 }
 
+extern struct pernet_operations __net_initdata loopback_net_ops;
 #endif /* __KERNEL__ */
 
 #endif	/* _LINUX_DEV_H */
diff --git a/net/core/dev.c b/net/core/dev.c
index 9475f3e624a..811507c3980 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4904,6 +4904,18 @@ static int __init net_dev_init(void)
 	if (register_pernet_subsys(&netdev_net_ops))
 		goto out;
 
+	/* The loopback device is special if any other network devices
+	 * is present in a network namespace the loopback device must
+	 * be present. Since we now dynamically allocate and free the
+	 * loopback device ensure this invariant is maintained by
+	 * keeping the loopback device as the first device on the
+	 * list of network devices.  Ensuring the loopback devices
+	 * is the first device that appears and the last network device
+	 * that disappears.
+	 */
+	if (register_pernet_device(&loopback_net_ops))
+		goto out;
+
 	if (register_pernet_device(&default_device_ops))
 		goto out;
 
-- 
cgit v1.2.3-70-g09d2


From 3d8160b1493bcadca74fbb635d79b3928b8999cf Mon Sep 17 00:00:00 2001
From: "David S. Miller" <davem@davemloft.net>
Date: Fri, 7 Nov 2008 22:52:14 -0800
Subject: Revert "net: Guaranetee the proper ordering of the loopback device."

This reverts commit ae33bc40c0d96d02f51a996482ea7e41c5152695.
---
 drivers/net/loopback.c    | 13 +++++++++++--
 include/linux/netdevice.h |  1 -
 net/core/dev.c            | 12 ------------
 3 files changed, 11 insertions(+), 15 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index c4516b580ba..91d08585a6d 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -204,8 +204,17 @@ static __net_exit void loopback_net_exit(struct net *net)
 	unregister_netdev(dev);
 }
 
-/* Registered in net/core/dev.c */
-struct pernet_operations __net_initdata loopback_net_ops = {
+static struct pernet_operations __net_initdata loopback_net_ops = {
        .init = loopback_net_init,
        .exit = loopback_net_exit,
 };
+
+static int __init loopback_init(void)
+{
+	return register_pernet_device(&loopback_net_ops);
+}
+
+/* Loopback is special. It should be initialized before any other network
+ * device and network subsystem.
+ */
+fs_initcall(loopback_init);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 12d7f4469dc..f1b0dbe5846 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1766,7 +1766,6 @@ static inline int skb_bond_should_drop(struct sk_buff *skb)
 	return 0;
 }
 
-extern struct pernet_operations __net_initdata loopback_net_ops;
 #endif /* __KERNEL__ */
 
 #endif	/* _LINUX_DEV_H */
diff --git a/net/core/dev.c b/net/core/dev.c
index e0dc67a789b..2306d56fbb5 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4909,18 +4909,6 @@ static int __init net_dev_init(void)
 	if (register_pernet_subsys(&netdev_net_ops))
 		goto out;
 
-	/* The loopback device is special if any other network devices
-	 * is present in a network namespace the loopback device must
-	 * be present. Since we now dynamically allocate and free the
-	 * loopback device ensure this invariant is maintained by
-	 * keeping the loopback device as the first device on the
-	 * list of network devices.  Ensuring the loopback devices
-	 * is the first device that appears and the last network device
-	 * that disappears.
-	 */
-	if (register_pernet_device(&loopback_net_ops))
-		goto out;
-
 	if (register_pernet_device(&default_device_ops))
 		goto out;
 
-- 
cgit v1.2.3-70-g09d2


From 505d4f73dda9e20d59da05008f1f5eb432613e71 Mon Sep 17 00:00:00 2001
From: "Eric W. Biederman" <ebiederm@maxwell.aristanetworks.com>
Date: Fri, 7 Nov 2008 22:54:20 -0800
Subject: net: Guaranetee the proper ordering of the loopback device. v2

I was recently hunting a bug that occurred in network namespace
cleanup.  In looking at the code it became apparrent that we have
and will continue to have cases where if we have anything going
on in a network namespace there will be assumptions that the
loopback device is present.   Things like sending igmp unsubscribe
messages when we bring down network devices invokes the routing
code which assumes that at least the loopback driver is present.

Therefore to avoid magic initcall ordering hackery that is hard
to follow and hard to get right insert a call to register the
loopback device directly from net_dev_init().    This guarantes
that the loopback device is the first device registered and
the last network device to go away.

But do it carefully so we register the loopback device after
we clear dev_boot_phase.

Signed-off-by: Eric W. Biederman <ebiederm@maxwell.aristanetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/loopback.c    | 13 ++-----------
 include/linux/netdevice.h |  1 +
 net/core/dev.c            | 22 +++++++++++++++++-----
 3 files changed, 20 insertions(+), 16 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index 91d08585a6d..c4516b580ba 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -204,17 +204,8 @@ static __net_exit void loopback_net_exit(struct net *net)
 	unregister_netdev(dev);
 }
 
-static struct pernet_operations __net_initdata loopback_net_ops = {
+/* Registered in net/core/dev.c */
+struct pernet_operations __net_initdata loopback_net_ops = {
        .init = loopback_net_init,
        .exit = loopback_net_exit,
 };
-
-static int __init loopback_init(void)
-{
-	return register_pernet_device(&loopback_net_ops);
-}
-
-/* Loopback is special. It should be initialized before any other network
- * device and network subsystem.
- */
-fs_initcall(loopback_init);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index f1b0dbe5846..12d7f4469dc 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1766,6 +1766,7 @@ static inline int skb_bond_should_drop(struct sk_buff *skb)
 	return 0;
 }
 
+extern struct pernet_operations __net_initdata loopback_net_ops;
 #endif /* __KERNEL__ */
 
 #endif	/* _LINUX_DEV_H */
diff --git a/net/core/dev.c b/net/core/dev.c
index 2306d56fbb5..31568b2068a 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4909,9 +4909,6 @@ static int __init net_dev_init(void)
 	if (register_pernet_subsys(&netdev_net_ops))
 		goto out;
 
-	if (register_pernet_device(&default_device_ops))
-		goto out;
-
 	/*
 	 *	Initialise the packet receive queues.
 	 */
@@ -4928,10 +4925,25 @@ static int __init net_dev_init(void)
 		queue->backlog.weight = weight_p;
 	}
 
-	netdev_dma_register();
-
 	dev_boot_phase = 0;
 
+	/* The loopback device is special if any other network devices
+	 * is present in a network namespace the loopback device must
+	 * be present. Since we now dynamically allocate and free the
+	 * loopback device ensure this invariant is maintained by
+	 * keeping the loopback device as the first device on the
+	 * list of network devices.  Ensuring the loopback devices
+	 * is the first device that appears and the last network device
+	 * that disappears.
+	 */
+	if (register_pernet_device(&loopback_net_ops))
+		goto out;
+
+	if (register_pernet_device(&default_device_ops))
+		goto out;
+
+	netdev_dma_register();
+
 	open_softirq(NET_TX_SOFTIRQ, net_tx_action);
 	open_softirq(NET_RX_SOFTIRQ, net_rx_action);
 
-- 
cgit v1.2.3-70-g09d2


From d314774cf2cd5dfeb39a00d37deee65d4c627927 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Wed, 19 Nov 2008 21:32:24 -0800
Subject: netdev: network device operations infrastructure

This patch changes the network device internal API to move adminstrative
operations out of the network device structure and into a separate structure.

This patch involves some hackery to maintain compatablity between the
new and old model, so all 300+ drivers don't have to be changed at once.
For drivers that aren't converted yet, the netdevice_ops virt function list
still resides in the net_device structure. For old protocols, the new
net_device_ops are copied out to the old net_device pointers.

After the transistion is completed the nag message can be changed to
an WARN_ON, and the compatiablity code can be made configurable.

Some function pointers aren't moved:
* destructor can't be in net_device_ops because
  it may need to be referenced after the module is unloaded.
* neighbor setup is manipulated in a couple of places that need special
  consideration
* hard_start_xmit is in the fast path for transmit.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 232 +++++++++++++++++++++++++++++++++-------------
 net/Kconfig               |   3 +
 net/core/dev.c            | 109 +++++++++++++++-------
 net/core/netpoll.c        |   7 +-
 net/core/rtnetlink.c      |   9 +-
 net/sched/sch_generic.c   |   4 +-
 6 files changed, 259 insertions(+), 105 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 12d7f4469dc..9060f5f3517 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -451,6 +451,131 @@ struct netdev_queue {
 	struct Qdisc		*qdisc_sleeping;
 } ____cacheline_aligned_in_smp;
 
+
+/*
+ * This structure defines the management hooks for network devices.
+ * The following hooks can bed defined and are optonal (can be null)
+ * unless otherwise noted.
+ *
+ * int (*ndo_init)(struct net_device *dev);
+ *     This function is called once when network device is registered.
+ *     The network device can use this to any late stage initializaton
+ *     or semantic validattion. It can fail with an error code which will
+ *     be propogated back to register_netdev
+ *
+ * void (*ndo_uninit)(struct net_device *dev);
+ *     This function is called when device is unregistered or when registration
+ *     fails. It is not called if init fails.
+ *
+ * int (*ndo_open)(struct net_device *dev);
+ *     This function is called when network device transistions to the up
+ *     state.
+ *
+ * int (*ndo_stop)(struct net_device *dev);
+ *     This function is called when network device transistions to the down
+ *     state.
+ *
+ * void (*ndo_change_rx_flags)(struct net_device *dev, int flags);
+ *	This function is called to allow device receiver to make
+ *	changes to configuration when multicast or promiscious is enabled.
+ *
+ * void (*ndo_set_rx_mode)(struct net_device *dev);
+ *	This function is called device changes address list filtering.
+ *
+ * void (*ndo_set_multicast_list)(struct net_device *dev);
+ *	This function is called when the multicast address list changes.
+ *
+ * int (*ndo_set_mac_address)(struct net_device *dev, void *addr);
+ *	This function  is called when the Media Access Control address
+ *	needs to be changed. If not this interface is not defined, the
+ *	mac address can not be changed.
+ *
+ * int (*ndo_validate_addr)(struct net_device *dev);
+ *	Test if Media Access Control address is valid for the device.
+ *
+ * int (*ndo_do_ioctl)(struct net_device *dev, struct ifreq *ifr, int cmd);
+ *	Called when a user request an ioctl which can't be handled by
+ *	the generic interface code. If not defined ioctl's return
+ *	not supported error code.
+ *
+ * int (*ndo_set_config)(struct net_device *dev, struct ifmap *map);
+ *	Used to set network devices bus interface parameters. This interface
+ *	is retained for legacy reason, new devices should use the bus
+ *	interface (PCI) for low level management.
+ *
+ * int (*ndo_change_mtu)(struct net_device *dev, int new_mtu);
+ *	Called when a user wants to change the Maximum Transfer Unit
+ *	of a device. If not defined, any request to change MTU will
+ *	will return an error.
+ *
+ * void (*ndo_tx_timeout) (struct net_device *dev);
+ *	Callback uses when the transmitter has not made any progress
+ *	for dev->watchdog ticks.
+ *
+ * struct net_device_stats* (*get_stats)(struct net_device *dev);
+ *	Called when a user wants to get the network device usage
+ *	statistics. If not defined, the counters in dev->stats will
+ *	be used.
+ *
+ * void (*ndo_vlan_rx_register)(struct net_device *dev, struct vlan_group *grp);
+ *	If device support VLAN receive accleration
+ *	(ie. dev->features & NETIF_F_HW_VLAN_RX), then this function is called
+ *	when vlan groups for the device changes.  Note: grp is NULL
+ *	if no vlan's groups are being used.
+ *
+ * void (*ndo_vlan_rx_add_vid)(struct net_device *dev, unsigned short vid);
+ *	If device support VLAN filtering (dev->features & NETIF_F_HW_VLAN_FILTER)
+ *	this function is called when a VLAN id is registered.
+ *
+ * void (*ndo_vlan_rx_kill_vid)(struct net_device *dev, unsigned short vid);
+ *	If device support VLAN filtering (dev->features & NETIF_F_HW_VLAN_FILTER)
+ *	this function is called when a VLAN id is unregistered.
+ *
+ * void (*ndo_poll_controller)(struct net_device *dev);
+ */
+struct net_device_ops {
+	int			(*ndo_init)(struct net_device *dev);
+	void			(*ndo_uninit)(struct net_device *dev);
+	int			(*ndo_open)(struct net_device *dev);
+	int			(*ndo_stop)(struct net_device *dev);
+#define HAVE_CHANGE_RX_FLAGS
+	void			(*ndo_change_rx_flags)(struct net_device *dev,
+						       int flags);
+#define HAVE_SET_RX_MODE
+	void			(*ndo_set_rx_mode)(struct net_device *dev);
+#define HAVE_MULTICAST
+	void			(*ndo_set_multicast_list)(struct net_device *dev);
+#define HAVE_SET_MAC_ADDR
+	int			(*ndo_set_mac_address)(struct net_device *dev,
+						       void *addr);
+#define HAVE_VALIDATE_ADDR
+	int			(*ndo_validate_addr)(struct net_device *dev);
+#define HAVE_PRIVATE_IOCTL
+	int			(*ndo_do_ioctl)(struct net_device *dev,
+					        struct ifreq *ifr, int cmd);
+#define HAVE_SET_CONFIG
+	int			(*ndo_set_config)(struct net_device *dev,
+					          struct ifmap *map);
+#define HAVE_CHANGE_MTU
+	int			(*ndo_change_mtu)(struct net_device *dev, int new_mtu);
+
+#define HAVE_TX_TIMEOUT
+	void			(*ndo_tx_timeout) (struct net_device *dev);
+
+	struct net_device_stats* (*ndo_get_stats)(struct net_device *dev);
+
+	void			(*ndo_vlan_rx_register)(struct net_device *dev,
+						        struct vlan_group *grp);
+	void			(*ndo_vlan_rx_add_vid)(struct net_device *dev,
+						       unsigned short vid);
+	void			(*ndo_vlan_rx_kill_vid)(struct net_device *dev,
+						        unsigned short vid);
+#ifdef CONFIG_NET_POLL_CONTROLLER
+#define HAVE_NETDEV_POLL
+	void                    (*ndo_poll_controller)(struct net_device *dev);
+#endif
+};
+
 /*
  *	The DEVICE structure.
  *	Actually, this whole structure is a big mistake.  It mixes I/O
@@ -498,11 +623,6 @@ struct net_device
 #ifdef CONFIG_NETPOLL
 	struct list_head	napi_list;
 #endif
-	
-	/* The device initialization function. Called only once. */
-	int			(*init)(struct net_device *dev);
-
-	/* ------- Fields preinitialized in Space.c finish here ------- */
 
 	/* Net device features */
 	unsigned long		features;
@@ -546,15 +666,13 @@ struct net_device
 	 * for all in netdev_increment_features.
 	 */
 #define NETIF_F_ONE_FOR_ALL	(NETIF_F_GSO_SOFTWARE | NETIF_F_GSO_ROBUST | \
-				 NETIF_F_SG | NETIF_F_HIGHDMA | \
+				 NETIF_F_SG | NETIF_F_HIGHDMA |		\
 				 NETIF_F_FRAGLIST)
 
 	/* Interface index. Unique device identifier	*/
 	int			ifindex;
 	int			iflink;
 
-
-	struct net_device_stats* (*get_stats)(struct net_device *dev);
 	struct net_device_stats	stats;
 
 #ifdef CONFIG_WIRELESS_EXT
@@ -564,18 +682,13 @@ struct net_device
 	/* Instance data managed by the core of Wireless Extensions. */
 	struct iw_public_data *	wireless_data;
 #endif
+	/* Management operations */
+	const struct net_device_ops *netdev_ops;
 	const struct ethtool_ops *ethtool_ops;
 
 	/* Hardware header description */
 	const struct header_ops *header_ops;
 
-	/*
-	 * This marks the end of the "visible" part of the structure. All
-	 * fields hereafter are internal to the system, and may change at
-	 * will (read: may be cleaned up at will).
-	 */
-
-
 	unsigned int		flags;	/* interface flags (a la BSD)	*/
 	unsigned short		gflags;
         unsigned short          priv_flags; /* Like 'flags' but invisible to userspace. */
@@ -634,7 +747,7 @@ struct net_device
 	unsigned long		last_rx;	/* Time of last Rx	*/
 	/* Interface address info used in eth_type_trans() */
 	unsigned char		dev_addr[MAX_ADDR_LEN];	/* hw address, (before bcast 
-							because most packets are unicast) */
+							   because most packets are unicast) */
 
 	unsigned char		broadcast[MAX_ADDR_LEN];	/* hw bcast add	*/
 
@@ -648,6 +761,10 @@ struct net_device
 	/* Number of TX queues currently active in device  */
 	unsigned int		real_num_tx_queues;
 
+	/* Map buffer to appropriate transmit queue */
+	u16			(*select_queue)(struct net_device *dev,
+						struct sk_buff *skb);
+
 	unsigned long		tx_queue_len;	/* Max frames per queue allowed */
 	spinlock_t		tx_global_lock;
 /*
@@ -662,9 +779,6 @@ struct net_device
 	int			watchdog_timeo; /* used by dev_watchdog() */
 	struct timer_list	watchdog_timer;
 
-/*
- * refcnt is a very hot point, so align it on SMP
- */
 	/* Number of references to this device */
 	atomic_t		refcnt ____cacheline_aligned_in_smp;
 
@@ -683,56 +797,14 @@ struct net_device
 	       NETREG_RELEASED,		/* called free_netdev */
 	} reg_state;
 
-	/* Called after device is detached from network. */
-	void			(*uninit)(struct net_device *dev);
-	/* Called after last user reference disappears. */
-	void			(*destructor)(struct net_device *dev);
+	/* Called from unregister, can be used to call free_netdev */
+	void (*destructor)(struct net_device *dev);
 
-	/* Pointers to interface service routines.	*/
-	int			(*open)(struct net_device *dev);
-	int			(*stop)(struct net_device *dev);
-#define HAVE_NETDEV_POLL
-#define HAVE_CHANGE_RX_FLAGS
-	void			(*change_rx_flags)(struct net_device *dev,
-						   int flags);
-#define HAVE_SET_RX_MODE
-	void			(*set_rx_mode)(struct net_device *dev);
-#define HAVE_MULTICAST			 
-	void			(*set_multicast_list)(struct net_device *dev);
-#define HAVE_SET_MAC_ADDR  		 
-	int			(*set_mac_address)(struct net_device *dev,
-						   void *addr);
-#define HAVE_VALIDATE_ADDR
-	int			(*validate_addr)(struct net_device *dev);
-#define HAVE_PRIVATE_IOCTL
-	int			(*do_ioctl)(struct net_device *dev,
-					    struct ifreq *ifr, int cmd);
-#define HAVE_SET_CONFIG
-	int			(*set_config)(struct net_device *dev,
-					      struct ifmap *map);
-#define HAVE_CHANGE_MTU
-	int			(*change_mtu)(struct net_device *dev, int new_mtu);
+	int (*neigh_setup)(struct net_device *dev, struct neigh_parms *);
 
-#define HAVE_TX_TIMEOUT
-	void			(*tx_timeout) (struct net_device *dev);
-
-	void			(*vlan_rx_register)(struct net_device *dev,
-						    struct vlan_group *grp);
-	void			(*vlan_rx_add_vid)(struct net_device *dev,
-						   unsigned short vid);
-	void			(*vlan_rx_kill_vid)(struct net_device *dev,
-						    unsigned short vid);
-
-	int			(*neigh_setup)(struct net_device *dev, struct neigh_parms *);
 #ifdef CONFIG_NETPOLL
 	struct netpoll_info	*npinfo;
 #endif
-#ifdef CONFIG_NET_POLL_CONTROLLER
-	void                    (*poll_controller)(struct net_device *dev);
-#endif
-
-	u16			(*select_queue)(struct net_device *dev,
-						struct sk_buff *skb);
 
 #ifdef CONFIG_NET_NS
 	/* Network namespace this network device is inside */
@@ -763,6 +835,38 @@ struct net_device
 	/* for setting kernel sock attribute on TCP connection setup */
 #define GSO_MAX_SIZE		65536
 	unsigned int		gso_max_size;
+
+#ifdef CONFIG_COMPAT_NET_DEV_OPS
+	struct {
+		int			(*init)(struct net_device *dev);
+		void			(*uninit)(struct net_device *dev);
+		int			(*open)(struct net_device *dev);
+		int			(*stop)(struct net_device *dev);
+		void			(*change_rx_flags)(struct net_device *dev,
+							   int flags);
+		void			(*set_rx_mode)(struct net_device *dev);
+		void			(*set_multicast_list)(struct net_device *dev);
+		int			(*set_mac_address)(struct net_device *dev,
+							   void *addr);
+		int			(*validate_addr)(struct net_device *dev);
+		int			(*do_ioctl)(struct net_device *dev,
+						    struct ifreq *ifr, int cmd);
+		int			(*set_config)(struct net_device *dev,
+						      struct ifmap *map);
+		int			(*change_mtu)(struct net_device *dev, int new_mtu);
+		void			(*tx_timeout) (struct net_device *dev);
+		struct net_device_stats* (*get_stats)(struct net_device *dev);
+		void			(*vlan_rx_register)(struct net_device *dev,
+							    struct vlan_group *grp);
+		void			(*vlan_rx_add_vid)(struct net_device *dev,
+							   unsigned short vid);
+		void			(*vlan_rx_kill_vid)(struct net_device *dev,
+							    unsigned short vid);
+#ifdef CONFIG_NET_POLL_CONTROLLER
+		void                    (*poll_controller)(struct net_device *dev);
+#endif
+#endif
+	};
 };
 #define to_net_dev(d) container_of(d, struct net_device, dev)
 
diff --git a/net/Kconfig b/net/Kconfig
index 8c3d97ca0d9..4e2e40ba8ba 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -32,6 +32,9 @@ config NET_NS
 	  Allow user space to create what appear to be multiple instances
 	  of the network stack.
 
+config COMPAT_NET_DEV_OPS
+       def_bool y
+
 source "net/packet/Kconfig"
 source "net/unix/Kconfig"
 source "net/xfrm/Kconfig"
diff --git a/net/core/dev.c b/net/core/dev.c
index e08c0fcd603..ca14ab407b3 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1059,6 +1059,7 @@ void dev_load(struct net *net, const char *name)
  */
 int dev_open(struct net_device *dev)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
 	int ret = 0;
 
 	ASSERT_RTNL();
@@ -1081,11 +1082,11 @@ int dev_open(struct net_device *dev)
 	 */
 	set_bit(__LINK_STATE_START, &dev->state);
 
-	if (dev->validate_addr)
-		ret = dev->validate_addr(dev);
+	if (ops->ndo_validate_addr)
+		ret = ops->ndo_validate_addr(dev);
 
-	if (!ret && dev->open)
-		ret = dev->open(dev);
+	if (!ret && ops->ndo_open)
+		ret = ops->ndo_open(dev);
 
 	/*
 	 *	If it went open OK then:
@@ -1129,6 +1130,7 @@ int dev_open(struct net_device *dev)
  */
 int dev_close(struct net_device *dev)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
 	ASSERT_RTNL();
 
 	might_sleep();
@@ -1161,8 +1163,8 @@ int dev_close(struct net_device *dev)
 	 *	We allow it to be called even after a DETACH hot-plug
 	 *	event.
 	 */
-	if (dev->stop)
-		dev->stop(dev);
+	if (ops->ndo_stop)
+		ops->ndo_stop(dev);
 
 	/*
 	 *	Device is now down.
@@ -2930,8 +2932,10 @@ int netdev_set_master(struct net_device *slave, struct net_device *master)
 
 static void dev_change_rx_flags(struct net_device *dev, int flags)
 {
-	if (dev->flags & IFF_UP && dev->change_rx_flags)
-		dev->change_rx_flags(dev, flags);
+	const struct net_device_ops *ops = dev->netdev_ops;
+
+	if ((dev->flags & IFF_UP) && ops->ndo_change_rx_flags)
+		ops->ndo_change_rx_flags(dev, flags);
 }
 
 static int __dev_set_promiscuity(struct net_device *dev, int inc)
@@ -3051,6 +3055,8 @@ int dev_set_allmulti(struct net_device *dev, int inc)
  */
 void __dev_set_rx_mode(struct net_device *dev)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
+
 	/* dev_open will call this function so the list will stay sane. */
 	if (!(dev->flags&IFF_UP))
 		return;
@@ -3058,8 +3064,8 @@ void __dev_set_rx_mode(struct net_device *dev)
 	if (!netif_device_present(dev))
 		return;
 
-	if (dev->set_rx_mode)
-		dev->set_rx_mode(dev);
+	if (ops->ndo_set_rx_mode)
+		ops->ndo_set_rx_mode(dev);
 	else {
 		/* Unicast addresses changes may only happen under the rtnl,
 		 * therefore calling __dev_set_promiscuity here is safe.
@@ -3072,8 +3078,8 @@ void __dev_set_rx_mode(struct net_device *dev)
 			dev->uc_promisc = 0;
 		}
 
-		if (dev->set_multicast_list)
-			dev->set_multicast_list(dev);
+		if (ops->ndo_set_multicast_list)
+			ops->ndo_set_multicast_list(dev);
 	}
 }
 
@@ -3432,6 +3438,7 @@ int dev_change_flags(struct net_device *dev, unsigned flags)
  */
 int dev_set_mtu(struct net_device *dev, int new_mtu)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
 	int err;
 
 	if (new_mtu == dev->mtu)
@@ -3445,10 +3452,11 @@ int dev_set_mtu(struct net_device *dev, int new_mtu)
 		return -ENODEV;
 
 	err = 0;
-	if (dev->change_mtu)
-		err = dev->change_mtu(dev, new_mtu);
+	if (ops->ndo_change_mtu)
+		err = ops->ndo_change_mtu(dev, new_mtu);
 	else
 		dev->mtu = new_mtu;
+
 	if (!err && dev->flags & IFF_UP)
 		call_netdevice_notifiers(NETDEV_CHANGEMTU, dev);
 	return err;
@@ -3463,15 +3471,16 @@ int dev_set_mtu(struct net_device *dev, int new_mtu)
  */
 int dev_set_mac_address(struct net_device *dev, struct sockaddr *sa)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
 	int err;
 
-	if (!dev->set_mac_address)
+	if (!ops->ndo_set_mac_address)
 		return -EOPNOTSUPP;
 	if (sa->sa_family != dev->type)
 		return -EINVAL;
 	if (!netif_device_present(dev))
 		return -ENODEV;
-	err = dev->set_mac_address(dev, sa);
+	err = ops->ndo_set_mac_address(dev, sa);
 	if (!err)
 		call_netdevice_notifiers(NETDEV_CHANGEADDR, dev);
 	return err;
@@ -3551,6 +3560,7 @@ static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd)
 {
 	int err;
 	struct net_device *dev = __dev_get_by_name(net, ifr->ifr_name);
+	const struct net_device_ops *ops = dev->netdev_ops;
 
 	if (!dev)
 		return -ENODEV;
@@ -3578,15 +3588,15 @@ static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd)
 			return 0;
 
 		case SIOCSIFMAP:
-			if (dev->set_config) {
+			if (ops->ndo_set_config) {
 				if (!netif_device_present(dev))
 					return -ENODEV;
-				return dev->set_config(dev, &ifr->ifr_map);
+				return ops->ndo_set_config(dev, &ifr->ifr_map);
 			}
 			return -EOPNOTSUPP;
 
 		case SIOCADDMULTI:
-			if ((!dev->set_multicast_list && !dev->set_rx_mode) ||
+			if ((!ops->ndo_set_multicast_list && !ops->ndo_set_rx_mode) ||
 			    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
 				return -EINVAL;
 			if (!netif_device_present(dev))
@@ -3595,7 +3605,7 @@ static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd)
 					  dev->addr_len, 1);
 
 		case SIOCDELMULTI:
-			if ((!dev->set_multicast_list && !dev->set_rx_mode) ||
+			if ((!ops->ndo_set_multicast_list && !ops->ndo_set_rx_mode) ||
 			    ifr->ifr_hwaddr.sa_family != AF_UNSPEC)
 				return -EINVAL;
 			if (!netif_device_present(dev))
@@ -3633,10 +3643,9 @@ static int dev_ifsioc(struct net *net, struct ifreq *ifr, unsigned int cmd)
 			    cmd == SIOCBRDELIF ||
 			    cmd == SIOCWANDEV) {
 				err = -EOPNOTSUPP;
-				if (dev->do_ioctl) {
+				if (ops->ndo_do_ioctl) {
 					if (netif_device_present(dev))
-						err = dev->do_ioctl(dev, ifr,
-								    cmd);
+						err = ops->ndo_do_ioctl(dev, ifr, cmd);
 					else
 						err = -ENODEV;
 				}
@@ -3897,8 +3906,8 @@ static void rollback_registered(struct net_device *dev)
 	 */
 	dev_addr_discard(dev);
 
-	if (dev->uninit)
-		dev->uninit(dev);
+	if (dev->netdev_ops->ndo_uninit)
+		dev->netdev_ops->ndo_uninit(dev);
 
 	/* Notifier chain MUST detach us from master device. */
 	WARN_ON(dev->master);
@@ -3988,7 +3997,7 @@ int register_netdevice(struct net_device *dev)
 	struct hlist_head *head;
 	struct hlist_node *p;
 	int ret;
-	struct net *net;
+	struct net *net = dev_net(dev);
 
 	BUG_ON(dev_boot_phase);
 	ASSERT_RTNL();
@@ -3997,8 +4006,7 @@ int register_netdevice(struct net_device *dev)
 
 	/* When net_device's are persistent, this will be fatal. */
 	BUG_ON(dev->reg_state != NETREG_UNINITIALIZED);
-	BUG_ON(!dev_net(dev));
-	net = dev_net(dev);
+	BUG_ON(!net);
 
 	spin_lock_init(&dev->addr_list_lock);
 	netdev_set_addr_lockdep_class(dev);
@@ -4006,9 +4014,46 @@ int register_netdevice(struct net_device *dev)
 
 	dev->iflink = -1;
 
+#ifdef CONFIG_COMPAT_NET_DEV_OPS
+	/* Netdevice_ops API compatiability support.
+	 * This is temporary until all network devices are converted.
+	 */
+	if (dev->netdev_ops) {
+		const struct net_device_ops *ops = dev->netdev_ops;
+
+		dev->init = ops->ndo_init;
+		dev->uninit = ops->ndo_uninit;
+		dev->open = ops->ndo_open;
+		dev->change_rx_flags = ops->ndo_change_rx_flags;
+		dev->set_rx_mode = ops->ndo_set_rx_mode;
+		dev->set_multicast_list = ops->ndo_set_multicast_list;
+		dev->set_mac_address = ops->ndo_set_mac_address;
+		dev->validate_addr = ops->ndo_validate_addr;
+		dev->do_ioctl = ops->ndo_do_ioctl;
+		dev->set_config = ops->ndo_set_config;
+		dev->change_mtu = ops->ndo_change_mtu;
+		dev->tx_timeout = ops->ndo_tx_timeout;
+		dev->get_stats = ops->ndo_get_stats;
+		dev->vlan_rx_register = ops->ndo_vlan_rx_register;
+		dev->vlan_rx_add_vid = ops->ndo_vlan_rx_add_vid;
+		dev->vlan_rx_kill_vid = ops->ndo_vlan_rx_kill_vid;
+#ifdef CONFIG_NET_POLL_CONTROLLER
+		dev->poll_controller = ops->ndo_poll_controller;
+#endif
+	} else {
+		char drivername[64];
+		pr_info("%s (%s): not using net_device_ops yet\n",
+			dev->name, netdev_drivername(dev, drivername, 64));
+
+		/* This works only because net_device_ops and the
+		   compatiablity structure are the same. */
+		dev->netdev_ops = (void *) &(dev->init);
+	}
+#endif
+
 	/* Init, if this function is available */
-	if (dev->init) {
-		ret = dev->init(dev);
+	if (dev->netdev_ops->ndo_init) {
+		ret = dev->netdev_ops->ndo_init(dev);
 		if (ret) {
 			if (ret > 0)
 				ret = -EIO;
@@ -4086,8 +4131,8 @@ out:
 	return ret;
 
 err_uninit:
-	if (dev->uninit)
-		dev->uninit(dev);
+	if (dev->netdev_ops->ndo_uninit)
+		dev->netdev_ops->ndo_uninit(dev);
 	goto out;
 }
 
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index fc4e28e23b8..630df603444 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -172,12 +172,13 @@ static void service_arp_queue(struct netpoll_info *npi)
 void netpoll_poll(struct netpoll *np)
 {
 	struct net_device *dev = np->dev;
+	const struct net_device_ops *ops = dev->netdev_ops;
 
-	if (!dev || !netif_running(dev) || !dev->poll_controller)
+	if (!dev || !netif_running(dev) || !ops->ndo_poll_controller)
 		return;
 
 	/* Process pending work on NIC */
-	dev->poll_controller(dev);
+	ops->ndo_poll_controller(dev);
 
 	poll_napi(dev);
 
@@ -694,7 +695,7 @@ int netpoll_setup(struct netpoll *np)
 		atomic_inc(&npinfo->refcnt);
 	}
 
-	if (!ndev->poll_controller) {
+	if (!ndev->netdev_ops->ndo_poll_controller) {
 		printk(KERN_ERR "%s: %s doesn't support polling, aborting.\n",
 		       np->name, np->dev_name);
 		err = -ENOTSUPP;
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 4dfb6b4d455..6f8e0778e56 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -762,6 +762,7 @@ static int validate_linkmsg(struct net_device *dev, struct nlattr *tb[])
 static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
 		      struct nlattr **tb, char *ifname, int modified)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
 	int send_addr_notify = 0;
 	int err;
 
@@ -783,7 +784,7 @@ static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
 		struct rtnl_link_ifmap *u_map;
 		struct ifmap k_map;
 
-		if (!dev->set_config) {
+		if (!ops->ndo_set_config) {
 			err = -EOPNOTSUPP;
 			goto errout;
 		}
@@ -801,7 +802,7 @@ static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
 		k_map.dma = (unsigned char) u_map->dma;
 		k_map.port = (unsigned char) u_map->port;
 
-		err = dev->set_config(dev, &k_map);
+		err = ops->ndo_set_config(dev, &k_map);
 		if (err < 0)
 			goto errout;
 
@@ -812,7 +813,7 @@ static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
 		struct sockaddr *sa;
 		int len;
 
-		if (!dev->set_mac_address) {
+		if (!ops->ndo_set_mac_address) {
 			err = -EOPNOTSUPP;
 			goto errout;
 		}
@@ -831,7 +832,7 @@ static int do_setlink(struct net_device *dev, struct ifinfomsg *ifm,
 		sa->sa_family = dev->type;
 		memcpy(sa->sa_data, nla_data(tb[IFLA_ADDRESS]),
 		       dev->addr_len);
-		err = dev->set_mac_address(dev, sa);
+		err = ops->ndo_set_mac_address(dev, sa);
 		kfree(sa);
 		if (err)
 			goto errout;
diff --git a/net/sched/sch_generic.c b/net/sched/sch_generic.c
index 80c8f3dbbea..95ab55c064f 100644
--- a/net/sched/sch_generic.c
+++ b/net/sched/sch_generic.c
@@ -224,7 +224,7 @@ static void dev_watchdog(unsigned long arg)
 				char drivername[64];
 				WARN_ONCE(1, KERN_INFO "NETDEV WATCHDOG: %s (%s): transmit timed out\n",
 				       dev->name, netdev_drivername(dev, drivername, 64));
-				dev->tx_timeout(dev);
+				dev->netdev_ops->ndo_tx_timeout(dev);
 			}
 			if (!mod_timer(&dev->watchdog_timer,
 				       round_jiffies(jiffies +
@@ -239,7 +239,7 @@ static void dev_watchdog(unsigned long arg)
 
 void __netdev_watchdog_up(struct net_device *dev)
 {
-	if (dev->tx_timeout) {
+	if (dev->netdev_ops->ndo_tx_timeout) {
 		if (dev->watchdog_timeo <= 0)
 			dev->watchdog_timeo = 5*HZ;
 		if (!mod_timer(&dev->watchdog_timer,
-- 
cgit v1.2.3-70-g09d2


From eeda3fd64f75bcbfaa70ce946513abaf3f23b8e0 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Wed, 19 Nov 2008 21:40:23 -0800
Subject: netdev: introduce dev_get_stats()

In order for the network device ops get_stats call to be immutable, the handling
of the default internal network device stats block has to be changed. Add a new
helper function which replaces the old use of internal_get_stats.

Note: change return code to make it clear that the caller should not
go changing the returned statistics.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 arch/s390/appldata/appldata_net_sum.c |  4 ++--
 drivers/net/bonding/bond_main.c       |  5 +++--
 drivers/net/sfc/ethtool.c             |  2 +-
 drivers/parisc/led.c                  |  4 ++--
 include/linux/netdevice.h             |  4 +++-
 net/core/dev.c                        | 23 ++++++++++++++++++-----
 net/core/net-sysfs.c                  |  3 +--
 net/core/rtnetlink.c                  |  6 +++---
 8 files changed, 33 insertions(+), 18 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/arch/s390/appldata/appldata_net_sum.c b/arch/s390/appldata/appldata_net_sum.c
index 3b746556e1a..fa741f84c5b 100644
--- a/arch/s390/appldata/appldata_net_sum.c
+++ b/arch/s390/appldata/appldata_net_sum.c
@@ -67,7 +67,6 @@ static void appldata_get_net_sum_data(void *data)
 	int i;
 	struct appldata_net_sum_data *net_data;
 	struct net_device *dev;
-	struct net_device_stats *stats;
 	unsigned long rx_packets, tx_packets, rx_bytes, tx_bytes, rx_errors,
 			tx_errors, rx_dropped, tx_dropped, collisions;
 
@@ -86,7 +85,8 @@ static void appldata_get_net_sum_data(void *data)
 	collisions = 0;
 	read_lock(&dev_base_lock);
 	for_each_netdev(&init_net, dev) {
-		stats = dev->get_stats(dev);
+		const struct net_device_stats *stats = dev_get_stats(dev);
+
 		rx_packets += stats->rx_packets;
 		tx_packets += stats->tx_packets;
 		rx_bytes   += stats->rx_bytes;
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index a08ea480805..db5f5c24a25 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -3899,7 +3899,7 @@ static int bond_close(struct net_device *bond_dev)
 static struct net_device_stats *bond_get_stats(struct net_device *bond_dev)
 {
 	struct bonding *bond = netdev_priv(bond_dev);
-	struct net_device_stats *stats = &(bond->stats), *sstats;
+	struct net_device_stats *stats = &bond->stats;
 	struct net_device_stats local_stats;
 	struct slave *slave;
 	int i;
@@ -3909,7 +3909,8 @@ 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);
+		const struct net_device_stats *sstats = dev_get_stats(slave->dev);
+
 		local_stats.rx_packets += sstats->rx_packets;
 		local_stats.rx_bytes += sstats->rx_bytes;
 		local_stats.rx_errors += sstats->rx_errors;
diff --git a/drivers/net/sfc/ethtool.c b/drivers/net/sfc/ethtool.c
index abd8fcd6e62..cd92c4d8dbc 100644
--- a/drivers/net/sfc/ethtool.c
+++ b/drivers/net/sfc/ethtool.c
@@ -426,7 +426,7 @@ static void efx_ethtool_get_stats(struct net_device *net_dev,
 	EFX_BUG_ON_PARANOID(stats->n_stats != EFX_ETHTOOL_NUM_STATS);
 
 	/* Update MAC and NIC statistics */
-	net_dev->get_stats(net_dev);
+	dev_get_stats(net_dev);
 
 	/* Fill detailed statistics buffer */
 	for (i = 0; i < EFX_ETHTOOL_NUM_STATS; i++) {
diff --git a/drivers/parisc/led.c b/drivers/parisc/led.c
index f9b12664f9f..454b6532e40 100644
--- a/drivers/parisc/led.c
+++ b/drivers/parisc/led.c
@@ -360,13 +360,13 @@ static __inline__ int led_get_net_activity(void)
 	read_lock(&dev_base_lock);
 	rcu_read_lock();
 	for_each_netdev(&init_net, dev) {
-	    struct net_device_stats *stats;
+	    const struct net_device_stats *stats;
 	    struct in_device *in_dev = __in_dev_get_rcu(dev);
 	    if (!in_dev || !in_dev->ifa_list)
 		continue;
 	    if (ipv4_is_loopback(in_dev->ifa_list->ifa_local))
 		continue;
-	    stats = dev->get_stats(dev);
+	    stats = dev_get_stats(dev);
 	    rx_total += stats->rx_packets;
 	    tx_total += stats->tx_packets;
 	}
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 9060f5f3517..981a089d514 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -864,9 +864,9 @@ struct net_device
 							    unsigned short vid);
 #ifdef CONFIG_NET_POLL_CONTROLLER
 		void                    (*poll_controller)(struct net_device *dev);
-#endif
 #endif
 	};
+#endif
 };
 #define to_net_dev(d) container_of(d, struct net_device, dev)
 
@@ -1780,6 +1780,8 @@ extern void		netdev_features_change(struct net_device *dev);
 /* Load a device via the kmod */
 extern void		dev_load(struct net *net, const char *name);
 extern void		dev_mcast_init(void);
+extern const struct net_device_stats *dev_get_stats(struct net_device *dev);
+
 extern int		netdev_max_backlog;
 extern int		weight_p;
 extern int		netdev_set_master(struct net_device *dev, struct net_device *master);
diff --git a/net/core/dev.c b/net/core/dev.c
index ca14ab407b3..8843f4e3f5e 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -2620,7 +2620,7 @@ void dev_seq_stop(struct seq_file *seq, void *v)
 
 static void dev_seq_printf_stats(struct seq_file *seq, struct net_device *dev)
 {
-	struct net_device_stats *stats = dev->get_stats(dev);
+	const struct net_device_stats *stats = dev_get_stats(dev);
 
 	seq_printf(seq, "%6s:%8lu %7lu %4lu %4lu %4lu %5lu %10lu %9lu "
 		   "%8lu %7lu %4lu %4lu %4lu %5lu %7lu %10lu\n",
@@ -4288,10 +4288,24 @@ void netdev_run_todo(void)
 	}
 }
 
-static struct net_device_stats *internal_stats(struct net_device *dev)
-{
-	return &dev->stats;
+/**
+ *	dev_get_stats	- get network device statistics
+ *	@dev: device to get statistics from
+ *
+ *	Get network statistics from device. The device driver may provide
+ *	its own method by setting dev->netdev_ops->get_stats; otherwise
+ *	the internal statistics structure is used.
+ */
+const struct net_device_stats *dev_get_stats(struct net_device *dev)
+ {
+	const struct net_device_ops *ops = dev->netdev_ops;
+
+	if (ops->ndo_get_stats)
+		return ops->ndo_get_stats(dev);
+	else
+		return &dev->stats;
 }
+EXPORT_SYMBOL(dev_get_stats);
 
 static void netdev_init_one_queue(struct net_device *dev,
 				  struct netdev_queue *queue,
@@ -4370,7 +4384,6 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
 
 	netdev_init_queues(dev);
 
-	dev->get_stats = internal_stats;
 	netpoll_netdev_init(dev);
 	setup(dev);
 	strcpy(dev->name, name);
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index 146dcfeb060..afd42d71732 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -270,7 +270,6 @@ static ssize_t netstat_show(const struct device *d,
 			    unsigned long offset)
 {
 	struct net_device *dev = to_net_dev(d);
-	struct net_device_stats *stats;
 	ssize_t ret = -EINVAL;
 
 	WARN_ON(offset > sizeof(struct net_device_stats) ||
@@ -278,7 +277,7 @@ static ssize_t netstat_show(const struct device *d,
 
 	read_lock(&dev_base_lock);
 	if (dev_isalive(dev)) {
-		stats = dev->get_stats(dev);
+		const struct net_device_stats *stats = dev_get_stats(dev);
 		ret = sprintf(buf, fmt_ulong,
 			      *(unsigned long *)(((u8 *) stats) + offset));
 	}
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 6f8e0778e56..790dd205bb5 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -551,7 +551,7 @@ static void set_operstate(struct net_device *dev, unsigned char transition)
 }
 
 static void copy_rtnl_link_stats(struct rtnl_link_stats *a,
-				 struct net_device_stats *b)
+				 const struct net_device_stats *b)
 {
 	a->rx_packets = b->rx_packets;
 	a->tx_packets = b->tx_packets;
@@ -609,7 +609,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 	struct netdev_queue *txq;
 	struct ifinfomsg *ifm;
 	struct nlmsghdr *nlh;
-	struct net_device_stats *stats;
+	const struct net_device_stats *stats;
 	struct nlattr *attr;
 
 	nlh = nlmsg_put(skb, pid, seq, type, sizeof(*ifm), flags);
@@ -666,7 +666,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 	if (attr == NULL)
 		goto nla_put_failure;
 
-	stats = dev->get_stats(dev);
+	stats = dev_get_stats(dev);
 	copy_rtnl_link_stats(nla_data(attr), stats);
 
 	if (dev->rtnl_link_ops) {
-- 
cgit v1.2.3-70-g09d2


From 008298231abbeb91bc7be9e8b078607b816d1a4a Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Thu, 20 Nov 2008 20:14:53 -0800
Subject: netdev: add more functions to netdevice ops

This patch moves neigh_setup and hard_start_xmit into the network device ops
structure. For bisection, fix all the previously converted drivers as well.
Bonding driver took the biggest hit on this.

Added a prefetch of the hard_start_xmit in the fast path to try and reduce
any impact this would have.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/8139cp.c            |  2 +-
 drivers/net/8139too.c           |  2 +-
 drivers/net/acenic.c            |  4 ++-
 drivers/net/atl1e/atl1e_main.c  |  3 ++-
 drivers/net/atlx/atl1.c         |  4 +--
 drivers/net/atlx/atl2.c         |  2 +-
 drivers/net/bonding/bond_main.c | 56 +++++++++++++++++++++++++++++++++--------
 drivers/net/chelsio/cxgb2.c     |  6 ++---
 drivers/net/cxgb3/cxgb3_main.c  |  1 -
 drivers/net/e100.c              |  2 +-
 drivers/net/e1000/e1000_main.c  |  2 +-
 drivers/net/e1000e/netdev.c     |  2 +-
 drivers/net/enic/enic_main.c    |  2 +-
 drivers/net/forcedeth.c         | 22 +++++++++++++---
 drivers/net/ifb.c               |  4 +--
 drivers/net/igb/igb_main.c      |  2 +-
 drivers/net/ixgb/ixgb_main.c    |  2 +-
 drivers/net/ixgbe/ixgbe_main.c  |  2 +-
 drivers/net/loopback.c          |  2 +-
 drivers/net/macvlan.c           |  4 +--
 drivers/net/niu.c               |  2 +-
 drivers/net/ppp_generic.c       |  5 ++--
 drivers/net/r8169.c             |  2 +-
 drivers/net/skge.c              |  2 +-
 drivers/net/sky2.c              |  3 ++-
 drivers/net/tg3.c               | 45 ++++++++++++++++++++++-----------
 drivers/net/tun.c               |  4 +--
 drivers/net/veth.c              |  2 +-
 drivers/net/via-velocity.c      |  2 +-
 include/linux/netdevice.h       | 39 ++++++++++++++++++----------
 net/bridge/br_device.c          | 10 ++++----
 net/bridge/br_if.c              |  2 +-
 net/core/dev.c                  | 12 ++++++---
 net/core/neighbour.c            |  6 ++---
 net/core/netpoll.c              |  6 +++--
 net/core/pktgen.c               |  8 +++---
 36 files changed, 183 insertions(+), 93 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/net/8139cp.c b/drivers/net/8139cp.c
index 13f75b67872..f6d9d1353dd 100644
--- a/drivers/net/8139cp.c
+++ b/drivers/net/8139cp.c
@@ -1824,6 +1824,7 @@ static const struct net_device_ops cp_netdev_ops = {
 	.ndo_set_multicast_list	= cp_set_rx_mode,
 	.ndo_get_stats		= cp_get_stats,
 	.ndo_do_ioctl		= cp_ioctl,
+	.ndo_start_xmit		= cp_start_xmit,
 	.ndo_tx_timeout		= cp_tx_timeout,
 #if CP_VLAN_TAG_USED
 	.ndo_vlan_rx_register	= cp_vlan_rx_register,
@@ -1949,7 +1950,6 @@ static int cp_init_one (struct pci_dev *pdev, const struct pci_device_id *ent)
 	memcpy(dev->perm_addr, dev->dev_addr, dev->addr_len);
 
 	dev->netdev_ops = &cp_netdev_ops;
-	dev->hard_start_xmit = cp_start_xmit;
 	netif_napi_add(dev, &cp->napi, cp_rx_poll, 16);
 	dev->ethtool_ops = &cp_ethtool_ops;
 	dev->watchdog_timeo = TX_TIMEOUT;
diff --git a/drivers/net/8139too.c b/drivers/net/8139too.c
index f8866552386..445a479db79 100644
--- a/drivers/net/8139too.c
+++ b/drivers/net/8139too.c
@@ -921,6 +921,7 @@ static const struct net_device_ops rtl8139_netdev_ops = {
 	.ndo_stop		= rtl8139_close,
 	.ndo_get_stats		= rtl8139_get_stats,
 	.ndo_validate_addr	= eth_validate_addr,
+	.ndo_start_xmit		= rtl8139_start_xmit,
 	.ndo_set_multicast_list	= rtl8139_set_rx_mode,
 	.ndo_do_ioctl		= netdev_ioctl,
 	.ndo_tx_timeout		= rtl8139_tx_timeout,
@@ -992,7 +993,6 @@ static int __devinit rtl8139_init_one (struct pci_dev *pdev,
 	dev->netdev_ops = &rtl8139_netdev_ops;
 	dev->ethtool_ops = &rtl8139_ethtool_ops;
 	dev->watchdog_timeo = TX_TIMEOUT;
-	dev->hard_start_xmit = rtl8139_start_xmit;
 	netif_napi_add(dev, &tp->napi, rtl8139_poll, 64);
 
 	/* note: the hardware is not capable of sg/csum/highdma, however
diff --git a/drivers/net/acenic.c b/drivers/net/acenic.c
index 309a90ea921..21d24320210 100644
--- a/drivers/net/acenic.c
+++ b/drivers/net/acenic.c
@@ -455,10 +455,13 @@ static const struct net_device_ops ace_netdev_ops = {
 	.ndo_stop		= ace_close,
 	.ndo_tx_timeout		= ace_watchdog,
 	.ndo_get_stats		= ace_get_stats,
+	.ndo_start_xmit		= ace_start_xmit,
 	.ndo_set_multicast_list	= ace_set_multicast_list,
 	.ndo_set_mac_address	= ace_set_mac_addr,
 	.ndo_change_mtu		= ace_change_mtu,
+#if ACENIC_DO_VLAN
 	.ndo_vlan_rx_register	= ace_vlan_rx_register,
+#endif
 };
 
 static int __devinit acenic_probe_one(struct pci_dev *pdev,
@@ -489,7 +492,6 @@ static int __devinit acenic_probe_one(struct pci_dev *pdev,
 	dev->watchdog_timeo = 5*HZ;
 
 	dev->netdev_ops = &ace_netdev_ops;
-	dev->hard_start_xmit = &ace_start_xmit;
 	SET_ETHTOOL_OPS(dev, &ace_ethtool_ops);
 
 	/* we only display this string ONCE */
diff --git a/drivers/net/atl1e/atl1e_main.c b/drivers/net/atl1e/atl1e_main.c
index a815fffc2a5..98b2a7a466b 100644
--- a/drivers/net/atl1e/atl1e_main.c
+++ b/drivers/net/atl1e/atl1e_main.c
@@ -2256,6 +2256,7 @@ static void atl1e_shutdown(struct pci_dev *pdev)
 static const struct net_device_ops atl1e_netdev_ops = {
 	.ndo_open		= atl1e_open,
 	.ndo_stop		= atl1e_close,
+	.ndo_start_xmit		= atl1e_xmit_frame,
 	.ndo_get_stats		= atl1e_get_stats,
 	.ndo_set_multicast_list	= atl1e_set_multi,
 	.ndo_validate_addr	= eth_validate_addr,
@@ -2277,7 +2278,7 @@ static int atl1e_init_netdev(struct net_device *netdev, struct pci_dev *pdev)
 
 	netdev->irq  = pdev->irq;
 	netdev->netdev_ops = &atl1e_netdev_ops;
-	netdev->hard_start_xmit = atl1e_xmit_frame,
+
 	netdev->watchdog_timeo = AT_TX_WATCHDOG;
 	atl1e_set_ethtool_ops(netdev);
 
diff --git a/drivers/net/atlx/atl1.c b/drivers/net/atlx/atl1.c
index 7a0fb04e348..aef7e47fdd2 100644
--- a/drivers/net/atlx/atl1.c
+++ b/drivers/net/atlx/atl1.c
@@ -2883,12 +2883,13 @@ static void atl1_poll_controller(struct net_device *netdev)
 static const struct net_device_ops atl1_netdev_ops = {
 	.ndo_open		= atl1_open,
 	.ndo_stop		= atl1_close,
+	.ndo_start_xmit		= atl1_xmit_frame,
 	.ndo_set_multicast_list	= atlx_set_multi,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_mac_address	= atl1_set_mac,
 	.ndo_change_mtu		= atl1_change_mtu,
 	.ndo_do_ioctl		= atlx_ioctl,
-	.ndo_tx_timeout	= atlx_tx_timeout,
+	.ndo_tx_timeout		= atlx_tx_timeout,
 	.ndo_vlan_rx_register	= atlx_vlan_rx_register,
 #ifdef CONFIG_NET_POLL_CONTROLLER
 	.ndo_poll_controller	= atl1_poll_controller,
@@ -2983,7 +2984,6 @@ static int __devinit atl1_probe(struct pci_dev *pdev,
 	adapter->mii.reg_num_mask = 0x1f;
 
 	netdev->netdev_ops = &atl1_netdev_ops;
-	netdev->hard_start_xmit = &atl1_xmit_frame;
 	netdev->watchdog_timeo = 5 * HZ;
 
 	netdev->ethtool_ops = &atl1_ethtool_ops;
diff --git a/drivers/net/atlx/atl2.c b/drivers/net/atlx/atl2.c
index b8d585722e1..0326a84503e 100644
--- a/drivers/net/atlx/atl2.c
+++ b/drivers/net/atlx/atl2.c
@@ -1315,6 +1315,7 @@ static void atl2_poll_controller(struct net_device *netdev)
 static const struct net_device_ops atl2_netdev_ops = {
 	.ndo_open		= atl2_open,
 	.ndo_stop		= atl2_close,
+	.ndo_start_xmit		= atl2_xmit_frame,
 	.ndo_set_multicast_list	= atl2_set_multi,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_mac_address	= atl2_set_mac,
@@ -1400,7 +1401,6 @@ static int __devinit atl2_probe(struct pci_dev *pdev,
 
 	atl2_setup_pcicmd(pdev);
 
-	netdev->hard_start_xmit = &atl2_xmit_frame;
 	netdev->netdev_ops = &atl2_netdev_ops;
 	atl2_set_ethtool_ops(netdev);
 	netdev->watchdog_timeo = 5 * HZ;
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 614656c8187..a339a805273 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -1377,14 +1377,12 @@ done:
 	return 0;
 }
 
-
 static void bond_setup_by_slave(struct net_device *bond_dev,
 				struct net_device *slave_dev)
 {
 	struct bonding *bond = netdev_priv(bond_dev);
 
-	bond_dev->neigh_setup           = slave_dev->neigh_setup;
-	bond_dev->header_ops		= slave_dev->header_ops;
+	bond_dev->header_ops	    = slave_dev->header_ops;
 
 	bond_dev->type		    = slave_dev->type;
 	bond_dev->hard_header_len   = slave_dev->hard_header_len;
@@ -4124,6 +4122,20 @@ static void bond_set_multicast_list(struct net_device *bond_dev)
 	read_unlock(&bond->lock);
 }
 
+static int bond_neigh_setup(struct net_device *dev, struct neigh_parms *parms)
+{
+	struct bonding *bond = netdev_priv(dev);
+	struct slave *slave = bond->first_slave;
+
+	if (slave) {
+		const struct net_device_ops *slave_ops
+			= slave->dev->netdev_ops;
+		if (slave_ops->ndo_neigh_setup)
+			return slave_ops->ndo_neigh_setup(dev, parms);
+	}
+	return 0;
+}
+
 /*
  * Change the MTU of all of a master's slaves to match the master
  */
@@ -4490,6 +4502,35 @@ static void bond_set_xmit_hash_policy(struct bonding *bond)
 	}
 }
 
+static int bond_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	const struct bonding *bond = netdev_priv(dev);
+
+	switch (bond->params.mode) {
+	case BOND_MODE_ROUNDROBIN:
+		return bond_xmit_roundrobin(skb, dev);
+	case BOND_MODE_ACTIVEBACKUP:
+		return bond_xmit_activebackup(skb, dev);
+	case BOND_MODE_XOR:
+		return bond_xmit_xor(skb, dev);
+	case BOND_MODE_BROADCAST:
+		return bond_xmit_broadcast(skb, dev);
+	case BOND_MODE_8023AD:
+		return bond_3ad_xmit_xor(skb, dev);
+	case BOND_MODE_ALB:
+	case BOND_MODE_TLB:
+		return bond_alb_xmit(skb, dev);
+	default:
+		/* Should never happen, mode already checked */
+		printk(KERN_ERR DRV_NAME ": %s: Error: Unknown bonding mode %d\n",
+		     dev->name, bond->params.mode);
+		WARN_ON_ONCE(1);
+		dev_kfree_skb(skb);
+		return NETDEV_TX_OK;
+	}
+}
+
+
 /*
  * set bond mode specific net device operations
  */
@@ -4499,28 +4540,22 @@ void bond_set_mode_ops(struct bonding *bond, int mode)
 
 	switch (mode) {
 	case BOND_MODE_ROUNDROBIN:
-		bond_dev->hard_start_xmit = bond_xmit_roundrobin;
 		break;
 	case BOND_MODE_ACTIVEBACKUP:
-		bond_dev->hard_start_xmit = bond_xmit_activebackup;
 		break;
 	case BOND_MODE_XOR:
-		bond_dev->hard_start_xmit = bond_xmit_xor;
 		bond_set_xmit_hash_policy(bond);
 		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;
 		bond_set_xmit_hash_policy(bond);
 		break;
 	case BOND_MODE_ALB:
 		bond_set_master_alb_flags(bond);
 		/* FALLTHRU */
 	case BOND_MODE_TLB:
-		bond_dev->hard_start_xmit = bond_alb_xmit;
 		break;
 	default:
 		/* Should never happen, mode already checked */
@@ -4553,12 +4588,13 @@ static const struct ethtool_ops bond_ethtool_ops = {
 static const struct net_device_ops bond_netdev_ops = {
 	.ndo_open		= bond_open,
 	.ndo_stop		= bond_close,
+	.ndo_start_xmit		= bond_start_xmit,
 	.ndo_get_stats		= bond_get_stats,
 	.ndo_do_ioctl		= bond_do_ioctl,
 	.ndo_set_multicast_list	= bond_set_multicast_list,
 	.ndo_change_mtu		= bond_change_mtu,
-	.ndo_validate_addr	= NULL,
 	.ndo_set_mac_address 	= bond_set_mac_address,
+	.ndo_neigh_setup	= bond_neigh_setup,
 	.ndo_vlan_rx_register	= bond_vlan_rx_register,
 	.ndo_vlan_rx_add_vid 	= bond_vlan_rx_add_vid,
 	.ndo_vlan_rx_kill_vid	= bond_vlan_rx_kill_vid,
diff --git a/drivers/net/chelsio/cxgb2.c b/drivers/net/chelsio/cxgb2.c
index 482741797eb..9b6011e7678 100644
--- a/drivers/net/chelsio/cxgb2.c
+++ b/drivers/net/chelsio/cxgb2.c
@@ -915,7 +915,7 @@ static int t1_set_mac_addr(struct net_device *dev, void *p)
 }
 
 #if defined(CONFIG_VLAN_8021Q) || defined(CONFIG_VLAN_8021Q_MODULE)
-static void vlan_rx_register(struct net_device *dev,
+static void t1_vlan_rx_register(struct net_device *dev,
 				   struct vlan_group *grp)
 {
 	struct adapter *adapter = dev->ml_priv;
@@ -1013,6 +1013,7 @@ void t1_fatal_err(struct adapter *adapter)
 static const struct net_device_ops cxgb_netdev_ops = {
 	.ndo_open		= cxgb_open,
 	.ndo_stop		= cxgb_close,
+	.ndo_start_xmit		= t1_start_xmit,
 	.ndo_get_stats		= t1_get_stats,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_multicast_list	= t1_set_rxmode,
@@ -1020,7 +1021,7 @@ static const struct net_device_ops cxgb_netdev_ops = {
 	.ndo_change_mtu		= t1_change_mtu,
 	.ndo_set_mac_address	= t1_set_mac_addr,
 #if defined(CONFIG_VLAN_8021Q) || defined(CONFIG_VLAN_8021Q_MODULE)
-	.ndo_vlan_rx_register	= vlan_rx_register,
+	.ndo_vlan_rx_register	= t1_vlan_rx_register,
 #endif
 #ifdef CONFIG_NET_POLL_CONTROLLER
 	.ndo_poll_controller	= t1_netpoll,
@@ -1157,7 +1158,6 @@ static int __devinit init_one(struct pci_dev *pdev,
 		}
 
 		netdev->netdev_ops = &cxgb_netdev_ops;
-		netdev->hard_start_xmit = t1_start_xmit;
 		netdev->hard_header_len += (adapter->flags & TSO_CAPABLE) ?
 			sizeof(struct cpl_tx_pkt_lso) : sizeof(struct cpl_tx_pkt);
 
diff --git a/drivers/net/cxgb3/cxgb3_main.c b/drivers/net/cxgb3/cxgb3_main.c
index a9479be53ec..cd9fcaca70f 100644
--- a/drivers/net/cxgb3/cxgb3_main.c
+++ b/drivers/net/cxgb3/cxgb3_main.c
@@ -2955,7 +2955,6 @@ static int __devinit init_one(struct pci_dev *pdev,
 
 		netdev->features |= NETIF_F_HW_VLAN_TX | NETIF_F_HW_VLAN_RX;
 		netdev->netdev_ops = &cxgb_netdev_ops;
-		netdev->hard_start_xmit = t3_eth_xmit;
 		SET_ETHTOOL_OPS(netdev, &cxgb_ethtool_ops);
 	}
 
diff --git a/drivers/net/e100.c b/drivers/net/e100.c
index 5894716de19..2001a63794f 100644
--- a/drivers/net/e100.c
+++ b/drivers/net/e100.c
@@ -2615,6 +2615,7 @@ static int e100_close(struct net_device *netdev)
 static const struct net_device_ops e100_netdev_ops = {
 	.ndo_open		= e100_open,
 	.ndo_stop		= e100_close,
+	.ndo_start_xmit		= e100_xmit_frame,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_multicast_list	= e100_set_multicast_list,
 	.ndo_set_mac_address	= e100_set_mac_address,
@@ -2640,7 +2641,6 @@ static int __devinit e100_probe(struct pci_dev *pdev,
 	}
 
 	netdev->netdev_ops = &e100_netdev_ops;
-	netdev->hard_start_xmit = e100_xmit_frame;
 	SET_ETHTOOL_OPS(netdev, &e100_ethtool_ops);
 	netdev->watchdog_timeo = E100_WATCHDOG_PERIOD;
 	strncpy(netdev->name, pci_name(pdev), sizeof(netdev->name) - 1);
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index debbba390d4..5c098c9d584 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -891,6 +891,7 @@ static int e1000_is_need_ioport(struct pci_dev *pdev)
 static const struct net_device_ops e1000_netdev_ops = {
 	.ndo_open		= e1000_open,
 	.ndo_stop		= e1000_close,
+	.ndo_start_xmit		= e1000_xmit_frame,
 	.ndo_get_stats		= e1000_get_stats,
 	.ndo_set_rx_mode	= e1000_set_rx_mode,
 	.ndo_set_mac_address	= e1000_set_mac,
@@ -1001,7 +1002,6 @@ static int __devinit e1000_probe(struct pci_dev *pdev,
 	}
 
 	netdev->netdev_ops = &e1000_netdev_ops;
-	netdev->hard_start_xmit = &e1000_xmit_frame;
 	e1000_set_ethtool_ops(netdev);
 	netdev->watchdog_timeo = 5 * HZ;
 	netif_napi_add(netdev, &adapter->napi, e1000_clean, 64);
diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
index ced839e4cae..cc0502bbb9f 100644
--- a/drivers/net/e1000e/netdev.c
+++ b/drivers/net/e1000e/netdev.c
@@ -4707,6 +4707,7 @@ static void e1000_eeprom_checks(struct e1000_adapter *adapter)
 static const struct net_device_ops e1000e_netdev_ops = {
 	.ndo_open		= e1000_open,
 	.ndo_stop		= e1000_close,
+	.ndo_start_xmit		= e1000_xmit_frame,
 	.ndo_get_stats		= e1000_get_stats,
 	.ndo_set_multicast_list	= e1000_set_multi,
 	.ndo_set_mac_address	= e1000_set_mac,
@@ -4822,7 +4823,6 @@ static int __devinit e1000_probe(struct pci_dev *pdev,
 
 	/* construct the net_device struct */
 	netdev->netdev_ops		= &e1000e_netdev_ops;
-	netdev->hard_start_xmit		= &e1000_xmit_frame;
 	e1000e_set_ethtool_ops(netdev);
 	netdev->watchdog_timeo		= 5 * HZ;
 	netif_napi_add(netdev, &adapter->napi, e1000_clean, 64);
diff --git a/drivers/net/enic/enic_main.c b/drivers/net/enic/enic_main.c
index 40f8c88b166..1c409df735d 100644
--- a/drivers/net/enic/enic_main.c
+++ b/drivers/net/enic/enic_main.c
@@ -1593,6 +1593,7 @@ static void enic_iounmap(struct enic *enic)
 static const struct net_device_ops enic_netdev_ops = {
 	.ndo_open		= enic_open,
 	.ndo_stop		= enic_stop,
+	.ndo_start_xmit		= enic_hard_start_xmit,
 	.ndo_get_stats		= enic_get_stats,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_multicast_list	= enic_set_multicast_list,
@@ -1830,7 +1831,6 @@ static int __devinit enic_probe(struct pci_dev *pdev,
 	}
 
 	netdev->netdev_ops = &enic_netdev_ops;
-	netdev->hard_start_xmit = enic_hard_start_xmit;
 	netdev->watchdog_timeo = 2 * HZ;
 	netdev->ethtool_ops = &enic_ethtool_ops;
 
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index dd2e1f670b0..0d7e5750245 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -5412,6 +5412,23 @@ static const struct net_device_ops nv_netdev_ops = {
 	.ndo_open		= nv_open,
 	.ndo_stop		= nv_close,
 	.ndo_get_stats		= nv_get_stats,
+	.ndo_start_xmit		= nv_start_xmit,
+	.ndo_tx_timeout		= nv_tx_timeout,
+	.ndo_change_mtu		= nv_change_mtu,
+	.ndo_validate_addr	= eth_validate_addr,
+	.ndo_set_mac_address	= nv_set_mac_address,
+	.ndo_set_multicast_list	= nv_set_multicast,
+	.ndo_vlan_rx_register	= nv_vlan_rx_register,
+#ifdef CONFIG_NET_POLL_CONTROLLER
+	.ndo_poll_controller	= nv_poll_controller,
+#endif
+};
+
+static const struct net_device_ops nv_netdev_ops_optimized = {
+	.ndo_open		= nv_open,
+	.ndo_stop		= nv_close,
+	.ndo_get_stats		= nv_get_stats,
+	.ndo_start_xmit		= nv_start_xmit_optimized,
 	.ndo_tx_timeout		= nv_tx_timeout,
 	.ndo_change_mtu		= nv_change_mtu,
 	.ndo_validate_addr	= eth_validate_addr,
@@ -5592,11 +5609,10 @@ static int __devinit nv_probe(struct pci_dev *pci_dev, const struct pci_device_i
 		goto out_freering;
 
 	if (!nv_optimized(np))
-		dev->hard_start_xmit = nv_start_xmit;
+		dev->netdev_ops = &nv_netdev_ops;
 	else
-		dev->hard_start_xmit = nv_start_xmit_optimized;
+		dev->netdev_ops = &nv_netdev_ops_optimized;
 
-	dev->netdev_ops = &nv_netdev_ops;
 #ifdef CONFIG_FORCEDETH_NAPI
 	netif_napi_add(dev, &np->napi, nv_napi_poll, RX_WORK_PER_LOOP);
 #endif
diff --git a/drivers/net/ifb.c b/drivers/net/ifb.c
index 363a166df8f..60a26300193 100644
--- a/drivers/net/ifb.c
+++ b/drivers/net/ifb.c
@@ -138,15 +138,15 @@ resched:
 }
 
 static const struct net_device_ops ifb_netdev_ops = {
-	.ndo_validate_addr = eth_validate_addr,
 	.ndo_open	= ifb_open,
 	.ndo_stop	= ifb_close,
+	.ndo_start_xmit	= ifb_xmit,
+	.ndo_validate_addr = eth_validate_addr,
 };
 
 static void ifb_setup(struct net_device *dev)
 {
 	/* Initialize the device structure. */
-	dev->hard_start_xmit = ifb_xmit;
 	dev->destructor = free_netdev;
 	dev->netdev_ops = &ifb_netdev_ops;
 
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index ceb0a045879..eca5684d565 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -953,6 +953,7 @@ static int igb_is_need_ioport(struct pci_dev *pdev)
 static const struct net_device_ops igb_netdev_ops = {
 	.ndo_open 		= igb_open,
 	.ndo_stop		= igb_close,
+	.ndo_start_xmit		= igb_xmit_frame_adv,
 	.ndo_get_stats		= igb_get_stats,
 	.ndo_set_multicast_list	= igb_set_multi,
 	.ndo_set_mac_address	= igb_set_mac,
@@ -1080,7 +1081,6 @@ static int __devinit igb_probe(struct pci_dev *pdev,
 	netdev->netdev_ops = &igb_netdev_ops;
 	igb_set_ethtool_ops(netdev);
 	netdev->watchdog_timeo = 5 * HZ;
-	netdev->hard_start_xmit = &igb_xmit_frame_adv;
 
 	strncpy(netdev->name, pci_name(pdev), sizeof(netdev->name) - 1);
 
diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
index 3ca9daa70b3..a04e3892ddf 100644
--- a/drivers/net/ixgb/ixgb_main.c
+++ b/drivers/net/ixgb/ixgb_main.c
@@ -324,6 +324,7 @@ ixgb_reset(struct ixgb_adapter *adapter)
 static const struct net_device_ops ixgb_netdev_ops = {
 	.ndo_open 		= ixgb_open,
 	.ndo_stop		= ixgb_close,
+	.ndo_start_xmit		= ixgb_xmit_frame,
 	.ndo_get_stats		= ixgb_get_stats,
 	.ndo_set_multicast_list	= ixgb_set_multi,
 	.ndo_validate_addr	= eth_validate_addr,
@@ -414,7 +415,6 @@ ixgb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	}
 
 	netdev->netdev_ops = &ixgb_netdev_ops;
-	netdev->hard_start_xmit = &ixgb_xmit_frame;
 	ixgb_set_ethtool_ops(netdev);
 	netdev->watchdog_timeo = 5 * HZ;
 	netif_napi_add(netdev, &adapter->napi, ixgb_clean, 64);
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 7ad07a00680..40108523377 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -3728,6 +3728,7 @@ static int ixgbe_link_config(struct ixgbe_hw *hw)
 static const struct net_device_ops ixgbe_netdev_ops = {
 	.ndo_open 		= ixgbe_open,
 	.ndo_stop		= ixgbe_close,
+	.ndo_start_xmit		= ixgbe_xmit_frame,
 	.ndo_get_stats		= ixgbe_get_stats,
 	.ndo_set_multicast_list	= ixgbe_set_rx_mode,
 	.ndo_validate_addr	= eth_validate_addr,
@@ -3824,7 +3825,6 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	}
 
 	netdev->netdev_ops = &ixgbe_netdev_ops;
-	netdev->hard_start_xmit = &ixgbe_xmit_frame;
 	ixgbe_set_ethtool_ops(netdev);
 	netdev->watchdog_timeo = 5 * HZ;
 	strcpy(netdev->name, pci_name(pdev));
diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c
index 958450124de..b7d438a367f 100644
--- a/drivers/net/loopback.c
+++ b/drivers/net/loopback.c
@@ -145,6 +145,7 @@ static void loopback_dev_free(struct net_device *dev)
 
 static const struct net_device_ops loopback_ops = {
 	.ndo_init      = loopback_dev_init,
+	.ndo_start_xmit= loopback_xmit,
 	.ndo_get_stats = loopback_get_stats,
 };
 
@@ -155,7 +156,6 @@ static const struct net_device_ops loopback_ops = {
 static void loopback_setup(struct net_device *dev)
 {
 	dev->mtu		= (16 * 1024) + 20 + 20 + 12;
-	dev->hard_start_xmit	= loopback_xmit;
 	dev->hard_header_len	= ETH_HLEN;	/* 14	*/
 	dev->addr_len		= ETH_ALEN;	/* 6	*/
 	dev->tx_queue_len	= 0;
diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c
index d00ea444e0a..e8879217a1d 100644
--- a/drivers/net/macvlan.c
+++ b/drivers/net/macvlan.c
@@ -140,7 +140,7 @@ static struct sk_buff *macvlan_handle_frame(struct sk_buff *skb)
 	return NULL;
 }
 
-static int macvlan_hard_start_xmit(struct sk_buff *skb, struct net_device *dev)
+static int macvlan_start_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	const struct macvlan_dev *vlan = netdev_priv(dev);
 	unsigned int len = skb->len;
@@ -365,6 +365,7 @@ static const struct net_device_ops macvlan_netdev_ops = {
 	.ndo_init		= macvlan_init,
 	.ndo_open		= macvlan_open,
 	.ndo_stop		= macvlan_stop,
+	.ndo_start_xmit		= macvlan_start_xmit,
 	.ndo_change_mtu		= macvlan_change_mtu,
 	.ndo_change_rx_flags	= macvlan_change_rx_flags,
 	.ndo_set_mac_address	= macvlan_set_mac_address,
@@ -377,7 +378,6 @@ static void macvlan_setup(struct net_device *dev)
 	ether_setup(dev);
 
 	dev->netdev_ops		= &macvlan_netdev_ops;
-	dev->hard_start_xmit	= macvlan_hard_start_xmit;
 	dev->destructor		= free_netdev;
 	dev->header_ops		= &macvlan_hard_header_ops,
 	dev->ethtool_ops	= &macvlan_ethtool_ops;
diff --git a/drivers/net/niu.c b/drivers/net/niu.c
index 318537efd58..a8d10630f80 100644
--- a/drivers/net/niu.c
+++ b/drivers/net/niu.c
@@ -8892,6 +8892,7 @@ static struct net_device * __devinit niu_alloc_and_init(
 static const struct net_device_ops niu_netdev_ops = {
 	.ndo_open		= niu_open,
 	.ndo_stop		= niu_close,
+	.ndo_start_xmit		= niu_start_xmit,
 	.ndo_get_stats		= niu_get_stats,
 	.ndo_set_multicast_list	= niu_set_rx_mode,
 	.ndo_validate_addr	= eth_validate_addr,
@@ -8904,7 +8905,6 @@ static const struct net_device_ops niu_netdev_ops = {
 static void __devinit niu_assign_netdev_ops(struct net_device *dev)
 {
 	dev->netdev_ops = &niu_netdev_ops;
-	dev->hard_start_xmit = niu_start_xmit;
 	dev->ethtool_ops = &niu_ethtool_ops;
 	dev->watchdog_timeo = NIU_TX_TIMEOUT;
 }
diff --git a/drivers/net/ppp_generic.c b/drivers/net/ppp_generic.c
index bad99e8cac3..1b15a088a3b 100644
--- a/drivers/net/ppp_generic.c
+++ b/drivers/net/ppp_generic.c
@@ -972,7 +972,8 @@ ppp_net_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
 }
 
 static const struct net_device_ops ppp_netdev_ops = {
-	.ndo_do_ioctl = ppp_net_ioctl,
+	.ndo_start_xmit = ppp_start_xmit,
+	.ndo_do_ioctl   = ppp_net_ioctl,
 };
 
 static void ppp_setup(struct net_device *dev)
@@ -2437,8 +2438,6 @@ ppp_create_interface(int unit, int *retp)
 	skb_queue_head_init(&ppp->mrq);
 #endif /* CONFIG_PPP_MULTILINK */
 
-	dev->hard_start_xmit = ppp_start_xmit;
-
 	ret = -EEXIST;
 	mutex_lock(&all_ppp_mutex);
 	if (unit < 0)
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index bac58ca628d..dddf6aeff49 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -1927,6 +1927,7 @@ static const struct net_device_ops rtl8169_netdev_ops = {
 	.ndo_open		= rtl8169_open,
 	.ndo_stop		= rtl8169_close,
 	.ndo_get_stats		= rtl8169_get_stats,
+	.ndo_start_xmit		= rtl8169_start_xmit,
 	.ndo_tx_timeout		= rtl8169_tx_timeout,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_change_mtu		= rtl8169_change_mtu,
@@ -2125,7 +2126,6 @@ rtl8169_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 		dev->dev_addr[i] = RTL_R8(MAC0 + i);
 	memcpy(dev->perm_addr, dev->dev_addr, dev->addr_len);
 
-	dev->hard_start_xmit = rtl8169_start_xmit;
 	SET_ETHTOOL_OPS(dev, &rtl8169_ethtool_ops);
 	dev->watchdog_timeo = RTL8169_TX_TIMEOUT;
 	dev->irq = pdev->irq;
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
index 93c1b1d9296..f73ee797400 100644
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -3805,6 +3805,7 @@ static __exit void skge_debug_cleanup(void)
 static const struct net_device_ops skge_netdev_ops = {
 	.ndo_open		= skge_up,
 	.ndo_stop		= skge_down,
+	.ndo_start_xmit		= skge_xmit_frame,
 	.ndo_do_ioctl		= skge_ioctl,
 	.ndo_get_stats		= skge_get_stats,
 	.ndo_tx_timeout		= skge_tx_timeout,
@@ -3831,7 +3832,6 @@ static struct net_device *skge_devinit(struct skge_hw *hw, int port,
 	}
 
 	SET_NETDEV_DEV(dev, &hw->pdev->dev);
-	dev->hard_start_xmit = skge_xmit_frame;
 	dev->netdev_ops = &skge_netdev_ops;
 	dev->ethtool_ops = &skge_ethtool_ops;
 	dev->watchdog_timeo = TX_WATCHDOG;
diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c
index 251505125cb..3668e81e474 100644
--- a/drivers/net/sky2.c
+++ b/drivers/net/sky2.c
@@ -4047,6 +4047,7 @@ static const struct net_device_ops sky2_netdev_ops[2] = {
   {
 	.ndo_open		= sky2_up,
 	.ndo_stop		= sky2_down,
+	.ndo_start_xmit		= sky2_xmit_frame,
 	.ndo_do_ioctl		= sky2_ioctl,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_mac_address	= sky2_set_mac_address,
@@ -4063,6 +4064,7 @@ static const struct net_device_ops sky2_netdev_ops[2] = {
   {
 	.ndo_open		= sky2_up,
 	.ndo_stop		= sky2_down,
+	.ndo_start_xmit		= sky2_xmit_frame,
 	.ndo_do_ioctl		= sky2_ioctl,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_mac_address	= sky2_set_mac_address,
@@ -4090,7 +4092,6 @@ static __devinit struct net_device *sky2_init_netdev(struct sky2_hw *hw,
 
 	SET_NETDEV_DEV(dev, &hw->pdev->dev);
 	dev->irq = hw->pdev->irq;
-	dev->hard_start_xmit = sky2_xmit_frame;
 	SET_ETHTOOL_OPS(dev, &sky2_ethtool_ops);
 	dev->watchdog_timeo = TX_WATCHDOG;
 	dev->netdev_ops = &sky2_netdev_ops[port];
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 4b97cb60136..9ba18e1bc34 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -12614,19 +12614,6 @@ static int __devinit tg3_get_invariants(struct tg3 *tp)
 	else
 		tp->tg3_flags &= ~TG3_FLAG_POLL_SERDES;
 
-	/* All chips before 5787 can get confused if TX buffers
-	 * straddle the 4GB address boundary in some cases.
-	 */
-	if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5755 ||
-	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5787 ||
-	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5784 ||
-	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5761 ||
-	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5785 ||
-	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5906)
-		tp->dev->hard_start_xmit = tg3_start_xmit;
-	else
-		tp->dev->hard_start_xmit = tg3_start_xmit_dma_bug;
-
 	tp->rx_offset = 2;
 	if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5701 &&
 	    (tp->tg3_flags & TG3_FLAG_PCIX_MODE) != 0)
@@ -13346,6 +13333,26 @@ static void __devinit tg3_init_coal(struct tg3 *tp)
 static const struct net_device_ops tg3_netdev_ops = {
 	.ndo_open		= tg3_open,
 	.ndo_stop		= tg3_close,
+	.ndo_start_xmit		= tg3_start_xmit,
+	.ndo_get_stats		= tg3_get_stats,
+	.ndo_validate_addr	= eth_validate_addr,
+	.ndo_set_multicast_list	= tg3_set_rx_mode,
+	.ndo_set_mac_address	= tg3_set_mac_addr,
+	.ndo_do_ioctl		= tg3_ioctl,
+	.ndo_tx_timeout		= tg3_tx_timeout,
+	.ndo_change_mtu		= tg3_change_mtu,
+#if TG3_VLAN_TAG_USED
+	.ndo_vlan_rx_register	= tg3_vlan_rx_register,
+#endif
+#ifdef CONFIG_NET_POLL_CONTROLLER
+	.ndo_poll_controller	= tg3_poll_controller,
+#endif
+};
+
+static const struct net_device_ops tg3_netdev_ops_dma_bug = {
+	.ndo_open		= tg3_open,
+	.ndo_stop		= tg3_close,
+	.ndo_start_xmit		= tg3_start_xmit_dma_bug,
 	.ndo_get_stats		= tg3_get_stats,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_multicast_list	= tg3_set_rx_mode,
@@ -13475,7 +13482,6 @@ static int __devinit tg3_init_one(struct pci_dev *pdev,
 	tp->rx_jumbo_pending = TG3_DEF_RX_JUMBO_RING_PENDING;
 	tp->tx_pending = TG3_DEF_TX_RING_PENDING;
 
-	dev->netdev_ops = &tg3_netdev_ops;
 	netif_napi_add(dev, &tp->napi, tg3_poll, 64);
 	dev->ethtool_ops = &tg3_ethtool_ops;
 	dev->watchdog_timeo = TG3_TX_TIMEOUT;
@@ -13488,6 +13494,17 @@ static int __devinit tg3_init_one(struct pci_dev *pdev,
 		goto err_out_iounmap;
 	}
 
+	if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5755 ||
+	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5787 ||
+	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5784 ||
+	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5761 ||
+	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5785 ||
+	    GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5906)
+		dev->netdev_ops = &tg3_netdev_ops;
+	else
+		dev->netdev_ops = &tg3_netdev_ops_dma_bug;
+
+
 	/* The EPB bridge inside 5714, 5715, and 5780 and any
 	 * device behind the EPB cannot support DMA addresses > 40-bit.
 	 * On 64-bit systems with IOMMU, use 40-bit dma_mask.
diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index b4c94144475..fd0b11ea556 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -308,13 +308,14 @@ tun_net_change_mtu(struct net_device *dev, int new_mtu)
 static const struct net_device_ops tun_netdev_ops = {
 	.ndo_open		= tun_net_open,
 	.ndo_stop		= tun_net_close,
+	.ndo_start_xmit		= tun_net_xmit,
 	.ndo_change_mtu		= tun_net_change_mtu,
-
 };
 
 static const struct net_device_ops tap_netdev_ops = {
 	.ndo_open		= tun_net_open,
 	.ndo_stop		= tun_net_close,
+	.ndo_start_xmit		= tun_net_xmit,
 	.ndo_change_mtu		= tun_net_change_mtu,
 	.ndo_set_multicast_list	= tun_net_mclist,
 	.ndo_set_mac_address	= eth_mac_addr,
@@ -691,7 +692,6 @@ static void tun_setup(struct net_device *dev)
 	tun->owner = -1;
 	tun->group = -1;
 
-	dev->hard_start_xmit = tun_net_xmit;
 	dev->ethtool_ops = &tun_ethtool_ops;
 	dev->destructor = free_netdev;
 	dev->features |= NETIF_F_NETNS_LOCAL;
diff --git a/drivers/net/veth.c b/drivers/net/veth.c
index 4f93a55aaaa..852d0e7c4e6 100644
--- a/drivers/net/veth.c
+++ b/drivers/net/veth.c
@@ -265,6 +265,7 @@ static void veth_dev_free(struct net_device *dev)
 static const struct net_device_ops veth_netdev_ops = {
 	.ndo_init	= veth_dev_init,
 	.ndo_open	= veth_open,
+	.ndo_start_xmit = veth_xmit,
 	.ndo_get_stats	= veth_get_stats,
 };
 
@@ -273,7 +274,6 @@ static void veth_setup(struct net_device *dev)
 	ether_setup(dev);
 
 	dev->netdev_ops = &veth_netdev_ops;
-	dev->hard_start_xmit = veth_xmit;
 	dev->ethtool_ops = &veth_ethtool_ops;
 	dev->features |= NETIF_F_LLTX;
 	dev->destructor = veth_dev_free;
diff --git a/drivers/net/via-velocity.c b/drivers/net/via-velocity.c
index 033e63a6843..58e25d090ae 100644
--- a/drivers/net/via-velocity.c
+++ b/drivers/net/via-velocity.c
@@ -852,6 +852,7 @@ static int velocity_soft_reset(struct velocity_info *vptr)
 static const struct net_device_ops velocity_netdev_ops = {
 	.ndo_open		= velocity_open,
 	.ndo_stop		= velocity_close,
+	.ndo_start_xmit		= velocity_xmit,
 	.ndo_get_stats		= velocity_get_stats,
 	.ndo_validate_addr	= eth_validate_addr,
 	.ndo_set_multicast_list	= velocity_set_multi,
@@ -971,7 +972,6 @@ static int __devinit velocity_found1(struct pci_dev *pdev, const struct pci_devi
 	vptr->phy_id = MII_GET_PHY_ID(vptr->mac_regs);
 
 	dev->irq = pdev->irq;
-	dev->hard_start_xmit = velocity_xmit;
 	dev->netdev_ops = &velocity_netdev_ops;
 	dev->ethtool_ops = &velocity_ethtool_ops;
 
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 981a089d514..d8fb23679ee 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -454,8 +454,8 @@ struct netdev_queue {
 
 /*
  * This structure defines the management hooks for network devices.
- * The following hooks can bed defined and are optonal (can be null)
- * unless otherwise noted.
+ * The following hooks can be defined; unless noted otherwise, they are
+ * optional and can be filled with a null pointer.
  *
  * int (*ndo_init)(struct net_device *dev);
  *     This function is called once when network device is registered.
@@ -475,6 +475,15 @@ struct netdev_queue {
  *     This function is called when network device transistions to the down
  *     state.
  *
+ * int (*ndo_hard_start_xmit)(struct sk_buff *skb, struct net_device *dev);
+ *	Called when a packet needs to be transmitted.
+ *	Must return NETDEV_TX_OK , NETDEV_TX_BUSY, or NETDEV_TX_LOCKED,
+ *	Required can not be NULL.
+ *
+ * u16 (*ndo_select_queue)(struct net_device *dev, struct sk_buff *skb);
+ *	Called to decide which queue to when device supports multiple
+ *	transmit queues.
+ *
  * void (*ndo_change_rx_flags)(struct net_device *dev, int flags);
  *	This function is called to allow device receiver to make
  *	changes to configuration when multicast or promiscious is enabled.
@@ -508,7 +517,7 @@ struct netdev_queue {
  *	of a device. If not defined, any request to change MTU will
  *	will return an error.
  *
- * void (*ndo_tx_timeout) (struct net_device *dev);
+ * void (*ndo_tx_timeout)(struct net_device *dev);
  *	Callback uses when the transmitter has not made any progress
  *	for dev->watchdog ticks.
  *
@@ -538,6 +547,10 @@ struct net_device_ops {
 	void			(*ndo_uninit)(struct net_device *dev);
 	int			(*ndo_open)(struct net_device *dev);
 	int			(*ndo_stop)(struct net_device *dev);
+	int			(*ndo_start_xmit) (struct sk_buff *skb,
+						   struct net_device *dev);
+	u16			(*ndo_select_queue)(struct net_device *dev,
+						    struct sk_buff *skb);
 #define HAVE_CHANGE_RX_FLAGS
 	void			(*ndo_change_rx_flags)(struct net_device *dev,
 						       int flags);
@@ -557,8 +570,10 @@ struct net_device_ops {
 	int			(*ndo_set_config)(struct net_device *dev,
 					          struct ifmap *map);
 #define HAVE_CHANGE_MTU
-	int			(*ndo_change_mtu)(struct net_device *dev, int new_mtu);
-
+	int			(*ndo_change_mtu)(struct net_device *dev,
+						  int new_mtu);
+	int			(*ndo_neigh_setup)(struct net_device *dev,
+						   struct neigh_parms *);
 #define HAVE_TX_TIMEOUT
 	void			(*ndo_tx_timeout) (struct net_device *dev);
 
@@ -761,18 +776,12 @@ struct net_device
 	/* Number of TX queues currently active in device  */
 	unsigned int		real_num_tx_queues;
 
-	/* Map buffer to appropriate transmit queue */
-	u16			(*select_queue)(struct net_device *dev,
-						struct sk_buff *skb);
-
 	unsigned long		tx_queue_len;	/* Max frames per queue allowed */
 	spinlock_t		tx_global_lock;
 /*
  * One part is mostly used on xmit path (device)
  */
 	void			*priv;	/* pointer to private data	*/
-	int			(*hard_start_xmit) (struct sk_buff *skb,
-						    struct net_device *dev);
 	/* These may be needed for future network-power-down code. */
 	unsigned long		trans_start;	/* Time (in jiffies) of last Tx	*/
 
@@ -800,8 +809,6 @@ struct net_device
 	/* Called from unregister, can be used to call free_netdev */
 	void (*destructor)(struct net_device *dev);
 
-	int (*neigh_setup)(struct net_device *dev, struct neigh_parms *);
-
 #ifdef CONFIG_NETPOLL
 	struct netpoll_info	*npinfo;
 #endif
@@ -842,6 +849,10 @@ struct net_device
 		void			(*uninit)(struct net_device *dev);
 		int			(*open)(struct net_device *dev);
 		int			(*stop)(struct net_device *dev);
+		int			(*hard_start_xmit) (struct sk_buff *skb,
+							    struct net_device *dev);
+		u16			(*select_queue)(struct net_device *dev,
+							struct sk_buff *skb);
 		void			(*change_rx_flags)(struct net_device *dev,
 							   int flags);
 		void			(*set_rx_mode)(struct net_device *dev);
@@ -854,6 +865,8 @@ struct net_device
 		int			(*set_config)(struct net_device *dev,
 						      struct ifmap *map);
 		int			(*change_mtu)(struct net_device *dev, int new_mtu);
+		int			(*neigh_setup)(struct net_device *dev,
+						       struct neigh_parms *);
 		void			(*tx_timeout) (struct net_device *dev);
 		struct net_device_stats* (*get_stats)(struct net_device *dev);
 		void			(*vlan_rx_register)(struct net_device *dev,
diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c
index 920ce334839..18538d7460d 100644
--- a/net/bridge/br_device.c
+++ b/net/bridge/br_device.c
@@ -163,10 +163,11 @@ static const struct ethtool_ops br_ethtool_ops = {
 static const struct net_device_ops br_netdev_ops = {
 	.ndo_open		 = br_dev_open,
 	.ndo_stop		 = br_dev_stop,
-	.ndo_set_mac_address = br_set_mac_address,
-	.ndo_set_multicast_list = br_dev_set_multicast_list,
-	.ndo_change_mtu	 = br_change_mtu,
-	.ndo_do_ioctl	= br_dev_ioctl,
+	.ndo_start_xmit		 = br_dev_xmit,
+	.ndo_set_mac_address	 = br_set_mac_address,
+	.ndo_set_multicast_list	 = br_dev_set_multicast_list,
+	.ndo_change_mtu		 = br_change_mtu,
+	.ndo_do_ioctl		 = br_dev_ioctl,
 };
 
 void br_dev_setup(struct net_device *dev)
@@ -175,7 +176,6 @@ void br_dev_setup(struct net_device *dev)
 	ether_setup(dev);
 
 	dev->netdev_ops = &br_netdev_ops;
-	dev->hard_start_xmit = br_dev_xmit;
 	dev->destructor = free_netdev;
 	SET_ETHTOOL_OPS(dev, &br_ethtool_ops);
 	dev->tx_queue_len = 0;
diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
index ee3a8dd13f5..727c5c510a6 100644
--- a/net/bridge/br_if.c
+++ b/net/bridge/br_if.c
@@ -373,7 +373,7 @@ int br_add_if(struct net_bridge *br, struct net_device *dev)
 	if (dev->flags & IFF_LOOPBACK || dev->type != ARPHRD_ETHER)
 		return -EINVAL;
 
-	if (dev->hard_start_xmit == br_dev_xmit)
+	if (dev->netdev_ops->ndo_start_xmit == br_dev_xmit)
 		return -ELOOP;
 
 	if (dev->br_port != NULL)
diff --git a/net/core/dev.c b/net/core/dev.c
index 8843f4e3f5e..4615e9a443a 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1660,6 +1660,9 @@ static int dev_gso_segment(struct sk_buff *skb)
 int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
 			struct netdev_queue *txq)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
+
+	prefetch(&dev->netdev_ops->ndo_start_xmit);
 	if (likely(!skb->next)) {
 		if (!list_empty(&ptype_all))
 			dev_queue_xmit_nit(skb, dev);
@@ -1671,7 +1674,7 @@ int dev_hard_start_xmit(struct sk_buff *skb, struct net_device *dev,
 				goto gso;
 		}
 
-		return dev->hard_start_xmit(skb, dev);
+		return ops->ndo_start_xmit(skb, dev);
 	}
 
 gso:
@@ -1681,7 +1684,7 @@ gso:
 
 		skb->next = nskb->next;
 		nskb->next = NULL;
-		rc = dev->hard_start_xmit(nskb, dev);
+		rc = ops->ndo_start_xmit(nskb, dev);
 		if (unlikely(rc)) {
 			nskb->next = skb->next;
 			skb->next = nskb;
@@ -1755,10 +1758,11 @@ static u16 simple_tx_hash(struct net_device *dev, struct sk_buff *skb)
 static struct netdev_queue *dev_pick_tx(struct net_device *dev,
 					struct sk_buff *skb)
 {
+	const struct net_device_ops *ops = dev->netdev_ops;
 	u16 queue_index = 0;
 
-	if (dev->select_queue)
-		queue_index = dev->select_queue(dev, skb);
+	if (ops->ndo_select_queue)
+		queue_index = ops->ndo_select_queue(dev, skb);
 	else if (dev->real_num_tx_queues > 1)
 		queue_index = simple_tx_hash(dev, skb);
 
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index cca6a55909e..9c3717a23cf 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -1327,9 +1327,9 @@ struct neigh_parms *neigh_parms_alloc(struct net_device *dev,
 				      struct neigh_table *tbl)
 {
 	struct neigh_parms *p, *ref;
-	struct net *net;
+	struct net *net = dev_net(dev);
+	const struct net_device_ops *ops = dev->netdev_ops;
 
-	net = dev_net(dev);
 	ref = lookup_neigh_params(tbl, net, 0);
 	if (!ref)
 		return NULL;
@@ -1341,7 +1341,7 @@ struct neigh_parms *neigh_parms_alloc(struct net_device *dev,
 		p->reachable_time =
 				neigh_rand_reach_time(p->base_reachable_time);
 
-		if (dev->neigh_setup && dev->neigh_setup(dev, p)) {
+		if (ops->ndo_neigh_setup && ops->ndo_neigh_setup(dev, p)) {
 			kfree(p);
 			return NULL;
 		}
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 630df603444..96fb0519eb7 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -58,6 +58,7 @@ static void queue_process(struct work_struct *work)
 
 	while ((skb = skb_dequeue(&npinfo->txq))) {
 		struct net_device *dev = skb->dev;
+		const struct net_device_ops *ops = dev->netdev_ops;
 		struct netdev_queue *txq;
 
 		if (!netif_device_present(dev) || !netif_running(dev)) {
@@ -71,7 +72,7 @@ static void queue_process(struct work_struct *work)
 		__netif_tx_lock(txq, smp_processor_id());
 		if (netif_tx_queue_stopped(txq) ||
 		    netif_tx_queue_frozen(txq) ||
-		    dev->hard_start_xmit(skb, dev) != NETDEV_TX_OK) {
+		    ops->ndo_start_xmit(skb, dev) != NETDEV_TX_OK) {
 			skb_queue_head(&npinfo->txq, skb);
 			__netif_tx_unlock(txq);
 			local_irq_restore(flags);
@@ -273,6 +274,7 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb)
 	int status = NETDEV_TX_BUSY;
 	unsigned long tries;
 	struct net_device *dev = np->dev;
+	const struct net_device_ops *ops = dev->netdev_ops;
 	struct netpoll_info *npinfo = np->dev->npinfo;
 
 	if (!npinfo || !netif_running(dev) || !netif_device_present(dev)) {
@@ -293,7 +295,7 @@ static void netpoll_send_skb(struct netpoll *np, struct sk_buff *skb)
 		     tries > 0; --tries) {
 			if (__netif_tx_trylock(txq)) {
 				if (!netif_tx_queue_stopped(txq))
-					status = dev->hard_start_xmit(skb, dev);
+					status = ops->ndo_start_xmit(skb, dev);
 				__netif_tx_unlock(txq);
 
 				if (status == NETDEV_TX_OK)
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index 4e77914c4d4..15e0c2c7aac 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -3352,14 +3352,14 @@ static void pktgen_rem_thread(struct pktgen_thread *t)
 
 static __inline__ void pktgen_xmit(struct pktgen_dev *pkt_dev)
 {
-	struct net_device *odev = NULL;
+	struct net_device *odev = pkt_dev->odev;
+	int (*xmit)(struct sk_buff *, struct net_device *)
+		= odev->netdev_ops->ndo_start_xmit;
 	struct netdev_queue *txq;
 	__u64 idle_start = 0;
 	u16 queue_map;
 	int ret;
 
-	odev = pkt_dev->odev;
-
 	if (pkt_dev->delay_us || pkt_dev->delay_ns) {
 		u64 now;
 
@@ -3440,7 +3440,7 @@ static __inline__ void pktgen_xmit(struct pktgen_dev *pkt_dev)
 
 		atomic_inc(&(pkt_dev->skb->users));
 	      retry_now:
-		ret = odev->hard_start_xmit(pkt_dev->skb, odev);
+		ret = (*xmit)(pkt_dev->skb, odev);
 		if (likely(ret == NETDEV_TX_OK)) {
 			pkt_dev->last_ok = 1;
 			pkt_dev->sofar++;
-- 
cgit v1.2.3-70-g09d2


From 2f90b8657ec942d1880f720e0177ee71df7c8e3c Mon Sep 17 00:00:00 2001
From: Alexander Duyck <alexander.h.duyck@intel.com>
Date: Thu, 20 Nov 2008 20:52:10 -0800
Subject: ixgbe: this patch adds support for DCB to the kernel and ixgbe driver

This adds support for Data Center Bridging (DCB) features in the ixgbe
driver and adds an rtnetlink interface for configuring DCB to the
kernel.  The DCB feature support included are Priority Grouping (PG) -
which allows bandwidth guarantees to be allocated to groups to traffic
based on the 802.1q priority, and Priority Based Flow Control (PFC) -
which introduces a new MAC control PAUSE frame which works at
granularity of the 802.1p priority instead of the link (IEEE 802.3x).

Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/Kconfig                 |  10 +
 drivers/net/ixgbe/Makefile          |   2 +
 drivers/net/ixgbe/ixgbe.h           |  25 +-
 drivers/net/ixgbe/ixgbe_dcb.c       | 332 +++++++++++++++++
 drivers/net/ixgbe/ixgbe_dcb.h       | 157 ++++++++
 drivers/net/ixgbe/ixgbe_dcb_82598.c | 398 ++++++++++++++++++++
 drivers/net/ixgbe/ixgbe_dcb_82598.h |  94 +++++
 drivers/net/ixgbe/ixgbe_dcb_nl.c    | 356 ++++++++++++++++++
 drivers/net/ixgbe/ixgbe_ethtool.c   |  30 +-
 drivers/net/ixgbe/ixgbe_main.c      | 200 +++++++++-
 include/linux/dcbnl.h               | 230 ++++++++++++
 include/linux/netdevice.h           |   8 +
 include/linux/rtnetlink.h           |   5 +
 include/net/dcbnl.h                 |  44 +++
 net/Kconfig                         |   1 +
 net/Makefile                        |   3 +
 net/dcb/Kconfig                     |  12 +
 net/dcb/Makefile                    |   1 +
 net/dcb/dcbnl.c                     | 704 ++++++++++++++++++++++++++++++++++++
 19 files changed, 2593 insertions(+), 19 deletions(-)
 create mode 100644 drivers/net/ixgbe/ixgbe_dcb.c
 create mode 100644 drivers/net/ixgbe/ixgbe_dcb.h
 create mode 100644 drivers/net/ixgbe/ixgbe_dcb_82598.c
 create mode 100644 drivers/net/ixgbe/ixgbe_dcb_82598.h
 create mode 100644 drivers/net/ixgbe/ixgbe_dcb_nl.c
 create mode 100644 include/linux/dcbnl.h
 create mode 100644 include/net/dcbnl.h
 create mode 100644 net/dcb/Kconfig
 create mode 100644 net/dcb/Makefile
 create mode 100644 net/dcb/dcbnl.c

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index afa206590ad..efd461d7c2b 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2451,6 +2451,16 @@ config IXGBE_DCA
 	  driver.  DCA is a method for warming the CPU cache before data
 	  is used, with the intent of lessening the impact of cache misses.
 
+config IXGBE_DCBNL
+	bool "Data Center Bridging (DCB) Support"
+	default n
+	depends on IXGBE && DCBNL
+	---help---
+	  Say Y here if you want to use Data Center Bridging (DCB) in the
+	  driver.
+
+	  If unsure, say N.
+
 config IXGB
 	tristate "Intel(R) PRO/10GbE support"
 	depends on PCI
diff --git a/drivers/net/ixgbe/Makefile b/drivers/net/ixgbe/Makefile
index ccd83d9f579..3228e508e62 100644
--- a/drivers/net/ixgbe/Makefile
+++ b/drivers/net/ixgbe/Makefile
@@ -34,3 +34,5 @@ obj-$(CONFIG_IXGBE) += ixgbe.o
 
 ixgbe-objs := ixgbe_main.o ixgbe_common.o ixgbe_ethtool.o \
               ixgbe_82598.o ixgbe_phy.o
+
+ixgbe-$(CONFIG_IXGBE_DCBNL) +=  ixgbe_dcb.o ixgbe_dcb_82598.o ixgbe_dcb_nl.o
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index 132854f646b..796f189f387 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -35,7 +35,7 @@
 
 #include "ixgbe_type.h"
 #include "ixgbe_common.h"
-
+#include "ixgbe_dcb.h"
 #ifdef CONFIG_IXGBE_DCA
 #include <linux/dca.h>
 #endif
@@ -84,6 +84,7 @@
 #define IXGBE_TX_FLAGS_TSO		(u32)(1 << 2)
 #define IXGBE_TX_FLAGS_IPV4		(u32)(1 << 3)
 #define IXGBE_TX_FLAGS_VLAN_MASK	0xffff0000
+#define IXGBE_TX_FLAGS_VLAN_PRIO_MASK   0x0000e000
 #define IXGBE_TX_FLAGS_VLAN_SHIFT	16
 
 #define IXGBE_MAX_LRO_DESCRIPTORS       8
@@ -134,7 +135,7 @@ struct ixgbe_ring {
 
 	u16 reg_idx; /* holds the special value that gets the hardware register
 		      * offset associated with this ring, which is different
-		      * for DCE and RSS modes */
+		      * for DCB and RSS modes */
 
 #ifdef CONFIG_IXGBE_DCA
 	/* cpu for tx queue */
@@ -152,8 +153,10 @@ struct ixgbe_ring {
 	u16 rx_buf_len;
 };
 
+#define RING_F_DCB  0
 #define RING_F_VMDQ 1
 #define RING_F_RSS  2
+#define IXGBE_MAX_DCB_INDICES   8
 #define IXGBE_MAX_RSS_INDICES  16
 #define IXGBE_MAX_VMDQ_INDICES 16
 struct ixgbe_ring_feature {
@@ -164,6 +167,10 @@ struct ixgbe_ring_feature {
 #define MAX_RX_QUEUES 64
 #define MAX_TX_QUEUES 32
 
+#define MAX_RX_PACKET_BUFFERS ((adapter->flags & IXGBE_FLAG_DCB_ENABLED) \
+                              ? 8 : 1)
+#define MAX_TX_PACKET_BUFFERS MAX_RX_PACKET_BUFFERS
+
 /* MAX_MSIX_Q_VECTORS of these are allocated,
  * but we only use one per queue-specific vector.
  */
@@ -215,6 +222,9 @@ struct ixgbe_adapter {
 	struct work_struct reset_task;
 	struct ixgbe_q_vector q_vector[MAX_MSIX_Q_VECTORS];
 	char name[MAX_MSIX_COUNT][IFNAMSIZ + 5];
+	struct ixgbe_dcb_config dcb_cfg;
+	struct ixgbe_dcb_config temp_dcb_cfg;
+	u8 dcb_set_bitmap;
 
 	/* Interrupt Throttle Rate */
 	u32 itr_setting;
@@ -270,6 +280,7 @@ struct ixgbe_adapter {
 #define IXGBE_FLAG_FAN_FAIL_CAPABLE             (u32)(1 << 20)
 #define IXGBE_FLAG_NEED_LINK_UPDATE             (u32)(1 << 22)
 #define IXGBE_FLAG_IN_WATCHDOG_TASK             (u32)(1 << 23)
+#define IXGBE_FLAG_DCB_ENABLED                  (u32)(1 << 24)
 
 /* default to trying for four seconds */
 #define IXGBE_TRY_LINK_TIMEOUT (4 * HZ)
@@ -313,6 +324,12 @@ enum ixgbe_boards {
 };
 
 extern struct ixgbe_info ixgbe_82598_info;
+#ifdef CONFIG_IXGBE_DCBNL
+extern struct dcbnl_rtnl_ops dcbnl_ops;
+extern int ixgbe_copy_dcb_cfg(struct ixgbe_dcb_config *src_dcb_cfg,
+                              struct ixgbe_dcb_config *dst_dcb_cfg,
+                              int tc_max);
+#endif
 
 extern char ixgbe_driver_name[];
 extern const char ixgbe_driver_version[];
@@ -327,5 +344,9 @@ extern int ixgbe_setup_tx_resources(struct ixgbe_adapter *, struct ixgbe_ring *)
 extern void ixgbe_free_rx_resources(struct ixgbe_adapter *, struct ixgbe_ring *);
 extern void ixgbe_free_tx_resources(struct ixgbe_adapter *, struct ixgbe_ring *);
 extern void ixgbe_update_stats(struct ixgbe_adapter *adapter);
+extern void ixgbe_reset_interrupt_capability(struct ixgbe_adapter *adapter);
+extern int ixgbe_init_interrupt_scheme(struct ixgbe_adapter *adapter);
+void ixgbe_napi_add_all(struct ixgbe_adapter *adapter);
+void ixgbe_napi_del_all(struct ixgbe_adapter *adapter);
 
 #endif /* _IXGBE_H_ */
diff --git a/drivers/net/ixgbe/ixgbe_dcb.c b/drivers/net/ixgbe/ixgbe_dcb.c
new file mode 100644
index 00000000000..e2e28ac63de
--- /dev/null
+++ b/drivers/net/ixgbe/ixgbe_dcb.c
@@ -0,0 +1,332 @@
+/*******************************************************************************
+
+  Intel 10 Gigabit PCI Express Linux driver
+  Copyright(c) 1999 - 2007 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  Linux NICS <linux.nics@intel.com>
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+
+#include "ixgbe.h"
+#include "ixgbe_type.h"
+#include "ixgbe_dcb.h"
+#include "ixgbe_dcb_82598.h"
+
+/**
+ * ixgbe_dcb_config - Struct containing DCB settings.
+ * @dcb_config: Pointer to DCB config structure
+ *
+ * This function checks DCB rules for DCB settings.
+ * The following rules are checked:
+ * 1. The sum of bandwidth percentages of all Bandwidth Groups must total 100%.
+ * 2. The sum of bandwidth percentages of all Traffic Classes within a Bandwidth
+ *    Group must total 100.
+ * 3. A Traffic Class should not be set to both Link Strict Priority
+ *    and Group Strict Priority.
+ * 4. Link strict Bandwidth Groups can only have link strict traffic classes
+ *    with zero bandwidth.
+ */
+s32 ixgbe_dcb_check_config(struct ixgbe_dcb_config *dcb_config)
+{
+	struct tc_bw_alloc *p;
+	s32 ret_val = 0;
+	u8 i, j, bw = 0, bw_id;
+	u8 bw_sum[2][MAX_BW_GROUP];
+	bool link_strict[2][MAX_BW_GROUP];
+
+	memset(bw_sum, 0, sizeof(bw_sum));
+	memset(link_strict, 0, sizeof(link_strict));
+
+	/* First Tx, then Rx */
+	for (i = 0; i < 2; i++) {
+		/* Check each traffic class for rule violation */
+		for (j = 0; j < MAX_TRAFFIC_CLASS; j++) {
+			p = &dcb_config->tc_config[j].path[i];
+
+			bw = p->bwg_percent;
+			bw_id = p->bwg_id;
+
+			if (bw_id >= MAX_BW_GROUP) {
+				ret_val = DCB_ERR_CONFIG;
+				goto err_config;
+			}
+			if (p->prio_type == prio_link) {
+				link_strict[i][bw_id] = true;
+				/* Link strict should have zero bandwidth */
+				if (bw) {
+					ret_val = DCB_ERR_LS_BW_NONZERO;
+					goto err_config;
+				}
+			} else if (!bw) {
+				/*
+				 * Traffic classes without link strict
+				 * should have non-zero bandwidth.
+				 */
+				ret_val = DCB_ERR_TC_BW_ZERO;
+				goto err_config;
+			}
+			bw_sum[i][bw_id] += bw;
+		}
+
+		bw = 0;
+
+		/* Check each bandwidth group for rule violation */
+		for (j = 0; j < MAX_BW_GROUP; j++) {
+			bw += dcb_config->bw_percentage[i][j];
+			/*
+			 * Sum of bandwidth percentages of all traffic classes
+			 * within a Bandwidth Group must total 100 except for
+			 * link strict group (zero bandwidth).
+			 */
+			if (link_strict[i][j]) {
+				if (bw_sum[i][j]) {
+					/*
+					 * Link strict group should have zero
+					 * bandwidth.
+					 */
+					ret_val = DCB_ERR_LS_BWG_NONZERO;
+					goto err_config;
+				}
+			} else if (bw_sum[i][j] != BW_PERCENT &&
+				   bw_sum[i][j] != 0) {
+				ret_val = DCB_ERR_TC_BW;
+				goto err_config;
+			}
+		}
+
+		if (bw != BW_PERCENT) {
+			ret_val = DCB_ERR_BW_GROUP;
+			goto err_config;
+		}
+	}
+
+err_config:
+	return ret_val;
+}
+
+/**
+ * ixgbe_dcb_calculate_tc_credits - Calculates traffic class credits
+ * @ixgbe_dcb_config: Struct containing DCB settings.
+ * @direction: Configuring either Tx or Rx.
+ *
+ * This function calculates the credits allocated to each traffic class.
+ * It should be called only after the rules are checked by
+ * ixgbe_dcb_check_config().
+ */
+s32 ixgbe_dcb_calculate_tc_credits(struct ixgbe_dcb_config *dcb_config,
+                                   u8 direction)
+{
+	struct tc_bw_alloc *p;
+	s32 ret_val = 0;
+	/* Initialization values default for Tx settings */
+	u32 credit_refill       = 0;
+	u32 credit_max          = 0;
+	u16 link_percentage     = 0;
+	u8  bw_percent          = 0;
+	u8  i;
+
+	if (dcb_config == NULL) {
+		ret_val = DCB_ERR_CONFIG;
+		goto out;
+	}
+
+	/* Find out the link percentage for each TC first */
+	for (i = 0; i < MAX_TRAFFIC_CLASS; i++) {
+		p = &dcb_config->tc_config[i].path[direction];
+		bw_percent = dcb_config->bw_percentage[direction][p->bwg_id];
+
+		link_percentage = p->bwg_percent;
+		/* Must be careful of integer division for very small nums */
+		link_percentage = (link_percentage * bw_percent) / 100;
+		if (p->bwg_percent > 0 && link_percentage == 0)
+			link_percentage = 1;
+
+		/* Save link_percentage for reference */
+		p->link_percent = (u8)link_percentage;
+
+		/* Calculate credit refill and save it */
+		credit_refill = link_percentage * MINIMUM_CREDIT_REFILL;
+		p->data_credits_refill = (u16)credit_refill;
+
+		/* Calculate maximum credit for the TC */
+		credit_max = (link_percentage * MAX_CREDIT) / 100;
+
+		/*
+		 * Adjustment based on rule checking, if the percentage
+		 * of a TC is too small, the maximum credit may not be
+		 * enough to send out a jumbo frame in data plane arbitration.
+		 */
+		if (credit_max && (credit_max < MINIMUM_CREDIT_FOR_JUMBO))
+			credit_max = MINIMUM_CREDIT_FOR_JUMBO;
+
+		if (direction == DCB_TX_CONFIG) {
+			/*
+			 * Adjustment based on rule checking, if the
+			 * percentage of a TC is too small, the maximum
+			 * credit may not be enough to send out a TSO
+			 * packet in descriptor plane arbitration.
+			 */
+			if (credit_max &&
+			    (credit_max < MINIMUM_CREDIT_FOR_TSO))
+				credit_max = MINIMUM_CREDIT_FOR_TSO;
+
+			dcb_config->tc_config[i].desc_credits_max =
+				(u16)credit_max;
+		}
+
+		p->data_credits_max = (u16)credit_max;
+	}
+
+out:
+	return ret_val;
+}
+
+/**
+ * ixgbe_dcb_get_tc_stats - Returns status of each traffic class
+ * @hw: pointer to hardware structure
+ * @stats: pointer to statistics structure
+ * @tc_count:  Number of elements in bwg_array.
+ *
+ * This function returns the status data for each of the Traffic Classes in use.
+ */
+s32 ixgbe_dcb_get_tc_stats(struct ixgbe_hw *hw, struct ixgbe_hw_stats *stats,
+                           u8 tc_count)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_get_tc_stats_82598(hw, stats, tc_count);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_get_pfc_stats - Returns CBFC status of each traffic class
+ * hw - pointer to hardware structure
+ * stats - pointer to statistics structure
+ * tc_count -  Number of elements in bwg_array.
+ *
+ * This function returns the CBFC status data for each of the Traffic Classes.
+ */
+s32 ixgbe_dcb_get_pfc_stats(struct ixgbe_hw *hw, struct ixgbe_hw_stats *stats,
+                            u8 tc_count)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_get_pfc_stats_82598(hw, stats, tc_count);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_config_rx_arbiter - Config Rx arbiter
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Rx Data Arbiter and credits for each traffic class.
+ */
+s32 ixgbe_dcb_config_rx_arbiter(struct ixgbe_hw *hw,
+                                struct ixgbe_dcb_config *dcb_config)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_config_rx_arbiter_82598(hw, dcb_config);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_config_tx_desc_arbiter - Config Tx Desc arbiter
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Tx Descriptor Arbiter and credits for each traffic class.
+ */
+s32 ixgbe_dcb_config_tx_desc_arbiter(struct ixgbe_hw *hw,
+                                     struct ixgbe_dcb_config *dcb_config)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_config_tx_desc_arbiter_82598(hw, dcb_config);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_config_tx_data_arbiter - Config Tx data arbiter
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Tx Data Arbiter and credits for each traffic class.
+ */
+s32 ixgbe_dcb_config_tx_data_arbiter(struct ixgbe_hw *hw,
+                                     struct ixgbe_dcb_config *dcb_config)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_config_tx_data_arbiter_82598(hw, dcb_config);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_config_pfc - Config priority flow control
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Priority Flow Control for each traffic class.
+ */
+s32 ixgbe_dcb_config_pfc(struct ixgbe_hw *hw,
+                         struct ixgbe_dcb_config *dcb_config)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_config_pfc_82598(hw, dcb_config);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_config_tc_stats - Config traffic class statistics
+ * @hw: pointer to hardware structure
+ *
+ * Configure queue statistics registers, all queues belonging to same traffic
+ * class uses a single set of queue statistics counters.
+ */
+s32 ixgbe_dcb_config_tc_stats(struct ixgbe_hw *hw)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_config_tc_stats_82598(hw);
+	return ret;
+}
+
+/**
+ * ixgbe_dcb_hw_config - Config and enable DCB
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure dcb settings and enable dcb mode.
+ */
+s32 ixgbe_dcb_hw_config(struct ixgbe_hw *hw,
+                        struct ixgbe_dcb_config *dcb_config)
+{
+	s32 ret = 0;
+	if (hw->mac.type == ixgbe_mac_82598EB)
+		ret = ixgbe_dcb_hw_config_82598(hw, dcb_config);
+	return ret;
+}
+
diff --git a/drivers/net/ixgbe/ixgbe_dcb.h b/drivers/net/ixgbe/ixgbe_dcb.h
new file mode 100644
index 00000000000..62dfd243bed
--- /dev/null
+++ b/drivers/net/ixgbe/ixgbe_dcb.h
@@ -0,0 +1,157 @@
+/*******************************************************************************
+
+  Intel 10 Gigabit PCI Express Linux driver
+  Copyright(c) 1999 - 2007 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  Linux NICS <linux.nics@intel.com>
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#ifndef _DCB_CONFIG_H_
+#define _DCB_CONFIG_H_
+
+#include "ixgbe_type.h"
+
+/* DCB data structures */
+
+#define IXGBE_MAX_PACKET_BUFFERS 8
+#define MAX_USER_PRIORITY        8
+#define MAX_TRAFFIC_CLASS        8
+#define MAX_BW_GROUP             8
+#define BW_PERCENT               100
+
+#define DCB_TX_CONFIG            0
+#define DCB_RX_CONFIG            1
+
+/* DCB error Codes */
+#define DCB_SUCCESS              0
+#define DCB_ERR_CONFIG           -1
+#define DCB_ERR_PARAM            -2
+
+/* Transmit and receive Errors */
+/* Error in bandwidth group allocation */
+#define DCB_ERR_BW_GROUP        -3
+/* Error in traffic class bandwidth allocation */
+#define DCB_ERR_TC_BW           -4
+/* Traffic class has both link strict and group strict enabled */
+#define DCB_ERR_LS_GS           -5
+/* Link strict traffic class has non zero bandwidth */
+#define DCB_ERR_LS_BW_NONZERO   -6
+/* Link strict bandwidth group has non zero bandwidth */
+#define DCB_ERR_LS_BWG_NONZERO  -7
+/*  Traffic class has zero bandwidth */
+#define DCB_ERR_TC_BW_ZERO      -8
+
+#define DCB_NOT_IMPLEMENTED      0x7FFFFFFF
+
+struct dcb_pfc_tc_debug {
+	u8  tc;
+	u8  pause_status;
+	u64 pause_quanta;
+};
+
+enum strict_prio_type {
+	prio_none = 0,
+	prio_group,
+	prio_link
+};
+
+/* Traffic class bandwidth allocation per direction */
+struct tc_bw_alloc {
+	u8 bwg_id;		  /* Bandwidth Group (BWG) ID */
+	u8 bwg_percent;		  /* % of BWG's bandwidth */
+	u8 link_percent;	  /* % of link bandwidth */
+	u8 up_to_tc_bitmap;	  /* User Priority to Traffic Class mapping */
+	u16 data_credits_refill;  /* Credit refill amount in 64B granularity */
+	u16 data_credits_max;	  /* Max credits for a configured packet buffer
+				   * in 64B granularity.*/
+	enum strict_prio_type prio_type; /* Link or Group Strict Priority */
+};
+
+enum dcb_pfc_type {
+	pfc_disabled = 0,
+	pfc_enabled_full,
+	pfc_enabled_tx,
+	pfc_enabled_rx
+};
+
+/* Traffic class configuration */
+struct tc_configuration {
+	struct tc_bw_alloc path[2]; /* One each for Tx/Rx */
+	enum dcb_pfc_type  dcb_pfc; /* Class based flow control setting */
+
+	u16 desc_credits_max; /* For Tx Descriptor arbitration */
+	u8 tc; /* Traffic class (TC) */
+};
+
+enum dcb_rx_pba_cfg {
+	pba_equal,     /* PBA[0-7] each use 64KB FIFO */
+	pba_80_48      /* PBA[0-3] each use 80KB, PBA[4-7] each use 48KB */
+};
+
+struct ixgbe_dcb_config {
+	struct tc_configuration tc_config[MAX_TRAFFIC_CLASS];
+	u8     bw_percentage[2][MAX_BW_GROUP]; /* One each for Tx/Rx */
+
+	bool  round_robin_enable;
+
+	enum dcb_rx_pba_cfg rx_pba_cfg;
+
+	u32  dcb_cfg_version; /* Not used...OS-specific? */
+	u32  link_speed; /* For bandwidth allocation validation purpose */
+};
+
+/* DCB driver APIs */
+
+/* DCB rule checking function.*/
+s32 ixgbe_dcb_check_config(struct ixgbe_dcb_config *config);
+
+/* DCB credits calculation */
+s32 ixgbe_dcb_calculate_tc_credits(struct ixgbe_dcb_config *, u8);
+
+/* DCB PFC functions */
+s32 ixgbe_dcb_config_pfc(struct ixgbe_hw *, struct ixgbe_dcb_config *g);
+s32 ixgbe_dcb_get_pfc_stats(struct ixgbe_hw *, struct ixgbe_hw_stats *, u8);
+
+/* DCB traffic class stats */
+s32 ixgbe_dcb_config_tc_stats(struct ixgbe_hw *);
+s32 ixgbe_dcb_get_tc_stats(struct ixgbe_hw *, struct ixgbe_hw_stats *, u8);
+
+/* DCB config arbiters */
+s32 ixgbe_dcb_config_tx_desc_arbiter(struct ixgbe_hw *,
+                                     struct ixgbe_dcb_config *);
+s32 ixgbe_dcb_config_tx_data_arbiter(struct ixgbe_hw *,
+                                     struct ixgbe_dcb_config *);
+s32 ixgbe_dcb_config_rx_arbiter(struct ixgbe_hw *, struct ixgbe_dcb_config *);
+
+/* DCB hw initialization */
+s32 ixgbe_dcb_hw_config(struct ixgbe_hw *, struct ixgbe_dcb_config *);
+
+/* DCB definitions for credit calculation */
+#define MAX_CREDIT_REFILL       511  /* 0x1FF * 64B = 32704B */
+#define MINIMUM_CREDIT_REFILL   5    /* 5*64B = 320B */
+#define MINIMUM_CREDIT_FOR_JUMBO 145  /* 145= UpperBound((9*1024+54)/64B) for 9KB jumbo frame */
+#define DCB_MAX_TSO_SIZE        (32*1024) /* MAX TSO packet size supported in DCB mode */
+#define MINIMUM_CREDIT_FOR_TSO  (DCB_MAX_TSO_SIZE/64 + 1) /* 513 for 32KB TSO packet */
+#define MAX_CREDIT              4095 /* Maximum credit supported: 256KB * 1204 / 64B */
+
+#endif /* _DCB_CONFIG_H */
diff --git a/drivers/net/ixgbe/ixgbe_dcb_82598.c b/drivers/net/ixgbe/ixgbe_dcb_82598.c
new file mode 100644
index 00000000000..fce6867a451
--- /dev/null
+++ b/drivers/net/ixgbe/ixgbe_dcb_82598.c
@@ -0,0 +1,398 @@
+/*******************************************************************************
+
+  Intel 10 Gigabit PCI Express Linux driver
+  Copyright(c) 1999 - 2007 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  Linux NICS <linux.nics@intel.com>
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#include "ixgbe.h"
+#include "ixgbe_type.h"
+#include "ixgbe_dcb.h"
+#include "ixgbe_dcb_82598.h"
+
+/**
+ * ixgbe_dcb_get_tc_stats_82598 - Return status data for each traffic class
+ * @hw: pointer to hardware structure
+ * @stats: pointer to statistics structure
+ * @tc_count:  Number of elements in bwg_array.
+ *
+ * This function returns the status data for each of the Traffic Classes in use.
+ */
+s32 ixgbe_dcb_get_tc_stats_82598(struct ixgbe_hw *hw,
+                                 struct ixgbe_hw_stats *stats,
+                                 u8 tc_count)
+{
+	int tc;
+
+	if (tc_count > MAX_TRAFFIC_CLASS)
+		return DCB_ERR_PARAM;
+
+	/* Statistics pertaining to each traffic class */
+	for (tc = 0; tc < tc_count; tc++) {
+		/* Transmitted Packets */
+		stats->qptc[tc] += IXGBE_READ_REG(hw, IXGBE_QPTC(tc));
+		/* Transmitted Bytes */
+		stats->qbtc[tc] += IXGBE_READ_REG(hw, IXGBE_QBTC(tc));
+		/* Received Packets */
+		stats->qprc[tc] += IXGBE_READ_REG(hw, IXGBE_QPRC(tc));
+		/* Received Bytes */
+		stats->qbrc[tc] += IXGBE_READ_REG(hw, IXGBE_QBRC(tc));
+	}
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_get_pfc_stats_82598 - Returns CBFC status data
+ * @hw: pointer to hardware structure
+ * @stats: pointer to statistics structure
+ * @tc_count:  Number of elements in bwg_array.
+ *
+ * This function returns the CBFC status data for each of the Traffic Classes.
+ */
+s32 ixgbe_dcb_get_pfc_stats_82598(struct ixgbe_hw *hw,
+                                  struct ixgbe_hw_stats *stats,
+                                  u8 tc_count)
+{
+	int tc;
+
+	if (tc_count > MAX_TRAFFIC_CLASS)
+		return DCB_ERR_PARAM;
+
+	for (tc = 0; tc < tc_count; tc++) {
+		/* Priority XOFF Transmitted */
+		stats->pxofftxc[tc] += IXGBE_READ_REG(hw, IXGBE_PXOFFTXC(tc));
+		/* Priority XOFF Received */
+		stats->pxoffrxc[tc] += IXGBE_READ_REG(hw, IXGBE_PXOFFRXC(tc));
+	}
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_config_packet_buffers_82598 - Configure packet buffers
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure packet buffers for DCB mode.
+ */
+s32 ixgbe_dcb_config_packet_buffers_82598(struct ixgbe_hw *hw,
+                                          struct ixgbe_dcb_config *dcb_config)
+{
+	s32 ret_val = 0;
+	u32 value = IXGBE_RXPBSIZE_64KB;
+	u8  i = 0;
+
+	/* Setup Rx packet buffer sizes */
+	switch (dcb_config->rx_pba_cfg) {
+	case pba_80_48:
+		/* Setup the first four at 80KB */
+		value = IXGBE_RXPBSIZE_80KB;
+		for (; i < 4; i++)
+			IXGBE_WRITE_REG(hw, IXGBE_RXPBSIZE(i), value);
+		/* Setup the last four at 48KB...don't re-init i */
+		value = IXGBE_RXPBSIZE_48KB;
+		/* Fall Through */
+	case pba_equal:
+	default:
+		for (; i < IXGBE_MAX_PACKET_BUFFERS; i++)
+			IXGBE_WRITE_REG(hw, IXGBE_RXPBSIZE(i), value);
+
+		/* Setup Tx packet buffer sizes */
+		for (i = 0; i < IXGBE_MAX_PACKET_BUFFERS; i++) {
+			IXGBE_WRITE_REG(hw, IXGBE_TXPBSIZE(i),
+					IXGBE_TXPBSIZE_40KB);
+		}
+		break;
+	}
+
+	return ret_val;
+}
+
+/**
+ * ixgbe_dcb_config_rx_arbiter_82598 - Config Rx data arbiter
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Rx Data Arbiter and credits for each traffic class.
+ */
+s32 ixgbe_dcb_config_rx_arbiter_82598(struct ixgbe_hw *hw,
+                                      struct ixgbe_dcb_config *dcb_config)
+{
+	struct tc_bw_alloc    *p;
+	u32    reg           = 0;
+	u32    credit_refill = 0;
+	u32    credit_max    = 0;
+	u8     i             = 0;
+
+	reg = IXGBE_READ_REG(hw, IXGBE_RUPPBMR) | IXGBE_RUPPBMR_MQA;
+	IXGBE_WRITE_REG(hw, IXGBE_RUPPBMR, reg);
+
+	reg = IXGBE_READ_REG(hw, IXGBE_RMCS);
+	/* Enable Arbiter */
+	reg &= ~IXGBE_RMCS_ARBDIS;
+	/* Enable Receive Recycle within the BWG */
+	reg |= IXGBE_RMCS_RRM;
+	/* Enable Deficit Fixed Priority arbitration*/
+	reg |= IXGBE_RMCS_DFP;
+
+	IXGBE_WRITE_REG(hw, IXGBE_RMCS, reg);
+
+	/* Configure traffic class credits and priority */
+	for (i = 0; i < MAX_TRAFFIC_CLASS; i++) {
+		p = &dcb_config->tc_config[i].path[DCB_RX_CONFIG];
+		credit_refill = p->data_credits_refill;
+		credit_max    = p->data_credits_max;
+
+		reg = credit_refill | (credit_max << IXGBE_RT2CR_MCL_SHIFT);
+
+		if (p->prio_type == prio_link)
+			reg |= IXGBE_RT2CR_LSP;
+
+		IXGBE_WRITE_REG(hw, IXGBE_RT2CR(i), reg);
+	}
+
+	reg = IXGBE_READ_REG(hw, IXGBE_RDRXCTL);
+	reg |= IXGBE_RDRXCTL_RDMTS_1_2;
+	reg |= IXGBE_RDRXCTL_MPBEN;
+	reg |= IXGBE_RDRXCTL_MCEN;
+	IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, reg);
+
+	reg = IXGBE_READ_REG(hw, IXGBE_RXCTRL);
+	/* Make sure there is enough descriptors before arbitration */
+	reg &= ~IXGBE_RXCTRL_DMBYPS;
+	IXGBE_WRITE_REG(hw, IXGBE_RXCTRL, reg);
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_config_tx_desc_arbiter_82598 - Config Tx Desc. arbiter
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Tx Descriptor Arbiter and credits for each traffic class.
+ */
+s32 ixgbe_dcb_config_tx_desc_arbiter_82598(struct ixgbe_hw *hw,
+                                           struct ixgbe_dcb_config *dcb_config)
+{
+	struct tc_bw_alloc *p;
+	u32    reg, max_credits;
+	u8     i;
+
+	reg = IXGBE_READ_REG(hw, IXGBE_DPMCS);
+
+	/* Enable arbiter */
+	reg &= ~IXGBE_DPMCS_ARBDIS;
+	if (!(dcb_config->round_robin_enable)) {
+		/* Enable DFP and Recycle mode */
+		reg |= (IXGBE_DPMCS_TDPAC | IXGBE_DPMCS_TRM);
+	}
+	reg |= IXGBE_DPMCS_TSOEF;
+	/* Configure Max TSO packet size 34KB including payload and headers */
+	reg |= (0x4 << IXGBE_DPMCS_MTSOS_SHIFT);
+
+	IXGBE_WRITE_REG(hw, IXGBE_DPMCS, reg);
+
+	/* Configure traffic class credits and priority */
+	for (i = 0; i < MAX_TRAFFIC_CLASS; i++) {
+		p = &dcb_config->tc_config[i].path[DCB_TX_CONFIG];
+		max_credits = dcb_config->tc_config[i].desc_credits_max;
+		reg = max_credits << IXGBE_TDTQ2TCCR_MCL_SHIFT;
+		reg |= p->data_credits_refill;
+		reg |= (u32)(p->bwg_id) << IXGBE_TDTQ2TCCR_BWG_SHIFT;
+
+		if (p->prio_type == prio_group)
+			reg |= IXGBE_TDTQ2TCCR_GSP;
+
+		if (p->prio_type == prio_link)
+			reg |= IXGBE_TDTQ2TCCR_LSP;
+
+		IXGBE_WRITE_REG(hw, IXGBE_TDTQ2TCCR(i), reg);
+	}
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_config_tx_data_arbiter_82598 - Config Tx data arbiter
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Tx Data Arbiter and credits for each traffic class.
+ */
+s32 ixgbe_dcb_config_tx_data_arbiter_82598(struct ixgbe_hw *hw,
+                                           struct ixgbe_dcb_config *dcb_config)
+{
+	struct tc_bw_alloc *p;
+	u32 reg;
+	u8 i;
+
+	reg = IXGBE_READ_REG(hw, IXGBE_PDPMCS);
+	/* Enable Data Plane Arbiter */
+	reg &= ~IXGBE_PDPMCS_ARBDIS;
+	/* Enable DFP and Transmit Recycle Mode */
+	reg |= (IXGBE_PDPMCS_TPPAC | IXGBE_PDPMCS_TRM);
+
+	IXGBE_WRITE_REG(hw, IXGBE_PDPMCS, reg);
+
+	/* Configure traffic class credits and priority */
+	for (i = 0; i < MAX_TRAFFIC_CLASS; i++) {
+		p = &dcb_config->tc_config[i].path[DCB_TX_CONFIG];
+		reg = p->data_credits_refill;
+		reg |= (u32)(p->data_credits_max) << IXGBE_TDPT2TCCR_MCL_SHIFT;
+		reg |= (u32)(p->bwg_id) << IXGBE_TDPT2TCCR_BWG_SHIFT;
+
+		if (p->prio_type == prio_group)
+			reg |= IXGBE_TDPT2TCCR_GSP;
+
+		if (p->prio_type == prio_link)
+			reg |= IXGBE_TDPT2TCCR_LSP;
+
+		IXGBE_WRITE_REG(hw, IXGBE_TDPT2TCCR(i), reg);
+	}
+
+	/* Enable Tx packet buffer division */
+	reg = IXGBE_READ_REG(hw, IXGBE_DTXCTL);
+	reg |= IXGBE_DTXCTL_ENDBUBD;
+	IXGBE_WRITE_REG(hw, IXGBE_DTXCTL, reg);
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_config_pfc_82598 - Config priority flow control
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure Priority Flow Control for each traffic class.
+ */
+s32 ixgbe_dcb_config_pfc_82598(struct ixgbe_hw *hw,
+                               struct ixgbe_dcb_config *dcb_config)
+{
+	u32 reg, rx_pba_size;
+	u8  i;
+
+	/* Enable Transmit Priority Flow Control */
+	reg = IXGBE_READ_REG(hw, IXGBE_RMCS);
+	reg &= ~IXGBE_RMCS_TFCE_802_3X;
+	/* correct the reporting of our flow control status */
+	hw->fc.type = ixgbe_fc_none;
+	reg |= IXGBE_RMCS_TFCE_PRIORITY;
+	IXGBE_WRITE_REG(hw, IXGBE_RMCS, reg);
+
+	/* Enable Receive Priority Flow Control */
+	reg = IXGBE_READ_REG(hw, IXGBE_FCTRL);
+	reg &= ~IXGBE_FCTRL_RFCE;
+	reg |= IXGBE_FCTRL_RPFCE;
+	IXGBE_WRITE_REG(hw, IXGBE_FCTRL, reg);
+
+	/*
+	 * Configure flow control thresholds and enable priority flow control
+	 * for each traffic class.
+	 */
+	for (i = 0; i < MAX_TRAFFIC_CLASS; i++) {
+		if (dcb_config->rx_pba_cfg == pba_equal) {
+			rx_pba_size = IXGBE_RXPBSIZE_64KB;
+		} else {
+			rx_pba_size = (i < 4) ? IXGBE_RXPBSIZE_80KB
+					      : IXGBE_RXPBSIZE_48KB;
+		}
+
+		reg = ((rx_pba_size >> 5) &  0xFFF0);
+		if (dcb_config->tc_config[i].dcb_pfc == pfc_enabled_tx ||
+		    dcb_config->tc_config[i].dcb_pfc == pfc_enabled_full)
+			reg |= IXGBE_FCRTL_XONE;
+
+		IXGBE_WRITE_REG(hw, IXGBE_FCRTL(i), reg);
+
+		reg = ((rx_pba_size >> 2) & 0xFFF0);
+		if (dcb_config->tc_config[i].dcb_pfc == pfc_enabled_tx ||
+		    dcb_config->tc_config[i].dcb_pfc == pfc_enabled_full)
+			reg |= IXGBE_FCRTH_FCEN;
+
+		IXGBE_WRITE_REG(hw, IXGBE_FCRTH(i), reg);
+	}
+
+	/* Configure pause time */
+	for (i = 0; i < (MAX_TRAFFIC_CLASS >> 1); i++)
+		IXGBE_WRITE_REG(hw, IXGBE_FCTTV(i), 0x68006800);
+
+	/* Configure flow control refresh threshold value */
+	IXGBE_WRITE_REG(hw, IXGBE_FCRTV, 0x3400);
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_config_tc_stats_82598 - Configure traffic class statistics
+ * @hw: pointer to hardware structure
+ *
+ * Configure queue statistics registers, all queues belonging to same traffic
+ * class uses a single set of queue statistics counters.
+ */
+s32 ixgbe_dcb_config_tc_stats_82598(struct ixgbe_hw *hw)
+{
+	u32 reg = 0;
+	u8  i   = 0;
+	u8  j   = 0;
+
+	/* Receive Queues stats setting -  8 queues per statistics reg */
+	for (i = 0, j = 0; i < 15 && j < 8; i = i + 2, j++) {
+		reg = IXGBE_READ_REG(hw, IXGBE_RQSMR(i));
+		reg |= ((0x1010101) * j);
+		IXGBE_WRITE_REG(hw, IXGBE_RQSMR(i), reg);
+		reg = IXGBE_READ_REG(hw, IXGBE_RQSMR(i + 1));
+		reg |= ((0x1010101) * j);
+		IXGBE_WRITE_REG(hw, IXGBE_RQSMR(i + 1), reg);
+	}
+	/* Transmit Queues stats setting -  4 queues per statistics reg */
+	for (i = 0; i < 8; i++) {
+		reg = IXGBE_READ_REG(hw, IXGBE_TQSMR(i));
+		reg |= ((0x1010101) * i);
+		IXGBE_WRITE_REG(hw, IXGBE_TQSMR(i), reg);
+	}
+
+	return 0;
+}
+
+/**
+ * ixgbe_dcb_hw_config_82598 - Config and enable DCB
+ * @hw: pointer to hardware structure
+ * @dcb_config: pointer to ixgbe_dcb_config structure
+ *
+ * Configure dcb settings and enable dcb mode.
+ */
+s32 ixgbe_dcb_hw_config_82598(struct ixgbe_hw *hw,
+                              struct ixgbe_dcb_config *dcb_config)
+{
+	ixgbe_dcb_config_packet_buffers_82598(hw, dcb_config);
+	ixgbe_dcb_config_rx_arbiter_82598(hw, dcb_config);
+	ixgbe_dcb_config_tx_desc_arbiter_82598(hw, dcb_config);
+	ixgbe_dcb_config_tx_data_arbiter_82598(hw, dcb_config);
+	ixgbe_dcb_config_pfc_82598(hw, dcb_config);
+	ixgbe_dcb_config_tc_stats_82598(hw);
+
+	return 0;
+}
diff --git a/drivers/net/ixgbe/ixgbe_dcb_82598.h b/drivers/net/ixgbe/ixgbe_dcb_82598.h
new file mode 100644
index 00000000000..1e6a313719d
--- /dev/null
+++ b/drivers/net/ixgbe/ixgbe_dcb_82598.h
@@ -0,0 +1,94 @@
+/*******************************************************************************
+
+  Intel 10 Gigabit PCI Express Linux driver
+  Copyright(c) 1999 - 2007 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  Linux NICS <linux.nics@intel.com>
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#ifndef _DCB_82598_CONFIG_H_
+#define _DCB_82598_CONFIG_H_
+
+/* DCB register definitions */
+
+#define IXGBE_DPMCS_MTSOS_SHIFT 16
+#define IXGBE_DPMCS_TDPAC       0x00000001 /* 0 Round Robin, 1 DFP - Deficit Fixed Priority */
+#define IXGBE_DPMCS_TRM         0x00000010 /* Transmit Recycle Mode */
+#define IXGBE_DPMCS_ARBDIS      0x00000040 /* DCB arbiter disable */
+#define IXGBE_DPMCS_TSOEF       0x00080000 /* TSO Expand Factor: 0=x4, 1=x2 */
+
+#define IXGBE_RUPPBMR_MQA       0x80000000 /* Enable UP to queue mapping */
+
+#define IXGBE_RT2CR_MCL_SHIFT   12 /* Offset to Max Credit Limit setting */
+#define IXGBE_RT2CR_LSP         0x80000000 /* LSP enable bit */
+
+#define IXGBE_RDRXCTL_MPBEN     0x00000010 /* DMA config for multiple packet buffers enable */
+#define IXGBE_RDRXCTL_MCEN      0x00000040 /* DMA config for multiple cores (RSS) enable */
+
+#define IXGBE_TDTQ2TCCR_MCL_SHIFT   12
+#define IXGBE_TDTQ2TCCR_BWG_SHIFT   9
+#define IXGBE_TDTQ2TCCR_GSP     0x40000000
+#define IXGBE_TDTQ2TCCR_LSP     0x80000000
+
+#define IXGBE_TDPT2TCCR_MCL_SHIFT   12
+#define IXGBE_TDPT2TCCR_BWG_SHIFT   9
+#define IXGBE_TDPT2TCCR_GSP     0x40000000
+#define IXGBE_TDPT2TCCR_LSP     0x80000000
+
+#define IXGBE_PDPMCS_TPPAC      0x00000020 /* 0 Round Robin, 1 for DFP - Deficit Fixed Priority */
+#define IXGBE_PDPMCS_ARBDIS     0x00000040 /* Arbiter disable */
+#define IXGBE_PDPMCS_TRM        0x00000100 /* Transmit Recycle Mode enable */
+
+#define IXGBE_DTXCTL_ENDBUBD    0x00000004 /* Enable DBU buffer division */
+
+#define IXGBE_TXPBSIZE_40KB     0x0000A000 /* 40KB Packet Buffer */
+#define IXGBE_RXPBSIZE_48KB     0x0000C000 /* 48KB Packet Buffer */
+#define IXGBE_RXPBSIZE_64KB     0x00010000 /* 64KB Packet Buffer */
+#define IXGBE_RXPBSIZE_80KB     0x00014000 /* 80KB Packet Buffer */
+
+#define IXGBE_RDRXCTL_RDMTS_1_2 0x00000000
+
+/* DCB hardware-specific driver APIs */
+
+/* DCB PFC functions */
+s32 ixgbe_dcb_config_pfc_82598(struct ixgbe_hw *, struct ixgbe_dcb_config *);
+s32 ixgbe_dcb_get_pfc_stats_82598(struct ixgbe_hw *, struct ixgbe_hw_stats *,
+                                  u8);
+
+/* DCB traffic class stats */
+s32 ixgbe_dcb_config_tc_stats_82598(struct ixgbe_hw *);
+s32 ixgbe_dcb_get_tc_stats_82598(struct ixgbe_hw *, struct ixgbe_hw_stats *,
+                                 u8);
+
+/* DCB config arbiters */
+s32 ixgbe_dcb_config_tx_desc_arbiter_82598(struct ixgbe_hw *,
+                                           struct ixgbe_dcb_config *);
+s32 ixgbe_dcb_config_tx_data_arbiter_82598(struct ixgbe_hw *,
+                                           struct ixgbe_dcb_config *);
+s32 ixgbe_dcb_config_rx_arbiter_82598(struct ixgbe_hw *,
+                                      struct ixgbe_dcb_config *);
+
+/* DCB hw initialization */
+s32 ixgbe_dcb_hw_config_82598(struct ixgbe_hw *, struct ixgbe_dcb_config *);
+
+#endif /* _DCB_82598_CONFIG_H */
diff --git a/drivers/net/ixgbe/ixgbe_dcb_nl.c b/drivers/net/ixgbe/ixgbe_dcb_nl.c
new file mode 100644
index 00000000000..50bff2af6b0
--- /dev/null
+++ b/drivers/net/ixgbe/ixgbe_dcb_nl.c
@@ -0,0 +1,356 @@
+/*******************************************************************************
+
+  Intel 10 Gigabit PCI Express Linux driver
+  Copyright(c) 1999 - 2008 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  Linux NICS <linux.nics@intel.com>
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#include "ixgbe.h"
+#include <linux/dcbnl.h>
+
+/* Callbacks for DCB netlink in the kernel */
+#define BIT_DCB_MODE	0x01
+#define BIT_PFC		0x02
+#define BIT_PG_RX	0x04
+#define BIT_PG_TX	0x08
+
+int ixgbe_copy_dcb_cfg(struct ixgbe_dcb_config *src_dcb_cfg,
+                       struct ixgbe_dcb_config *dst_dcb_cfg, int tc_max)
+{
+	struct tc_configuration *src_tc_cfg = NULL;
+	struct tc_configuration *dst_tc_cfg = NULL;
+	int i;
+
+	if (!src_dcb_cfg || !dst_dcb_cfg)
+		return -EINVAL;
+
+	for (i = DCB_PG_ATTR_TC_0; i < tc_max + DCB_PG_ATTR_TC_0; i++) {
+		src_tc_cfg = &src_dcb_cfg->tc_config[i - DCB_PG_ATTR_TC_0];
+		dst_tc_cfg = &dst_dcb_cfg->tc_config[i - DCB_PG_ATTR_TC_0];
+
+		dst_tc_cfg->path[DCB_TX_CONFIG].prio_type =
+				src_tc_cfg->path[DCB_TX_CONFIG].prio_type;
+
+		dst_tc_cfg->path[DCB_TX_CONFIG].bwg_id =
+				src_tc_cfg->path[DCB_TX_CONFIG].bwg_id;
+
+		dst_tc_cfg->path[DCB_TX_CONFIG].bwg_percent =
+				src_tc_cfg->path[DCB_TX_CONFIG].bwg_percent;
+
+		dst_tc_cfg->path[DCB_TX_CONFIG].up_to_tc_bitmap =
+				src_tc_cfg->path[DCB_TX_CONFIG].up_to_tc_bitmap;
+
+		dst_tc_cfg->path[DCB_RX_CONFIG].prio_type =
+				src_tc_cfg->path[DCB_RX_CONFIG].prio_type;
+
+		dst_tc_cfg->path[DCB_RX_CONFIG].bwg_id =
+				src_tc_cfg->path[DCB_RX_CONFIG].bwg_id;
+
+		dst_tc_cfg->path[DCB_RX_CONFIG].bwg_percent =
+				src_tc_cfg->path[DCB_RX_CONFIG].bwg_percent;
+
+		dst_tc_cfg->path[DCB_RX_CONFIG].up_to_tc_bitmap =
+				src_tc_cfg->path[DCB_RX_CONFIG].up_to_tc_bitmap;
+	}
+
+	for (i = DCB_PG_ATTR_BW_ID_0; i < DCB_PG_ATTR_BW_ID_MAX; i++) {
+		dst_dcb_cfg->bw_percentage[DCB_TX_CONFIG]
+			[i-DCB_PG_ATTR_BW_ID_0] = src_dcb_cfg->bw_percentage
+				[DCB_TX_CONFIG][i-DCB_PG_ATTR_BW_ID_0];
+		dst_dcb_cfg->bw_percentage[DCB_RX_CONFIG]
+			[i-DCB_PG_ATTR_BW_ID_0] = src_dcb_cfg->bw_percentage
+				[DCB_RX_CONFIG][i-DCB_PG_ATTR_BW_ID_0];
+	}
+
+	for (i = DCB_PFC_UP_ATTR_0; i < DCB_PFC_UP_ATTR_MAX; i++) {
+		dst_dcb_cfg->tc_config[i - DCB_PFC_UP_ATTR_0].dcb_pfc =
+			src_dcb_cfg->tc_config[i - DCB_PFC_UP_ATTR_0].dcb_pfc;
+	}
+
+	return 0;
+}
+
+static u8 ixgbe_dcbnl_get_state(struct net_device *netdev)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	DPRINTK(DRV, INFO, "Get DCB Admin Mode.\n");
+
+	return !!(adapter->flags & IXGBE_FLAG_DCB_ENABLED);
+}
+
+static u16 ixgbe_dcb_select_queue(struct net_device *dev, struct sk_buff *skb)
+{
+	/* All traffic should default to class 0 */
+	return 0;
+}
+
+static void ixgbe_dcbnl_set_state(struct net_device *netdev, u8 state)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	DPRINTK(DRV, INFO, "Set DCB Admin Mode.\n");
+
+	if (state > 0) {
+		/* Turn on DCB */
+		if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+			return;
+		} else {
+			if (netif_running(netdev))
+				netdev->stop(netdev);
+			ixgbe_reset_interrupt_capability(adapter);
+			ixgbe_napi_del_all(adapter);
+			kfree(adapter->tx_ring);
+			kfree(adapter->rx_ring);
+			adapter->tx_ring = NULL;
+			adapter->rx_ring = NULL;
+			netdev->select_queue = &ixgbe_dcb_select_queue;
+
+			adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
+			adapter->flags |= IXGBE_FLAG_DCB_ENABLED;
+			ixgbe_init_interrupt_scheme(adapter);
+			ixgbe_napi_add_all(adapter);
+			if (netif_running(netdev))
+				netdev->open(netdev);
+		}
+	} else {
+		/* Turn off DCB */
+		if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+			if (netif_running(netdev))
+				netdev->stop(netdev);
+			ixgbe_reset_interrupt_capability(adapter);
+			ixgbe_napi_del_all(adapter);
+			kfree(adapter->tx_ring);
+			kfree(adapter->rx_ring);
+			adapter->tx_ring = NULL;
+			adapter->rx_ring = NULL;
+			netdev->select_queue = NULL;
+
+			adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED;
+			adapter->flags |= IXGBE_FLAG_RSS_ENABLED;
+			ixgbe_init_interrupt_scheme(adapter);
+			ixgbe_napi_add_all(adapter);
+			if (netif_running(netdev))
+				netdev->open(netdev);
+		} else {
+			return;
+		}
+	}
+}
+
+static void ixgbe_dcbnl_get_perm_hw_addr(struct net_device *netdev,
+					 u8 *perm_addr)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+	int i;
+
+	for (i = 0; i < netdev->addr_len; i++)
+		perm_addr[i] = adapter->hw.mac.perm_addr[i];
+}
+
+static void ixgbe_dcbnl_set_pg_tc_cfg_tx(struct net_device *netdev, int tc,
+                                         u8 prio, u8 bwg_id, u8 bw_pct,
+                                         u8 up_map)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	if (prio != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[0].prio_type = prio;
+	if (bwg_id != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[0].bwg_id = bwg_id;
+	if (bw_pct != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[0].bwg_percent =
+			bw_pct;
+	if (up_map != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[0].up_to_tc_bitmap =
+			up_map;
+
+	if ((adapter->temp_dcb_cfg.tc_config[tc].path[0].prio_type !=
+	     adapter->dcb_cfg.tc_config[tc].path[0].prio_type) ||
+	    (adapter->temp_dcb_cfg.tc_config[tc].path[0].bwg_id !=
+	     adapter->dcb_cfg.tc_config[tc].path[0].bwg_id) ||
+	    (adapter->temp_dcb_cfg.tc_config[tc].path[0].bwg_percent !=
+	     adapter->dcb_cfg.tc_config[tc].path[0].bwg_percent) ||
+	    (adapter->temp_dcb_cfg.tc_config[tc].path[0].up_to_tc_bitmap !=
+	     adapter->dcb_cfg.tc_config[tc].path[0].up_to_tc_bitmap))
+		adapter->dcb_set_bitmap |= BIT_PG_TX;
+}
+
+static void ixgbe_dcbnl_set_pg_bwg_cfg_tx(struct net_device *netdev, int bwg_id,
+                                          u8 bw_pct)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	adapter->temp_dcb_cfg.bw_percentage[0][bwg_id] = bw_pct;
+
+	if (adapter->temp_dcb_cfg.bw_percentage[0][bwg_id] !=
+	    adapter->dcb_cfg.bw_percentage[0][bwg_id])
+		adapter->dcb_set_bitmap |= BIT_PG_RX;
+}
+
+static void ixgbe_dcbnl_set_pg_tc_cfg_rx(struct net_device *netdev, int tc,
+                                         u8 prio, u8 bwg_id, u8 bw_pct,
+                                         u8 up_map)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	if (prio != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[1].prio_type = prio;
+	if (bwg_id != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[1].bwg_id = bwg_id;
+	if (bw_pct != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[1].bwg_percent =
+			bw_pct;
+	if (up_map != DCB_ATTR_VALUE_UNDEFINED)
+		adapter->temp_dcb_cfg.tc_config[tc].path[1].up_to_tc_bitmap =
+			up_map;
+
+	if ((adapter->temp_dcb_cfg.tc_config[tc].path[1].prio_type !=
+	     adapter->dcb_cfg.tc_config[tc].path[1].prio_type) ||
+	    (adapter->temp_dcb_cfg.tc_config[tc].path[1].bwg_id !=
+	     adapter->dcb_cfg.tc_config[tc].path[1].bwg_id) ||
+	    (adapter->temp_dcb_cfg.tc_config[tc].path[1].bwg_percent !=
+	     adapter->dcb_cfg.tc_config[tc].path[1].bwg_percent) ||
+	    (adapter->temp_dcb_cfg.tc_config[tc].path[1].up_to_tc_bitmap !=
+	     adapter->dcb_cfg.tc_config[tc].path[1].up_to_tc_bitmap))
+		adapter->dcb_set_bitmap |= BIT_PG_RX;
+}
+
+static void ixgbe_dcbnl_set_pg_bwg_cfg_rx(struct net_device *netdev, int bwg_id,
+                                          u8 bw_pct)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	adapter->temp_dcb_cfg.bw_percentage[1][bwg_id] = bw_pct;
+
+	if (adapter->temp_dcb_cfg.bw_percentage[1][bwg_id] !=
+	    adapter->dcb_cfg.bw_percentage[1][bwg_id])
+		adapter->dcb_set_bitmap |= BIT_PG_RX;
+}
+
+static void ixgbe_dcbnl_get_pg_tc_cfg_tx(struct net_device *netdev, int tc,
+                                         u8 *prio, u8 *bwg_id, u8 *bw_pct,
+                                         u8 *up_map)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	*prio = adapter->dcb_cfg.tc_config[tc].path[0].prio_type;
+	*bwg_id = adapter->dcb_cfg.tc_config[tc].path[0].bwg_id;
+	*bw_pct = adapter->dcb_cfg.tc_config[tc].path[0].bwg_percent;
+	*up_map = adapter->dcb_cfg.tc_config[tc].path[0].up_to_tc_bitmap;
+}
+
+static void ixgbe_dcbnl_get_pg_bwg_cfg_tx(struct net_device *netdev, int bwg_id,
+                                          u8 *bw_pct)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	*bw_pct = adapter->dcb_cfg.bw_percentage[0][bwg_id];
+}
+
+static void ixgbe_dcbnl_get_pg_tc_cfg_rx(struct net_device *netdev, int tc,
+                                         u8 *prio, u8 *bwg_id, u8 *bw_pct,
+                                         u8 *up_map)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	*prio = adapter->dcb_cfg.tc_config[tc].path[1].prio_type;
+	*bwg_id = adapter->dcb_cfg.tc_config[tc].path[1].bwg_id;
+	*bw_pct = adapter->dcb_cfg.tc_config[tc].path[1].bwg_percent;
+	*up_map = adapter->dcb_cfg.tc_config[tc].path[1].up_to_tc_bitmap;
+}
+
+static void ixgbe_dcbnl_get_pg_bwg_cfg_rx(struct net_device *netdev, int bwg_id,
+                                          u8 *bw_pct)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	*bw_pct = adapter->dcb_cfg.bw_percentage[1][bwg_id];
+}
+
+static void ixgbe_dcbnl_set_pfc_cfg(struct net_device *netdev, int priority,
+                                    u8 setting)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	adapter->temp_dcb_cfg.tc_config[priority].dcb_pfc = setting;
+	if (adapter->temp_dcb_cfg.tc_config[priority].dcb_pfc !=
+	    adapter->dcb_cfg.tc_config[priority].dcb_pfc)
+		adapter->dcb_set_bitmap |= BIT_PFC;
+}
+
+static void ixgbe_dcbnl_get_pfc_cfg(struct net_device *netdev, int priority,
+                                    u8 *setting)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+
+	*setting = adapter->dcb_cfg.tc_config[priority].dcb_pfc;
+}
+
+static u8 ixgbe_dcbnl_set_all(struct net_device *netdev)
+{
+	struct ixgbe_adapter *adapter = netdev_priv(netdev);
+	int ret;
+
+	if (!adapter->dcb_set_bitmap)
+		return 1;
+
+	while (test_and_set_bit(__IXGBE_RESETTING, &adapter->state))
+		msleep(1);
+
+	if (netif_running(netdev))
+		ixgbe_down(adapter);
+
+	ret = ixgbe_copy_dcb_cfg(&adapter->temp_dcb_cfg, &adapter->dcb_cfg,
+				 adapter->ring_feature[RING_F_DCB].indices);
+	if (ret) {
+		clear_bit(__IXGBE_RESETTING, &adapter->state);
+		return ret;
+	}
+
+	if (netif_running(netdev))
+		ixgbe_up(adapter);
+
+	adapter->dcb_set_bitmap = 0x00;
+	clear_bit(__IXGBE_RESETTING, &adapter->state);
+	return ret;
+}
+
+struct dcbnl_rtnl_ops dcbnl_ops = {
+	.getstate	= ixgbe_dcbnl_get_state,
+	.setstate	= ixgbe_dcbnl_set_state,
+	.getpermhwaddr	= ixgbe_dcbnl_get_perm_hw_addr,
+	.setpgtccfgtx	= ixgbe_dcbnl_set_pg_tc_cfg_tx,
+	.setpgbwgcfgtx	= ixgbe_dcbnl_set_pg_bwg_cfg_tx,
+	.setpgtccfgrx	= ixgbe_dcbnl_set_pg_tc_cfg_rx,
+	.setpgbwgcfgrx	= ixgbe_dcbnl_set_pg_bwg_cfg_rx,
+	.getpgtccfgtx	= ixgbe_dcbnl_get_pg_tc_cfg_tx,
+	.getpgbwgcfgtx	= ixgbe_dcbnl_get_pg_bwg_cfg_tx,
+	.getpgtccfgrx	= ixgbe_dcbnl_get_pg_tc_cfg_rx,
+	.getpgbwgcfgrx	= ixgbe_dcbnl_get_pg_bwg_cfg_rx,
+	.setpfccfg	= ixgbe_dcbnl_set_pfc_cfg,
+	.getpfccfg	= ixgbe_dcbnl_get_pfc_cfg,
+	.setall		= ixgbe_dcbnl_set_all
+};
+
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index a610016a017..aaa4404e7c5 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -97,9 +97,18 @@ static struct ixgbe_stats ixgbe_gstrings_stats[] = {
 	((((struct ixgbe_adapter *)netdev_priv(netdev))->num_tx_queues + \
 	((struct ixgbe_adapter *)netdev_priv(netdev))->num_rx_queues) * \
 	(sizeof(struct ixgbe_queue_stats) / sizeof(u64)))
-#define IXGBE_STATS_LEN (IXGBE_GLOBAL_STATS_LEN + IXGBE_QUEUE_STATS_LEN)
 #define IXGBE_GLOBAL_STATS_LEN ARRAY_SIZE(ixgbe_gstrings_stats)
-#define IXGBE_STATS_LEN (IXGBE_GLOBAL_STATS_LEN + IXGBE_QUEUE_STATS_LEN)
+#define IXGBE_PB_STATS_LEN ( \
+                 (((struct ixgbe_adapter *)netdev->priv)->flags & \
+                 IXGBE_FLAG_DCB_ENABLED) ? \
+                 (sizeof(((struct ixgbe_adapter *)0)->stats.pxonrxc) + \
+                  sizeof(((struct ixgbe_adapter *)0)->stats.pxontxc) + \
+                  sizeof(((struct ixgbe_adapter *)0)->stats.pxoffrxc) + \
+                  sizeof(((struct ixgbe_adapter *)0)->stats.pxofftxc)) \
+                  / sizeof(u64) : 0)
+#define IXGBE_STATS_LEN (IXGBE_GLOBAL_STATS_LEN + \
+                         IXGBE_PB_STATS_LEN + \
+                         IXGBE_QUEUE_STATS_LEN)
 
 static int ixgbe_get_settings(struct net_device *netdev,
                               struct ethtool_cmd *ecmd)
@@ -831,6 +840,16 @@ static void ixgbe_get_ethtool_stats(struct net_device *netdev,
 			data[i + k] = queue_stat[k];
 		i += k;
 	}
+	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+		for (j = 0; j < MAX_TX_PACKET_BUFFERS; j++) {
+			data[i++] = adapter->stats.pxontxc[j];
+			data[i++] = adapter->stats.pxofftxc[j];
+		}
+		for (j = 0; j < MAX_RX_PACKET_BUFFERS; j++) {
+			data[i++] = adapter->stats.pxonrxc[j];
+			data[i++] = adapter->stats.pxoffrxc[j];
+		}
+	}
 }
 
 static void ixgbe_get_strings(struct net_device *netdev, u32 stringset,
@@ -859,6 +878,13 @@ static void ixgbe_get_strings(struct net_device *netdev, u32 stringset,
 			sprintf(p, "rx_queue_%u_bytes", i);
 			p += ETH_GSTRING_LEN;
 		}
+		if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+			for (i = 0; i < MAX_TX_PACKET_BUFFERS; i++) {
+				sprintf(p, "tx_pb_%u_pxon", i);
+			}
+			for (i = 0; i < MAX_RX_PACKET_BUFFERS; i++) {
+			}
+		}
 		/* BUG_ON(p - data != IXGBE_STATS_LEN * ETH_GSTRING_LEN); */
 		break;
 	}
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 40108523377..91dde9cdab6 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -404,7 +404,7 @@ static void ixgbe_receive_skb(struct ixgbe_adapter *adapter,
 
 	if (adapter->netdev->features & NETIF_F_LRO &&
 	    skb->ip_summed == CHECKSUM_UNNECESSARY) {
-		if (adapter->vlgrp && is_vlan)
+		if (adapter->vlgrp && is_vlan && (tag != 0))
 			lro_vlan_hwaccel_receive_skb(&ring->lro_mgr, skb,
 			                             adapter->vlgrp, tag,
 			                             rx_desc);
@@ -413,12 +413,12 @@ static void ixgbe_receive_skb(struct ixgbe_adapter *adapter,
 		ring->lro_used = true;
 	} else {
 		if (!(adapter->flags & IXGBE_FLAG_IN_NETPOLL)) {
-			if (adapter->vlgrp && is_vlan)
+			if (adapter->vlgrp && is_vlan && (tag != 0))
 				vlan_hwaccel_receive_skb(skb, adapter->vlgrp, tag);
 			else
 				netif_receive_skb(skb);
 		} else {
-			if (adapter->vlgrp && is_vlan)
+			if (adapter->vlgrp && is_vlan && (tag != 0))
 				vlan_hwaccel_rx(skb, adapter->vlgrp, tag);
 			else
 				netif_rx(skb);
@@ -1670,10 +1670,12 @@ static void ixgbe_configure_rx(struct ixgbe_adapter *adapter)
 	 * effects of setting this bit are only that SRRCTL must be
 	 * fully programmed [0..15]
 	 */
-	rdrxctl = IXGBE_READ_REG(hw, IXGBE_RDRXCTL);
-	rdrxctl |= IXGBE_RDRXCTL_MVMEN;
-	IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, rdrxctl);
-
+	if (adapter->flags &
+	    (IXGBE_FLAG_RSS_ENABLED | IXGBE_FLAG_VMDQ_ENABLED)) {
+		rdrxctl = IXGBE_READ_REG(hw, IXGBE_RDRXCTL);
+		rdrxctl |= IXGBE_RDRXCTL_MVMEN;
+		IXGBE_WRITE_REG(hw, IXGBE_RDRXCTL, rdrxctl);
+	}
 
 	if (adapter->flags & IXGBE_FLAG_RSS_ENABLED) {
 		/* Fill out redirection table */
@@ -1732,6 +1734,16 @@ static void ixgbe_vlan_rx_register(struct net_device *netdev,
 		ixgbe_irq_disable(adapter);
 	adapter->vlgrp = grp;
 
+	/*
+	 * For a DCB driver, always enable VLAN tag stripping so we can
+	 * still receive traffic from a DCB-enabled host even if we're
+	 * not in DCB mode.
+	 */
+	ctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_VLNCTRL);
+	ctrl |= IXGBE_VLNCTRL_VME;
+	ctrl &= ~IXGBE_VLNCTRL_CFIEN;
+	IXGBE_WRITE_REG(&adapter->hw, IXGBE_VLNCTRL, ctrl);
+
 	if (grp) {
 		/* enable VLAN tag insert/strip */
 		ctrl = IXGBE_READ_REG(&adapter->hw, IXGBE_VLNCTRL);
@@ -1896,6 +1908,44 @@ static void ixgbe_napi_disable_all(struct ixgbe_adapter *adapter)
 	}
 }
 
+#ifdef CONFIG_IXGBE_DCBNL
+/*
+ * ixgbe_configure_dcb - Configure DCB hardware
+ * @adapter: ixgbe adapter struct
+ *
+ * This is called by the driver on open to configure the DCB hardware.
+ * This is also called by the gennetlink interface when reconfiguring
+ * the DCB state.
+ */
+static void ixgbe_configure_dcb(struct ixgbe_adapter *adapter)
+{
+	struct ixgbe_hw *hw = &adapter->hw;
+	u32 txdctl, vlnctrl;
+	int i, j;
+
+	ixgbe_dcb_check_config(&adapter->dcb_cfg);
+	ixgbe_dcb_calculate_tc_credits(&adapter->dcb_cfg, DCB_TX_CONFIG);
+	ixgbe_dcb_calculate_tc_credits(&adapter->dcb_cfg, DCB_RX_CONFIG);
+
+	/* reconfigure the hardware */
+	ixgbe_dcb_hw_config(&adapter->hw, &adapter->dcb_cfg);
+
+	for (i = 0; i < adapter->num_tx_queues; i++) {
+		j = adapter->tx_ring[i].reg_idx;
+		txdctl = IXGBE_READ_REG(hw, IXGBE_TXDCTL(j));
+		/* PThresh workaround for Tx hang with DFP enabled. */
+		txdctl |= 32;
+		IXGBE_WRITE_REG(hw, IXGBE_TXDCTL(j), txdctl);
+	}
+	/* Enable VLAN tag insert/strip */
+	vlnctrl = IXGBE_READ_REG(hw, IXGBE_VLNCTRL);
+	vlnctrl |= IXGBE_VLNCTRL_VME | IXGBE_VLNCTRL_VFE;
+	vlnctrl &= ~IXGBE_VLNCTRL_CFIEN;
+	IXGBE_WRITE_REG(hw, IXGBE_VLNCTRL, vlnctrl);
+	hw->mac.ops.set_vfta(&adapter->hw, 0, 0, true);
+}
+
+#endif
 static void ixgbe_configure(struct ixgbe_adapter *adapter)
 {
 	struct net_device *netdev = adapter->netdev;
@@ -1904,6 +1954,16 @@ static void ixgbe_configure(struct ixgbe_adapter *adapter)
 	ixgbe_set_rx_mode(netdev);
 
 	ixgbe_restore_vlan(adapter);
+#ifdef CONFIG_IXGBE_DCBNL
+	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+		netif_set_gso_max_size(netdev, 32768);
+		ixgbe_configure_dcb(adapter);
+	} else {
+		netif_set_gso_max_size(netdev, 65536);
+	}
+#else
+	netif_set_gso_max_size(netdev, 65536);
+#endif
 
 	ixgbe_configure_tx(adapter);
 	ixgbe_configure_rx(adapter);
@@ -1995,9 +2055,6 @@ static int ixgbe_up_complete(struct ixgbe_adapter *adapter)
 
 	ixgbe_irq_enable(adapter);
 
-	/* enable transmits */
-	netif_tx_start_all_queues(netdev);
-
 	/* bring the link up in the watchdog, this could race with our first
 	 * link up interrupt but shouldn't be a problem */
 	adapter->flags |= IXGBE_FLAG_NEED_LINK_UPDATE;
@@ -2260,6 +2317,11 @@ static void ixgbe_reset_task(struct work_struct *work)
 	struct ixgbe_adapter *adapter;
 	adapter = container_of(work, struct ixgbe_adapter, reset_task);
 
+	/* If we're already down or resetting, just bail */
+	if (test_bit(__IXGBE_DOWN, &adapter->state) ||
+	    test_bit(__IXGBE_RESETTING, &adapter->state))
+		return;
+
 	adapter->tx_timeout_count++;
 
 	ixgbe_reinit_locked(adapter);
@@ -2269,15 +2331,31 @@ static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter)
 {
 	int nrq = 1, ntq = 1;
 	int feature_mask = 0, rss_i, rss_m;
+	int dcb_i, dcb_m;
 
 	/* Number of supported queues */
 	switch (adapter->hw.mac.type) {
 	case ixgbe_mac_82598EB:
+		dcb_i = adapter->ring_feature[RING_F_DCB].indices;
+		dcb_m = 0;
 		rss_i = adapter->ring_feature[RING_F_RSS].indices;
 		rss_m = 0;
 		feature_mask |= IXGBE_FLAG_RSS_ENABLED;
+		feature_mask |= IXGBE_FLAG_DCB_ENABLED;
 
 		switch (adapter->flags & feature_mask) {
+		case (IXGBE_FLAG_RSS_ENABLED | IXGBE_FLAG_DCB_ENABLED):
+			dcb_m = 0x7 << 3;
+			rss_i = min(8, rss_i);
+			rss_m = 0x7;
+			nrq = dcb_i * rss_i;
+			ntq = min(MAX_TX_QUEUES, dcb_i * rss_i);
+			break;
+		case (IXGBE_FLAG_DCB_ENABLED):
+			dcb_m = 0x7 << 3;
+			nrq = dcb_i;
+			ntq = dcb_i;
+			break;
 		case (IXGBE_FLAG_RSS_ENABLED):
 			rss_m = 0xF;
 			nrq = rss_i;
@@ -2285,6 +2363,8 @@ static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter)
 			break;
 		case 0:
 		default:
+			dcb_i = 0;
+			dcb_m = 0;
 			rss_i = 0;
 			rss_m = 0;
 			nrq = 1;
@@ -2292,6 +2372,12 @@ static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter)
 			break;
 		}
 
+		/* Sanity check, we should never have zero queues */
+		nrq = (nrq ?:1);
+		ntq = (ntq ?:1);
+
+		adapter->ring_feature[RING_F_DCB].indices = dcb_i;
+		adapter->ring_feature[RING_F_DCB].mask = dcb_m;
 		adapter->ring_feature[RING_F_RSS].indices = rss_i;
 		adapter->ring_feature[RING_F_RSS].mask = rss_m;
 		break;
@@ -2343,6 +2429,7 @@ static void ixgbe_acquire_msix_vectors(struct ixgbe_adapter *adapter,
 		adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED;
 		kfree(adapter->msix_entries);
 		adapter->msix_entries = NULL;
+		adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED;
 		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
 		ixgbe_set_num_queues(adapter);
 	} else {
@@ -2362,15 +2449,42 @@ static void __devinit ixgbe_cache_ring_register(struct ixgbe_adapter *adapter)
 {
 	int feature_mask = 0, rss_i;
 	int i, txr_idx, rxr_idx;
+	int dcb_i;
 
 	/* Number of supported queues */
 	switch (adapter->hw.mac.type) {
 	case ixgbe_mac_82598EB:
+		dcb_i = adapter->ring_feature[RING_F_DCB].indices;
 		rss_i = adapter->ring_feature[RING_F_RSS].indices;
 		txr_idx = 0;
 		rxr_idx = 0;
+		feature_mask |= IXGBE_FLAG_DCB_ENABLED;
 		feature_mask |= IXGBE_FLAG_RSS_ENABLED;
 		switch (adapter->flags & feature_mask) {
+		case (IXGBE_FLAG_RSS_ENABLED | IXGBE_FLAG_DCB_ENABLED):
+			for (i = 0; i < dcb_i; i++) {
+				int j;
+				/* Rx first */
+				for (j = 0; j < adapter->num_rx_queues; j++) {
+					adapter->rx_ring[rxr_idx].reg_idx =
+						i << 3 | j;
+					rxr_idx++;
+				}
+				/* Tx now */
+				for (j = 0; j < adapter->num_tx_queues; j++) {
+					adapter->tx_ring[txr_idx].reg_idx =
+						i << 2 | (j >> 1);
+					if (j & 1)
+						txr_idx++;
+				}
+			}
+		case (IXGBE_FLAG_DCB_ENABLED):
+			/* the number of queues is assumed to be symmetric */
+			for (i = 0; i < dcb_i; i++) {
+				adapter->rx_ring[i].reg_idx = i << 3;
+				adapter->tx_ring[i].reg_idx = i << 2;
+			}
+			break;
 		case (IXGBE_FLAG_RSS_ENABLED):
 			for (i = 0; i < adapter->num_rx_queues; i++)
 				adapter->rx_ring[i].reg_idx = i;
@@ -2395,7 +2509,7 @@ static void __devinit ixgbe_cache_ring_register(struct ixgbe_adapter *adapter)
  * number of queues at compile-time.  The polling_netdev array is
  * intended for Multiqueue, but should work fine with a single queue.
  **/
-static int __devinit ixgbe_alloc_queues(struct ixgbe_adapter *adapter)
+static int ixgbe_alloc_queues(struct ixgbe_adapter *adapter)
 {
 	int i;
 
@@ -2465,6 +2579,7 @@ static int __devinit ixgbe_set_interrupt_capability(struct ixgbe_adapter
 	adapter->msix_entries = kcalloc(v_budget,
 	                                sizeof(struct msix_entry), GFP_KERNEL);
 	if (!adapter->msix_entries) {
+		adapter->flags &= ~IXGBE_FLAG_DCB_ENABLED;
 		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
 		ixgbe_set_num_queues(adapter);
 		kfree(adapter->tx_ring);
@@ -2505,7 +2620,7 @@ out:
 	return err;
 }
 
-static void ixgbe_reset_interrupt_capability(struct ixgbe_adapter *adapter)
+void ixgbe_reset_interrupt_capability(struct ixgbe_adapter *adapter)
 {
 	if (adapter->flags & IXGBE_FLAG_MSIX_ENABLED) {
 		adapter->flags &= ~IXGBE_FLAG_MSIX_ENABLED;
@@ -2529,7 +2644,7 @@ static void ixgbe_reset_interrupt_capability(struct ixgbe_adapter *adapter)
  * - Hardware queue count (num_*_queues)
  *   - defined by miscellaneous hardware support/features (RSS, etc.)
  **/
-static int __devinit ixgbe_init_interrupt_scheme(struct ixgbe_adapter *adapter)
+int ixgbe_init_interrupt_scheme(struct ixgbe_adapter *adapter)
 {
 	int err;
 
@@ -2577,6 +2692,10 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
 	struct ixgbe_hw *hw = &adapter->hw;
 	struct pci_dev *pdev = adapter->pdev;
 	unsigned int rss;
+#ifdef CONFIG_IXGBE_DCBNL
+	int j;
+	struct tc_configuration *tc;
+#endif
 
 	/* PCI config space info */
 
@@ -2590,6 +2709,27 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
 	rss = min(IXGBE_MAX_RSS_INDICES, (int)num_online_cpus());
 	adapter->ring_feature[RING_F_RSS].indices = rss;
 	adapter->flags |= IXGBE_FLAG_RSS_ENABLED;
+	adapter->ring_feature[RING_F_DCB].indices = IXGBE_MAX_DCB_INDICES;
+
+#ifdef CONFIG_IXGBE_DCBNL
+	/* Configure DCB traffic classes */
+	for (j = 0; j < MAX_TRAFFIC_CLASS; j++) {
+		tc = &adapter->dcb_cfg.tc_config[j];
+		tc->path[DCB_TX_CONFIG].bwg_id = 0;
+		tc->path[DCB_TX_CONFIG].bwg_percent = 12 + (j & 1);
+		tc->path[DCB_RX_CONFIG].bwg_id = 0;
+		tc->path[DCB_RX_CONFIG].bwg_percent = 12 + (j & 1);
+		tc->dcb_pfc = pfc_disabled;
+	}
+	adapter->dcb_cfg.bw_percentage[DCB_TX_CONFIG][0] = 100;
+	adapter->dcb_cfg.bw_percentage[DCB_RX_CONFIG][0] = 100;
+	adapter->dcb_cfg.rx_pba_cfg = pba_equal;
+	adapter->dcb_cfg.round_robin_enable = false;
+	adapter->dcb_set_bitmap = 0x00;
+	ixgbe_copy_dcb_cfg(&adapter->dcb_cfg, &adapter->temp_dcb_cfg,
+	                   adapter->ring_feature[RING_F_DCB].indices);
+
+#endif
 	if (hw->mac.ops.get_media_type &&
 	    (hw->mac.ops.get_media_type(hw) == ixgbe_media_type_copper))
 		adapter->flags |= IXGBE_FLAG_FAN_FAIL_CAPABLE;
@@ -2967,7 +3107,7 @@ static int ixgbe_close(struct net_device *netdev)
  * @adapter: private struct
  * helper function to napi_add each possible q_vector->napi
  */
-static void ixgbe_napi_add_all(struct ixgbe_adapter *adapter)
+void ixgbe_napi_add_all(struct ixgbe_adapter *adapter)
 {
 	int q_idx, q_vectors;
 	int (*poll)(struct napi_struct *, int);
@@ -2988,7 +3128,7 @@ static void ixgbe_napi_add_all(struct ixgbe_adapter *adapter)
 	}
 }
 
-static void ixgbe_napi_del_all(struct ixgbe_adapter *adapter)
+void ixgbe_napi_del_all(struct ixgbe_adapter *adapter)
 {
 	int q_idx;
 	int q_vectors = adapter->num_msix_vectors - NON_Q_VECTORS;
@@ -3109,6 +3249,18 @@ void ixgbe_update_stats(struct ixgbe_adapter *adapter)
 		adapter->stats.mpc[i] += mpc;
 		total_mpc += adapter->stats.mpc[i];
 		adapter->stats.rnbc[i] += IXGBE_READ_REG(hw, IXGBE_RNBC(i));
+		adapter->stats.qptc[i] += IXGBE_READ_REG(hw, IXGBE_QPTC(i));
+		adapter->stats.qbtc[i] += IXGBE_READ_REG(hw, IXGBE_QBTC(i));
+		adapter->stats.qprc[i] += IXGBE_READ_REG(hw, IXGBE_QPRC(i));
+		adapter->stats.qbrc[i] += IXGBE_READ_REG(hw, IXGBE_QBRC(i));
+		adapter->stats.pxonrxc[i] += IXGBE_READ_REG(hw,
+		                                            IXGBE_PXONRXC(i));
+		adapter->stats.pxontxc[i] += IXGBE_READ_REG(hw,
+		                                            IXGBE_PXONTXC(i));
+		adapter->stats.pxoffrxc[i] += IXGBE_READ_REG(hw,
+		                                            IXGBE_PXOFFRXC(i));
+		adapter->stats.pxofftxc[i] += IXGBE_READ_REG(hw,
+		                                            IXGBE_PXOFFTXC(i));
 	}
 	adapter->stats.gprc += IXGBE_READ_REG(hw, IXGBE_GPRC);
 	/* work around hardware counting issue */
@@ -3248,6 +3400,7 @@ static void ixgbe_watchdog_task(struct work_struct *work)
 			         (FLOW_TX ? "TX" : "None"))));
 
 			netif_carrier_on(netdev);
+			netif_tx_wake_all_queues(netdev);
 		} else {
 			/* Force detection of hung controller */
 			adapter->detect_tx_hung = true;
@@ -3258,6 +3411,7 @@ static void ixgbe_watchdog_task(struct work_struct *work)
 		if (netif_carrier_ok(netdev)) {
 			DPRINTK(LINK, INFO, "NIC Link is Down\n");
 			netif_carrier_off(netdev);
+			netif_tx_stop_all_queues(netdev);
 		}
 	}
 
@@ -3604,6 +3758,14 @@ static int ixgbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
 
 	if (adapter->vlgrp && vlan_tx_tag_present(skb)) {
 		tx_flags |= vlan_tx_tag_get(skb);
+		if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+			tx_flags &= ~IXGBE_TX_FLAGS_VLAN_PRIO_MASK;
+			tx_flags |= (skb->queue_mapping << 13);
+		}
+		tx_flags <<= IXGBE_TX_FLAGS_VLAN_SHIFT;
+		tx_flags |= IXGBE_TX_FLAGS_VLAN;
+	} else if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
+		tx_flags |= (skb->queue_mapping << 13);
 		tx_flags <<= IXGBE_TX_FLAGS_VLAN_SHIFT;
 		tx_flags |= IXGBE_TX_FLAGS_VLAN;
 	}
@@ -3878,6 +4040,13 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	netdev->vlan_features |= NETIF_F_IP_CSUM;
 	netdev->vlan_features |= NETIF_F_SG;
 
+	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED)
+		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
+
+#ifdef CONFIG_IXGBE_DCBNL
+	netdev->dcbnl_ops = &dcbnl_ops;
+#endif
+
 	if (pci_using_dac)
 		netdev->features |= NETIF_F_HIGHDMA;
 
@@ -3946,6 +4115,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	}
 
 	netif_carrier_off(netdev);
+	netif_tx_stop_all_queues(netdev);
 
 	ixgbe_napi_add_all(adapter);
 
diff --git a/include/linux/dcbnl.h b/include/linux/dcbnl.h
new file mode 100644
index 00000000000..32d32c1ee41
--- /dev/null
+++ b/include/linux/dcbnl.h
@@ -0,0 +1,230 @@
+/*
+ * Copyright (c) 2008, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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.
+ *
+ * Author: Lucy Liu <lucy.liu@intel.com>
+ */
+
+#ifndef __LINUX_DCBNL_H__
+#define __LINUX_DCBNL_H__
+
+#define DCB_PROTO_VERSION 1
+
+struct dcbmsg {
+	unsigned char      dcb_family;
+	__u8               cmd;
+	__u16              dcb_pad;
+};
+
+/**
+ * enum dcbnl_commands - supported DCB commands
+ *
+ * @DCB_CMD_UNDEFINED: unspecified command to catch errors
+ * @DCB_CMD_GSTATE: request the state of DCB in the device
+ * @DCB_CMD_SSTATE: set the state of DCB in the device
+ * @DCB_CMD_PGTX_GCFG: request the priority group configuration for Tx
+ * @DCB_CMD_PGTX_SCFG: set the priority group configuration for Tx
+ * @DCB_CMD_PGRX_GCFG: request the priority group configuration for Rx
+ * @DCB_CMD_PGRX_SCFG: set the priority group configuration for Rx
+ * @DCB_CMD_PFC_GCFG: request the priority flow control configuration
+ * @DCB_CMD_PFC_SCFG: set the priority flow control configuration
+ * @DCB_CMD_SET_ALL: apply all changes to the underlying device
+ * @DCB_CMD_GPERM_HWADDR: get the permanent MAC address of the underlying
+ *                        device.  Only useful when using bonding.
+ */
+enum dcbnl_commands {
+	DCB_CMD_UNDEFINED,
+
+	DCB_CMD_GSTATE,
+	DCB_CMD_SSTATE,
+
+	DCB_CMD_PGTX_GCFG,
+	DCB_CMD_PGTX_SCFG,
+	DCB_CMD_PGRX_GCFG,
+	DCB_CMD_PGRX_SCFG,
+
+	DCB_CMD_PFC_GCFG,
+	DCB_CMD_PFC_SCFG,
+
+	DCB_CMD_SET_ALL,
+	DCB_CMD_GPERM_HWADDR,
+
+	__DCB_CMD_ENUM_MAX,
+	DCB_CMD_MAX = __DCB_CMD_ENUM_MAX - 1,
+};
+
+
+/**
+ * enum dcbnl_attrs - DCB top-level netlink attributes
+ *
+ * @DCB_ATTR_UNDEFINED: unspecified attribute to catch errors
+ * @DCB_ATTR_IFNAME: interface name of the underlying device (NLA_STRING)
+ * @DCB_ATTR_STATE: enable state of DCB in the device (NLA_U8)
+ * @DCB_ATTR_PFC_STATE: enable state of PFC in the device (NLA_U8)
+ * @DCB_ATTR_PFC_CFG: priority flow control configuration (NLA_NESTED)
+ * @DCB_ATTR_NUM_TC: number of traffic classes supported in the device (NLA_U8)
+ * @DCB_ATTR_PG_CFG: priority group configuration (NLA_NESTED)
+ * @DCB_ATTR_SET_ALL: bool to commit changes to hardware or not (NLA_U8)
+ * @DCB_ATTR_PERM_HWADDR: MAC address of the physical device (NLA_NESTED)
+ */
+enum dcbnl_attrs {
+	DCB_ATTR_UNDEFINED,
+
+	DCB_ATTR_IFNAME,
+	DCB_ATTR_STATE,
+	DCB_ATTR_PFC_STATE,
+	DCB_ATTR_PFC_CFG,
+	DCB_ATTR_NUM_TC,
+	DCB_ATTR_PG_CFG,
+	DCB_ATTR_SET_ALL,
+	DCB_ATTR_PERM_HWADDR,
+
+	__DCB_ATTR_ENUM_MAX,
+	DCB_ATTR_MAX = __DCB_ATTR_ENUM_MAX - 1,
+};
+
+/**
+ * enum dcbnl_pfc_attrs - DCB Priority Flow Control user priority nested attrs
+ *
+ * @DCB_PFC_UP_ATTR_UNDEFINED: unspecified attribute to catch errors
+ * @DCB_PFC_UP_ATTR_0: Priority Flow Control value for User Priority 0 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_1: Priority Flow Control value for User Priority 1 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_2: Priority Flow Control value for User Priority 2 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_3: Priority Flow Control value for User Priority 3 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_4: Priority Flow Control value for User Priority 4 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_5: Priority Flow Control value for User Priority 5 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_6: Priority Flow Control value for User Priority 6 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_7: Priority Flow Control value for User Priority 7 (NLA_U8)
+ * @DCB_PFC_UP_ATTR_MAX: highest attribute number currently defined
+ * @DCB_PFC_UP_ATTR_ALL: apply to all priority flow control attrs (NLA_FLAG)
+ *
+ */
+enum dcbnl_pfc_up_attrs {
+	DCB_PFC_UP_ATTR_UNDEFINED,
+
+	DCB_PFC_UP_ATTR_0,
+	DCB_PFC_UP_ATTR_1,
+	DCB_PFC_UP_ATTR_2,
+	DCB_PFC_UP_ATTR_3,
+	DCB_PFC_UP_ATTR_4,
+	DCB_PFC_UP_ATTR_5,
+	DCB_PFC_UP_ATTR_6,
+	DCB_PFC_UP_ATTR_7,
+	DCB_PFC_UP_ATTR_ALL,
+
+	__DCB_PFC_UP_ATTR_ENUM_MAX,
+	DCB_PFC_UP_ATTR_MAX = __DCB_PFC_UP_ATTR_ENUM_MAX - 1,
+};
+
+/**
+ * enum dcbnl_pg_attrs - DCB Priority Group attributes
+ *
+ * @DCB_PG_ATTR_UNDEFINED: unspecified attribute to catch errors
+ * @DCB_PG_ATTR_TC_0: Priority Group Traffic Class 0 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_1: Priority Group Traffic Class 1 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_2: Priority Group Traffic Class 2 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_3: Priority Group Traffic Class 3 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_4: Priority Group Traffic Class 4 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_5: Priority Group Traffic Class 5 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_6: Priority Group Traffic Class 6 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_7: Priority Group Traffic Class 7 configuration (NLA_NESTED)
+ * @DCB_PG_ATTR_TC_MAX: highest attribute number currently defined
+ * @DCB_PG_ATTR_TC_ALL: apply to all traffic classes (NLA_NESTED)
+ * @DCB_PG_ATTR_BW_ID_0: Percent of link bandwidth for Priority Group 0 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_1: Percent of link bandwidth for Priority Group 1 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_2: Percent of link bandwidth for Priority Group 2 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_3: Percent of link bandwidth for Priority Group 3 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_4: Percent of link bandwidth for Priority Group 4 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_5: Percent of link bandwidth for Priority Group 5 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_6: Percent of link bandwidth for Priority Group 6 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_7: Percent of link bandwidth for Priority Group 7 (NLA_U8)
+ * @DCB_PG_ATTR_BW_ID_MAX: highest attribute number currently defined
+ * @DCB_PG_ATTR_BW_ID_ALL: apply to all priority groups (NLA_FLAG)
+ *
+ */
+enum dcbnl_pg_attrs {
+	DCB_PG_ATTR_UNDEFINED,
+
+	DCB_PG_ATTR_TC_0,
+	DCB_PG_ATTR_TC_1,
+	DCB_PG_ATTR_TC_2,
+	DCB_PG_ATTR_TC_3,
+	DCB_PG_ATTR_TC_4,
+	DCB_PG_ATTR_TC_5,
+	DCB_PG_ATTR_TC_6,
+	DCB_PG_ATTR_TC_7,
+	DCB_PG_ATTR_TC_MAX,
+	DCB_PG_ATTR_TC_ALL,
+
+	DCB_PG_ATTR_BW_ID_0,
+	DCB_PG_ATTR_BW_ID_1,
+	DCB_PG_ATTR_BW_ID_2,
+	DCB_PG_ATTR_BW_ID_3,
+	DCB_PG_ATTR_BW_ID_4,
+	DCB_PG_ATTR_BW_ID_5,
+	DCB_PG_ATTR_BW_ID_6,
+	DCB_PG_ATTR_BW_ID_7,
+	DCB_PG_ATTR_BW_ID_MAX,
+	DCB_PG_ATTR_BW_ID_ALL,
+
+	__DCB_PG_ATTR_ENUM_MAX,
+	DCB_PG_ATTR_MAX = __DCB_PG_ATTR_ENUM_MAX - 1,
+};
+
+/**
+ * enum dcbnl_tc_attrs - DCB Traffic Class attributes
+ *
+ * @DCB_TC_ATTR_PARAM_UNDEFINED: unspecified attribute to catch errors
+ * @DCB_TC_ATTR_PARAM_PGID: (NLA_U8) Priority group the traffic class belongs to
+ *                          Valid values are:  0-7
+ * @DCB_TC_ATTR_PARAM_UP_MAPPING: (NLA_U8) Traffic class to user priority map
+ *                                Some devices may not support changing the
+ *                                user priority map of a TC.
+ * @DCB_TC_ATTR_PARAM_STRICT_PRIO: (NLA_U8) Strict priority setting
+ *                                 0 - none
+ *                                 1 - group strict
+ *                                 2 - link strict
+ * @DCB_TC_ATTR_PARAM_BW_PCT: optional - (NLA_U8) If supported by the device and
+ *                            not configured to use link strict priority,
+ *                            this is the percentage of bandwidth of the
+ *                            priority group this traffic class belongs to
+ * @DCB_TC_ATTR_PARAM_ALL: (NLA_FLAG) all traffic class parameters
+ *
+ */
+enum dcbnl_tc_attrs {
+	DCB_TC_ATTR_PARAM_UNDEFINED,
+
+	DCB_TC_ATTR_PARAM_PGID,
+	DCB_TC_ATTR_PARAM_UP_MAPPING,
+	DCB_TC_ATTR_PARAM_STRICT_PRIO,
+	DCB_TC_ATTR_PARAM_BW_PCT,
+	DCB_TC_ATTR_PARAM_ALL,
+
+	__DCB_TC_ATTR_PARAM_ENUM_MAX,
+	DCB_TC_ATTR_PARAM_MAX = __DCB_TC_ATTR_PARAM_ENUM_MAX - 1,
+};
+
+/**
+ * enum dcb_general_attr_values - general DCB attribute values
+ *
+ * @DCB_ATTR_UNDEFINED: value used to indicate an attribute is not supported
+ *
+ */
+enum dcb_general_attr_values {
+	DCB_ATTR_VALUE_UNDEFINED = 0xff
+};
+
+
+#endif /* __LINUX_DCBNL_H__ */
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index d8fb23679ee..6095af572df 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -43,6 +43,9 @@
 
 #include <net/net_namespace.h>
 #include <net/dsa.h>
+#ifdef CONFIG_DCBNL
+#include <net/dcbnl.h>
+#endif
 
 struct vlan_group;
 struct ethtool_ops;
@@ -843,6 +846,11 @@ struct net_device
 #define GSO_MAX_SIZE		65536
 	unsigned int		gso_max_size;
 
+#ifdef CONFIG_DCBNL
+	/* Data Center Bridging netlink ops */
+	struct dcbnl_rtnl_ops *dcbnl_ops;
+#endif
+
 #ifdef CONFIG_COMPAT_NET_DEV_OPS
 	struct {
 		int			(*init)(struct net_device *dev);
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index 2b3d51c6ec9..e88f7058b3a 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -107,6 +107,11 @@ enum {
 	RTM_GETADDRLABEL,
 #define RTM_GETADDRLABEL RTM_GETADDRLABEL
 
+	RTM_GETDCB = 78,
+#define RTM_GETDCB RTM_GETDCB
+	RTM_SETDCB,
+#define RTM_SETDCB RTM_SETDCB
+
 	__RTM_MAX,
 #define RTM_MAX		(((__RTM_MAX + 3) & ~3) - 1)
 };
diff --git a/include/net/dcbnl.h b/include/net/dcbnl.h
new file mode 100644
index 00000000000..0ef0c5a46d8
--- /dev/null
+++ b/include/net/dcbnl.h
@@ -0,0 +1,44 @@
+/*
+ * Copyright (c) 2008, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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.
+ *
+ * Author: Lucy Liu <lucy.liu@intel.com>
+ */
+
+#ifndef __NET_DCBNL_H__
+#define __NET_DCBNL_H__
+
+/*
+ * Ops struct for the netlink callbacks.  Used by DCB-enabled drivers through
+ * the netdevice struct.
+ */
+struct dcbnl_rtnl_ops {
+	u8   (*getstate)(struct net_device *);
+	void (*setstate)(struct net_device *, u8);
+	void (*getpermhwaddr)(struct net_device *, u8 *);
+	void (*setpgtccfgtx)(struct net_device *, int, u8, u8, u8, u8);
+	void (*setpgbwgcfgtx)(struct net_device *, int, u8);
+	void (*setpgtccfgrx)(struct net_device *, int, u8, u8, u8, u8);
+	void (*setpgbwgcfgrx)(struct net_device *, int, u8);
+	void (*getpgtccfgtx)(struct net_device *, int, u8 *, u8 *, u8 *, u8 *);
+	void (*getpgbwgcfgtx)(struct net_device *, int, u8 *);
+	void (*getpgtccfgrx)(struct net_device *, int, u8 *, u8 *, u8 *, u8 *);
+	void (*getpgbwgcfgrx)(struct net_device *, int, u8 *);
+	void (*setpfccfg)(struct net_device *, int, u8);
+	void (*getpfccfg)(struct net_device *, int, u8 *);
+	u8   (*setall)(struct net_device *);
+};
+
+#endif /* __NET_DCBNL_H__ */
diff --git a/net/Kconfig b/net/Kconfig
index 4e2e40ba8ba..c7d01c3a23c 100644
--- a/net/Kconfig
+++ b/net/Kconfig
@@ -194,6 +194,7 @@ source "net/lapb/Kconfig"
 source "net/econet/Kconfig"
 source "net/wanrouter/Kconfig"
 source "net/sched/Kconfig"
+source "net/dcb/Kconfig"
 
 menu "Network testing"
 
diff --git a/net/Makefile b/net/Makefile
index 27d1f10dc0e..83b064651f1 100644
--- a/net/Makefile
+++ b/net/Makefile
@@ -57,6 +57,9 @@ obj-$(CONFIG_NETLABEL)		+= netlabel/
 obj-$(CONFIG_IUCV)		+= iucv/
 obj-$(CONFIG_RFKILL)		+= rfkill/
 obj-$(CONFIG_NET_9P)		+= 9p/
+ifeq ($(CONFIG_DCBNL),y)
+obj-$(CONFIG_DCB)		+= dcb/
+endif
 
 ifeq ($(CONFIG_NET),y)
 obj-$(CONFIG_SYSCTL)		+= sysctl_net.o
diff --git a/net/dcb/Kconfig b/net/dcb/Kconfig
new file mode 100644
index 00000000000..bdf38802d33
--- /dev/null
+++ b/net/dcb/Kconfig
@@ -0,0 +1,12 @@
+config DCB
+        tristate "Data Center Bridging support"
+
+config DCBNL
+	bool "Data Center Bridging netlink interface support"
+	depends on DCB
+	default n
+	---help---
+	  This option turns on the netlink interface
+	  (dcbnl) for Data Center Bridging capable devices.
+
+	  If unsure, say N.
diff --git a/net/dcb/Makefile b/net/dcb/Makefile
new file mode 100644
index 00000000000..9930f4cde81
--- /dev/null
+++ b/net/dcb/Makefile
@@ -0,0 +1 @@
+obj-$(CONFIG_DCB) += dcbnl.o
diff --git a/net/dcb/dcbnl.c b/net/dcb/dcbnl.c
new file mode 100644
index 00000000000..516e8be83d7
--- /dev/null
+++ b/net/dcb/dcbnl.c
@@ -0,0 +1,704 @@
+/*
+ * Copyright (c) 2008, Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope 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.
+ *
+ * Author: Lucy Liu <lucy.liu@intel.com>
+ */
+
+#include <linux/netdevice.h>
+#include <linux/netlink.h>
+#include <net/netlink.h>
+#include <net/rtnetlink.h>
+#include <linux/dcbnl.h>
+#include <linux/rtnetlink.h>
+#include <net/sock.h>
+
+/**
+ * Data Center Bridging (DCB) is a collection of Ethernet enhancements
+ * intended to allow network traffic with differing requirements
+ * (highly reliable, no drops vs. best effort vs. low latency) to operate
+ * and co-exist on Ethernet.  Current DCB features are:
+ *
+ * Enhanced Transmission Selection (aka Priority Grouping [PG]) - provides a
+ *   framework for assigning bandwidth guarantees to traffic classes.
+ *
+ * Priority-based Flow Control (PFC) - provides a flow control mechanism which
+ *   can work independently for each 802.1p priority.
+ *
+ * Congestion Notification - provides a mechanism for end-to-end congestion
+ *   control for protocols which do not have built-in congestion management.
+ *
+ * More information about the emerging standards for these Ethernet features
+ * can be found at: http://www.ieee802.org/1/pages/dcbridges.html
+ *
+ * This file implements an rtnetlink interface to allow configuration of DCB
+ * features for capable devices.
+ */
+
+MODULE_AUTHOR("Lucy Liu, <lucy.liu@intel.com>");
+MODULE_DESCRIPTION("Data Center Bridging generic netlink interface");
+MODULE_LICENSE("GPL");
+
+/**************** DCB attribute policies *************************************/
+
+/* DCB netlink attributes policy */
+static struct nla_policy dcbnl_rtnl_policy[DCB_ATTR_MAX + 1] = {
+	[DCB_ATTR_IFNAME]    = {.type = NLA_STRING, .len = IFNAMSIZ - 1},
+	[DCB_ATTR_STATE]     = {.type = NLA_U8},
+	[DCB_ATTR_PFC_CFG]   = {.type = NLA_NESTED},
+	[DCB_ATTR_PG_CFG]    = {.type = NLA_NESTED},
+	[DCB_ATTR_SET_ALL]   = {.type = NLA_U8},
+	[DCB_ATTR_PERM_HWADDR] = {.type = NLA_FLAG},
+};
+
+/* DCB priority flow control to User Priority nested attributes */
+static struct nla_policy dcbnl_pfc_up_nest[DCB_PFC_UP_ATTR_MAX + 1] = {
+	[DCB_PFC_UP_ATTR_0]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_1]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_2]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_3]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_4]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_5]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_6]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_7]   = {.type = NLA_U8},
+	[DCB_PFC_UP_ATTR_ALL] = {.type = NLA_FLAG},
+};
+
+/* DCB priority grouping nested attributes */
+static struct nla_policy dcbnl_pg_nest[DCB_PG_ATTR_MAX + 1] = {
+	[DCB_PG_ATTR_TC_0]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_1]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_2]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_3]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_4]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_5]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_6]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_7]      = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_TC_ALL]    = {.type = NLA_NESTED},
+	[DCB_PG_ATTR_BW_ID_0]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_1]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_2]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_3]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_4]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_5]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_6]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_7]   = {.type = NLA_U8},
+	[DCB_PG_ATTR_BW_ID_ALL] = {.type = NLA_FLAG},
+};
+
+/* DCB traffic class nested attributes. */
+static struct nla_policy dcbnl_tc_param_nest[DCB_TC_ATTR_PARAM_MAX + 1] = {
+	[DCB_TC_ATTR_PARAM_PGID]            = {.type = NLA_U8},
+	[DCB_TC_ATTR_PARAM_UP_MAPPING]      = {.type = NLA_U8},
+	[DCB_TC_ATTR_PARAM_STRICT_PRIO]     = {.type = NLA_U8},
+	[DCB_TC_ATTR_PARAM_BW_PCT]          = {.type = NLA_U8},
+	[DCB_TC_ATTR_PARAM_ALL]             = {.type = NLA_FLAG},
+};
+
+
+/* standard netlink reply call */
+static int dcbnl_reply(u8 value, u8 event, u8 cmd, u8 attr, u32 pid,
+                       u32 seq, u16 flags)
+{
+	struct sk_buff *dcbnl_skb;
+	struct dcbmsg *dcb;
+	struct nlmsghdr *nlh;
+	int ret = -EINVAL;
+
+	dcbnl_skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+	if (!dcbnl_skb)
+		return ret;
+
+	nlh = NLMSG_NEW(dcbnl_skb, pid, seq, event, sizeof(*dcb), flags);
+
+	dcb = NLMSG_DATA(nlh);
+	dcb->dcb_family = AF_UNSPEC;
+	dcb->cmd = cmd;
+	dcb->dcb_pad = 0;
+
+	ret = nla_put_u8(dcbnl_skb, attr, value);
+	if (ret)
+		goto err;
+
+	/* end the message, assign the nlmsg_len. */
+	nlmsg_end(dcbnl_skb, nlh);
+	ret = rtnl_unicast(dcbnl_skb, &init_net, pid);
+	if (ret)
+		goto err;
+
+	return 0;
+nlmsg_failure:
+err:
+	kfree(dcbnl_skb);
+	return ret;
+}
+
+static int dcbnl_getstate(struct net_device *netdev, struct nlattr **tb,
+                          u32 pid, u32 seq, u16 flags)
+{
+	int ret = -EINVAL;
+
+	/* if (!tb[DCB_ATTR_STATE] || !netdev->dcbnl_ops->getstate) */
+	if (!netdev->dcbnl_ops->getstate)
+		return ret;
+
+	ret = dcbnl_reply(netdev->dcbnl_ops->getstate(netdev), RTM_GETDCB,
+	                  DCB_CMD_GSTATE, DCB_ATTR_STATE, pid, seq, flags);
+
+	return ret;
+}
+
+static int dcbnl_getpfccfg(struct net_device *netdev, struct nlattr **tb,
+                           u32 pid, u32 seq, u16 flags)
+{
+	struct sk_buff *dcbnl_skb;
+	struct nlmsghdr *nlh;
+	struct dcbmsg *dcb;
+	struct nlattr *data[DCB_PFC_UP_ATTR_MAX + 1], *nest;
+	u8 value;
+	int ret = -EINVAL;
+	int i;
+	int getall = 0;
+
+	if (!tb[DCB_ATTR_PFC_CFG] || !netdev->dcbnl_ops->getpfccfg)
+		return ret;
+
+	ret = nla_parse_nested(data, DCB_PFC_UP_ATTR_MAX,
+	                       tb[DCB_ATTR_PFC_CFG],
+	                       dcbnl_pfc_up_nest);
+	if (ret)
+		goto err_out;
+
+	dcbnl_skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+	if (!dcbnl_skb)
+		goto err_out;
+
+	nlh = NLMSG_NEW(dcbnl_skb, pid, seq, RTM_GETDCB, sizeof(*dcb), flags);
+
+	dcb = NLMSG_DATA(nlh);
+	dcb->dcb_family = AF_UNSPEC;
+	dcb->cmd = DCB_CMD_PFC_GCFG;
+
+	nest = nla_nest_start(dcbnl_skb, DCB_ATTR_PFC_CFG);
+	if (!nest)
+		goto err;
+
+	if (data[DCB_PFC_UP_ATTR_ALL])
+		getall = 1;
+
+	for (i = DCB_PFC_UP_ATTR_0; i <= DCB_PFC_UP_ATTR_7; i++) {
+		if (!getall && !data[i])
+			continue;
+
+		netdev->dcbnl_ops->getpfccfg(netdev, i - DCB_PFC_UP_ATTR_0,
+		                             &value);
+		ret = nla_put_u8(dcbnl_skb, i, value);
+
+		if (ret) {
+			nla_nest_cancel(dcbnl_skb, nest);
+			goto err;
+		}
+	}
+	nla_nest_end(dcbnl_skb, nest);
+
+	nlmsg_end(dcbnl_skb, nlh);
+
+	ret = rtnl_unicast(dcbnl_skb, &init_net, pid);
+	if (ret)
+		goto err;
+
+	return 0;
+nlmsg_failure:
+err:
+	kfree(dcbnl_skb);
+err_out:
+	return -EINVAL;
+}
+
+static int dcbnl_getperm_hwaddr(struct net_device *netdev, struct nlattr **tb,
+                                u32 pid, u32 seq, u16 flags)
+{
+	struct sk_buff *dcbnl_skb;
+	struct nlmsghdr *nlh;
+	struct dcbmsg *dcb;
+	u8 perm_addr[MAX_ADDR_LEN];
+	int ret = -EINVAL;
+
+	if (!netdev->dcbnl_ops->getpermhwaddr)
+		return ret;
+
+	dcbnl_skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+	if (!dcbnl_skb)
+		goto err_out;
+
+	nlh = NLMSG_NEW(dcbnl_skb, pid, seq, RTM_GETDCB, sizeof(*dcb), flags);
+
+	dcb = NLMSG_DATA(nlh);
+	dcb->dcb_family = AF_UNSPEC;
+	dcb->cmd = DCB_CMD_GPERM_HWADDR;
+
+	netdev->dcbnl_ops->getpermhwaddr(netdev, perm_addr);
+
+	ret = nla_put(dcbnl_skb, DCB_ATTR_PERM_HWADDR, sizeof(perm_addr),
+	              perm_addr);
+
+	nlmsg_end(dcbnl_skb, nlh);
+
+	ret = rtnl_unicast(dcbnl_skb, &init_net, pid);
+	if (ret)
+		goto err;
+
+	return 0;
+
+nlmsg_failure:
+err:
+	kfree(dcbnl_skb);
+err_out:
+	return -EINVAL;
+}
+
+static int __dcbnl_pg_getcfg(struct net_device *netdev, struct nlattr **tb,
+                             u32 pid, u32 seq, u16 flags, int dir)
+{
+	struct sk_buff *dcbnl_skb;
+	struct nlmsghdr *nlh;
+	struct dcbmsg *dcb;
+	struct nlattr *pg_nest, *param_nest, *data;
+	struct nlattr *pg_tb[DCB_PG_ATTR_MAX + 1];
+	struct nlattr *param_tb[DCB_TC_ATTR_PARAM_MAX + 1];
+	u8 prio, pgid, tc_pct, up_map;
+	int ret  = -EINVAL;
+	int getall = 0;
+	int i;
+
+	if (!tb[DCB_ATTR_PG_CFG] ||
+	    !netdev->dcbnl_ops->getpgtccfgtx ||
+	    !netdev->dcbnl_ops->getpgtccfgrx ||
+	    !netdev->dcbnl_ops->getpgbwgcfgtx ||
+	    !netdev->dcbnl_ops->getpgbwgcfgrx)
+		return ret;
+
+	ret = nla_parse_nested(pg_tb, DCB_PG_ATTR_MAX,
+	                       tb[DCB_ATTR_PG_CFG], dcbnl_pg_nest);
+
+	if (ret)
+		goto err_out;
+
+	dcbnl_skb = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+	if (!dcbnl_skb)
+		goto err_out;
+
+	nlh = NLMSG_NEW(dcbnl_skb, pid, seq, RTM_GETDCB, sizeof(*dcb), flags);
+
+	dcb = NLMSG_DATA(nlh);
+	dcb->dcb_family = AF_UNSPEC;
+	dcb->cmd = (dir) ? DCB_CMD_PGRX_GCFG : DCB_CMD_PGTX_GCFG;
+
+	pg_nest = nla_nest_start(dcbnl_skb, DCB_ATTR_PG_CFG);
+	if (!pg_nest)
+		goto err;
+
+	if (pg_tb[DCB_PG_ATTR_TC_ALL])
+		getall = 1;
+
+	for (i = DCB_PG_ATTR_TC_0; i <= DCB_PG_ATTR_TC_7; i++) {
+		if (!getall && !pg_tb[i])
+			continue;
+
+		if (pg_tb[DCB_PG_ATTR_TC_ALL])
+			data = pg_tb[DCB_PG_ATTR_TC_ALL];
+		else
+			data = pg_tb[i];
+		ret = nla_parse_nested(param_tb, DCB_TC_ATTR_PARAM_MAX,
+				       data, dcbnl_tc_param_nest);
+		if (ret)
+			goto err_pg;
+
+		param_nest = nla_nest_start(dcbnl_skb, i);
+		if (!param_nest)
+			goto err_pg;
+
+		pgid = DCB_ATTR_VALUE_UNDEFINED;
+		prio = DCB_ATTR_VALUE_UNDEFINED;
+		tc_pct = DCB_ATTR_VALUE_UNDEFINED;
+		up_map = DCB_ATTR_VALUE_UNDEFINED;
+
+		if (dir) {
+			/* Rx */
+			netdev->dcbnl_ops->getpgtccfgrx(netdev,
+						i - DCB_PG_ATTR_TC_0, &prio,
+						&pgid, &tc_pct, &up_map);
+		} else {
+			/* Tx */
+			netdev->dcbnl_ops->getpgtccfgtx(netdev,
+						i - DCB_PG_ATTR_TC_0, &prio,
+						&pgid, &tc_pct, &up_map);
+		}
+
+		if (param_tb[DCB_TC_ATTR_PARAM_PGID] ||
+		    param_tb[DCB_TC_ATTR_PARAM_ALL]) {
+			ret = nla_put_u8(dcbnl_skb,
+			                 DCB_TC_ATTR_PARAM_PGID, pgid);
+			if (ret)
+				goto err_param;
+		}
+		if (param_tb[DCB_TC_ATTR_PARAM_UP_MAPPING] ||
+		    param_tb[DCB_TC_ATTR_PARAM_ALL]) {
+			ret = nla_put_u8(dcbnl_skb,
+			                 DCB_TC_ATTR_PARAM_UP_MAPPING, up_map);
+			if (ret)
+				goto err_param;
+		}
+		if (param_tb[DCB_TC_ATTR_PARAM_STRICT_PRIO] ||
+		    param_tb[DCB_TC_ATTR_PARAM_ALL]) {
+			ret = nla_put_u8(dcbnl_skb,
+			                 DCB_TC_ATTR_PARAM_STRICT_PRIO, prio);
+			if (ret)
+				goto err_param;
+		}
+		if (param_tb[DCB_TC_ATTR_PARAM_BW_PCT] ||
+		    param_tb[DCB_TC_ATTR_PARAM_ALL]) {
+			ret = nla_put_u8(dcbnl_skb, DCB_TC_ATTR_PARAM_BW_PCT,
+			                 tc_pct);
+			if (ret)
+				goto err_param;
+		}
+		nla_nest_end(dcbnl_skb, param_nest);
+	}
+
+	if (pg_tb[DCB_PG_ATTR_BW_ID_ALL])
+		getall = 1;
+	else
+		getall = 0;
+
+	for (i = DCB_PG_ATTR_BW_ID_0; i <= DCB_PG_ATTR_BW_ID_7; i++) {
+		if (!getall && !pg_tb[i])
+			continue;
+
+		tc_pct = DCB_ATTR_VALUE_UNDEFINED;
+
+		if (dir) {
+			/* Rx */
+			netdev->dcbnl_ops->getpgbwgcfgrx(netdev,
+					i - DCB_PG_ATTR_BW_ID_0, &tc_pct);
+		} else {
+			/* Tx */
+			netdev->dcbnl_ops->getpgbwgcfgtx(netdev,
+					i - DCB_PG_ATTR_BW_ID_0, &tc_pct);
+		}
+		ret = nla_put_u8(dcbnl_skb, i, tc_pct);
+
+		if (ret)
+			goto err_pg;
+	}
+
+	nla_nest_end(dcbnl_skb, pg_nest);
+
+	nlmsg_end(dcbnl_skb, nlh);
+
+	ret = rtnl_unicast(dcbnl_skb, &init_net, pid);
+	if (ret)
+		goto err;
+
+	return 0;
+
+err_param:
+	nla_nest_cancel(dcbnl_skb, param_nest);
+err_pg:
+	nla_nest_cancel(dcbnl_skb, pg_nest);
+nlmsg_failure:
+err:
+	kfree(dcbnl_skb);
+err_out:
+	ret  = -EINVAL;
+	return ret;
+}
+
+static int dcbnl_pgtx_getcfg(struct net_device *netdev, struct nlattr **tb,
+                             u32 pid, u32 seq, u16 flags)
+{
+	return __dcbnl_pg_getcfg(netdev, tb, pid, seq, flags, 0);
+}
+
+static int dcbnl_pgrx_getcfg(struct net_device *netdev, struct nlattr **tb,
+                             u32 pid, u32 seq, u16 flags)
+{
+	return __dcbnl_pg_getcfg(netdev, tb, pid, seq, flags, 1);
+}
+
+static int dcbnl_setstate(struct net_device *netdev, struct nlattr **tb,
+                          u32 pid, u32 seq, u16 flags)
+{
+	int ret = -EINVAL;
+	u8 value;
+
+	if (!tb[DCB_ATTR_STATE] || !netdev->dcbnl_ops->setstate)
+		return ret;
+
+	value = nla_get_u8(tb[DCB_ATTR_STATE]);
+
+	netdev->dcbnl_ops->setstate(netdev, value);
+
+	ret = dcbnl_reply(0, RTM_SETDCB, DCB_CMD_SSTATE, DCB_ATTR_STATE,
+	                  pid, seq, flags);
+
+	return ret;
+}
+
+static int dcbnl_setpfccfg(struct net_device *netdev, struct nlattr **tb,
+                           u32 pid, u32 seq, u16 flags)
+{
+	struct nlattr *data[DCB_PFC_UP_ATTR_MAX + 1];
+	int i;
+	int ret = -EINVAL;
+	u8 value;
+
+	if (!tb[DCB_ATTR_PFC_CFG] || !netdev->dcbnl_ops->setpfccfg)
+		return ret;
+
+	ret = nla_parse_nested(data, DCB_PFC_UP_ATTR_MAX,
+	                       tb[DCB_ATTR_PFC_CFG],
+	                       dcbnl_pfc_up_nest);
+	if (ret)
+		goto err;
+
+	for (i = DCB_PFC_UP_ATTR_0; i <= DCB_PFC_UP_ATTR_7; i++) {
+		if (data[i] == NULL)
+			continue;
+		value = nla_get_u8(data[i]);
+		netdev->dcbnl_ops->setpfccfg(netdev,
+			data[i]->nla_type - DCB_PFC_UP_ATTR_0, value);
+	}
+
+	ret = dcbnl_reply(0, RTM_SETDCB, DCB_CMD_PFC_SCFG, DCB_ATTR_PFC_CFG,
+	                  pid, seq, flags);
+err:
+	return ret;
+}
+
+static int dcbnl_setall(struct net_device *netdev, struct nlattr **tb,
+                        u32 pid, u32 seq, u16 flags)
+{
+	int ret = -EINVAL;
+
+	if (!tb[DCB_ATTR_SET_ALL] || !netdev->dcbnl_ops->setall)
+		return ret;
+
+	ret = dcbnl_reply(netdev->dcbnl_ops->setall(netdev), RTM_SETDCB,
+	                  DCB_CMD_SET_ALL, DCB_ATTR_SET_ALL, pid, seq, flags);
+
+	return ret;
+}
+
+static int __dcbnl_pg_setcfg(struct net_device *netdev, struct nlattr **tb,
+                             u32 pid, u32 seq, u16 flags, int dir)
+{
+	struct nlattr *pg_tb[DCB_PG_ATTR_MAX + 1];
+	struct nlattr *param_tb[DCB_TC_ATTR_PARAM_MAX + 1];
+	int ret = -EINVAL;
+	int i;
+	u8 pgid;
+	u8 up_map;
+	u8 prio;
+	u8 tc_pct;
+
+	if (!tb[DCB_ATTR_PG_CFG] ||
+	    !netdev->dcbnl_ops->setpgtccfgtx ||
+	    !netdev->dcbnl_ops->setpgtccfgrx ||
+	    !netdev->dcbnl_ops->setpgbwgcfgtx ||
+	    !netdev->dcbnl_ops->setpgbwgcfgrx)
+		return ret;
+
+	ret = nla_parse_nested(pg_tb, DCB_PG_ATTR_MAX,
+	                       tb[DCB_ATTR_PG_CFG], dcbnl_pg_nest);
+	if (ret)
+		goto err;
+
+	for (i = DCB_PG_ATTR_TC_0; i <= DCB_PG_ATTR_TC_7; i++) {
+		if (!pg_tb[i])
+			continue;
+
+		ret = nla_parse_nested(param_tb, DCB_TC_ATTR_PARAM_MAX,
+		                       pg_tb[i], dcbnl_tc_param_nest);
+		if (ret)
+			goto err;
+
+		pgid = DCB_ATTR_VALUE_UNDEFINED;
+		prio = DCB_ATTR_VALUE_UNDEFINED;
+		tc_pct = DCB_ATTR_VALUE_UNDEFINED;
+		up_map = DCB_ATTR_VALUE_UNDEFINED;
+
+		if (param_tb[DCB_TC_ATTR_PARAM_STRICT_PRIO])
+			prio =
+			    nla_get_u8(param_tb[DCB_TC_ATTR_PARAM_STRICT_PRIO]);
+
+		if (param_tb[DCB_TC_ATTR_PARAM_PGID])
+			pgid = nla_get_u8(param_tb[DCB_TC_ATTR_PARAM_PGID]);
+
+		if (param_tb[DCB_TC_ATTR_PARAM_BW_PCT])
+			tc_pct = nla_get_u8(param_tb[DCB_TC_ATTR_PARAM_BW_PCT]);
+
+		if (param_tb[DCB_TC_ATTR_PARAM_UP_MAPPING])
+			up_map =
+			     nla_get_u8(param_tb[DCB_TC_ATTR_PARAM_UP_MAPPING]);
+
+		/* dir: Tx = 0, Rx = 1 */
+		if (dir) {
+			/* Rx */
+			netdev->dcbnl_ops->setpgtccfgrx(netdev,
+				i - DCB_PG_ATTR_TC_0,
+				prio, pgid, tc_pct, up_map);
+		} else {
+			/* Tx */
+			netdev->dcbnl_ops->setpgtccfgtx(netdev,
+				i - DCB_PG_ATTR_TC_0,
+				prio, pgid, tc_pct, up_map);
+		}
+	}
+
+	for (i = DCB_PG_ATTR_BW_ID_0; i <= DCB_PG_ATTR_BW_ID_7; i++) {
+		if (!pg_tb[i])
+			continue;
+
+		tc_pct = nla_get_u8(pg_tb[i]);
+
+		/* dir: Tx = 0, Rx = 1 */
+		if (dir) {
+			/* Rx */
+			netdev->dcbnl_ops->setpgbwgcfgrx(netdev,
+					 i - DCB_PG_ATTR_BW_ID_0, tc_pct);
+		} else {
+			/* Tx */
+			netdev->dcbnl_ops->setpgbwgcfgtx(netdev,
+					 i - DCB_PG_ATTR_BW_ID_0, tc_pct);
+		}
+	}
+
+	ret = dcbnl_reply(0, RTM_SETDCB,
+			  (dir ? DCB_CMD_PGRX_SCFG : DCB_CMD_PGTX_SCFG),
+			  DCB_ATTR_PG_CFG, pid, seq, flags);
+
+err:
+	return ret;
+}
+
+static int dcbnl_pgtx_setcfg(struct net_device *netdev, struct nlattr **tb,
+                             u32 pid, u32 seq, u16 flags)
+{
+	return __dcbnl_pg_setcfg(netdev, tb, pid, seq, flags, 0);
+}
+
+static int dcbnl_pgrx_setcfg(struct net_device *netdev, struct nlattr **tb,
+                             u32 pid, u32 seq, u16 flags)
+{
+	return __dcbnl_pg_setcfg(netdev, tb, pid, seq, flags, 1);
+}
+
+static int dcb_doit(struct sk_buff *skb, struct nlmsghdr *nlh, void *arg)
+{
+	struct net *net = sock_net(skb->sk);
+	struct net_device *netdev;
+	struct dcbmsg  *dcb = (struct dcbmsg *)NLMSG_DATA(nlh);
+	struct nlattr *tb[DCB_ATTR_MAX + 1];
+	u32 pid = skb ? NETLINK_CB(skb).pid : 0;
+	int ret = -EINVAL;
+
+	if (net != &init_net)
+		return -EINVAL;
+
+	ret = nlmsg_parse(nlh, sizeof(*dcb), tb, DCB_ATTR_MAX,
+			  dcbnl_rtnl_policy);
+	if (ret < 0)
+		return ret;
+
+	if (!tb[DCB_ATTR_IFNAME])
+		return -EINVAL;
+
+	netdev = dev_get_by_name(&init_net, nla_data(tb[DCB_ATTR_IFNAME]));
+	if (!netdev)
+		return -EINVAL;
+
+	if (!netdev->dcbnl_ops)
+		goto errout;
+
+	switch (dcb->cmd) {
+	case DCB_CMD_GSTATE:
+		ret = dcbnl_getstate(netdev, tb, pid, nlh->nlmsg_seq,
+		                     nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_PFC_GCFG:
+		ret = dcbnl_getpfccfg(netdev, tb, pid, nlh->nlmsg_seq,
+		                      nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_GPERM_HWADDR:
+		ret = dcbnl_getperm_hwaddr(netdev, tb, pid, nlh->nlmsg_seq,
+		                           nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_PGTX_GCFG:
+		ret = dcbnl_pgtx_getcfg(netdev, tb, pid, nlh->nlmsg_seq,
+		                        nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_PGRX_GCFG:
+		ret = dcbnl_pgrx_getcfg(netdev, tb, pid, nlh->nlmsg_seq,
+		                        nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_SSTATE:
+		ret = dcbnl_setstate(netdev, tb, pid, nlh->nlmsg_seq,
+		                     nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_PFC_SCFG:
+		ret = dcbnl_setpfccfg(netdev, tb, pid, nlh->nlmsg_seq,
+		                      nlh->nlmsg_flags);
+		goto out;
+
+	case DCB_CMD_SET_ALL:
+		ret = dcbnl_setall(netdev, tb, pid, nlh->nlmsg_seq,
+		                   nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_PGTX_SCFG:
+		ret = dcbnl_pgtx_setcfg(netdev, tb, pid, nlh->nlmsg_seq,
+		                        nlh->nlmsg_flags);
+		goto out;
+	case DCB_CMD_PGRX_SCFG:
+		ret = dcbnl_pgrx_setcfg(netdev, tb, pid, nlh->nlmsg_seq,
+		                        nlh->nlmsg_flags);
+		goto out;
+	default:
+		goto errout;
+	}
+errout:
+	ret = -EINVAL;
+out:
+	dev_put(netdev);
+	return ret;
+}
+
+static int __init dcbnl_init(void)
+{
+	rtnl_register(PF_UNSPEC, RTM_GETDCB, dcb_doit, NULL);
+	rtnl_register(PF_UNSPEC, RTM_SETDCB, dcb_doit, NULL);
+
+	return 0;
+}
+module_init(dcbnl_init);
+
+static void __exit dcbnl_exit(void)
+{
+	rtnl_unregister(PF_UNSPEC, RTM_GETDCB);
+	rtnl_unregister(PF_UNSPEC, RTM_SETDCB);
+}
+module_exit(dcbnl_exit);
+
+
-- 
cgit v1.2.3-70-g09d2


From 47fd5b8373ecc6bf5473e4139b62b06425448252 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <shemminger@vyatta.com>
Date: Tue, 25 Nov 2008 00:20:43 -0800
Subject: netdev: add HAVE_NET_DEVICE_OPS

As a concession to vendors who have to deal with one source for different
kernel versions, add a HAVE_NET_DEVICE_OPS so they don't end up hard
coding ifdef against kernel version.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 1 +
 1 file changed, 1 insertion(+)

(limited to 'include/linux/netdevice.h')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 6095af572df..76a89f8e6a1 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -545,6 +545,7 @@ struct netdev_queue {
  *
  * void (*ndo_poll_controller)(struct net_device *dev);
  */
+#define HAVE_NET_DEVICE_OPS
 struct net_device_ops {
 	int			(*ndo_init)(struct net_device *dev);
 	void			(*ndo_uninit)(struct net_device *dev);
-- 
cgit v1.2.3-70-g09d2


From 7a6b6f515f77d1c62a2f383b6dce18cb0af0cf4f Mon Sep 17 00:00:00 2001
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 25 Nov 2008 01:02:08 -0800
Subject: DCB: fix kconfig option

Since the netlink option for DCB is necessary to actually be useful,
simplified the Kconfig option.  In addition, added useful help text for the
Kconfig option.

Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/net/Kconfig            |  4 ++--
 drivers/net/ixgbe/Makefile     |  2 +-
 drivers/net/ixgbe/ixgbe.h      |  2 +-
 drivers/net/ixgbe/ixgbe_main.c | 10 +++++-----
 include/linux/netdevice.h      |  4 ++--
 net/Makefile                   |  4 ++--
 net/dcb/Kconfig                | 24 +++++++++++++++++-------
 net/dcb/dcbnl.c                |  2 +-
 8 files changed, 31 insertions(+), 21 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index efd461d7c2b..75f9f7f2fbe 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2451,10 +2451,10 @@ config IXGBE_DCA
 	  driver.  DCA is a method for warming the CPU cache before data
 	  is used, with the intent of lessening the impact of cache misses.
 
-config IXGBE_DCBNL
+config IXGBE_DCB
 	bool "Data Center Bridging (DCB) Support"
 	default n
-	depends on IXGBE && DCBNL
+	depends on IXGBE && DCB
 	---help---
 	  Say Y here if you want to use Data Center Bridging (DCB) in the
 	  driver.
diff --git a/drivers/net/ixgbe/Makefile b/drivers/net/ixgbe/Makefile
index 3228e508e62..6e7ef765bcd 100644
--- a/drivers/net/ixgbe/Makefile
+++ b/drivers/net/ixgbe/Makefile
@@ -35,4 +35,4 @@ obj-$(CONFIG_IXGBE) += ixgbe.o
 ixgbe-objs := ixgbe_main.o ixgbe_common.o ixgbe_ethtool.o \
               ixgbe_82598.o ixgbe_phy.o
 
-ixgbe-$(CONFIG_IXGBE_DCBNL) +=  ixgbe_dcb.o ixgbe_dcb_82598.o ixgbe_dcb_nl.o
+ixgbe-$(CONFIG_IXGBE_DCB) +=  ixgbe_dcb.o ixgbe_dcb_82598.o ixgbe_dcb_nl.o
diff --git a/drivers/net/ixgbe/ixgbe.h b/drivers/net/ixgbe/ixgbe.h
index 9509aee2d36..6cbf26e3db8 100644
--- a/drivers/net/ixgbe/ixgbe.h
+++ b/drivers/net/ixgbe/ixgbe.h
@@ -327,7 +327,7 @@ enum ixgbe_boards {
 };
 
 extern struct ixgbe_info ixgbe_82598_info;
-#ifdef CONFIG_IXGBE_DCBNL
+#ifdef CONFIG_IXGBE_DCB
 extern struct dcbnl_rtnl_ops dcbnl_ops;
 extern int ixgbe_copy_dcb_cfg(struct ixgbe_dcb_config *src_dcb_cfg,
                               struct ixgbe_dcb_config *dst_dcb_cfg,
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 6620397a1fb..667a6463193 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -1914,7 +1914,7 @@ static void ixgbe_napi_disable_all(struct ixgbe_adapter *adapter)
 	}
 }
 
-#ifdef CONFIG_IXGBE_DCBNL
+#ifdef CONFIG_IXGBE_DCB
 /*
  * ixgbe_configure_dcb - Configure DCB hardware
  * @adapter: ixgbe adapter struct
@@ -1960,7 +1960,7 @@ static void ixgbe_configure(struct ixgbe_adapter *adapter)
 	ixgbe_set_rx_mode(netdev);
 
 	ixgbe_restore_vlan(adapter);
-#ifdef CONFIG_IXGBE_DCBNL
+#ifdef CONFIG_IXGBE_DCB
 	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED) {
 		netif_set_gso_max_size(netdev, 32768);
 		ixgbe_configure_dcb(adapter);
@@ -2749,7 +2749,7 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
 	struct ixgbe_hw *hw = &adapter->hw;
 	struct pci_dev *pdev = adapter->pdev;
 	unsigned int rss;
-#ifdef CONFIG_IXGBE_DCBNL
+#ifdef CONFIG_IXGBE_DCB
 	int j;
 	struct tc_configuration *tc;
 #endif
@@ -2768,7 +2768,7 @@ static int __devinit ixgbe_sw_init(struct ixgbe_adapter *adapter)
 	adapter->flags |= IXGBE_FLAG_RSS_ENABLED;
 	adapter->ring_feature[RING_F_DCB].indices = IXGBE_MAX_DCB_INDICES;
 
-#ifdef CONFIG_IXGBE_DCBNL
+#ifdef CONFIG_IXGBE_DCB
 	/* Configure DCB traffic classes */
 	for (j = 0; j < MAX_TRAFFIC_CLASS; j++) {
 		tc = &adapter->dcb_cfg.tc_config[j];
@@ -4120,7 +4120,7 @@ static int __devinit ixgbe_probe(struct pci_dev *pdev,
 	if (adapter->flags & IXGBE_FLAG_DCB_ENABLED)
 		adapter->flags &= ~IXGBE_FLAG_RSS_ENABLED;
 
-#ifdef CONFIG_IXGBE_DCBNL
+#ifdef CONFIG_IXGBE_DCB
 	netdev->dcbnl_ops = &dcbnl_ops;
 #endif
 
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 76a89f8e6a1..0df0db068ac 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -43,7 +43,7 @@
 
 #include <net/net_namespace.h>
 #include <net/dsa.h>
-#ifdef CONFIG_DCBNL
+#ifdef CONFIG_DCB
 #include <net/dcbnl.h>
 #endif
 
@@ -847,7 +847,7 @@ struct net_device
 #define GSO_MAX_SIZE		65536
 	unsigned int		gso_max_size;
 
-#ifdef CONFIG_DCBNL
+#ifdef CONFIG_DCB
 	/* Data Center Bridging netlink ops */
 	struct dcbnl_rtnl_ops *dcbnl_ops;
 #endif
diff --git a/net/Makefile b/net/Makefile
index e5af3dc3a03..ba4460432b7 100644
--- a/net/Makefile
+++ b/net/Makefile
@@ -56,8 +56,8 @@ obj-$(CONFIG_NETLABEL)		+= netlabel/
 obj-$(CONFIG_IUCV)		+= iucv/
 obj-$(CONFIG_RFKILL)		+= rfkill/
 obj-$(CONFIG_NET_9P)		+= 9p/
-ifeq ($(CONFIG_DCBNL),y)
-obj-$(CONFIG_DCB)		+= dcb/
+ifneq ($(CONFIG_DCB),)
+obj-y				+= dcb/
 endif
 
 ifeq ($(CONFIG_NET),y)
diff --git a/net/dcb/Kconfig b/net/dcb/Kconfig
index bdf38802d33..4066d59c8de 100644
--- a/net/dcb/Kconfig
+++ b/net/dcb/Kconfig
@@ -1,12 +1,22 @@
 config DCB
-        tristate "Data Center Bridging support"
-
-config DCBNL
-	bool "Data Center Bridging netlink interface support"
-	depends on DCB
+	bool "Data Center Bridging support"
 	default n
 	---help---
-	  This option turns on the netlink interface
-	  (dcbnl) for Data Center Bridging capable devices.
+	  This enables support for configuring Data Center Bridging (DCB)
+	  features on DCB capable Ethernet adapters via rtnetlink.  Say 'Y'
+	  if you have a DCB capable Ethernet adapter which supports this
+	  interface and you are connected to a DCB capable switch.
+
+	  DCB is a collection of Ethernet enhancements which allow DCB capable
+	  NICs and switches to support network traffic with differing
+	  requirements (highly reliable, no drops vs. best effort vs. low
+	  latency) to co-exist on Ethernet.
+
+	  DCB features include:
+	    Enhanced Transmission Selection (aka Priority Grouping) - provides a
+	      framework for assigning bandwidth guarantees to traffic classes.
+	    Priority-based Flow Control (PFC) - a MAC control pause frame which
+	      works at the granularity of the 802.1p priority instead of the
+	      link (802.3x).
 
 	  If unsure, say N.
diff --git a/net/dcb/dcbnl.c b/net/dcb/dcbnl.c
index b2bda3f610d..79a351d323a 100644
--- a/net/dcb/dcbnl.c
+++ b/net/dcb/dcbnl.c
@@ -48,7 +48,7 @@
  */
 
 MODULE_AUTHOR("Lucy Liu, <lucy.liu@intel.com>");
-MODULE_DESCRIPTION("Data Center Bridging generic netlink interface");
+MODULE_DESCRIPTION("Data Center Bridging netlink interface");
 MODULE_LICENSE("GPL");
 
 /**************** DCB attribute policies *************************************/
-- 
cgit v1.2.3-70-g09d2


From b74ca3a896b9ab5f952bc440154758e708c48884 Mon Sep 17 00:00:00 2001
From: Wang Chen <wangchen@cn.fujitsu.com>
Date: Mon, 8 Dec 2008 01:14:16 -0800
Subject: netdevice: Kill netdev->priv

This is the last shoot of this series.
After I removing all directly reference of netdev->priv, I am killing
"priv" of "struct net_device" and fixing relative comments/docs.

Anyone will not be allowed to reference netdev->priv directly.
If you want to reference the memory of private data, use netdev_priv()
instead.
If the private data is not allocted when alloc_netdev(), use
netdev->ml_priv to point that memory after you creating that private
data.

Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 Documentation/networking/driver.txt     | 2 +-
 Documentation/networking/netdevices.txt | 2 +-
 drivers/net/3c501.h                     | 2 +-
 drivers/net/atp.c                       | 2 +-
 drivers/net/eexpress.c                  | 2 +-
 drivers/net/forcedeth.c                 | 4 ++--
 drivers/net/lance.c                     | 2 +-
 drivers/net/myri_sbus.c                 | 2 +-
 drivers/net/pci-skeleton.c              | 2 +-
 drivers/net/sun3_82586.c                | 2 +-
 drivers/net/sunbmac.c                   | 2 +-
 drivers/net/tokenring/3c359.c           | 5 +++--
 drivers/net/via-rhine.c                 | 9 +++++----
 drivers/net/wireless/strip.c            | 2 +-
 include/linux/hdlc.h                    | 2 +-
 include/linux/netdevice.h               | 1 -
 net/atm/mpc.c                           | 4 ++--
 net/core/dev.c                          | 6 ------
 18 files changed, 24 insertions(+), 29 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/Documentation/networking/driver.txt b/Documentation/networking/driver.txt
index ea72d2e66ca..03283daa64f 100644
--- a/Documentation/networking/driver.txt
+++ b/Documentation/networking/driver.txt
@@ -13,7 +13,7 @@ Transmit path guidelines:
 	static int drv_hard_start_xmit(struct sk_buff *skb,
 		   		       struct net_device *dev)
 	{
-		struct drv *dp = dev->priv;
+		struct drv *dp = netdev_priv(dev);
 
 		lock_tx(dp);
 		...
diff --git a/Documentation/networking/netdevices.txt b/Documentation/networking/netdevices.txt
index d0f71fc7f78..a2ab6a0b116 100644
--- a/Documentation/networking/netdevices.txt
+++ b/Documentation/networking/netdevices.txt
@@ -18,7 +18,7 @@ There are routines in net_init.c to handle the common cases of
 alloc_etherdev, alloc_netdev.  These reserve extra space for driver
 private data which gets freed when the network device is freed. If
 separately allocated data is attached to the network device
-(dev->priv) then it is up to the module exit handler to free that.
+(netdev_priv(dev)) then it is up to the module exit handler to free that.
 
 MTU
 ===
diff --git a/drivers/net/3c501.h b/drivers/net/3c501.h
index cfec64efff7..f40b0493337 100644
--- a/drivers/net/3c501.h
+++ b/drivers/net/3c501.h
@@ -23,7 +23,7 @@ static const struct ethtool_ops netdev_ethtool_ops;
 static int el_debug = EL_DEBUG;
 
 /*
- *	Board-specific info in dev->priv.
+ *	Board-specific info in netdev_priv(dev).
  */
 
 struct net_local
diff --git a/drivers/net/atp.c b/drivers/net/atp.c
index 7028b276dfd..1d6b74c5d6c 100644
--- a/drivers/net/atp.c
+++ b/drivers/net/atp.c
@@ -420,7 +420,7 @@ static unsigned short __init eeprom_op(long ioaddr, u32 cmd)
    registers that "should" only need to be set once at boot, so that
    there is non-reboot way to recover if something goes wrong.
 
-   This is an attachable device: if there is no dev->priv entry then it wasn't
+   This is an attachable device: if there is no private entry then it wasn't
    probed for at boot-time, and we need to probe for it again.
    */
 static int net_open(struct net_device *dev)
diff --git a/drivers/net/eexpress.c b/drivers/net/eexpress.c
index a125e41240f..9ff3f2f5e38 100644
--- a/drivers/net/eexpress.c
+++ b/drivers/net/eexpress.c
@@ -1046,7 +1046,7 @@ static void eexp_hw_tx_pio(struct net_device *dev, unsigned short *buf,
 /*
  * Sanity check the suspected EtherExpress card
  * Read hardware address, reset card, size memory and initialize buffer
- * memory pointers. These are held in dev->priv, in case someone has more
+ * memory pointers. These are held in netdev_priv(), in case someone has more
  * than one card in a machine.
  */
 
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index 12384df8cb2..1f2b24743ee 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -712,12 +712,12 @@ struct nv_skb_map {
 
 /*
  * SMP locking:
- * All hardware access under dev->priv->lock, except the performance
+ * All hardware access under netdev_priv(dev)->lock, except the performance
  * critical parts:
  * - rx is (pseudo-) lockless: it relies on the single-threading provided
  *	by the arch code for interrupts.
  * - tx setup is lockless: it relies on netif_tx_lock. Actual submission
- *	needs dev->priv->lock :-(
+ *	needs netdev_priv(dev)->lock :-(
  * - set_multicast_list: preparation lockless, relies on netif_tx_lock.
  */
 
diff --git a/drivers/net/lance.c b/drivers/net/lance.c
index e81b6113ed9..d7afb938ea6 100644
--- a/drivers/net/lance.c
+++ b/drivers/net/lance.c
@@ -519,7 +519,7 @@ static int __init lance_probe1(struct net_device *dev, int ioaddr, int irq, int
 		}
 	}
 
-	/* We can't allocate dev->priv from alloc_etherdev() because it must
+	/* We can't allocate private data from alloc_etherdev() because it must
 	   a ISA DMA-able region. */
 	chipname = chip_table[lance_version].name;
 	printk("%s: %s at %#3x, ", dev->name, chipname, ioaddr);
diff --git a/drivers/net/myri_sbus.c b/drivers/net/myri_sbus.c
index 6833f65f8ae..899ed065a14 100644
--- a/drivers/net/myri_sbus.c
+++ b/drivers/net/myri_sbus.c
@@ -1091,7 +1091,7 @@ static int __devinit myri_sbus_probe(struct of_device *op, const struct of_devic
 err_free_irq:
 	free_irq(dev->irq, dev);
 err:
-	/* This will also free the co-allocated 'dev->priv' */
+	/* This will also free the co-allocated private data*/
 	free_netdev(dev);
 	return -ENODEV;
 }
diff --git a/drivers/net/pci-skeleton.c b/drivers/net/pci-skeleton.c
index b23b5c397b1..c95fd72c3bb 100644
--- a/drivers/net/pci-skeleton.c
+++ b/drivers/net/pci-skeleton.c
@@ -781,7 +781,7 @@ static int __devinit netdrv_init_one (struct pci_dev *pdev,
 	dev->irq = pdev->irq;
 	dev->base_addr = (unsigned long) ioaddr;
 
-	/* dev->priv/tp zeroed and aligned in alloc_etherdev */
+	/* netdev_priv()/tp zeroed and aligned in alloc_etherdev */
 	tp = netdev_priv(dev);
 
 	/* note: tp->chipset set in netdrv_init_board */
diff --git a/drivers/net/sun3_82586.c b/drivers/net/sun3_82586.c
index e8f97d5c9c2..e0d84772771 100644
--- a/drivers/net/sun3_82586.c
+++ b/drivers/net/sun3_82586.c
@@ -209,7 +209,7 @@ static int sun3_82586_open(struct net_device *dev)
 static int check586(struct net_device *dev,char *where,unsigned size)
 {
 	struct priv pb;
-	struct priv *p = /* (struct priv *) dev->priv*/ &pb;
+	struct priv *p = &pb;
 	char *iscp_addr;
 	int i;
 
diff --git a/drivers/net/sunbmac.c b/drivers/net/sunbmac.c
index 977b3e08bbf..7f69c7f176c 100644
--- a/drivers/net/sunbmac.c
+++ b/drivers/net/sunbmac.c
@@ -1233,7 +1233,7 @@ fail_and_cleanup:
 				  bp->bmac_block,
 				  bp->bblock_dvma);
 
-	/* This also frees the co-located 'dev->priv' */
+	/* This also frees the co-located private data */
 	free_netdev(dev);
 	return -ENODEV;
 }
diff --git a/drivers/net/tokenring/3c359.c b/drivers/net/tokenring/3c359.c
index e7a944657cf..43853e3b210 100644
--- a/drivers/net/tokenring/3c359.c
+++ b/drivers/net/tokenring/3c359.c
@@ -296,8 +296,9 @@ static int __devinit xl_probe(struct pci_dev *pdev,
 	} ; 
 
 	/* 
-	 * Allowing init_trdev to allocate the dev->priv structure will align xl_private
-   	 * on a 32 bytes boundary which we need for the rx/tx descriptors
+	 * Allowing init_trdev to allocate the private data will align
+	 * xl_private on a 32 bytes boundary which we need for the rx/tx
+	 * descriptors
 	 */
 
 	dev = alloc_trdev(sizeof(struct xl_private)) ; 
diff --git a/drivers/net/via-rhine.c b/drivers/net/via-rhine.c
index 93b74b7b707..8d405c83df8 100644
--- a/drivers/net/via-rhine.c
+++ b/drivers/net/via-rhine.c
@@ -191,12 +191,13 @@ IIId. Synchronization
 
 The driver runs as two independent, single-threaded flows of control. One
 is the send-packet routine, which enforces single-threaded use by the
-dev->priv->lock spinlock. The other thread is the interrupt handler, which
-is single threaded by the hardware and interrupt handling software.
+netdev_priv(dev)->lock spinlock. The other thread is the interrupt handler,
+which is single threaded by the hardware and interrupt handling software.
 
 The send packet thread has partial control over the Tx ring. It locks the
-dev->priv->lock whenever it's queuing a Tx packet. If the next slot in the ring
-is not available it stops the transmit queue by calling netif_stop_queue.
+netdev_priv(dev)->lock whenever it's queuing a Tx packet. If the next slot in
+the ring is not available it stops the transmit queue by
+calling netif_stop_queue.
 
 The interrupt handler has exclusive control over the Rx ring and records stats
 from the Tx ring. After reaping the stats, it marks the Tx queue entry as
diff --git a/drivers/net/wireless/strip.c b/drivers/net/wireless/strip.c
index 692e6c5e009..dd0de3a9ed4 100644
--- a/drivers/net/wireless/strip.c
+++ b/drivers/net/wireless/strip.c
@@ -2494,7 +2494,7 @@ static void strip_dev_setup(struct net_device *dev)
 	dev->type = ARPHRD_METRICOM;	/* dtang */
 	dev->hard_header_len = sizeof(STRIP_Header);
 	/*
-	 *  dev->priv             Already holds a pointer to our struct strip
+	 *  netdev_priv(dev) Already holds a pointer to our struct strip
 	 */
 
 	*(MetricomAddress *) & dev->broadcast = broadcast_address;
diff --git a/include/linux/hdlc.h b/include/linux/hdlc.h
index e960faac609..fd47a151665 100644
--- a/include/linux/hdlc.h
+++ b/include/linux/hdlc.h
@@ -43,7 +43,7 @@ struct hdlc_proto {
 };
 
 
-/* Pointed to by dev->priv */
+/* Pointed to by netdev_priv(dev) */
 typedef struct hdlc_device {
 	/* used by HDLC layer to take control over HDLC device from hw driver*/
 	int (*attach)(struct net_device *dev,
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 0df0db068ac..47e73152831 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -785,7 +785,6 @@ struct net_device
 /*
  * One part is mostly used on xmit path (device)
  */
-	void			*priv;	/* pointer to private data	*/
 	/* These may be needed for future network-power-down code. */
 	unsigned long		trans_start;	/* Time (in jiffies) of last Tx	*/
 
diff --git a/net/atm/mpc.c b/net/atm/mpc.c
index 12e9ea371db..039d5cc72c3 100644
--- a/net/atm/mpc.c
+++ b/net/atm/mpc.c
@@ -341,8 +341,8 @@ static const char *mpoa_device_type_string(char type)
 }
 
 /*
- * lec device calls this via its dev->priv->lane2_ops->associate_indicator()
- * when it sees a TLV in LE_ARP packet.
+ * lec device calls this via its netdev_priv(dev)->lane2_ops
+ * ->associate_indicator() when it sees a TLV in LE_ARP packet.
  * We fill in the pointer above when we see a LANE2 lec initializing
  * See LANE2 spec 3.1.5
  *
diff --git a/net/core/dev.c b/net/core/dev.c
index 4615e9a443a..f54cac76438 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -4378,12 +4378,6 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
 	dev->num_tx_queues = queue_count;
 	dev->real_num_tx_queues = queue_count;
 
-	if (sizeof_priv) {
-		dev->priv = ((char *)dev +
-			     ((sizeof(struct net_device) + NETDEV_ALIGN_CONST)
-			      & ~NETDEV_ALIGN_CONST));
-	}
-
 	dev->gso_max_size = GSO_MAX_SIZE;
 
 	netdev_init_queues(dev);
-- 
cgit v1.2.3-70-g09d2


From 7b363e440021a1cf9ed76944b2685f48dacefb3e Mon Sep 17 00:00:00 2001
From: Neil Horman <nhorman@tuxdriver.com>
Date: Tue, 9 Dec 2008 23:22:26 -0800
Subject: netpoll: fix race on poll_list resulting in garbage entry

	A few months back a race was discused between the netpoll napi service
path, and the fast path through net_rx_action:
http://kerneltrap.org/mailarchive/linux-netdev/2007/10/16/345470

A patch was submitted for that bug, but I think we missed a case.

Consider the following scenario:

INITIAL STATE
CPU0 has one napi_struct A on its poll_list
CPU1 is calling netpoll_send_skb and needs to call poll_napi on the same
napi_struct A that CPU0 has on its list



CPU0						CPU1
net_rx_action					poll_napi
!list_empty (returns true)			locks poll_lock for A
						 poll_one_napi
						  napi->poll
						   netif_rx_complete
						    __napi_complete
						    (removes A from poll_list)
list_entry(list->next)


In the above scenario, net_rx_action assumes that the per-cpu poll_list is
exclusive to that cpu.  netpoll of course violates that, and because the netpoll
path can dequeue from the poll list, its possible for CPU0 to detect a non-empty
list at the top of the while loop in net_rx_action, but have it become empty by
the time it calls list_entry.  Since the poll_list isn't surrounded by any other
structure, the returned data from that list_entry call in this situation is
garbage, and any number of crashes can result based on what exactly that garbage
is.

Given that its not fasible for performance reasons to place exclusive locks
arround each cpus poll list to provide that mutal exclusion, I think the best
solution is modify the netpoll path in such a way that we continue to guarantee
that the poll_list for a cpu is in fact exclusive to that cpu.  To do this I've
implemented the patch below.  It adds an additional bit to the state field in
the napi_struct.  When executing napi->poll from the netpoll_path, this bit will
be set. When a driver calls netif_rx_complete, if that bit is set, it will not
remove the napi_struct from the poll_list.  That work will be saved for the next
iteration of net_rx_action.

I've tested this and it seems to work well.  About the biggest drawback I can
see to it is the fact that it might result in an extra loop through
net_rx_action in the event that the device is actually contended for (i.e. the
netpoll path actually preforms all the needed work no the device, and the call
to net_rx_action winds up doing nothing, except removing the napi_struct from
the poll_list.  However I think this is probably a small price to pay, given
that the alternative is a crash.

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 7 +++++++
 net/core/netpoll.c        | 2 ++
 2 files changed, 9 insertions(+)

(limited to 'include/linux/netdevice.h')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 9d77b1d7dca..e26f5495289 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -319,6 +319,7 @@ enum
 {
 	NAPI_STATE_SCHED,	/* Poll is scheduled */
 	NAPI_STATE_DISABLE,	/* Disable pending */
+	NAPI_STATE_NPSVC,	/* Netpoll - don't dequeue from poll_list */
 };
 
 extern void __napi_schedule(struct napi_struct *n);
@@ -1497,6 +1498,12 @@ static inline void netif_rx_complete(struct net_device *dev,
 {
 	unsigned long flags;
 
+	/*
+	 * don't let napi dequeue from the cpu poll list
+	 * just in case its running on a different cpu
+	 */
+	if (unlikely(test_bit(NAPI_STATE_NPSVC, &napi->state)))
+		return;
 	local_irq_save(flags);
 	__netif_rx_complete(dev, napi);
 	local_irq_restore(flags);
diff --git a/net/core/netpoll.c b/net/core/netpoll.c
index 6c7af390be0..dadac6281f2 100644
--- a/net/core/netpoll.c
+++ b/net/core/netpoll.c
@@ -133,9 +133,11 @@ static int poll_one_napi(struct netpoll_info *npinfo,
 
 	npinfo->rx_flags |= NETPOLL_RX_DROP;
 	atomic_inc(&trapped);
+	set_bit(NAPI_STATE_NPSVC, &napi->state);
 
 	work = napi->poll(napi, budget);
 
+	clear_bit(NAPI_STATE_NPSVC, &napi->state);
 	atomic_dec(&trapped);
 	npinfo->rx_flags &= ~NETPOLL_RX_DROP;
 
-- 
cgit v1.2.3-70-g09d2


From 1a881f27c50b4fbd6858a8696a189263621136b0 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 15 Dec 2008 23:27:47 -0800
Subject: net: Add frag_list support to GSO

This patch allows GSO to handle frag_list in a limited way for the
purposes of allowing packets merged by GRO to be refragmented on
output.

Most hardware won't (and aren't expected to) support handling GRO
frag_list packets directly.  Therefore we will perform GSO in
software for those cases.

However, for drivers that can support it (such as virtual NICs) we
may not have to segment the packets at all.

Whether the added overhead of GRO/GSO is worthwhile for bridges
and routers when weighed against the benefit of potentially
increasing the MTU within the host is still an open question.
However, for the case of host nodes this is undoubtedly a win.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h | 2 ++
 net/core/dev.c            | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index b60c26b7d31..bdf5465deb9 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1858,6 +1858,8 @@ static inline int netif_needs_gso(struct net_device *dev, struct sk_buff *skb)
 {
 	return skb_is_gso(skb) &&
 	       (!skb_gso_ok(skb, dev->features) ||
+	        (skb_shinfo(skb)->frag_list &&
+	         !(dev->features & NETIF_F_FRAGLIST)) ||
 		unlikely(skb->ip_summed != CHECKSUM_PARTIAL));
 }
 
diff --git a/net/core/dev.c b/net/core/dev.c
index f54cac76438..e415f0b0d0d 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1533,8 +1533,6 @@ struct sk_buff *skb_gso_segment(struct sk_buff *skb, int features)
 	__be16 type = skb->protocol;
 	int err;
 
-	BUG_ON(skb_shinfo(skb)->frag_list);
-
 	skb_reset_mac_header(skb);
 	skb->mac_len = skb->network_header - skb->mac_header;
 	__skb_pull(skb, skb->mac_len);
-- 
cgit v1.2.3-70-g09d2


From d565b0a1a9b6ee7dff46e1f68b26b526ac11ae50 Mon Sep 17 00:00:00 2001
From: Herbert Xu <herbert@gondor.apana.org.au>
Date: Mon, 15 Dec 2008 23:38:52 -0800
Subject: net: Add Generic Receive Offload infrastructure

This patch adds the top-level GRO (Generic Receive Offload) infrastructure.
This is pretty similar to LRO except that this is protocol-independent.
Instead of holding packets in an lro_mgr structure, they're now held in
napi_struct.

For drivers that intend to use this, they can set the NETIF_F_GRO bit and
call napi_gro_receive instead of netif_receive_skb or just call netif_rx.
The latter will call napi_receive_skb automatically.  When napi_gro_receive
is used, the driver must either call napi_complete/napi_rx_complete, or
call napi_gro_flush in softirq context if the driver uses the primitives
__napi_complete/__napi_rx_complete.

Protocols will set the gro_receive and gro_complete function pointers in
order to participate in this scheme.

In addition to the packet, gro_receive will get a list of currently held
packets.  Each packet in the list has a same_flow field which is non-zero
if it is a potential match for the new packet.  For each packet that may
match, they also have a flush field which is non-zero if the held packet
must not be merged with the new packet.

Once gro_receive has determined that the new skb matches a held packet,
the held packet may be processed immediately if the new skb cannot be
merged with it.  In this case gro_receive should return the pointer to
the existing skb in gro_list.  Otherwise the new skb should be merged into
the existing packet and NULL should be returned, unless the new skb makes
it impossible for any further merges to be made (e.g., FIN packet) where
the merged skb should be returned.

Whenever the skb is merged into an existing entry, the gro_receive
function should set NAPI_GRO_CB(skb)->same_flow.  Note that if an skb
merely matches an existing entry but can't be merged with it, then
this shouldn't be set.

If gro_receive finds it pointless to hold the new skb for future merging,
it should set NAPI_GRO_CB(skb)->flush.

Held packets will be flushed by napi_gro_flush which is called by
napi_complete and napi_rx_complete.

Currently held packets are stored in a singly liked list just like LRO.
The list is limited to a maximum of 8 entries.  In future, this may be
expanded to use a hash table to allow more flows to be held for merging.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/linux/netdevice.h |  80 +++++++------------
 include/linux/netpoll.h   |   5 --
 net/core/dev.c            | 193 +++++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 219 insertions(+), 59 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index bdf5465deb9..58856b6737f 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -314,8 +314,9 @@ struct napi_struct {
 	spinlock_t		poll_lock;
 	int			poll_owner;
 	struct net_device	*dev;
-	struct list_head	dev_list;
 #endif
+	struct list_head	dev_list;
+	struct sk_buff		*gro_list;
 };
 
 enum
@@ -376,22 +377,8 @@ static inline int napi_reschedule(struct napi_struct *napi)
  *
  * Mark NAPI processing as complete.
  */
-static inline void __napi_complete(struct napi_struct *n)
-{
-	BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
-	list_del(&n->poll_list);
-	smp_mb__before_clear_bit();
-	clear_bit(NAPI_STATE_SCHED, &n->state);
-}
-
-static inline void napi_complete(struct napi_struct *n)
-{
-	unsigned long flags;
-
-	local_irq_save(flags);
-	__napi_complete(n);
-	local_irq_restore(flags);
-}
+extern void __napi_complete(struct napi_struct *n);
+extern void napi_complete(struct napi_struct *n);
 
 /**
  *	napi_disable - prevent NAPI from scheduling
@@ -640,9 +627,7 @@ struct net_device
 	unsigned long		state;
 
 	struct list_head	dev_list;
-#ifdef CONFIG_NETPOLL
 	struct list_head	napi_list;
-#endif
 
 	/* Net device features */
 	unsigned long		features;
@@ -661,6 +646,7 @@ struct net_device
 #define NETIF_F_LLTX		4096	/* LockLess TX - deprecated. Please */
 					/* do not use LLTX in new drivers */
 #define NETIF_F_NETNS_LOCAL	8192	/* Does not change network namespaces */
+#define NETIF_F_GRO		16384	/* Generic receive offload */
 #define NETIF_F_LRO		32768	/* large receive offload */
 
 	/* Segmentation offload features */
@@ -984,22 +970,8 @@ static inline void *netdev_priv(const struct net_device *dev)
  * netif_napi_add() must be used to initialize a napi context prior to calling
  * *any* of the other napi related functions.
  */
-static inline void netif_napi_add(struct net_device *dev,
-				  struct napi_struct *napi,
-				  int (*poll)(struct napi_struct *, int),
-				  int weight)
-{
-	INIT_LIST_HEAD(&napi->poll_list);
-	napi->poll = poll;
-	napi->weight = weight;
-#ifdef CONFIG_NETPOLL
-	napi->dev = dev;
-	list_add(&napi->dev_list, &dev->napi_list);
-	spin_lock_init(&napi->poll_lock);
-	napi->poll_owner = -1;
-#endif
-	set_bit(NAPI_STATE_SCHED, &napi->state);
-}
+void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
+		    int (*poll)(struct napi_struct *, int), int weight);
 
 /**
  *  netif_napi_del - remove a napi context
@@ -1007,12 +979,20 @@ static inline void netif_napi_add(struct net_device *dev,
  *
  *  netif_napi_del() removes a napi context from the network device napi list
  */
-static inline void netif_napi_del(struct napi_struct *napi)
-{
-#ifdef CONFIG_NETPOLL
-	list_del(&napi->dev_list);
-#endif
-}
+void netif_napi_del(struct napi_struct *napi);
+
+struct napi_gro_cb {
+	/* This is non-zero if the packet may be of the same flow. */
+	int same_flow;
+
+	/* This is non-zero if the packet cannot be merged with the new skb. */
+	int flush;
+
+	/* Number of segments aggregated. */
+	int count;
+};
+
+#define NAPI_GRO_CB(skb) ((struct napi_gro_cb *)(skb)->cb)
 
 struct packet_type {
 	__be16			type;	/* This is really htons(ether_type). */
@@ -1024,6 +1004,9 @@ struct packet_type {
 	struct sk_buff		*(*gso_segment)(struct sk_buff *skb,
 						int features);
 	int			(*gso_send_check)(struct sk_buff *skb);
+	struct sk_buff		**(*gro_receive)(struct sk_buff **head,
+					       struct sk_buff *skb);
+	int			(*gro_complete)(struct sk_buff *skb);
 	void			*af_packet_priv;
 	struct list_head	list;
 };
@@ -1377,6 +1360,9 @@ extern int		netif_rx(struct sk_buff *skb);
 extern int		netif_rx_ni(struct sk_buff *skb);
 #define HAVE_NETIF_RECEIVE_SKB 1
 extern int		netif_receive_skb(struct sk_buff *skb);
+extern void		napi_gro_flush(struct napi_struct *napi);
+extern int		napi_gro_receive(struct napi_struct *napi,
+					 struct sk_buff *skb);
 extern void		netif_nit_deliver(struct sk_buff *skb);
 extern int		dev_valid_name(const char *name);
 extern int		dev_ioctl(struct net *net, unsigned int cmd, void __user *);
@@ -1621,17 +1607,7 @@ static inline void __netif_rx_complete(struct net_device *dev,
 static inline void netif_rx_complete(struct net_device *dev,
 				     struct napi_struct *napi)
 {
-	unsigned long flags;
-
-	/*
-	 * don't let napi dequeue from the cpu poll list
-	 * just in case its running on a different cpu
-	 */
-	if (unlikely(test_bit(NAPI_STATE_NPSVC, &napi->state)))
-		return;
-	local_irq_save(flags);
-	__netif_rx_complete(dev, napi);
-	local_irq_restore(flags);
+	napi_complete(napi);
 }
 
 static inline void __netif_tx_lock(struct netdev_queue *txq, int cpu)
diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h
index e3d79593fb3..e38d3c9dccd 100644
--- a/include/linux/netpoll.h
+++ b/include/linux/netpoll.h
@@ -94,11 +94,6 @@ static inline void netpoll_poll_unlock(void *have)
 	rcu_read_unlock();
 }
 
-static inline void netpoll_netdev_init(struct net_device *dev)
-{
-	INIT_LIST_HEAD(&dev->napi_list);
-}
-
 #else
 static inline int netpoll_rx(struct sk_buff *skb)
 {
diff --git a/net/core/dev.c b/net/core/dev.c
index e415f0b0d0d..d8d7d1fccde 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -129,6 +129,9 @@
 
 #include "net-sysfs.h"
 
+/* Instead of increasing this, you should create a hash table. */
+#define MAX_GRO_SKBS 8
+
 /*
  *	The list of packet types we will receive (as opposed to discard)
  *	and the routines to invoke.
@@ -2335,6 +2338,122 @@ static void flush_backlog(void *arg)
 		}
 }
 
+static int napi_gro_complete(struct sk_buff *skb)
+{
+	struct packet_type *ptype;
+	__be16 type = skb->protocol;
+	struct list_head *head = &ptype_base[ntohs(type) & PTYPE_HASH_MASK];
+	int err = -ENOENT;
+
+	if (!skb_shinfo(skb)->frag_list)
+		goto out;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(ptype, head, list) {
+		if (ptype->type != type || ptype->dev || !ptype->gro_complete)
+			continue;
+
+		err = ptype->gro_complete(skb);
+		break;
+	}
+	rcu_read_unlock();
+
+	if (err) {
+		WARN_ON(&ptype->list == head);
+		kfree_skb(skb);
+		return NET_RX_SUCCESS;
+	}
+
+out:
+	__skb_push(skb, -skb_network_offset(skb));
+	return netif_receive_skb(skb);
+}
+
+void napi_gro_flush(struct napi_struct *napi)
+{
+	struct sk_buff *skb, *next;
+
+	for (skb = napi->gro_list; skb; skb = next) {
+		next = skb->next;
+		skb->next = NULL;
+		napi_gro_complete(skb);
+	}
+
+	napi->gro_list = NULL;
+}
+EXPORT_SYMBOL(napi_gro_flush);
+
+int napi_gro_receive(struct napi_struct *napi, struct sk_buff *skb)
+{
+	struct sk_buff **pp = NULL;
+	struct packet_type *ptype;
+	__be16 type = skb->protocol;
+	struct list_head *head = &ptype_base[ntohs(type) & PTYPE_HASH_MASK];
+	int count = 0;
+	int mac_len;
+
+	if (!(skb->dev->features & NETIF_F_GRO))
+		goto normal;
+
+	rcu_read_lock();
+	list_for_each_entry_rcu(ptype, head, list) {
+		struct sk_buff *p;
+
+		if (ptype->type != type || ptype->dev || !ptype->gro_receive)
+			continue;
+
+		skb_reset_network_header(skb);
+		mac_len = skb->network_header - skb->mac_header;
+		skb->mac_len = mac_len;
+		NAPI_GRO_CB(skb)->same_flow = 0;
+		NAPI_GRO_CB(skb)->flush = 0;
+
+		for (p = napi->gro_list; p; p = p->next) {
+			count++;
+			NAPI_GRO_CB(p)->same_flow =
+				p->mac_len == mac_len &&
+				!memcmp(skb_mac_header(p), skb_mac_header(skb),
+					mac_len);
+			NAPI_GRO_CB(p)->flush = 0;
+		}
+
+		pp = ptype->gro_receive(&napi->gro_list, skb);
+		break;
+	}
+	rcu_read_unlock();
+
+	if (&ptype->list == head)
+		goto normal;
+
+	if (pp) {
+		struct sk_buff *nskb = *pp;
+
+		*pp = nskb->next;
+		nskb->next = NULL;
+		napi_gro_complete(nskb);
+		count--;
+	}
+
+	if (NAPI_GRO_CB(skb)->same_flow)
+		goto ok;
+
+	if (NAPI_GRO_CB(skb)->flush || count >= MAX_GRO_SKBS) {
+		__skb_push(skb, -skb_network_offset(skb));
+		goto normal;
+	}
+
+	NAPI_GRO_CB(skb)->count = 1;
+	skb->next = napi->gro_list;
+	napi->gro_list = skb;
+
+ok:
+	return NET_RX_SUCCESS;
+
+normal:
+	return netif_receive_skb(skb);
+}
+EXPORT_SYMBOL(napi_gro_receive);
+
 static int process_backlog(struct napi_struct *napi, int quota)
 {
 	int work = 0;
@@ -2354,9 +2473,11 @@ static int process_backlog(struct napi_struct *napi, int quota)
 		}
 		local_irq_enable();
 
-		netif_receive_skb(skb);
+		napi_gro_receive(napi, skb);
 	} while (++work < quota && jiffies == start_time);
 
+	napi_gro_flush(napi);
+
 	return work;
 }
 
@@ -2377,6 +2498,68 @@ void __napi_schedule(struct napi_struct *n)
 }
 EXPORT_SYMBOL(__napi_schedule);
 
+void __napi_complete(struct napi_struct *n)
+{
+	BUG_ON(!test_bit(NAPI_STATE_SCHED, &n->state));
+	BUG_ON(n->gro_list);
+
+	list_del(&n->poll_list);
+	smp_mb__before_clear_bit();
+	clear_bit(NAPI_STATE_SCHED, &n->state);
+}
+EXPORT_SYMBOL(__napi_complete);
+
+void napi_complete(struct napi_struct *n)
+{
+	unsigned long flags;
+
+	/*
+	 * don't let napi dequeue from the cpu poll list
+	 * just in case its running on a different cpu
+	 */
+	if (unlikely(test_bit(NAPI_STATE_NPSVC, &n->state)))
+		return;
+
+	napi_gro_flush(n);
+	local_irq_save(flags);
+	__napi_complete(n);
+	local_irq_restore(flags);
+}
+EXPORT_SYMBOL(napi_complete);
+
+void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
+		    int (*poll)(struct napi_struct *, int), int weight)
+{
+	INIT_LIST_HEAD(&napi->poll_list);
+	napi->gro_list = NULL;
+	napi->poll = poll;
+	napi->weight = weight;
+	list_add(&napi->dev_list, &dev->napi_list);
+#ifdef CONFIG_NETPOLL
+	napi->dev = dev;
+	spin_lock_init(&napi->poll_lock);
+	napi->poll_owner = -1;
+#endif
+	set_bit(NAPI_STATE_SCHED, &napi->state);
+}
+EXPORT_SYMBOL(netif_napi_add);
+
+void netif_napi_del(struct napi_struct *napi)
+{
+	struct sk_buff *skb, *next;
+
+	list_del(&napi->dev_list);
+
+	for (skb = napi->gro_list; skb; skb = next) {
+		next = skb->next;
+		skb->next = NULL;
+		kfree_skb(skb);
+	}
+
+	napi->gro_list = NULL;
+}
+EXPORT_SYMBOL(netif_napi_del);
+
 
 static void net_rx_action(struct softirq_action *h)
 {
@@ -4380,7 +4563,7 @@ struct net_device *alloc_netdev_mq(int sizeof_priv, const char *name,
 
 	netdev_init_queues(dev);
 
-	netpoll_netdev_init(dev);
+	INIT_LIST_HEAD(&dev->napi_list);
 	setup(dev);
 	strcpy(dev->name, name);
 	return dev;
@@ -4397,10 +4580,15 @@ EXPORT_SYMBOL(alloc_netdev_mq);
  */
 void free_netdev(struct net_device *dev)
 {
+	struct napi_struct *p, *n;
+
 	release_net(dev_net(dev));
 
 	kfree(dev->_tx);
 
+	list_for_each_entry_safe(p, n, &dev->napi_list, dev_list)
+		netif_napi_del(p);
+
 	/*  Compatibility with error handling in drivers */
 	if (dev->reg_state == NETREG_UNINITIALIZED) {
 		kfree((char *)dev - dev->padded);
@@ -4949,6 +5137,7 @@ static int __init net_dev_init(void)
 
 		queue->backlog.poll = process_backlog;
 		queue->backlog.weight = weight_p;
+		queue->backlog.gro_list = NULL;
 	}
 
 	dev_boot_phase = 0;
-- 
cgit v1.2.3-70-g09d2


From 908a7a16b852ffd618a9127be8d62432182d81b4 Mon Sep 17 00:00:00 2001
From: Neil Horman <nhorman@tuxdriver.com>
Date: Mon, 22 Dec 2008 20:43:12 -0800
Subject: net: Remove unused netdev arg from some NAPI interfaces.

When the napi api was changed to separate its 1:1 binding to the net_device
struct, the netif_rx_[prep|schedule|complete] api failed to remove the now
vestigual net_device structure parameter.  This patch cleans up that api by
properly removing it..

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 drivers/infiniband/hw/nes/nes_hw.c      |  2 +-
 drivers/infiniband/hw/nes/nes_nic.c     |  2 +-
 drivers/infiniband/ulp/ipoib/ipoib_ib.c |  6 +++---
 drivers/net/8139cp.c                    |  6 +++---
 drivers/net/8139too.c                   |  6 +++---
 drivers/net/amd8111e.c                  |  6 +++---
 drivers/net/arm/ep93xx_eth.c            |  6 +++---
 drivers/net/arm/ixp4xx_eth.c            |  6 +++---
 drivers/net/atl1e/atl1e_main.c          |  6 +++---
 drivers/net/b44.c                       |  6 +++---
 drivers/net/bnx2.c                      | 15 ++++++---------
 drivers/net/bnx2x_main.c                |  6 +++---
 drivers/net/cassini.c                   |  8 ++++----
 drivers/net/chelsio/sge.c               |  4 ++--
 drivers/net/cpmac.c                     | 10 +++++-----
 drivers/net/e100.c                      |  7 +++----
 drivers/net/e1000/e1000_main.c          | 10 +++++-----
 drivers/net/e1000e/netdev.c             | 14 +++++++-------
 drivers/net/ehea/ehea_main.c            |  6 +++---
 drivers/net/enic/enic_main.c            | 12 ++++++------
 drivers/net/epic100.c                   |  6 +++---
 drivers/net/forcedeth.c                 | 10 +++++-----
 drivers/net/fs_enet/fs_enet-main.c      |  4 ++--
 drivers/net/gianfar.c                   |  6 +++---
 drivers/net/ibmveth.c                   |  6 +++---
 drivers/net/igb/igb_main.c              | 12 ++++++------
 drivers/net/ixgb/ixgb_main.c            |  6 +++---
 drivers/net/ixgbe/ixgbe_main.c          | 12 ++++++------
 drivers/net/ixp2000/ixpdev.c            |  4 ++--
 drivers/net/jme.c                       |  1 -
 drivers/net/jme.h                       |  6 +++---
 drivers/net/korina.c                    |  4 ++--
 drivers/net/macb.c                      | 10 +++++-----
 drivers/net/mlx4/en_rx.c                |  4 ++--
 drivers/net/myri10ge/myri10ge.c         |  6 +++---
 drivers/net/natsemi.c                   |  6 +++---
 drivers/net/netxen/netxen_nic_main.c    |  2 +-
 drivers/net/niu.c                       |  6 +++---
 drivers/net/pasemi_mac.c                |  6 +++---
 drivers/net/pcnet32.c                   |  6 +++---
 drivers/net/qla3xxx.c                   |  6 +++---
 drivers/net/qlge/qlge_main.c            |  7 +++----
 drivers/net/r6040.c                     |  4 ++--
 drivers/net/r8169.c                     |  6 +++---
 drivers/net/s2io.c                      |  8 ++++----
 drivers/net/sb1250-mac.c                |  6 +++---
 drivers/net/sfc/efx.c                   |  2 +-
 drivers/net/sfc/efx.h                   |  2 +-
 drivers/net/skge.c                      |  6 +++---
 drivers/net/smsc911x.c                  |  2 +-
 drivers/net/smsc9420.c                  |  4 ++--
 drivers/net/spider_net.c                | 15 ++++++---------
 drivers/net/starfire.c                  |  6 +++---
 drivers/net/sungem.c                    |  6 +++---
 drivers/net/tc35815.c                   |  6 +++---
 drivers/net/tehuti.c                    |  7 +++----
 drivers/net/tg3.c                       | 14 +++++++-------
 drivers/net/tsi108_eth.c                |  6 +++---
 drivers/net/tulip/interrupt.c           |  8 ++++----
 drivers/net/typhoon.c                   |  7 +++----
 drivers/net/ucc_geth.c                  |  6 +++---
 drivers/net/via-rhine.c                 |  4 ++--
 drivers/net/virtio_net.c                | 12 ++++++------
 drivers/net/wan/hd64572.c               |  4 ++--
 drivers/net/xen-netfront.c              |  8 ++++----
 include/linux/netdevice.h               | 24 +++++++++---------------
 66 files changed, 218 insertions(+), 235 deletions(-)

(limited to 'include/linux/netdevice.h')

diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c
index 7c49cc882d7..735c125b48a 100644
--- a/drivers/infiniband/hw/nes/nes_hw.c
+++ b/drivers/infiniband/hw/nes/nes_hw.c
@@ -2541,7 +2541,7 @@ static void nes_nic_napi_ce_handler(struct nes_device *nesdev, struct nes_hw_nic
 {
 	struct nes_vnic *nesvnic = container_of(cq, struct nes_vnic, nic_cq);
 
-	netif_rx_schedule(nesdev->netdev[nesvnic->netdev_index], &nesvnic->napi);
+	netif_rx_schedule(&nesvnic->napi);
 }
 
 
diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 3c96203e0d9..80e7a4d98d5 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -112,7 +112,7 @@ static int nes_netdev_poll(struct napi_struct *napi, int budget)
 	nes_nic_ce_handler(nesdev, nescq);
 
 	if (nescq->cqes_pending == 0) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		/* clear out completed cqes and arm */
 		nes_write32(nesdev->regs+NES_CQE_ALLOC, NES_CQE_ALLOC_NOTIFY_NEXT |
 				nescq->cq_number | (nescq->cqe_allocs_pending << 16));
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index 28eb6f03c58..a1925810be3 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -446,11 +446,11 @@ poll_more:
 		if (dev->features & NETIF_F_LRO)
 			lro_flush_all(&priv->lro.lro_mgr);
 
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		if (unlikely(ib_req_notify_cq(priv->recv_cq,
 					      IB_CQ_NEXT_COMP |
 					      IB_CQ_REPORT_MISSED_EVENTS)) &&
-		    netif_rx_reschedule(dev, napi))
+		    netif_rx_reschedule(napi))
 			goto poll_more;
 	}
 
@@ -462,7 +462,7 @@ void ipoib_ib_completion(struct ib_cq *cq, void *dev_ptr)
 	struct net_device *dev = dev_ptr;
 	struct ipoib_dev_priv *priv = netdev_priv(dev);
 
-	netif_rx_schedule(dev, &priv->napi);
+	netif_rx_schedule(&priv->napi);
 }
 
 static void drain_tx_cq(struct net_device *dev)
diff --git a/drivers/net/8139cp.c b/drivers/net/8139cp.c
index f6d9d1353dd..dd7ac8290ae 100644
--- a/drivers/net/8139cp.c
+++ b/drivers/net/8139cp.c
@@ -604,7 +604,7 @@ rx_next:
 
 		spin_lock_irqsave(&cp->lock, flags);
 		cpw16_f(IntrMask, cp_intr_mask);
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 		spin_unlock_irqrestore(&cp->lock, flags);
 	}
 
@@ -641,9 +641,9 @@ static irqreturn_t cp_interrupt (int irq, void *dev_instance)
 	}
 
 	if (status & (RxOK | RxErr | RxEmpty | RxFIFOOvr))
-		if (netif_rx_schedule_prep(dev, &cp->napi)) {
+		if (netif_rx_schedule_prep(&cp->napi)) {
 			cpw16_f(IntrMask, cp_norx_intr_mask);
-			__netif_rx_schedule(dev, &cp->napi);
+			__netif_rx_schedule(&cp->napi);
 		}
 
 	if (status & (TxOK | TxErr | TxEmpty | SWInt))
diff --git a/drivers/net/8139too.c b/drivers/net/8139too.c
index 67bbf4f25be..fe370f80579 100644
--- a/drivers/net/8139too.c
+++ b/drivers/net/8139too.c
@@ -2128,7 +2128,7 @@ static int rtl8139_poll(struct napi_struct *napi, int budget)
 		 */
 		spin_lock_irqsave(&tp->lock, flags);
 		RTL_W16_F(IntrMask, rtl8139_intr_mask);
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 		spin_unlock_irqrestore(&tp->lock, flags);
 	}
 	spin_unlock(&tp->rx_lock);
@@ -2178,9 +2178,9 @@ static irqreturn_t rtl8139_interrupt (int irq, void *dev_instance)
 	/* Receive packets are processed by poll routine.
 	   If not running start it now. */
 	if (status & RxAckBits){
-		if (netif_rx_schedule_prep(dev, &tp->napi)) {
+		if (netif_rx_schedule_prep(&tp->napi)) {
 			RTL_W16_F (IntrMask, rtl8139_norx_intr_mask);
-			__netif_rx_schedule(dev, &tp->napi);
+			__netif_rx_schedule(&tp->napi);
 		}
 	}
 
diff --git a/drivers/net/amd8111e.c b/drivers/net/amd8111e.c
index 0bc4f54d5db..187ac6eb6e9 100644
--- a/drivers/net/amd8111e.c
+++ b/drivers/net/amd8111e.c
@@ -831,7 +831,7 @@ static int amd8111e_rx_poll(struct napi_struct *napi, int budget)
 	if (rx_pkt_limit > 0) {
 		/* Receive descriptor is empty now */
 		spin_lock_irqsave(&lp->lock, flags);
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 		writel(VAL0|RINTEN0, mmio + INTEN0);
 		writel(VAL2 | RDMD0, mmio + CMD0);
 		spin_unlock_irqrestore(&lp->lock, flags);
@@ -1170,11 +1170,11 @@ static irqreturn_t amd8111e_interrupt(int irq, void *dev_id)
 
 	/* Check if Receive Interrupt has occurred. */
 	if (intr0 & RINT0) {
-		if (netif_rx_schedule_prep(dev, &lp->napi)) {
+		if (netif_rx_schedule_prep(&lp->napi)) {
 			/* Disable receive interupts */
 			writel(RINTEN0, mmio + INTEN0);
 			/* Schedule a polling routine */
-			__netif_rx_schedule(dev, &lp->napi);
+			__netif_rx_schedule(&lp->napi);
 		} else if (intren0 & RINTEN0) {
 			printk("************Driver bug! \
 				interrupt while in poll\n");
diff --git a/drivers/net/arm/ep93xx_eth.c b/drivers/net/arm/ep93xx_eth.c
index 588c9739d13..6ecc600c1bc 100644
--- a/drivers/net/arm/ep93xx_eth.c
+++ b/drivers/net/arm/ep93xx_eth.c
@@ -298,7 +298,7 @@ poll_some_more:
 		int more = 0;
 
 		spin_lock_irq(&ep->rx_lock);
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 		wrl(ep, REG_INTEN, REG_INTEN_TX | REG_INTEN_RX);
 		if (ep93xx_have_more_rx(ep)) {
 			wrl(ep, REG_INTEN, REG_INTEN_TX);
@@ -415,9 +415,9 @@ static irqreturn_t ep93xx_irq(int irq, void *dev_id)
 
 	if (status & REG_INTSTS_RX) {
 		spin_lock(&ep->rx_lock);
-		if (likely(netif_rx_schedule_prep(dev, &ep->napi))) {
+		if (likely(netif_rx_schedule_prep(&ep->napi))) {
 			wrl(ep, REG_INTEN, REG_INTEN_TX);
-			__netif_rx_schedule(dev, &ep->napi);
+			__netif_rx_schedule(&ep->napi);
 		}
 		spin_unlock(&ep->rx_lock);
 	}
diff --git a/drivers/net/arm/ixp4xx_eth.c b/drivers/net/arm/ixp4xx_eth.c
index 14ffa2a6189..b03609f2e90 100644
--- a/drivers/net/arm/ixp4xx_eth.c
+++ b/drivers/net/arm/ixp4xx_eth.c
@@ -498,7 +498,7 @@ static void eth_rx_irq(void *pdev)
 	printk(KERN_DEBUG "%s: eth_rx_irq\n", dev->name);
 #endif
 	qmgr_disable_irq(port->plat->rxq);
-	netif_rx_schedule(dev, &port->napi);
+	netif_rx_schedule(&port->napi);
 }
 
 static int eth_poll(struct napi_struct *napi, int budget)
@@ -526,7 +526,7 @@ static int eth_poll(struct napi_struct *napi, int budget)
 			printk(KERN_DEBUG "%s: eth_poll netif_rx_complete\n",
 			       dev->name);
 #endif
-			netif_rx_complete(dev, napi);
+			netif_rx_complete(napi);
 			qmgr_enable_irq(rxq);
 			if (!qmgr_stat_empty(rxq) &&
 			    netif_rx_reschedule(dev, napi)) {
@@ -1025,7 +1025,7 @@ static int eth_open(struct net_device *dev)
 	}
 	ports_open++;
 	/* we may already have RX data, enables IRQ */
-	netif_rx_schedule(dev, &port->napi);
+	netif_rx_schedule(&port->napi);
 	return 0;
 }
 
diff --git a/drivers/net/atl1e/atl1e_main.c b/drivers/net/atl1e/atl1e_main.c
index 98b2a7a466b..a72a46145ed 100644
--- a/drivers/net/atl1e/atl1e_main.c
+++ b/drivers/net/atl1e/atl1e_main.c
@@ -1326,9 +1326,9 @@ static irqreturn_t atl1e_intr(int irq, void *data)
 			AT_WRITE_REG(hw, REG_IMR,
 				     IMR_NORMAL_MASK & ~ISR_RX_EVENT);
 			AT_WRITE_FLUSH(hw);
-			if (likely(netif_rx_schedule_prep(netdev,
+			if (likely(netif_rx_schedule_prep(
 				   &adapter->napi)))
-				__netif_rx_schedule(netdev, &adapter->napi);
+				__netif_rx_schedule(&adapter->napi);
 		}
 	} while (--max_ints > 0);
 	/* re-enable Interrupt*/
@@ -1515,7 +1515,7 @@ static int atl1e_clean(struct napi_struct *napi, int budget)
 	/* If no Tx and not enough Rx work done, exit the polling mode */
 	if (work_done < budget) {
 quit_polling:
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		imr_data = AT_READ_REG(&adapter->hw, REG_IMR);
 		AT_WRITE_REG(&adapter->hw, REG_IMR, imr_data | ISR_RX_EVENT);
 		/* test debug */
diff --git a/drivers/net/b44.c b/drivers/net/b44.c
index 2c7a32eb92a..934a95091dc 100644
--- a/drivers/net/b44.c
+++ b/drivers/net/b44.c
@@ -875,7 +875,7 @@ static int b44_poll(struct napi_struct *napi, int budget)
 	}
 
 	if (work_done < budget) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		b44_enable_ints(bp);
 	}
 
@@ -907,13 +907,13 @@ static irqreturn_t b44_interrupt(int irq, void *dev_id)
 			goto irq_ack;
 		}
 
-		if (netif_rx_schedule_prep(dev, &bp->napi)) {
+		if (netif_rx_schedule_prep(&bp->napi)) {
 			/* NOTE: These writes are posted by the readback of
 			 *       the ISTAT register below.
 			 */
 			bp->istat = istat;
 			__b44_disable_ints(bp);
-			__netif_rx_schedule(dev, &bp->napi);
+			__netif_rx_schedule(&bp->napi);
 		} else {
 			printk(KERN_ERR PFX "%s: Error, poll already scheduled\n",
 			       dev->name);
diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c
index 1a2780374a4..33d69ddc90a 100644
--- a/drivers/net/bnx2.c
+++ b/drivers/net/bnx2.c
@@ -3043,7 +3043,6 @@ bnx2_msi(int irq, void *dev_instance)
 {
 	struct bnx2_napi *bnapi = dev_instance;
 	struct bnx2 *bp = bnapi->bp;
-	struct net_device *dev = bp->dev;
 
 	prefetch(bnapi->status_blk.msi);
 	REG_WR(bp, BNX2_PCICFG_INT_ACK_CMD,
@@ -3054,7 +3053,7 @@ bnx2_msi(int irq, void *dev_instance)
 	if (unlikely(atomic_read(&bp->intr_sem) != 0))
 		return IRQ_HANDLED;
 
-	netif_rx_schedule(dev, &bnapi->napi);
+	netif_rx_schedule(&bnapi->napi);
 
 	return IRQ_HANDLED;
 }
@@ -3064,7 +3063,6 @@ bnx2_msi_1shot(int irq, void *dev_instance)
 {
 	struct bnx2_napi *bnapi = dev_instance;
 	struct bnx2 *bp = bnapi->bp;
-	struct net_device *dev = bp->dev;
 
 	prefetch(bnapi->status_blk.msi);
 
@@ -3072,7 +3070,7 @@ bnx2_msi_1shot(int irq, void *dev_instance)
 	if (unlikely(atomic_read(&bp->intr_sem) != 0))
 		return IRQ_HANDLED;
 
-	netif_rx_schedule(dev, &bnapi->napi);
+	netif_rx_schedule(&bnapi->napi);
 
 	return IRQ_HANDLED;
 }
@@ -3082,7 +3080,6 @@ bnx2_interrupt(int irq, void *dev_instance)
 {
 	struct bnx2_napi *bnapi = dev_instance;
 	struct bnx2 *bp = bnapi->bp;
-	struct net_device *dev = bp->dev;
 	struct status_block *sblk = bnapi->status_blk.msi;
 
 	/* When using INTx, it is possible for the interrupt to arrive
@@ -3109,9 +3106,9 @@ bnx2_interrupt(int irq, void *dev_instance)
 	if (unlikely(atomic_read(&bp->intr_sem) != 0))
 		return IRQ_HANDLED;
 
-	if (netif_rx_schedule_prep(dev, &bnapi->napi)) {
+	if (netif_rx_schedule_prep(&bnapi->napi)) {
 		bnapi->last_status_idx = sblk->status_idx;
-		__netif_rx_schedule(dev, &bnapi->napi);
+		__netif_rx_schedule(&bnapi->napi);
 	}
 
 	return IRQ_HANDLED;
@@ -3221,7 +3218,7 @@ static int bnx2_poll_msix(struct napi_struct *napi, int budget)
 		rmb();
 		if (likely(!bnx2_has_fast_work(bnapi))) {
 
-			netif_rx_complete(bp->dev, napi);
+			netif_rx_complete(napi);
 			REG_WR(bp, BNX2_PCICFG_INT_ACK_CMD, bnapi->int_num |
 			       BNX2_PCICFG_INT_ACK_CMD_INDEX_VALID |
 			       bnapi->last_status_idx);
@@ -3254,7 +3251,7 @@ static int bnx2_poll(struct napi_struct *napi, int budget)
 
 		rmb();
 		if (likely(!bnx2_has_work(bnapi))) {
-			netif_rx_complete(bp->dev, napi);
+			netif_rx_complete(napi);
 			if (likely(bp->flags & BNX2_FLAG_USING_MSI_OR_MSIX)) {
 				REG_WR(bp, BNX2_PCICFG_INT_ACK_CMD,
 				       BNX2_PCICFG_INT_ACK_CMD_INDEX_VALID |
diff --git a/drivers/net/bnx2x_main.c b/drivers/net/bnx2x_main.c
index 24d2ae8b74b..02ab9b0ea69 100644
--- a/drivers/net/bnx2x_main.c
+++ b/drivers/net/bnx2x_main.c
@@ -1615,7 +1615,7 @@ static irqreturn_t bnx2x_msix_fp_int(int irq, void *fp_cookie)
 	prefetch(&fp->status_blk->c_status_block.status_block_index);
 	prefetch(&fp->status_blk->u_status_block.status_block_index);
 
-	netif_rx_schedule(dev, &bnx2x_fp(bp, index, napi));
+	netif_rx_schedule(&bnx2x_fp(bp, index, napi));
 
 	return IRQ_HANDLED;
 }
@@ -1654,7 +1654,7 @@ static irqreturn_t bnx2x_interrupt(int irq, void *dev_instance)
 		prefetch(&fp->status_blk->c_status_block.status_block_index);
 		prefetch(&fp->status_blk->u_status_block.status_block_index);
 
-		netif_rx_schedule(dev, &bnx2x_fp(bp, 0, napi));
+		netif_rx_schedule(&bnx2x_fp(bp, 0, napi));
 
 		status &= ~mask;
 	}
@@ -9284,7 +9284,7 @@ static int bnx2x_poll(struct napi_struct *napi, int budget)
 #ifdef BNX2X_STOP_ON_ERROR
 poll_panic:
 #endif
-		netif_rx_complete(bp->dev, napi);
+		netif_rx_complete(napi);
 
 		bnx2x_ack_sb(bp, FP_SB_ID(fp), USTORM_ID,
 			     le16_to_cpu(fp->fp_u_idx), IGU_INT_NOP, 1);
diff --git a/drivers/net/cassini.c b/drivers/net/cassini.c
index 023d205e905..321f43d9f0e 100644
--- a/drivers/net/cassini.c
+++ b/drivers/net/cassini.c
@@ -2506,7 +2506,7 @@ static irqreturn_t cas_interruptN(int irq, void *dev_id)
 	if (status & INTR_RX_DONE_ALT) { /* handle rx separately */
 #ifdef USE_NAPI
 		cas_mask_intr(cp);
-		netif_rx_schedule(dev, &cp->napi);
+		netif_rx_schedule(&cp->napi);
 #else
 		cas_rx_ringN(cp, ring, 0);
 #endif
@@ -2557,7 +2557,7 @@ static irqreturn_t cas_interrupt1(int irq, void *dev_id)
 	if (status & INTR_RX_DONE_ALT) { /* handle rx separately */
 #ifdef USE_NAPI
 		cas_mask_intr(cp);
-		netif_rx_schedule(dev, &cp->napi);
+		netif_rx_schedule(&cp->napi);
 #else
 		cas_rx_ringN(cp, 1, 0);
 #endif
@@ -2613,7 +2613,7 @@ static irqreturn_t cas_interrupt(int irq, void *dev_id)
 	if (status & INTR_RX_DONE) {
 #ifdef USE_NAPI
 		cas_mask_intr(cp);
-		netif_rx_schedule(dev, &cp->napi);
+		netif_rx_schedule(&cp->napi);
 #else
 		cas_rx_ringN(cp, 0, 0);
 #endif
@@ -2691,7 +2691,7 @@ rx_comp:
 #endif
 	spin_unlock_irqrestore(&cp->lock, flags);
 	if (enable_intr) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		cas_unmask_intr(cp);
 	}
 	return credits;
diff --git a/drivers/net/chelsio/sge.c b/drivers/net/chelsio/sge.c
index 1da70070c2f..7896468dda1 100644
--- a/drivers/net/chelsio/sge.c
+++ b/drivers/net/chelsio/sge.c
@@ -1613,7 +1613,7 @@ int t1_poll(struct napi_struct *napi, int budget)
 	int work_done = process_responses(adapter, budget);
 
 	if (likely(work_done < budget)) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		writel(adapter->sge->respQ.cidx,
 		       adapter->regs + A_SG_SLEEPING);
 	}
@@ -1633,7 +1633,7 @@ irqreturn_t t1_interrupt(int irq, void *data)
 
 		if (napi_schedule_prep(&adapter->napi)) {
 			if (process_pure_responses(adapter))
-				__netif_rx_schedule(dev, &adapter->napi);
+				__netif_rx_schedule(&adapter->napi);
 			else {
 				/* no data, no NAPI needed */
 				writel(sge->respQ.cidx, adapter->regs + A_SG_SLEEPING);
diff --git a/drivers/net/cpmac.c b/drivers/net/cpmac.c
index d39a77cba1a..f66548751c3 100644
--- a/drivers/net/cpmac.c
+++ b/drivers/net/cpmac.c
@@ -428,7 +428,7 @@ static int cpmac_poll(struct napi_struct *napi, int budget)
 			printk(KERN_WARNING "%s: rx: polling, but no queue\n",
 			       priv->dev->name);
 		spin_unlock(&priv->rx_lock);
-		netif_rx_complete(priv->dev, napi);
+		netif_rx_complete(napi);
 		return 0;
 	}
 
@@ -514,7 +514,7 @@ static int cpmac_poll(struct napi_struct *napi, int budget)
 	if (processed == 0) {
 		/* we ran out of packets to read,
 		 * revert to interrupt-driven mode */
-		netif_rx_complete(priv->dev, napi);
+		netif_rx_complete(napi);
 		cpmac_write(priv->regs, CPMAC_RX_INT_ENABLE, 1);
 		return 0;
 	}
@@ -536,7 +536,7 @@ fatal_error:
 	}
 
 	spin_unlock(&priv->rx_lock);
-	netif_rx_complete(priv->dev, napi);
+	netif_rx_complete(napi);
 	netif_tx_stop_all_queues(priv->dev);
 	napi_disable(&priv->napi);
 
@@ -802,9 +802,9 @@ static irqreturn_t cpmac_irq(int irq, void *dev_id)
 
 	if (status & MAC_INT_RX) {
 		queue = (status >> 8) & 7;
-		if (netif_rx_schedule_prep(dev, &priv->napi)) {
+		if (netif_rx_schedule_prep(&priv->napi)) {
 			cpmac_write(priv->regs, CPMAC_RX_INT_CLEAR, 1 << queue);
-			__netif_rx_schedule(dev, &priv->napi);
+			__netif_rx_schedule(&priv->napi);
 		}
 	}
 
diff --git a/drivers/net/e100.c b/drivers/net/e100.c
index dce7ff28c3f..9f38b16ccbb 100644
--- a/drivers/net/e100.c
+++ b/drivers/net/e100.c
@@ -2049,9 +2049,9 @@ static irqreturn_t e100_intr(int irq, void *dev_id)
 	if(stat_ack & stat_ack_rnr)
 		nic->ru_running = RU_SUSPENDED;
 
-	if(likely(netif_rx_schedule_prep(netdev, &nic->napi))) {
+	if(likely(netif_rx_schedule_prep(&nic->napi))) {
 		e100_disable_irq(nic);
-		__netif_rx_schedule(netdev, &nic->napi);
+		__netif_rx_schedule(&nic->napi);
 	}
 
 	return IRQ_HANDLED;
@@ -2060,7 +2060,6 @@ static irqreturn_t e100_intr(int irq, void *dev_id)
 static int e100_poll(struct napi_struct *napi, int budget)
 {
 	struct nic *nic = container_of(napi, struct nic, napi);
-	struct net_device *netdev = nic->netdev;
 	unsigned int work_done = 0;
 
 	e100_rx_clean(nic, &work_done, budget);
@@ -2068,7 +2067,7 @@ static int e100_poll(struct napi_struct *napi, int budget)
 
 	/* If budget not fully consumed, exit the polling mode */
 	if (work_done < budget) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		e100_enable_irq(nic);
 	}
 
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index 116c96e0b11..26474c92193 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -3687,12 +3687,12 @@ static irqreturn_t e1000_intr_msi(int irq, void *data)
 			mod_timer(&adapter->watchdog_timer, jiffies + 1);
 	}
 
-	if (likely(netif_rx_schedule_prep(netdev, &adapter->napi))) {
+	if (likely(netif_rx_schedule_prep(&adapter->napi))) {
 		adapter->total_tx_bytes = 0;
 		adapter->total_tx_packets = 0;
 		adapter->total_rx_bytes = 0;
 		adapter->total_rx_packets = 0;
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	} else
 		e1000_irq_enable(adapter);
 
@@ -3747,12 +3747,12 @@ static irqreturn_t e1000_intr(int irq, void *data)
 		ew32(IMC, ~0);
 		E1000_WRITE_FLUSH();
 	}
-	if (likely(netif_rx_schedule_prep(netdev, &adapter->napi))) {
+	if (likely(netif_rx_schedule_prep(&adapter->napi))) {
 		adapter->total_tx_bytes = 0;
 		adapter->total_tx_packets = 0;
 		adapter->total_rx_bytes = 0;
 		adapter->total_rx_packets = 0;
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	} else
 		/* this really should not happen! if it does it is basically a
 		 * bug, but not a hard error, so enable ints and continue */
@@ -3793,7 +3793,7 @@ static int e1000_clean(struct napi_struct *napi, int budget)
 	if (work_done < budget) {
 		if (likely(adapter->itr_setting & 3))
 			e1000_set_itr(adapter);
-		netif_rx_complete(poll_dev, napi);
+		netif_rx_complete(napi);
 		e1000_irq_enable(adapter);
 	}
 
diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
index f7b05609073..d4639facd1b 100644
--- a/drivers/net/e1000e/netdev.c
+++ b/drivers/net/e1000e/netdev.c
@@ -1179,12 +1179,12 @@ static irqreturn_t e1000_intr_msi(int irq, void *data)
 			mod_timer(&adapter->watchdog_timer, jiffies + 1);
 	}
 
-	if (netif_rx_schedule_prep(netdev, &adapter->napi)) {
+	if (netif_rx_schedule_prep(&adapter->napi)) {
 		adapter->total_tx_bytes = 0;
 		adapter->total_tx_packets = 0;
 		adapter->total_rx_bytes = 0;
 		adapter->total_rx_packets = 0;
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	}
 
 	return IRQ_HANDLED;
@@ -1246,12 +1246,12 @@ static irqreturn_t e1000_intr(int irq, void *data)
 			mod_timer(&adapter->watchdog_timer, jiffies + 1);
 	}
 
-	if (netif_rx_schedule_prep(netdev, &adapter->napi)) {
+	if (netif_rx_schedule_prep(&adapter->napi)) {
 		adapter->total_tx_bytes = 0;
 		adapter->total_tx_packets = 0;
 		adapter->total_rx_bytes = 0;
 		adapter->total_rx_packets = 0;
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	}
 
 	return IRQ_HANDLED;
@@ -1320,10 +1320,10 @@ static irqreturn_t e1000_intr_msix_rx(int irq, void *data)
 		adapter->rx_ring->set_itr = 0;
 	}
 
-	if (netif_rx_schedule_prep(netdev, &adapter->napi)) {
+	if (netif_rx_schedule_prep(&adapter->napi)) {
 		adapter->total_rx_bytes = 0;
 		adapter->total_rx_packets = 0;
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	}
 	return IRQ_HANDLED;
 }
@@ -2028,7 +2028,7 @@ clean_rx:
 	if (work_done < budget) {
 		if (adapter->itr_setting & 3)
 			e1000_set_itr(adapter);
-		netif_rx_complete(poll_dev, napi);
+		netif_rx_complete(napi);
 		if (adapter->msix_entries)
 			ew32(IMS, adapter->rx_ring->ims_val);
 		else
diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c
index 44c9ae18383..035aa7dfc5c 100644
--- a/drivers/net/ehea/ehea_main.c
+++ b/drivers/net/ehea/ehea_main.c
@@ -830,7 +830,7 @@ static int ehea_poll(struct napi_struct *napi, int budget)
 	while ((rx != budget) || force_irq) {
 		pr->poll_counter = 0;
 		force_irq = 0;
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		ehea_reset_cq_ep(pr->recv_cq);
 		ehea_reset_cq_ep(pr->send_cq);
 		ehea_reset_cq_n1(pr->recv_cq);
@@ -859,7 +859,7 @@ static void ehea_netpoll(struct net_device *dev)
 	int i;
 
 	for (i = 0; i < port->num_def_qps; i++)
-		netif_rx_schedule(dev, &port->port_res[i].napi);
+		netif_rx_schedule(&port->port_res[i].napi);
 }
 #endif
 
@@ -867,7 +867,7 @@ static irqreturn_t ehea_recv_irq_handler(int irq, void *param)
 {
 	struct ehea_port_res *pr = param;
 
-	netif_rx_schedule(pr->port->netdev, &pr->napi);
+	netif_rx_schedule(&pr->napi);
 
 	return IRQ_HANDLED;
 }
diff --git a/drivers/net/enic/enic_main.c b/drivers/net/enic/enic_main.c
index deddd76a550..d039e16f276 100644
--- a/drivers/net/enic/enic_main.c
+++ b/drivers/net/enic/enic_main.c
@@ -411,8 +411,8 @@ static irqreturn_t enic_isr_legacy(int irq, void *data)
 	}
 
 	if (ENIC_TEST_INTR(pba, ENIC_INTX_WQ_RQ)) {
-		if (netif_rx_schedule_prep(netdev, &enic->napi))
-			__netif_rx_schedule(netdev, &enic->napi);
+		if (netif_rx_schedule_prep(&enic->napi))
+			__netif_rx_schedule(&enic->napi);
 	} else {
 		vnic_intr_unmask(&enic->intr[ENIC_INTX_WQ_RQ]);
 	}
@@ -440,7 +440,7 @@ static irqreturn_t enic_isr_msi(int irq, void *data)
 	 * writes).
 	 */
 
-	netif_rx_schedule(enic->netdev, &enic->napi);
+	netif_rx_schedule(&enic->napi);
 
 	return IRQ_HANDLED;
 }
@@ -450,7 +450,7 @@ static irqreturn_t enic_isr_msix_rq(int irq, void *data)
 	struct enic *enic = data;
 
 	/* schedule NAPI polling for RQ cleanup */
-	netif_rx_schedule(enic->netdev, &enic->napi);
+	netif_rx_schedule(&enic->napi);
 
 	return IRQ_HANDLED;
 }
@@ -1068,7 +1068,7 @@ static int enic_poll(struct napi_struct *napi, int budget)
 		if (netdev->features & NETIF_F_LRO)
 			lro_flush_all(&enic->lro_mgr);
 
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		vnic_intr_unmask(&enic->intr[ENIC_MSIX_RQ]);
 	}
 
@@ -1112,7 +1112,7 @@ static int enic_poll_msix(struct napi_struct *napi, int budget)
 		if (netdev->features & NETIF_F_LRO)
 			lro_flush_all(&enic->lro_mgr);
 
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		vnic_intr_unmask(&enic->intr[ENIC_MSIX_RQ]);
 	}
 
diff --git a/drivers/net/epic100.c b/drivers/net/epic100.c
index 4a951b8cb4d..f9b37c80dda 100644
--- a/drivers/net/epic100.c
+++ b/drivers/net/epic100.c
@@ -1109,9 +1109,9 @@ static irqreturn_t epic_interrupt(int irq, void *dev_instance)
 
 	if ((status & EpicNapiEvent) && !ep->reschedule_in_poll) {
 		spin_lock(&ep->napi_lock);
-		if (netif_rx_schedule_prep(dev, &ep->napi)) {
+		if (netif_rx_schedule_prep(&ep->napi)) {
 			epic_napi_irq_off(dev, ep);
-			__netif_rx_schedule(dev, &ep->napi);
+			__netif_rx_schedule(&ep->napi);
 		} else
 			ep->reschedule_in_poll++;
 		spin_unlock(&ep->napi_lock);
@@ -1288,7 +1288,7 @@ rx_action:
 
 		more = ep->reschedule_in_poll;
 		if (!more) {
-			__netif_rx_complete(dev, napi);
+			__netif_rx_complete(napi);
 			outl(EpicNapiEvent, ioaddr + INTSTAT);
 			epic_napi_irq_on(dev, ep);
 		} else
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index 1f2b24743ee..9fbfa856ae5 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -1760,7 +1760,7 @@ static void nv_do_rx_refill(unsigned long data)
 	struct fe_priv *np = netdev_priv(dev);
 
 	/* Just reschedule NAPI rx processing */
-	netif_rx_schedule(dev, &np->napi);
+	netif_rx_schedule(&np->napi);
 }
 #else
 static void nv_do_rx_refill(unsigned long data)
@@ -3403,7 +3403,7 @@ static irqreturn_t nv_nic_irq(int foo, void *data)
 
 #ifdef CONFIG_FORCEDETH_NAPI
 		if (events & NVREG_IRQ_RX_ALL) {
-			netif_rx_schedule(dev, &np->napi);
+			netif_rx_schedule(&np->napi);
 
 			/* Disable furthur receive irq's */
 			spin_lock(&np->lock);
@@ -3520,7 +3520,7 @@ static irqreturn_t nv_nic_irq_optimized(int foo, void *data)
 
 #ifdef CONFIG_FORCEDETH_NAPI
 		if (events & NVREG_IRQ_RX_ALL) {
-			netif_rx_schedule(dev, &np->napi);
+			netif_rx_schedule(&np->napi);
 
 			/* Disable furthur receive irq's */
 			spin_lock(&np->lock);
@@ -3678,7 +3678,7 @@ static int nv_napi_poll(struct napi_struct *napi, int budget)
 		/* re-enable receive interrupts */
 		spin_lock_irqsave(&np->lock, flags);
 
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 
 		np->irqmask |= NVREG_IRQ_RX_ALL;
 		if (np->msi_flags & NV_MSI_X_ENABLED)
@@ -3704,7 +3704,7 @@ static irqreturn_t nv_nic_irq_rx(int foo, void *data)
 	writel(NVREG_IRQ_RX_ALL, base + NvRegMSIXIrqStatus);
 
 	if (events) {
-		netif_rx_schedule(dev, &np->napi);
+		netif_rx_schedule(&np->napi);
 		/* disable receive interrupts on the nic */
 		writel(NVREG_IRQ_RX_ALL, base + NvRegIrqMask);
 		pci_push(base);
diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c
index df66d620b11..4e6a9195fe5 100644
--- a/drivers/net/fs_enet/fs_enet-main.c
+++ b/drivers/net/fs_enet/fs_enet-main.c
@@ -209,7 +209,7 @@ static int fs_enet_rx_napi(struct napi_struct *napi, int budget)
 
 	if (received < budget) {
 		/* done */
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		(*fep->ops->napi_enable_rx)(dev);
 	}
 	return received;
@@ -478,7 +478,7 @@ fs_enet_interrupt(int irq, void *dev_id)
 				/* NOTE: it is possible for FCCs in NAPI mode    */
 				/* to submit a spurious interrupt while in poll  */
 				if (napi_ok)
-					__netif_rx_schedule(dev, &fep->napi);
+					__netif_rx_schedule(&fep->napi);
 			}
 		}
 
diff --git a/drivers/net/gianfar.c b/drivers/net/gianfar.c
index 13f49643ba0..c672ecfc959 100644
--- a/drivers/net/gianfar.c
+++ b/drivers/net/gianfar.c
@@ -1607,9 +1607,9 @@ static int gfar_clean_tx_ring(struct net_device *dev)
 static void gfar_schedule_cleanup(struct net_device *dev)
 {
 	struct gfar_private *priv = netdev_priv(dev);
-	if (netif_rx_schedule_prep(dev, &priv->napi)) {
+	if (netif_rx_schedule_prep(&priv->napi)) {
 		gfar_write(&priv->regs->imask, IMASK_RTX_DISABLED);
-		__netif_rx_schedule(dev, &priv->napi);
+		__netif_rx_schedule(&priv->napi);
 	}
 }
 
@@ -1863,7 +1863,7 @@ static int gfar_poll(struct napi_struct *napi, int budget)
 		return budget;
 
 	if (rx_cleaned < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 		/* Clear the halt bit in RSTAT */
 		gfar_write(&priv->regs->rstat, RSTAT_CLEAR_RHALT);
diff --git a/drivers/net/ibmveth.c b/drivers/net/ibmveth.c
index 02ecfdb4df6..1f055a95508 100644
--- a/drivers/net/ibmveth.c
+++ b/drivers/net/ibmveth.c
@@ -1028,7 +1028,7 @@ static int ibmveth_poll(struct napi_struct *napi, int budget)
 
 		ibmveth_assert(lpar_rc == H_SUCCESS);
 
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 
 		if (ibmveth_rxq_pending_buffer(adapter) &&
 		    netif_rx_reschedule(netdev, napi)) {
@@ -1047,11 +1047,11 @@ static irqreturn_t ibmveth_interrupt(int irq, void *dev_instance)
 	struct ibmveth_adapter *adapter = netdev_priv(netdev);
 	unsigned long lpar_rc;
 
-	if (netif_rx_schedule_prep(netdev, &adapter->napi)) {
+	if (netif_rx_schedule_prep(&adapter->napi)) {
 		lpar_rc = h_vio_signal(adapter->vdev->unit_address,
 				       VIO_IRQ_DISABLE);
 		ibmveth_assert(lpar_rc == H_SUCCESS);
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	}
 	return IRQ_HANDLED;
 }
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index 25df7c93106..6a40d9486da 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -3347,8 +3347,8 @@ static irqreturn_t igb_msix_rx(int irq, void *data)
 
 	igb_write_itr(rx_ring);
 
-	if (netif_rx_schedule_prep(adapter->netdev, &rx_ring->napi))
-		__netif_rx_schedule(adapter->netdev, &rx_ring->napi);
+	if (netif_rx_schedule_prep(&rx_ring->napi))
+		__netif_rx_schedule(&rx_ring->napi);
 
 #ifdef CONFIG_IGB_DCA
 	if (adapter->flags & IGB_FLAG_DCA_ENABLED)
@@ -3500,7 +3500,7 @@ static irqreturn_t igb_intr_msi(int irq, void *data)
 			mod_timer(&adapter->watchdog_timer, jiffies + 1);
 	}
 
-	netif_rx_schedule(netdev, &adapter->rx_ring[0].napi);
+	netif_rx_schedule(&adapter->rx_ring[0].napi);
 
 	return IRQ_HANDLED;
 }
@@ -3538,7 +3538,7 @@ static irqreturn_t igb_intr(int irq, void *data)
 			mod_timer(&adapter->watchdog_timer, jiffies + 1);
 	}
 
-	netif_rx_schedule(netdev, &adapter->rx_ring[0].napi);
+	netif_rx_schedule(&adapter->rx_ring[0].napi);
 
 	return IRQ_HANDLED;
 }
@@ -3573,7 +3573,7 @@ static int igb_poll(struct napi_struct *napi, int budget)
 	    !netif_running(netdev)) {
 		if (adapter->itr_setting & 3)
 			igb_set_itr(adapter);
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		if (!test_bit(__IGB_DOWN, &adapter->state))
 			igb_irq_enable(adapter);
 		return 0;
@@ -3599,7 +3599,7 @@ static int igb_clean_rx_ring_msix(struct napi_struct *napi, int budget)
 
 	/* If not enough Rx work done, exit the polling mode */
 	if ((work_done == 0) || !netif_running(netdev)) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 
 		if (adapter->itr_setting & 3) {
 			if (adapter->num_rx_queues == 1)
diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
index 820a92cc7f6..679125b3bbc 100644
--- a/drivers/net/ixgb/ixgb_main.c
+++ b/drivers/net/ixgb/ixgb_main.c
@@ -1721,14 +1721,14 @@ ixgb_intr(int irq, void *data)
 		if (!test_bit(__IXGB_DOWN, &adapter->flags))
 			mod_timer(&adapter->watchdog_timer, jiffies);
 
-	if (netif_rx_schedule_prep(netdev, &adapter->napi)) {
+	if (netif_rx_schedule_prep(&adapter->napi)) {
 
 		/* Disable interrupts and register for poll. The flush
 		  of the posted write is intentionally left out.
 		*/
 
 		IXGB_WRITE_REG(&adapter->hw, IMC, ~0);
-		__netif_rx_schedule(netdev, &adapter->napi);
+		__netif_rx_schedule(&adapter->napi);
 	}
 	return IRQ_HANDLED;
 }
@@ -1750,7 +1750,7 @@ ixgb_clean(struct napi_struct *napi, int budget)
 
 	/* If budget not fully consumed, exit the polling mode */
 	if (work_done < budget) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		if (!test_bit(__IXGB_DOWN, &adapter->flags))
 			ixgb_irq_enable(adapter);
 	}
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 92b35cfc7a4..b6ae9f674ba 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -1012,7 +1012,7 @@ static irqreturn_t ixgbe_msix_clean_rx(int irq, void *data)
 	rx_ring = &(adapter->rx_ring[r_idx]);
 	/* disable interrupts on this vector only */
 	IXGBE_WRITE_REG(&adapter->hw, IXGBE_EIMC, rx_ring->v_idx);
-	netif_rx_schedule(adapter->netdev, &q_vector->napi);
+	netif_rx_schedule(&q_vector->napi);
 
 	return IRQ_HANDLED;
 }
@@ -1053,7 +1053,7 @@ static int ixgbe_clean_rxonly(struct napi_struct *napi, int budget)
 
 	/* If all Rx work done, exit the polling mode */
 	if (work_done < budget) {
-		netif_rx_complete(adapter->netdev, napi);
+		netif_rx_complete(napi);
 		if (adapter->itr_setting & 3)
 			ixgbe_set_itr_msix(q_vector);
 		if (!test_bit(__IXGBE_DOWN, &adapter->state))
@@ -1102,7 +1102,7 @@ static int ixgbe_clean_rxonly_many(struct napi_struct *napi, int budget)
 	rx_ring = &(adapter->rx_ring[r_idx]);
 	/* If all Rx work done, exit the polling mode */
 	if (work_done < budget) {
-		netif_rx_complete(adapter->netdev, napi);
+		netif_rx_complete(napi);
 		if (adapter->itr_setting & 3)
 			ixgbe_set_itr_msix(q_vector);
 		if (!test_bit(__IXGBE_DOWN, &adapter->state))
@@ -1378,13 +1378,13 @@ static irqreturn_t ixgbe_intr(int irq, void *data)
 
 	ixgbe_check_fan_failure(adapter, eicr);
 
-	if (netif_rx_schedule_prep(netdev, &adapter->q_vector[0].napi)) {
+	if (netif_rx_schedule_prep(&adapter->q_vector[0].napi)) {
 		adapter->tx_ring[0].total_packets = 0;
 		adapter->tx_ring[0].total_bytes = 0;
 		adapter->rx_ring[0].total_packets = 0;
 		adapter->rx_ring[0].total_bytes = 0;
 		/* would disable interrupts here but EIAM disabled it */
-		__netif_rx_schedule(netdev, &adapter->q_vector[0].napi);
+		__netif_rx_schedule(&adapter->q_vector[0].napi);
 	}
 
 	return IRQ_HANDLED;
@@ -2308,7 +2308,7 @@ static int ixgbe_poll(struct napi_struct *napi, int budget)
 
 	/* If budget not fully consumed, exit the polling mode */
 	if (work_done < budget) {
-		netif_rx_complete(adapter->netdev, napi);
+		netif_rx_complete(napi);
 		if (adapter->itr_setting & 3)
 			ixgbe_set_itr(adapter);
 		if (!test_bit(__IXGBE_DOWN, &adapter->state))
diff --git a/drivers/net/ixp2000/ixpdev.c b/drivers/net/ixp2000/ixpdev.c
index bd96dbc8e02..01474572056 100644
--- a/drivers/net/ixp2000/ixpdev.c
+++ b/drivers/net/ixp2000/ixpdev.c
@@ -141,7 +141,7 @@ static int ixpdev_poll(struct napi_struct *napi, int budget)
 			break;
 	} while (ixp2000_reg_read(IXP2000_IRQ_THD_RAW_STATUS_A_0) & 0x00ff);
 
-	netif_rx_complete(dev, napi);
+	netif_rx_complete(napi);
 	ixp2000_reg_write(IXP2000_IRQ_THD_ENABLE_SET_A_0, 0x00ff);
 
 	return rx;
@@ -204,7 +204,7 @@ static irqreturn_t ixpdev_interrupt(int irq, void *dev_id)
 
 		ixp2000_reg_wrb(IXP2000_IRQ_THD_ENABLE_CLEAR_A_0, 0x00ff);
 		if (likely(napi_schedule_prep(&ip->napi))) {
-			__netif_rx_schedule(dev, &ip->napi);
+			__netif_rx_schedule(&ip->napi);
 		} else {
 			printk(KERN_CRIT "ixp2000: irq while polling!!\n");
 		}
diff --git a/drivers/net/jme.c b/drivers/net/jme.c
index 15035cb1738..08b34051c64 100644
--- a/drivers/net/jme.c
+++ b/drivers/net/jme.c
@@ -1250,7 +1250,6 @@ static int
 jme_poll(JME_NAPI_HOLDER(holder), JME_NAPI_WEIGHT(budget))
 {
 	struct jme_adapter *jme = jme_napi_priv(holder);
-	struct net_device *netdev = jme->dev;
 	int rest;
 
 	rest = jme_process_receive(jme, JME_NAPI_WEIGHT_VAL(budget));
diff --git a/drivers/net/jme.h b/drivers/net/jme.h
index adaf3ddbf78..2d6f30e638f 100644
--- a/drivers/net/jme.h
+++ b/drivers/net/jme.h
@@ -398,15 +398,15 @@ struct jme_ring {
 #define JME_NAPI_WEIGHT(w) int w
 #define JME_NAPI_WEIGHT_VAL(w) w
 #define JME_NAPI_WEIGHT_SET(w, r)
-#define JME_RX_COMPLETE(dev, napis) netif_rx_complete(dev, napis)
+#define JME_RX_COMPLETE(dev, napis) netif_rx_complete(napis)
 #define JME_NAPI_ENABLE(priv) napi_enable(&priv->napi);
 #define JME_NAPI_DISABLE(priv) \
 	if (!napi_disable_pending(&priv->napi)) \
 		napi_disable(&priv->napi);
 #define JME_RX_SCHEDULE_PREP(priv) \
-	netif_rx_schedule_prep(priv->dev, &priv->napi)
+	netif_rx_schedule_prep(&priv->napi)
 #define JME_RX_SCHEDULE(priv) \
-	__netif_rx_schedule(priv->dev, &priv->napi);
+	__netif_rx_schedule(&priv->napi);
 
 /*
  * Jmac Adapter Private data
diff --git a/drivers/net/korina.c b/drivers/net/korina.c
index 63626953f07..4a5580c1126 100644
--- a/drivers/net/korina.c
+++ b/drivers/net/korina.c
@@ -327,7 +327,7 @@ static irqreturn_t korina_rx_dma_interrupt(int irq, void *dev_id)
 
 	dmas = readl(&lp->rx_dma_regs->dmas);
 	if (dmas & (DMA_STAT_DONE | DMA_STAT_HALT | DMA_STAT_ERR)) {
-		netif_rx_schedule_prep(dev, &lp->napi);
+		netif_rx_schedule_prep(&lp->napi);
 
 		dmasm = readl(&lp->rx_dma_regs->dmasm);
 		writel(dmasm | (DMA_STAT_DONE |
@@ -466,7 +466,7 @@ static int korina_poll(struct napi_struct *napi, int budget)
 
 	work_done = korina_rx(dev, budget);
 	if (work_done < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 		writel(readl(&lp->rx_dma_regs->dmasm) &
 			~(DMA_STAT_DONE | DMA_STAT_HALT | DMA_STAT_ERR),
diff --git a/drivers/net/macb.c b/drivers/net/macb.c
index 261b9507124..a04da4ecaa8 100644
--- a/drivers/net/macb.c
+++ b/drivers/net/macb.c
@@ -519,7 +519,7 @@ static int macb_poll(struct napi_struct *napi, int budget)
 		 * this function was called last time, and no packets
 		 * have been received since.
 		 */
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		goto out;
 	}
 
@@ -530,13 +530,13 @@ static int macb_poll(struct napi_struct *napi, int budget)
 		dev_warn(&bp->pdev->dev,
 			 "No RX buffers complete, status = %02lx\n",
 			 (unsigned long)status);
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		goto out;
 	}
 
 	work_done = macb_rx(bp, budget);
 	if (work_done < budget)
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 	/*
 	 * We've done what we can to clean the buffers. Make sure we
@@ -571,7 +571,7 @@ static irqreturn_t macb_interrupt(int irq, void *dev_id)
 		}
 
 		if (status & MACB_RX_INT_FLAGS) {
-			if (netif_rx_schedule_prep(dev, &bp->napi)) {
+			if (netif_rx_schedule_prep(&bp->napi)) {
 				/*
 				 * There's no point taking any more interrupts
 				 * until we have processed the buffers
@@ -579,7 +579,7 @@ static irqreturn_t macb_interrupt(int irq, void *dev_id)
 				macb_writel(bp, IDR, MACB_RX_INT_FLAGS);
 				dev_dbg(&bp->pdev->dev,
 					"scheduling RX softirq\n");
-				__netif_rx_schedule(dev, &bp->napi);
+				__netif_rx_schedule(&bp->napi);
 			}
 		}
 
diff --git a/drivers/net/mlx4/en_rx.c b/drivers/net/mlx4/en_rx.c
index ffe28089b68..c61b0bdca1a 100644
--- a/drivers/net/mlx4/en_rx.c
+++ b/drivers/net/mlx4/en_rx.c
@@ -814,7 +814,7 @@ void mlx4_en_rx_irq(struct mlx4_cq *mcq)
 	struct mlx4_en_priv *priv = netdev_priv(cq->dev);
 
 	if (priv->port_up)
-		netif_rx_schedule(cq->dev, &cq->napi);
+		netif_rx_schedule(&cq->napi);
 	else
 		mlx4_en_arm_cq(priv, cq);
 }
@@ -834,7 +834,7 @@ int mlx4_en_poll_rx_cq(struct napi_struct *napi, int budget)
 		INC_PERF_COUNTER(priv->pstats.napi_quota);
 	else {
 		/* Done for now */
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		mlx4_en_arm_cq(priv, cq);
 	}
 	return done;
diff --git a/drivers/net/myri10ge/myri10ge.c b/drivers/net/myri10ge/myri10ge.c
index f017c774e1a..378c89e6c7d 100644
--- a/drivers/net/myri10ge/myri10ge.c
+++ b/drivers/net/myri10ge/myri10ge.c
@@ -1515,7 +1515,7 @@ static int myri10ge_poll(struct napi_struct *napi, int budget)
 	work_done = myri10ge_clean_rx_done(ss, budget);
 
 	if (work_done < budget) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		put_be32(htonl(3), ss->irq_claim);
 	}
 	return work_done;
@@ -1533,7 +1533,7 @@ static irqreturn_t myri10ge_intr(int irq, void *arg)
 	/* an interrupt on a non-zero receive-only slice is implicitly
 	 * valid  since MSI-X irqs are not shared */
 	if ((mgp->dev->real_num_tx_queues == 1) && (ss != mgp->ss)) {
-		netif_rx_schedule(ss->dev, &ss->napi);
+		netif_rx_schedule(&ss->napi);
 		return (IRQ_HANDLED);
 	}
 
@@ -1544,7 +1544,7 @@ static irqreturn_t myri10ge_intr(int irq, void *arg)
 	/* low bit indicates receives are present, so schedule
 	 * napi poll handler */
 	if (stats->valid & 1)
-		netif_rx_schedule(ss->dev, &ss->napi);
+		netif_rx_schedule(&ss->napi);
 
 	if (!mgp->msi_enabled && !mgp->msix_enabled) {
 		put_be32(0, mgp->irq_deassert);
diff --git a/drivers/net/natsemi.c b/drivers/net/natsemi.c
index 9f81fcb9688..478edb92bca 100644
--- a/drivers/net/natsemi.c
+++ b/drivers/net/natsemi.c
@@ -2193,10 +2193,10 @@ static irqreturn_t intr_handler(int irq, void *dev_instance)
 
 	prefetch(&np->rx_skbuff[np->cur_rx % RX_RING_SIZE]);
 
-	if (netif_rx_schedule_prep(dev, &np->napi)) {
+	if (netif_rx_schedule_prep(&np->napi)) {
 		/* Disable interrupts and register for poll */
 		natsemi_irq_disable(dev);
-		__netif_rx_schedule(dev, &np->napi);
+		__netif_rx_schedule(&np->napi);
 	} else
 		printk(KERN_WARNING
 	       	       "%s: Ignoring interrupt, status %#08x, mask %#08x.\n",
@@ -2248,7 +2248,7 @@ static int natsemi_poll(struct napi_struct *napi, int budget)
 		np->intr_status = readl(ioaddr + IntrStatus);
 	} while (np->intr_status);
 
-	netif_rx_complete(dev, napi);
+	netif_rx_complete(napi);
 
 	/* Reenable interrupts providing nothing is trying to shut
 	 * the chip down. */
diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c
index 6876bfd4455..ba01524b553 100644
--- a/drivers/net/netxen/netxen_nic_main.c
+++ b/drivers/net/netxen/netxen_nic_main.c
@@ -1583,7 +1583,7 @@ static int netxen_nic_poll(struct napi_struct *napi, int budget)
 	}
 
 	if ((work_done < budget) && tx_complete) {
-		netif_rx_complete(adapter->netdev, &adapter->napi);
+		netif_rx_complete(&adapter->napi);
 		netxen_nic_enable_int(adapter);
 	}
 
diff --git a/drivers/net/niu.c b/drivers/net/niu.c
index f219f16ec97..5698c155bbf 100644
--- a/drivers/net/niu.c
+++ b/drivers/net/niu.c
@@ -3669,7 +3669,7 @@ static int niu_poll(struct napi_struct *napi, int budget)
 	work_done = niu_poll_core(np, lp, budget);
 
 	if (work_done < budget) {
-		netif_rx_complete(np->dev, napi);
+		netif_rx_complete(napi);
 		niu_ldg_rearm(np, lp, 1);
 	}
 	return work_done;
@@ -4088,12 +4088,12 @@ static void __niu_fastpath_interrupt(struct niu *np, int ldg, u64 v0)
 static void niu_schedule_napi(struct niu *np, struct niu_ldg *lp,
 			      u64 v0, u64 v1, u64 v2)
 {
-	if (likely(netif_rx_schedule_prep(np->dev, &lp->napi))) {
+	if (likely(netif_rx_schedule_prep(&lp->napi))) {
 		lp->v0 = v0;
 		lp->v1 = v1;
 		lp->v2 = v2;
 		__niu_fastpath_interrupt(np, lp->ldg_num, v0);
-		__netif_rx_schedule(np->dev, &lp->napi);
+		__netif_rx_schedule(&lp->napi);
 	}
 }
 
diff --git a/drivers/net/pasemi_mac.c b/drivers/net/pasemi_mac.c
index fcbf6ccd0a8..dcd19904561 100644
--- a/drivers/net/pasemi_mac.c
+++ b/drivers/net/pasemi_mac.c
@@ -971,7 +971,7 @@ static irqreturn_t pasemi_mac_rx_intr(int irq, void *data)
 	if (*chan->status & PAS_STATUS_ERROR)
 		reg |= PAS_IOB_DMA_RXCH_RESET_DINTC;
 
-	netif_rx_schedule(dev, &mac->napi);
+	netif_rx_schedule(&mac->napi);
 
 	write_iob_reg(PAS_IOB_DMA_RXCH_RESET(chan->chno), reg);
 
@@ -1011,7 +1011,7 @@ static irqreturn_t pasemi_mac_tx_intr(int irq, void *data)
 
 	mod_timer(&txring->clean_timer, jiffies + (TX_CLEAN_INTERVAL)*2);
 
-	netif_rx_schedule(mac->netdev, &mac->napi);
+	netif_rx_schedule(&mac->napi);
 
 	if (reg)
 		write_iob_reg(PAS_IOB_DMA_TXCH_RESET(chan->chno), reg);
@@ -1641,7 +1641,7 @@ static int pasemi_mac_poll(struct napi_struct *napi, int budget)
 	pkts = pasemi_mac_clean_rx(rx_ring(mac), budget);
 	if (pkts < budget) {
 		/* all done, no more packets present */
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 		pasemi_mac_restart_rx_intr(mac);
 		pasemi_mac_restart_tx_intr(mac);
diff --git a/drivers/net/pcnet32.c b/drivers/net/pcnet32.c
index f2b192c80e1..044b7b07f5f 100644
--- a/drivers/net/pcnet32.c
+++ b/drivers/net/pcnet32.c
@@ -1397,7 +1397,7 @@ static int pcnet32_poll(struct napi_struct *napi, int budget)
 	if (work_done < budget) {
 		spin_lock_irqsave(&lp->lock, flags);
 
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 
 		/* clear interrupt masks */
 		val = lp->a.read_csr(ioaddr, CSR3);
@@ -2586,14 +2586,14 @@ pcnet32_interrupt(int irq, void *dev_id)
 				       dev->name, csr0);
 			/* unlike for the lance, there is no restart needed */
 		}
-		if (netif_rx_schedule_prep(dev, &lp->napi)) {
+		if (netif_rx_schedule_prep(&lp->napi)) {
 			u16 val;
 			/* set interrupt masks */
 			val = lp->a.read_csr(ioaddr, CSR3);
 			val |= 0x5f00;
 			lp->a.write_csr(ioaddr, CSR3, val);
 			mmiowb();
-			__netif_rx_schedule(dev, &lp->napi);
+			__netif_rx_schedule(&lp->napi);
 			break;
 		}
 		csr0 = lp->a.read_csr(ioaddr, CSR0);
diff --git a/drivers/net/qla3xxx.c b/drivers/net/qla3xxx.c
index 6b7ed1a5b3b..33e8e62b450 100644
--- a/drivers/net/qla3xxx.c
+++ b/drivers/net/qla3xxx.c
@@ -2293,7 +2293,7 @@ static int ql_poll(struct napi_struct *napi, int budget)
 
 	if (tx_cleaned + rx_cleaned != budget) {
 		spin_lock_irqsave(&qdev->hw_lock, hw_flags);
-		__netif_rx_complete(ndev, napi);
+		__netif_rx_complete(napi);
 		ql_update_small_bufq_prod_index(qdev);
 		ql_update_lrg_bufq_prod_index(qdev);
 		writel(qdev->rsp_consumer_index,
@@ -2352,8 +2352,8 @@ static irqreturn_t ql3xxx_isr(int irq, void *dev_id)
 		spin_unlock(&qdev->adapter_lock);
 	} else if (value & ISP_IMR_DISABLE_CMPL_INT) {
 		ql_disable_interrupts(qdev);
-		if (likely(netif_rx_schedule_prep(ndev, &qdev->napi))) {
-			__netif_rx_schedule(ndev, &qdev->napi);
+		if (likely(netif_rx_schedule_prep(&qdev->napi))) {
+			__netif_rx_schedule(&qdev->napi);
 		}
 	} else {
 		return IRQ_NONE;
diff --git a/drivers/net/qlge/qlge_main.c b/drivers/net/qlge/qlge_main.c
index 225930fda5a..02147082786 100644
--- a/drivers/net/qlge/qlge_main.c
+++ b/drivers/net/qlge/qlge_main.c
@@ -1647,7 +1647,7 @@ static int ql_napi_poll_msix(struct napi_struct *napi, int budget)
 		rx_ring->cq_id);
 
 	if (work_done < budget) {
-		__netif_rx_complete(qdev->ndev, napi);
+		__netif_rx_complete(napi);
 		ql_enable_completion_interrupt(qdev, rx_ring->irq);
 	}
 	return work_done;
@@ -1733,7 +1733,7 @@ static irqreturn_t qlge_msix_rx_isr(int irq, void *dev_id)
 {
 	struct rx_ring *rx_ring = dev_id;
 	struct ql_adapter *qdev = rx_ring->qdev;
-	netif_rx_schedule(qdev->ndev, &rx_ring->napi);
+	netif_rx_schedule(&rx_ring->napi);
 	return IRQ_HANDLED;
 }
 
@@ -1819,8 +1819,7 @@ static irqreturn_t qlge_isr(int irq, void *dev_id)
 							      &rx_ring->rx_work,
 							      0);
 				else
-					netif_rx_schedule(qdev->ndev,
-							  &rx_ring->napi);
+					netif_rx_schedule(&rx_ring->napi);
 				work_done++;
 			}
 		}
diff --git a/drivers/net/r6040.c b/drivers/net/r6040.c
index aff1cc627c0..53bbddfc8c9 100644
--- a/drivers/net/r6040.c
+++ b/drivers/net/r6040.c
@@ -667,7 +667,7 @@ static int r6040_poll(struct napi_struct *napi, int budget)
 	work_done = r6040_rx(dev, budget);
 
 	if (work_done < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		/* Enable RX interrupt */
 		iowrite16(ioread16(ioaddr + MIER) | RX_INTS, ioaddr + MIER);
 	}
@@ -704,7 +704,7 @@ static irqreturn_t r6040_interrupt(int irq, void *dev_id)
 
 		/* Mask off RX interrupt */
 		misr &= ~RX_INTS;
-		netif_rx_schedule(dev, &lp->napi);
+		netif_rx_schedule(&lp->napi);
 	}
 
 	/* TX interrupt request */
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index dddf6aeff49..2c73ca606b3 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -3581,8 +3581,8 @@ static irqreturn_t rtl8169_interrupt(int irq, void *dev_instance)
 		RTL_W16(IntrMask, tp->intr_event & ~tp->napi_event);
 		tp->intr_mask = ~tp->napi_event;
 
-		if (likely(netif_rx_schedule_prep(dev, &tp->napi)))
-			__netif_rx_schedule(dev, &tp->napi);
+		if (likely(netif_rx_schedule_prep(&tp->napi)))
+			__netif_rx_schedule(&tp->napi);
 		else if (netif_msg_intr(tp)) {
 			printk(KERN_INFO "%s: interrupt %04x in poll\n",
 			       dev->name, status);
@@ -3603,7 +3603,7 @@ static int rtl8169_poll(struct napi_struct *napi, int budget)
 	rtl8169_tx_interrupt(dev, tp, ioaddr);
 
 	if (work_done < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		tp->intr_mask = 0xffff;
 		/*
 		 * 20040426: the barrier is not strictly required but the
diff --git a/drivers/net/s2io.c b/drivers/net/s2io.c
index 1b489df80fa..512861923c6 100644
--- a/drivers/net/s2io.c
+++ b/drivers/net/s2io.c
@@ -2852,7 +2852,7 @@ static int s2io_poll_msix(struct napi_struct *napi, int budget)
 	s2io_chk_rx_buffers(nic, ring);
 
 	if (pkts_processed < budget_org) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		/*Re Enable MSI-Rx Vector*/
 		addr = (u8 __iomem *)&bar0->xmsi_mask_reg;
 		addr += 7 - ring->ring_no;
@@ -2890,7 +2890,7 @@ static int s2io_poll_inta(struct napi_struct *napi, int budget)
 			break;
 	}
 	if (pkts_processed < budget_org) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		/* Re enable the Rx interrupts for the ring */
 		writeq(0, &bar0->rx_traffic_mask);
 		readl(&bar0->rx_traffic_mask);
@@ -4344,7 +4344,7 @@ static irqreturn_t s2io_msix_ring_handle(int irq, void *dev_id)
 		val8 = (ring->ring_no == 0) ? 0x7f : 0xff;
 		writeb(val8, addr);
 		val8 = readb(addr);
-		netif_rx_schedule(dev, &ring->napi);
+		netif_rx_schedule(&ring->napi);
 	} else {
 		rx_intr_handler(ring, 0);
 		s2io_chk_rx_buffers(sp, ring);
@@ -4791,7 +4791,7 @@ static irqreturn_t s2io_isr(int irq, void *dev_id)
 
 		if (config->napi) {
 			if (reason & GEN_INTR_RXTRAFFIC) {
-				netif_rx_schedule(dev, &sp->napi);
+				netif_rx_schedule(&sp->napi);
 				writeq(S2IO_MINUS_ONE, &bar0->rx_traffic_mask);
 				writeq(S2IO_MINUS_ONE, &bar0->rx_traffic_int);
 				readl(&bar0->rx_traffic_int);
diff --git a/drivers/net/sb1250-mac.c b/drivers/net/sb1250-mac.c
index 480caec1e02..31e38fae017 100644
--- a/drivers/net/sb1250-mac.c
+++ b/drivers/net/sb1250-mac.c
@@ -2039,9 +2039,9 @@ static irqreturn_t sbmac_intr(int irq,void *dev_instance)
 		sbdma_tx_process(sc,&(sc->sbm_txdma), 0);
 
 	if (isr & (M_MAC_INT_CHANNEL << S_MAC_RX_CH0)) {
-		if (netif_rx_schedule_prep(dev, &sc->napi)) {
+		if (netif_rx_schedule_prep(&sc->napi)) {
 			__raw_writeq(0, sc->sbm_imr);
-			__netif_rx_schedule(dev, &sc->napi);
+			__netif_rx_schedule(&sc->napi);
 			/* Depend on the exit from poll to reenable intr */
 		}
 		else {
@@ -2667,7 +2667,7 @@ static int sbmac_poll(struct napi_struct *napi, int budget)
 	sbdma_tx_process(sc, &(sc->sbm_txdma), 1);
 
 	if (work_done < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 #ifdef CONFIG_SBMAC_COALESCE
 		__raw_writeq(((M_MAC_INT_EOP_COUNT | M_MAC_INT_EOP_TIMER) << S_MAC_TX_CH0) |
diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c
index 086629c0fe5..42934ba2030 100644
--- a/drivers/net/sfc/efx.c
+++ b/drivers/net/sfc/efx.c
@@ -230,7 +230,7 @@ static int efx_poll(struct napi_struct *napi, int budget)
 		 * since efx_channel_processed() will have no effect if
 		 * interrupts have already been disabled.
 		 */
-		netif_rx_complete(napi_dev, napi);
+		netif_rx_complete(napi);
 		efx_channel_processed(channel);
 	}
 
diff --git a/drivers/net/sfc/efx.h b/drivers/net/sfc/efx.h
index dd0d45b9e71..0dd7a532c78 100644
--- a/drivers/net/sfc/efx.h
+++ b/drivers/net/sfc/efx.h
@@ -77,7 +77,7 @@ static inline void efx_schedule_channel(struct efx_channel *channel)
 		  channel->channel, raw_smp_processor_id());
 	channel->work_pending = true;
 
-	netif_rx_schedule(channel->napi_dev, &channel->napi_str);
+	netif_rx_schedule(&channel->napi_str);
 }
 
 #endif /* EFX_EFX_H */
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
index f73ee797400..c9dbb06f8c9 100644
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -3214,7 +3214,7 @@ static int skge_poll(struct napi_struct *napi, int to_do)
 		unsigned long flags;
 
 		spin_lock_irqsave(&hw->hw_lock, flags);
-		__netif_rx_complete(dev, napi);
+		__netif_rx_complete(napi);
 		hw->intr_mask |= napimask[skge->port];
 		skge_write32(hw, B0_IMSK, hw->intr_mask);
 		skge_read32(hw, B0_IMSK);
@@ -3377,7 +3377,7 @@ static irqreturn_t skge_intr(int irq, void *dev_id)
 	if (status & (IS_XA1_F|IS_R1_F)) {
 		struct skge_port *skge = netdev_priv(hw->dev[0]);
 		hw->intr_mask &= ~(IS_XA1_F|IS_R1_F);
-		netif_rx_schedule(hw->dev[0], &skge->napi);
+		netif_rx_schedule(&skge->napi);
 	}
 
 	if (status & IS_PA_TO_TX1)
@@ -3397,7 +3397,7 @@ static irqreturn_t skge_intr(int irq, void *dev_id)
 
 		if (status & (IS_XA2_F|IS_R2_F)) {
 			hw->intr_mask &= ~(IS_XA2_F|IS_R2_F);
-			netif_rx_schedule(hw->dev[1], &skge->napi);
+			netif_rx_schedule(&skge->napi);
 		}
 
 		if (status & IS_PA_TO_RX2) {
diff --git a/drivers/net/smsc911x.c b/drivers/net/smsc911x.c
index fa28542b47d..ecdde03d416 100644
--- a/drivers/net/smsc911x.c
+++ b/drivers/net/smsc911x.c
@@ -984,7 +984,7 @@ static int smsc911x_poll(struct napi_struct *napi, int budget)
 			/* We processed all packets available.  Tell NAPI it can
 			 * stop polling then re-enable rx interrupts */
 			smsc911x_reg_write(pdata, INT_STS, INT_STS_RSFL_);
-			netif_rx_complete(dev, napi);
+			netif_rx_complete(napi);
 			temp = smsc911x_reg_read(pdata, INT_EN);
 			temp |= INT_EN_RSFL_EN_;
 			smsc911x_reg_write(pdata, INT_EN, temp);
diff --git a/drivers/net/smsc9420.c b/drivers/net/smsc9420.c
index 940220f6092..27e017d9696 100644
--- a/drivers/net/smsc9420.c
+++ b/drivers/net/smsc9420.c
@@ -666,7 +666,7 @@ static irqreturn_t smsc9420_isr(int irq, void *dev_id)
 			smsc9420_pci_flush_write(pd);
 
 			ints_to_clear |= (DMAC_STS_RX_ | DMAC_STS_NIS_);
-			netif_rx_schedule(pd->dev, &pd->napi);
+			netif_rx_schedule(&pd->napi);
 		}
 
 		if (ints_to_clear)
@@ -889,7 +889,7 @@ static int smsc9420_rx_poll(struct napi_struct *napi, int budget)
 	smsc9420_pci_flush_write(pd);
 
 	if (work_done < budget) {
-		netif_rx_complete(dev, &pd->napi);
+		netif_rx_complete(&pd->napi);
 
 		/* re-enable RX DMA interrupts */
 		dma_intr_ena = smsc9420_reg_read(pd, DMAC_INTR_ENA);
diff --git a/drivers/net/spider_net.c b/drivers/net/spider_net.c
index 325fbc9612c..c5c123d3af5 100644
--- a/drivers/net/spider_net.c
+++ b/drivers/net/spider_net.c
@@ -1302,7 +1302,7 @@ static int spider_net_poll(struct napi_struct *napi, int budget)
 	/* if all packets are in the stack, enable interrupts and return 0 */
 	/* if not, return 1 */
 	if (packets_done < budget) {
-		netif_rx_complete(netdev, napi);
+		netif_rx_complete(napi);
 		spider_net_rx_irq_on(card);
 		card->ignore_rx_ramfull = 0;
 	}
@@ -1529,8 +1529,7 @@ spider_net_handle_error_irq(struct spider_net_card *card, u32 status_reg,
 			spider_net_refill_rx_chain(card);
 			spider_net_enable_rxdmac(card);
 			card->num_rx_ints ++;
-			netif_rx_schedule(card->netdev,
-					  &card->napi);
+			netif_rx_schedule(&card->napi);
 		}
 		show_error = 0;
 		break;
@@ -1550,8 +1549,7 @@ spider_net_handle_error_irq(struct spider_net_card *card, u32 status_reg,
 		spider_net_refill_rx_chain(card);
 		spider_net_enable_rxdmac(card);
 		card->num_rx_ints ++;
-		netif_rx_schedule(card->netdev,
-				  &card->napi);
+		netif_rx_schedule(&card->napi);
 		show_error = 0;
 		break;
 
@@ -1565,8 +1563,7 @@ spider_net_handle_error_irq(struct spider_net_card *card, u32 status_reg,
 		spider_net_refill_rx_chain(card);
 		spider_net_enable_rxdmac(card);
 		card->num_rx_ints ++;
-		netif_rx_schedule(card->netdev,
-				  &card->napi);
+		netif_rx_schedule(&card->napi);
 		show_error = 0;
 		break;
 
@@ -1660,11 +1657,11 @@ spider_net_interrupt(int irq, void *ptr)
 
 	if (status_reg & SPIDER_NET_RXINT ) {
 		spider_net_rx_irq_off(card);
-		netif_rx_schedule(netdev, &card->napi);
+		netif_rx_schedule(&card->napi);
 		card->num_rx_ints ++;
 	}
 	if (status_reg & SPIDER_NET_TXINT)
-		netif_rx_schedule(netdev, &card->napi);
+		netif_rx_schedule(&card->napi);
 
 	if (status_reg & SPIDER_NET_LINKINT)
 		spider_net_link_reset(netdev);
diff --git a/drivers/net/starfire.c b/drivers/net/starfire.c
index 0358809f409..d5b9dd842c6 100644
--- a/drivers/net/starfire.c
+++ b/drivers/net/starfire.c
@@ -1290,8 +1290,8 @@ static irqreturn_t intr_handler(int irq, void *dev_instance)
 		if (intr_status & (IntrRxDone | IntrRxEmpty)) {
 			u32 enable;
 
-			if (likely(netif_rx_schedule_prep(dev, &np->napi))) {
-				__netif_rx_schedule(dev, &np->napi);
+			if (likely(netif_rx_schedule_prep(&np->napi))) {
+				__netif_rx_schedule(&np->napi);
 				enable = readl(ioaddr + IntrEnable);
 				enable &= ~(IntrRxDone | IntrRxEmpty);
 				writel(enable, ioaddr + IntrEnable);
@@ -1530,7 +1530,7 @@ static int netdev_poll(struct napi_struct *napi, int budget)
 		intr_status = readl(ioaddr + IntrStatus);
 	} while (intr_status & (IntrRxDone | IntrRxEmpty));
 
-	netif_rx_complete(dev, napi);
+	netif_rx_complete(napi);
 	intr_status = readl(ioaddr + IntrEnable);
 	intr_status |= IntrRxDone | IntrRxEmpty;
 	writel(intr_status, ioaddr + IntrEnable);
diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c
index f4b0beec4d1..8a746041248 100644
--- a/drivers/net/sungem.c
+++ b/drivers/net/sungem.c
@@ -921,7 +921,7 @@ static int gem_poll(struct napi_struct *napi, int budget)
 		gp->status = readl(gp->regs + GREG_STAT);
 	} while (gp->status & GREG_STAT_NAPI);
 
-	__netif_rx_complete(dev, napi);
+	__netif_rx_complete(napi);
 	gem_enable_ints(gp);
 
 	spin_unlock_irqrestore(&gp->lock, flags);
@@ -944,7 +944,7 @@ static irqreturn_t gem_interrupt(int irq, void *dev_id)
 
 	spin_lock_irqsave(&gp->lock, flags);
 
-	if (netif_rx_schedule_prep(dev, &gp->napi)) {
+	if (netif_rx_schedule_prep(&gp->napi)) {
 		u32 gem_status = readl(gp->regs + GREG_STAT);
 
 		if (gem_status == 0) {
@@ -954,7 +954,7 @@ static irqreturn_t gem_interrupt(int irq, void *dev_id)
 		}
 		gp->status = gem_status;
 		gem_disable_ints(gp);
-		__netif_rx_schedule(dev, &gp->napi);
+		__netif_rx_schedule(&gp->napi);
 	}
 
 	spin_unlock_irqrestore(&gp->lock, flags);
diff --git a/drivers/net/tc35815.c b/drivers/net/tc35815.c
index 308f365270e..bcd0e60cbda 100644
--- a/drivers/net/tc35815.c
+++ b/drivers/net/tc35815.c
@@ -1609,8 +1609,8 @@ static irqreturn_t tc35815_interrupt(int irq, void *dev_id)
 	if (!(dmactl & DMA_IntMask)) {
 		/* disable interrupts */
 		tc_writel(dmactl | DMA_IntMask, &tr->DMA_Ctl);
-		if (netif_rx_schedule_prep(dev, &lp->napi))
-			__netif_rx_schedule(dev, &lp->napi);
+		if (netif_rx_schedule_prep(&lp->napi))
+			__netif_rx_schedule(&lp->napi);
 		else {
 			printk(KERN_ERR "%s: interrupt taken in poll\n",
 			       dev->name);
@@ -1919,7 +1919,7 @@ static int tc35815_poll(struct napi_struct *napi, int budget)
 	spin_unlock(&lp->lock);
 
 	if (received < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		/* enable interrupts */
 		tc_writel(tc_readl(&tr->DMA_Ctl) & ~DMA_IntMask, &tr->DMA_Ctl);
 	}
diff --git a/drivers/net/tehuti.c b/drivers/net/tehuti.c
index 5b83fbb0201..a10a83a11d9 100644
--- a/drivers/net/tehuti.c
+++ b/drivers/net/tehuti.c
@@ -265,8 +265,8 @@ static irqreturn_t bdx_isr_napi(int irq, void *dev)
 		bdx_isr_extra(priv, isr);
 
 	if (isr & (IR_RX_DESC_0 | IR_TX_FREE_0)) {
-		if (likely(netif_rx_schedule_prep(ndev, &priv->napi))) {
-			__netif_rx_schedule(ndev, &priv->napi);
+		if (likely(netif_rx_schedule_prep(&priv->napi))) {
+			__netif_rx_schedule(&priv->napi);
 			RET(IRQ_HANDLED);
 		} else {
 			/* NOTE: we get here if intr has slipped into window
@@ -289,7 +289,6 @@ static irqreturn_t bdx_isr_napi(int irq, void *dev)
 static int bdx_poll(struct napi_struct *napi, int budget)
 {
 	struct bdx_priv *priv = container_of(napi, struct bdx_priv, napi);
-	struct net_device *dev = priv->ndev;
 	int work_done;
 
 	ENTER;
@@ -303,7 +302,7 @@ static int bdx_poll(struct napi_struct *napi, int budget)
 		 * device lock and allow waiting tasks (eg rmmod) to advance) */
 		priv->napi_stop = 0;
 
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		bdx_enable_interrupts(priv);
 	}
 	return work_done;
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 7971d802508..04ae1e86aea 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -4451,7 +4451,7 @@ static int tg3_poll(struct napi_struct *napi, int budget)
 			sblk->status &= ~SD_STATUS_UPDATED;
 
 		if (likely(!tg3_has_work(tp))) {
-			netif_rx_complete(tp->dev, napi);
+			netif_rx_complete(napi);
 			tg3_restart_ints(tp);
 			break;
 		}
@@ -4461,7 +4461,7 @@ static int tg3_poll(struct napi_struct *napi, int budget)
 
 tx_recovery:
 	/* work_done is guaranteed to be less than budget. */
-	netif_rx_complete(tp->dev, napi);
+	netif_rx_complete(napi);
 	schedule_work(&tp->reset_task);
 	return work_done;
 }
@@ -4510,7 +4510,7 @@ static irqreturn_t tg3_msi_1shot(int irq, void *dev_id)
 	prefetch(&tp->rx_rcb[tp->rx_rcb_ptr]);
 
 	if (likely(!tg3_irq_sync(tp)))
-		netif_rx_schedule(dev, &tp->napi);
+		netif_rx_schedule(&tp->napi);
 
 	return IRQ_HANDLED;
 }
@@ -4535,7 +4535,7 @@ static irqreturn_t tg3_msi(int irq, void *dev_id)
 	 */
 	tw32_mailbox(MAILBOX_INTERRUPT_0 + TG3_64BIT_REG_LOW, 0x00000001);
 	if (likely(!tg3_irq_sync(tp)))
-		netif_rx_schedule(dev, &tp->napi);
+		netif_rx_schedule(&tp->napi);
 
 	return IRQ_RETVAL(1);
 }
@@ -4577,7 +4577,7 @@ static irqreturn_t tg3_interrupt(int irq, void *dev_id)
 	sblk->status &= ~SD_STATUS_UPDATED;
 	if (likely(tg3_has_work(tp))) {
 		prefetch(&tp->rx_rcb[tp->rx_rcb_ptr]);
-		netif_rx_schedule(dev, &tp->napi);
+		netif_rx_schedule(&tp->napi);
 	} else {
 		/* No work, shared interrupt perhaps?  re-enable
 		 * interrupts, and flush that PCI write
@@ -4623,7 +4623,7 @@ static irqreturn_t tg3_interrupt_tagged(int irq, void *dev_id)
 	tw32_mailbox_f(MAILBOX_INTERRUPT_0 + TG3_64BIT_REG_LOW, 0x00000001);
 	if (tg3_irq_sync(tp))
 		goto out;
-	if (netif_rx_schedule_prep(dev, &tp->napi)) {
+	if (netif_rx_schedule_prep(&tp->napi)) {
 		prefetch(&tp->rx_rcb[tp->rx_rcb_ptr]);
 		/* Update last_tag to mark that this status has been
 		 * seen. Because interrupt may be shared, we may be
@@ -4631,7 +4631,7 @@ static irqreturn_t tg3_interrupt_tagged(int irq, void *dev_id)
 		 * if tg3_poll() is not scheduled.
 		 */
 		tp->last_tag = sblk->status_tag;
-		__netif_rx_schedule(dev, &tp->napi);
+		__netif_rx_schedule(&tp->napi);
 	}
 out:
 	return IRQ_RETVAL(handled);
diff --git a/drivers/net/tsi108_eth.c b/drivers/net/tsi108_eth.c
index 271bc230c8a..75461dbd487 100644
--- a/drivers/net/tsi108_eth.c
+++ b/drivers/net/tsi108_eth.c
@@ -888,7 +888,7 @@ static int tsi108_poll(struct napi_struct *napi, int budget)
 
 	if (num_received < budget) {
 		data->rxpending = 0;
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 		TSI_WRITE(TSI108_EC_INTMASK,
 				     TSI_READ(TSI108_EC_INTMASK)
@@ -919,7 +919,7 @@ static void tsi108_rx_int(struct net_device *dev)
 	 * from tsi108_check_rxring().
 	 */
 
-	if (netif_rx_schedule_prep(dev, &data->napi)) {
+	if (netif_rx_schedule_prep(&data->napi)) {
 		/* Mask, rather than ack, the receive interrupts.  The ack
 		 * will happen in tsi108_poll().
 		 */
@@ -930,7 +930,7 @@ static void tsi108_rx_int(struct net_device *dev)
 				     | TSI108_INT_RXTHRESH |
 				     TSI108_INT_RXOVERRUN | TSI108_INT_RXERROR |
 				     TSI108_INT_RXWAIT);
-		__netif_rx_schedule(dev, &data->napi);
+		__netif_rx_schedule(&data->napi);
 	} else {
 		if (!netif_running(dev)) {
 			/* This can happen if an interrupt occurs while the
diff --git a/drivers/net/tulip/interrupt.c b/drivers/net/tulip/interrupt.c
index 739d610d18c..6c3428a37c0 100644
--- a/drivers/net/tulip/interrupt.c
+++ b/drivers/net/tulip/interrupt.c
@@ -103,7 +103,7 @@ void oom_timer(unsigned long data)
 {
         struct net_device *dev = (struct net_device *)data;
 	struct tulip_private *tp = netdev_priv(dev);
-	netif_rx_schedule(dev, &tp->napi);
+	netif_rx_schedule(&tp->napi);
 }
 
 int tulip_poll(struct napi_struct *napi, int budget)
@@ -300,7 +300,7 @@ int tulip_poll(struct napi_struct *napi, int budget)
 
          /* Remove us from polling list and enable RX intr. */
 
-         netif_rx_complete(dev, napi);
+         netif_rx_complete(napi);
          iowrite32(tulip_tbl[tp->chip_id].valid_intrs, tp->base_addr+CSR7);
 
          /* The last op happens after poll completion. Which means the following:
@@ -336,7 +336,7 @@ int tulip_poll(struct napi_struct *napi, int budget)
           * before we did netif_rx_complete(). See? We would lose it. */
 
          /* remove ourselves from the polling list */
-         netif_rx_complete(dev, napi);
+         netif_rx_complete(napi);
 
          return work_done;
 }
@@ -519,7 +519,7 @@ irqreturn_t tulip_interrupt(int irq, void *dev_instance)
 			rxd++;
 			/* Mask RX intrs and add the device to poll list. */
 			iowrite32(tulip_tbl[tp->chip_id].valid_intrs&~RxPollInt, ioaddr + CSR7);
-			netif_rx_schedule(dev, &tp->napi);
+			netif_rx_schedule(&tp->napi);
 
 			if (!(csr5&~(AbnormalIntr|NormalIntr|RxPollInt|TPLnkPass)))
                                break;
diff --git a/drivers/net/typhoon.c b/drivers/net/typhoon.c
index 5386d9b73e6..0009f4e3443 100644
--- a/drivers/net/typhoon.c
+++ b/drivers/net/typhoon.c
@@ -1755,7 +1755,6 @@ static int
 typhoon_poll(struct napi_struct *napi, int budget)
 {
 	struct typhoon *tp = container_of(napi, struct typhoon, napi);
-	struct net_device *dev = tp->dev;
 	struct typhoon_indexes *indexes = tp->indexes;
 	int work_done;
 
@@ -1784,7 +1783,7 @@ typhoon_poll(struct napi_struct *napi, int budget)
 	}
 
 	if (work_done < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		iowrite32(TYPHOON_INTR_NONE,
 				tp->ioaddr + TYPHOON_REG_INTR_MASK);
 		typhoon_post_pci_writes(tp->ioaddr);
@@ -1807,10 +1806,10 @@ typhoon_interrupt(int irq, void *dev_instance)
 
 	iowrite32(intr_status, ioaddr + TYPHOON_REG_INTR_STATUS);
 
-	if (netif_rx_schedule_prep(dev, &tp->napi)) {
+	if (netif_rx_schedule_prep(&tp->napi)) {
 		iowrite32(TYPHOON_INTR_ALL, ioaddr + TYPHOON_REG_INTR_MASK);
 		typhoon_post_pci_writes(ioaddr);
-		__netif_rx_schedule(dev, &tp->napi);
+		__netif_rx_schedule(&tp->napi);
 	} else {
 		printk(KERN_ERR "%s: Error, poll already scheduled\n",
                        dev->name);
diff --git a/drivers/net/ucc_geth.c b/drivers/net/ucc_geth.c
index 5c82f147f15..78a2ede19c5 100644
--- a/drivers/net/ucc_geth.c
+++ b/drivers/net/ucc_geth.c
@@ -3330,7 +3330,7 @@ static int ucc_geth_poll(struct napi_struct *napi, int budget)
 		struct ucc_fast_private *uccf;
 		u32 uccm;
 
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 		uccf = ugeth->uccf;
 		uccm = in_be32(uccf->p_uccm);
 		uccm |= UCCE_RX_EVENTS;
@@ -3364,10 +3364,10 @@ static irqreturn_t ucc_geth_irq_handler(int irq, void *info)
 
 	/* check for receive events that require processing */
 	if (ucce & UCCE_RX_EVENTS) {
-		if (netif_rx_schedule_prep(dev, &ugeth->napi)) {
+		if (netif_rx_schedule_prep(&ugeth->napi)) {
 			uccm &= ~UCCE_RX_EVENTS;
 			out_be32(uccf->p_uccm, uccm);
-			__netif_rx_schedule(dev, &ugeth->napi);
+			__netif_rx_schedule(&ugeth->napi);
 		}
 	}
 
diff --git a/drivers/net/via-rhine.c b/drivers/net/via-rhine.c
index 8d405c83df8..ac07cc6e3cb 100644
--- a/drivers/net/via-rhine.c
+++ b/drivers/net/via-rhine.c
@@ -589,7 +589,7 @@ static int rhine_napipoll(struct napi_struct *napi, int budget)
 	work_done = rhine_rx(dev, budget);
 
 	if (work_done < budget) {
-		netif_rx_complete(dev, napi);
+		netif_rx_complete(napi);
 
 		iowrite16(IntrRxDone | IntrRxErr | IntrRxEmpty| IntrRxOverflow |
 			  IntrRxDropped | IntrRxNoBuf | IntrTxAborted |
@@ -1318,7 +1318,7 @@ static irqreturn_t rhine_interrupt(int irq, void *dev_instance)
 				  IntrPCIErr | IntrStatsMax | IntrLinkChange,
 				  ioaddr + IntrEnable);
 
-			netif_rx_schedule(dev, &rp->napi);
+			netif_rx_schedule(&rp->napi);
 		}
 
 		if (intr_status & (IntrTxErrSummary | IntrTxDone)) {
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 71ca29cc184..b7004ff3645 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -374,9 +374,9 @@ static void skb_recv_done(struct virtqueue *rvq)
 {
 	struct virtnet_info *vi = rvq->vdev->priv;
 	/* Schedule NAPI, Suppress further interrupts if successful. */
-	if (netif_rx_schedule_prep(vi->dev, &vi->napi)) {
+	if (netif_rx_schedule_prep(&vi->napi)) {
 		rvq->vq_ops->disable_cb(rvq);
-		__netif_rx_schedule(vi->dev, &vi->napi);
+		__netif_rx_schedule(&vi->napi);
 	}
 }
 
@@ -402,11 +402,11 @@ again:
 
 	/* Out of packets? */
 	if (received < budget) {
-		netif_rx_complete(vi->dev, napi);
+		netif_rx_complete(napi);
 		if (unlikely(!vi->rvq->vq_ops->enable_cb(vi->rvq))
 		    && napi_schedule_prep(napi)) {
 			vi->rvq->vq_ops->disable_cb(vi->rvq);
-			__netif_rx_schedule(vi->dev, napi);
+			__netif_rx_schedule(napi);
 			goto again;
 		}
 	}
@@ -580,9 +580,9 @@ static int virtnet_open(struct net_device *dev)
 	 * won't get another interrupt, so process any outstanding packets
 	 * now.  virtnet_poll wants re-enable the queue, so we disable here.
 	 * We synchronize against interrupts via NAPI_STATE_SCHED */
-	if (netif_rx_schedule_prep(dev, &vi->napi)) {
+	if (netif_rx_schedule_prep(&vi->napi)) {
 		vi->rvq->vq_ops->disable_cb(vi->rvq);
-		__netif_rx_schedule(dev, &vi->napi);
+		__netif_rx_schedule(&vi->napi);
 	}
 	return 0;
 }
diff --git a/drivers/net/wan/hd64572.c b/drivers/net/wan/hd64572.c
index 0bcc0b5f22d..08b3536944f 100644
--- a/drivers/net/wan/hd64572.c
+++ b/drivers/net/wan/hd64572.c
@@ -341,7 +341,7 @@ static int sca_poll(struct napi_struct *napi, int budget)
 		received = sca_rx_done(port, budget);
 
 	if (received < budget) {
-		netif_rx_complete(port->netdev, napi);
+		netif_rx_complete(napi);
 		enable_intr(port);
 	}
 
@@ -359,7 +359,7 @@ static irqreturn_t sca_intr(int irq, void *dev_id)
 		if (port && (isr0 & (i ? 0x08002200 : 0x00080022))) {
 			handled = 1;
 			disable_intr(port);
-			netif_rx_schedule(port->netdev, &port->napi);
+			netif_rx_schedule(&port->napi);
 		}
 	}
 
diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
index fe376fde4e8..761635be910 100644
--- a/drivers/net/xen-netfront.c
+++ b/drivers/net/xen-netfront.c
@@ -196,7 +196,7 @@ static void rx_refill_timeout(unsigned long data)
 {
 	struct net_device *dev = (struct net_device *)data;
 	struct netfront_info *np = netdev_priv(dev);
-	netif_rx_schedule(dev, &np->napi);
+	netif_rx_schedule(&np->napi);
 }
 
 static int netfront_tx_slot_available(struct netfront_info *np)
@@ -328,7 +328,7 @@ static int xennet_open(struct net_device *dev)
 		xennet_alloc_rx_buffers(dev);
 		np->rx.sring->rsp_event = np->rx.rsp_cons + 1;
 		if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx))
-			netif_rx_schedule(dev, &np->napi);
+			netif_rx_schedule(&np->napi);
 	}
 	spin_unlock_bh(&np->rx_lock);
 
@@ -979,7 +979,7 @@ err:
 
 		RING_FINAL_CHECK_FOR_RESPONSES(&np->rx, more_to_do);
 		if (!more_to_do)
-			__netif_rx_complete(dev, napi);
+			__netif_rx_complete(napi);
 
 		local_irq_restore(flags);
 	}
@@ -1310,7 +1310,7 @@ static irqreturn_t xennet_interrupt(int irq, void *dev_id)
 		xennet_tx_buf_gc(dev);
 		/* Under tx_lock: protects access to rx shared-ring indexes. */
 		if (RING_HAS_UNCONSUMED_RESPONSES(&np->rx))
-			netif_rx_schedule(dev, &np->napi);
+			netif_rx_schedule(&np->napi);
 	}
 
 	spin_unlock_irqrestore(&np->tx_lock, flags);
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index 58856b6737f..41e1224651c 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1555,8 +1555,7 @@ static inline u32 netif_msg_init(int debug_value, int default_msg_enable_bits)
 }
 
 /* Test if receive needs to be scheduled but only if up */
-static inline int netif_rx_schedule_prep(struct net_device *dev,
-					 struct napi_struct *napi)
+static inline int netif_rx_schedule_prep(struct napi_struct *napi)
 {
 	return napi_schedule_prep(napi);
 }
@@ -1564,27 +1563,24 @@ static inline int netif_rx_schedule_prep(struct net_device *dev,
 /* Add interface to tail of rx poll list. This assumes that _prep has
  * already been called and returned 1.
  */
-static inline void __netif_rx_schedule(struct net_device *dev,
-				       struct napi_struct *napi)
+static inline void __netif_rx_schedule(struct napi_struct *napi)
 {
 	__napi_schedule(napi);
 }
 
 /* Try to reschedule poll. Called by irq handler. */
 
-static inline void netif_rx_schedule(struct net_device *dev,
-				     struct napi_struct *napi)
+static inline void netif_rx_schedule(struct napi_struct *napi)
 {
-	if (netif_rx_schedule_prep(dev, napi))
-		__netif_rx_schedule(dev, napi);
+	if (netif_rx_schedule_prep(napi))
+		__netif_rx_schedule(napi);
 }
 
 /* Try to reschedule poll. Called by dev->poll() after netif_rx_complete().  */
-static inline int netif_rx_reschedule(struct net_device *dev,
-				      struct napi_struct *napi)
+static inline int netif_rx_reschedule(struct napi_struct *napi)
 {
 	if (napi_schedule_prep(napi)) {
-		__netif_rx_schedule(dev, napi);
+		__netif_rx_schedule(napi);
 		return 1;
 	}
 	return 0;
@@ -1593,8 +1589,7 @@ static inline int netif_rx_reschedule(struct net_device *dev,
 /* same as netif_rx_complete, except that local_irq_save(flags)
  * has already been issued
  */
-static inline void __netif_rx_complete(struct net_device *dev,
-				       struct napi_struct *napi)
+static inline void __netif_rx_complete(struct napi_struct *napi)
 {
 	__napi_complete(napi);
 }
@@ -1604,8 +1599,7 @@ static inline void __netif_rx_complete(struct net_device *dev,
  * it completes the work. The device cannot be out of poll list at this
  * moment, it is BUG().
  */
-static inline void netif_rx_complete(struct net_device *dev,
-				     struct napi_struct *napi)
+static inline void netif_rx_complete(struct napi_struct *napi)
 {
 	napi_complete(napi);
 }
-- 
cgit v1.2.3-70-g09d2