[deliverable/linux.git] / drivers / net / macvlan.c

/*
 * Copyright (c) 2007 Patrick McHardy <kaber@trash.net>
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License as
 * published by the Free Software Foundation; either version 2 of
 * the License, or (at your option) any later version.
 *
 * The code this is based on carried the following copyright notice:
 * ---
 * (C) Copyright 2001-2006
 * Alex Zeffertt, Cambridge Broadband Ltd, ajz@cambridgebroadband.com
 * Re-worked by Ben Greear <greearb@candelatech.com>
 * ---
 */
#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/errno.h>
#include <linux/slab.h>
#include <linux/string.h>
#include <linux/rculist.h>
#include <linux/notifier.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/ethtool.h>
#include <linux/if_arp.h>
#include <linux/if_link.h>
#include <linux/if_macvlan.h>
#include <net/rtnetlink.h>

#define MACVLAN_HASH_SIZE	(1 << BITS_PER_BYTE)

struct macvlan_port {
	struct net_device	*dev;
	struct hlist_head	vlan_hash[MACVLAN_HASH_SIZE];
	struct list_head	vlans;
};

struct macvlan_dev {
	struct net_device	*dev;
	struct list_head	list;
	struct hlist_node	hlist;
	struct macvlan_port	*port;
	struct net_device	*lowerdev;
};


static struct macvlan_dev *macvlan_hash_lookup(const struct macvlan_port *port,
					       const unsigned char *addr)
{
	struct macvlan_dev *vlan;
	struct hlist_node *n;

	hlist_for_each_entry_rcu(vlan, n, &port->vlan_hash[addr[5]], hlist) {
		if (!compare_ether_addr_64bits(vlan->dev->dev_addr, addr))
			return vlan;
	}
	return NULL;
}

static void macvlan_hash_add(struct macvlan_dev *vlan)
{
	struct macvlan_port *port = vlan->port;
	const unsigned char *addr = vlan->dev->dev_addr;

	hlist_add_head_rcu(&vlan->hlist, &port->vlan_hash[addr[5]]);
}

static void macvlan_hash_del(struct macvlan_dev *vlan)
{
	hlist_del_rcu(&vlan->hlist);
	synchronize_rcu();
}

static void macvlan_hash_change_addr(struct macvlan_dev *vlan,
					const unsigned char *addr)
{
	macvlan_hash_del(vlan);
	/* Now that we are unhashed it is safe to change the device
	 * address without confusing packet delivery.
	 */
	memcpy(vlan->dev->dev_addr, addr, ETH_ALEN);
	macvlan_hash_add(vlan);
}

static int macvlan_addr_busy(const struct macvlan_port *port,
				const unsigned char *addr)
{
	/* Test to see if the specified multicast address is
	 * currently in use by the underlying device or
	 * another macvlan.
	 */
	if (!compare_ether_addr_64bits(port->dev->dev_addr, addr))
		return 1;

	if (macvlan_hash_lookup(port, addr))
		return 1;

	return 0;
}

static void macvlan_broadcast(struct sk_buff *skb,
			      const struct macvlan_port *port)
{
	const struct ethhdr *eth = eth_hdr(skb);
	const struct macvlan_dev *vlan;
	struct hlist_node *n;
	struct net_device *dev;
	struct sk_buff *nskb;
	unsigned int i;

	if (skb->protocol == htons(ETH_P_PAUSE))
		return;

	for (i = 0; i < MACVLAN_HASH_SIZE; i++) {
		hlist_for_each_entry_rcu(vlan, n, &port->vlan_hash[i], hlist) {
			dev = vlan->dev;

			nskb = skb_clone(skb, GFP_ATOMIC);
			if (nskb == NULL) {
				dev->stats.rx_errors++;
				dev->stats.rx_dropped++;
				continue;
			}

			dev->stats.rx_bytes += skb->len + ETH_HLEN;
			dev->stats.rx_packets++;
			dev->stats.multicast++;

			nskb->dev = dev;
			if (!compare_ether_addr_64bits(eth->h_dest, dev->broadcast))
				nskb->pkt_type = PACKET_BROADCAST;
			else
				nskb->pkt_type = PACKET_MULTICAST;

			netif_rx(nskb);
		}
	}
}

/* called under rcu_read_lock() from netif_receive_skb */
static struct sk_buff *macvlan_handle_frame(struct sk_buff *skb)
{
	const struct ethhdr *eth = eth_hdr(skb);
	const struct macvlan_port *port;
	const struct macvlan_dev *vlan;
	struct net_device *dev;

	port = rcu_dereference(skb->dev->macvlan_port);
	if (port == NULL)
		return skb;

	if (is_multicast_ether_addr(eth->h_dest)) {
		macvlan_broadcast(skb, port);
		return skb;
	}

	vlan = macvlan_hash_lookup(port, eth->h_dest);
	if (vlan == NULL)
		return skb;

	dev = vlan->dev;
	if (unlikely(!(dev->flags & IFF_UP))) {
		kfree_skb(skb);
		return NULL;
	}

	skb = skb_share_check(skb, GFP_ATOMIC);
	if (skb == NULL) {
		dev->stats.rx_errors++;
		dev->stats.rx_dropped++;
		return NULL;
	}

	dev->stats.rx_bytes += skb->len + ETH_HLEN;
	dev->stats.rx_packets++;

	skb->dev = dev;
	skb->pkt_type = PACKET_HOST;

	netif_rx(skb);
	return NULL;
}

static netdev_tx_t macvlan_start_xmit(struct sk_buff *skb,
				      struct net_device *dev)
{
	int i = skb_get_queue_mapping(skb);
	struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
	const struct macvlan_dev *vlan = netdev_priv(dev);
	unsigned int len = skb->len;
	int ret;

	skb->dev = vlan->lowerdev;
	ret = dev_queue_xmit(skb);

	if (likely(ret == NET_XMIT_SUCCESS)) {
		txq->tx_packets++;
		txq->tx_bytes += len;
	} else
		txq->tx_dropped++;

	return NETDEV_TX_OK;
}

static int macvlan_hard_header(struct sk_buff *skb, struct net_device *dev,
			       unsigned short type, const void *daddr,
			       const void *saddr, unsigned len)
{
	const struct macvlan_dev *vlan = netdev_priv(dev);
	struct net_device *lowerdev = vlan->lowerdev;

	return dev_hard_header(skb, lowerdev, type, daddr,
			       saddr ? : dev->dev_addr, len);
}

static const struct header_ops macvlan_hard_header_ops = {
	.create  	= macvlan_hard_header,
	.rebuild	= eth_rebuild_header,
	.parse		= eth_header_parse,
	.cache		= eth_header_cache,
	.cache_update	= eth_header_cache_update,
};

static int macvlan_open(struct net_device *dev)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	struct net_device *lowerdev = vlan->lowerdev;
	int err;

	err = -EBUSY;
	if (macvlan_addr_busy(vlan->port, dev->dev_addr))
		goto out;

	err = dev_unicast_add(lowerdev, dev->dev_addr);
	if (err < 0)
		goto out;
	if (dev->flags & IFF_ALLMULTI) {
		err = dev_set_allmulti(lowerdev, 1);
		if (err < 0)
			goto del_unicast;
	}
	macvlan_hash_add(vlan);
	return 0;

del_unicast:
	dev_unicast_delete(lowerdev, dev->dev_addr);
out:
	return err;
}

static int macvlan_stop(struct net_device *dev)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	struct net_device *lowerdev = vlan->lowerdev;

	dev_mc_unsync(lowerdev, dev);
	if (dev->flags & IFF_ALLMULTI)
		dev_set_allmulti(lowerdev, -1);

	dev_unicast_delete(lowerdev, dev->dev_addr);

	macvlan_hash_del(vlan);
	return 0;
}

static int macvlan_set_mac_address(struct net_device *dev, void *p)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	struct net_device *lowerdev = vlan->lowerdev;
	struct sockaddr *addr = p;
	int err;

	if (!is_valid_ether_addr(addr->sa_data))
		return -EADDRNOTAVAIL;

	if (!(dev->flags & IFF_UP)) {
		/* Just copy in the new address */
		memcpy(dev->dev_addr, addr->sa_data, ETH_ALEN);
	} else {
		/* Rehash and update the device filters */
		if (macvlan_addr_busy(vlan->port, addr->sa_data))
			return -EBUSY;

		err = dev_unicast_add(lowerdev, addr->sa_data);
		if (err)
			return err;

		dev_unicast_delete(lowerdev, dev->dev_addr);

		macvlan_hash_change_addr(vlan, addr->sa_data);
	}
	return 0;
}

static void macvlan_change_rx_flags(struct net_device *dev, int change)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	struct net_device *lowerdev = vlan->lowerdev;

	if (change & IFF_ALLMULTI)
		dev_set_allmulti(lowerdev, dev->flags & IFF_ALLMULTI ? 1 : -1);
}

static void macvlan_set_multicast_list(struct net_device *dev)
{
	struct macvlan_dev *vlan = netdev_priv(dev);

	dev_mc_sync(vlan->lowerdev, dev);
}

static int macvlan_change_mtu(struct net_device *dev, int new_mtu)
{
	struct macvlan_dev *vlan = netdev_priv(dev);

	if (new_mtu < 68 || vlan->lowerdev->mtu < new_mtu)
		return -EINVAL;
	dev->mtu = new_mtu;
	return 0;
}

/*
 * macvlan network devices have devices nesting below it and are a special
 * "super class" of normal network devices; split their locks off into a
 * separate class since they always nest.
 */
static struct lock_class_key macvlan_netdev_xmit_lock_key;
static struct lock_class_key macvlan_netdev_addr_lock_key;

#define MACVLAN_FEATURES \
	(NETIF_F_SG | NETIF_F_ALL_CSUM | NETIF_F_HIGHDMA | NETIF_F_FRAGLIST | \
	 NETIF_F_GSO | NETIF_F_TSO | NETIF_F_UFO | NETIF_F_GSO_ROBUST | \
	 NETIF_F_TSO_ECN | NETIF_F_TSO6)

#define MACVLAN_STATE_MASK \
	((1<<__LINK_STATE_NOCARRIER) | (1<<__LINK_STATE_DORMANT))

static void macvlan_set_lockdep_class_one(struct net_device *dev,
					  struct netdev_queue *txq,
					  void *_unused)
{
	lockdep_set_class(&txq->_xmit_lock,
			  &macvlan_netdev_xmit_lock_key);
}

static void macvlan_set_lockdep_class(struct net_device *dev)
{
	lockdep_set_class(&dev->addr_list_lock,
			  &macvlan_netdev_addr_lock_key);
	netdev_for_each_tx_queue(dev, macvlan_set_lockdep_class_one, NULL);
}

static int macvlan_init(struct net_device *dev)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	const struct net_device *lowerdev = vlan->lowerdev;

	dev->state		= (dev->state & ~MACVLAN_STATE_MASK) |
				  (lowerdev->state & MACVLAN_STATE_MASK);
	dev->features 		= lowerdev->features & MACVLAN_FEATURES;
	dev->gso_max_size	= lowerdev->gso_max_size;
	dev->iflink		= lowerdev->ifindex;
	dev->hard_header_len	= lowerdev->hard_header_len;

	macvlan_set_lockdep_class(dev);

	return 0;
}

static void macvlan_ethtool_get_drvinfo(struct net_device *dev,
					struct ethtool_drvinfo *drvinfo)
{
	snprintf(drvinfo->driver, 32, "macvlan");
	snprintf(drvinfo->version, 32, "0.1");
}

static u32 macvlan_ethtool_get_rx_csum(struct net_device *dev)
{
	const struct macvlan_dev *vlan = netdev_priv(dev);
	return dev_ethtool_get_rx_csum(vlan->lowerdev);
}

static int macvlan_ethtool_get_settings(struct net_device *dev,
					struct ethtool_cmd *cmd)
{
	const struct macvlan_dev *vlan = netdev_priv(dev);
	return dev_ethtool_get_settings(vlan->lowerdev, cmd);
}

static u32 macvlan_ethtool_get_flags(struct net_device *dev)
{
	const struct macvlan_dev *vlan = netdev_priv(dev);
	return dev_ethtool_get_flags(vlan->lowerdev);
}

static const struct ethtool_ops macvlan_ethtool_ops = {
	.get_link		= ethtool_op_get_link,
	.get_settings		= macvlan_ethtool_get_settings,
	.get_rx_csum		= macvlan_ethtool_get_rx_csum,
	.get_drvinfo		= macvlan_ethtool_get_drvinfo,
	.get_flags		= macvlan_ethtool_get_flags,
};

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,
	.ndo_set_multicast_list	= macvlan_set_multicast_list,
	.ndo_validate_addr	= eth_validate_addr,
};

static void macvlan_setup(struct net_device *dev)
{
	ether_setup(dev);

	dev->priv_flags	       &= ~IFF_XMIT_DST_RELEASE;
	dev->netdev_ops		= &macvlan_netdev_ops;
	dev->destructor		= free_netdev;
	dev->header_ops		= &macvlan_hard_header_ops,
	dev->ethtool_ops	= &macvlan_ethtool_ops;
	dev->tx_queue_len	= 0;
}

static int macvlan_port_create(struct net_device *dev)
{
	struct macvlan_port *port;
	unsigned int i;

	if (dev->type != ARPHRD_ETHER || dev->flags & IFF_LOOPBACK)
		return -EINVAL;

	port = kzalloc(sizeof(*port), GFP_KERNEL);
	if (port == NULL)
		return -ENOMEM;

	port->dev = dev;
	INIT_LIST_HEAD(&port->vlans);
	for (i = 0; i < MACVLAN_HASH_SIZE; i++)
		INIT_HLIST_HEAD(&port->vlan_hash[i]);
	rcu_assign_pointer(dev->macvlan_port, port);
	return 0;
}

static void macvlan_port_destroy(struct net_device *dev)
{
	struct macvlan_port *port = dev->macvlan_port;

	rcu_assign_pointer(dev->macvlan_port, NULL);
	synchronize_rcu();
	kfree(port);
}

static void macvlan_transfer_operstate(struct net_device *dev)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	const struct net_device *lowerdev = vlan->lowerdev;

	if (lowerdev->operstate == IF_OPER_DORMANT)
		netif_dormant_on(dev);
	else
		netif_dormant_off(dev);

	if (netif_carrier_ok(lowerdev)) {
		if (!netif_carrier_ok(dev))
			netif_carrier_on(dev);
	} else {
		if (netif_carrier_ok(dev))
			netif_carrier_off(dev);
	}
}

static int macvlan_validate(struct nlattr *tb[], struct nlattr *data[])
{
	if (tb[IFLA_ADDRESS]) {
		if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
			return -EINVAL;
		if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
			return -EADDRNOTAVAIL;
	}
	return 0;
}

static int macvlan_get_tx_queues(struct net *net,
				 struct nlattr *tb[],
				 unsigned int *num_tx_queues,
				 unsigned int *real_num_tx_queues)
{
	struct net_device *real_dev;

	if (!tb[IFLA_LINK])
		return -EINVAL;

	real_dev = __dev_get_by_index(net, nla_get_u32(tb[IFLA_LINK]));
	if (!real_dev)
		return -ENODEV;

	*num_tx_queues      = real_dev->num_tx_queues;
	*real_num_tx_queues = real_dev->real_num_tx_queues;
	return 0;
}

static int macvlan_newlink(struct net_device *dev,
			   struct nlattr *tb[], struct nlattr *data[])
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	struct macvlan_port *port;
	struct net_device *lowerdev;
	int err;

	if (!tb[IFLA_LINK])
		return -EINVAL;

	lowerdev = __dev_get_by_index(dev_net(dev), nla_get_u32(tb[IFLA_LINK]));
	if (lowerdev == NULL)
		return -ENODEV;

	/* When creating macvlans on top of other macvlans - use
	 * the real device as the lowerdev.
	 */
	if (lowerdev->rtnl_link_ops == dev->rtnl_link_ops) {
		struct macvlan_dev *lowervlan = netdev_priv(lowerdev);
		lowerdev = lowervlan->lowerdev;
	}

	if (!tb[IFLA_MTU])
		dev->mtu = lowerdev->mtu;
	else if (dev->mtu > lowerdev->mtu)
		return -EINVAL;

	if (!tb[IFLA_ADDRESS])
		random_ether_addr(dev->dev_addr);

	if (lowerdev->macvlan_port == NULL) {
		err = macvlan_port_create(lowerdev);
		if (err < 0)
			return err;
	}
	port = lowerdev->macvlan_port;

	vlan->lowerdev = lowerdev;
	vlan->dev      = dev;
	vlan->port     = port;

	err = register_netdevice(dev);
	if (err < 0)
		return err;

	list_add_tail(&vlan->list, &port->vlans);
	macvlan_transfer_operstate(dev);
	return 0;
}

static void macvlan_dellink(struct net_device *dev)
{
	struct macvlan_dev *vlan = netdev_priv(dev);
	struct macvlan_port *port = vlan->port;

	list_del(&vlan->list);
	unregister_netdevice(dev);

	if (list_empty(&port->vlans))
		macvlan_port_destroy(port->dev);
}

static struct rtnl_link_ops macvlan_link_ops __read_mostly = {
	.kind		= "macvlan",
	.priv_size	= sizeof(struct macvlan_dev),
	.get_tx_queues  = macvlan_get_tx_queues,
	.setup		= macvlan_setup,
	.validate	= macvlan_validate,
	.newlink	= macvlan_newlink,
	.dellink	= macvlan_dellink,
};

static int macvlan_device_event(struct notifier_block *unused,
				unsigned long event, void *ptr)
{
	struct net_device *dev = ptr;
	struct macvlan_dev *vlan, *next;
	struct macvlan_port *port;

	port = dev->macvlan_port;
	if (port == NULL)
		return NOTIFY_DONE;

	switch (event) {
	case NETDEV_CHANGE:
		list_for_each_entry(vlan, &port->vlans, list)
			macvlan_transfer_operstate(vlan->dev);
		break;
	case NETDEV_FEAT_CHANGE:
		list_for_each_entry(vlan, &port->vlans, list) {
			vlan->dev->features = dev->features & MACVLAN_FEATURES;
			vlan->dev->gso_max_size = dev->gso_max_size;
			netdev_features_change(vlan->dev);
		}
		break;
	case NETDEV_UNREGISTER:
		list_for_each_entry_safe(vlan, next, &port->vlans, list)
			macvlan_dellink(vlan->dev);
		break;
	}
	return NOTIFY_DONE;
}

static struct notifier_block macvlan_notifier_block __read_mostly = {
	.notifier_call	= macvlan_device_event,
};

static int __init macvlan_init_module(void)
{
	int err;

	register_netdevice_notifier(&macvlan_notifier_block);
	macvlan_handle_frame_hook = macvlan_handle_frame;

	err = rtnl_link_register(&macvlan_link_ops);
	if (err < 0)
		goto err1;
	return 0;
err1:
	macvlan_handle_frame_hook = NULL;
	unregister_netdevice_notifier(&macvlan_notifier_block);
	return err;
}

static void __exit macvlan_cleanup_module(void)
{
	rtnl_link_unregister(&macvlan_link_ops);
	macvlan_handle_frame_hook = NULL;
	unregister_netdevice_notifier(&macvlan_notifier_block);
}

module_init(macvlan_init_module);
module_exit(macvlan_cleanup_module);

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");
MODULE_DESCRIPTION("Driver for MAC address based VLANs");
MODULE_ALIAS_RTNL_LINK("macvlan");
Commit	Line	Data
b863ceb7 PM	1	/*
	2	* Copyright (c) 2007 Patrick McHardy <kaber@trash.net>
	3	*
	4	* This program is free software; you can redistribute it and/or
	5	* modify it under the terms of the GNU General Public License as
	6	* published by the Free Software Foundation; either version 2 of
	7	* the License, or (at your option) any later version.
	8	*
	9	* The code this is based on carried the following copyright notice:
	10	* ---
	11	* (C) Copyright 2001-2006
	12	* Alex Zeffertt, Cambridge Broadband Ltd, ajz@cambridgebroadband.com
	13	* Re-worked by Ben Greear <greearb@candelatech.com>
	14	* ---
	15	*/
	16	#include <linux/kernel.h>
	17	#include <linux/types.h>
	18	#include <linux/module.h>
	19	#include <linux/init.h>
	20	#include <linux/errno.h>
	21	#include <linux/slab.h>
	22	#include <linux/string.h>
82524746	23	#include <linux/rculist.h>
b863ceb7 PM	24	#include <linux/notifier.h>
	25	#include <linux/netdevice.h>
	26	#include <linux/etherdevice.h>
	27	#include <linux/ethtool.h>
	28	#include <linux/if_arp.h>
	29	#include <linux/if_link.h>
	30	#include <linux/if_macvlan.h>
	31	#include <net/rtnetlink.h>
	32
	33	#define MACVLAN_HASH_SIZE (1 << BITS_PER_BYTE)
	34
	35	struct macvlan_port {
	36	struct net_device *dev;
	37	struct hlist_head vlan_hash[MACVLAN_HASH_SIZE];
	38	struct list_head vlans;
	39	};
	40
	41	struct macvlan_dev {
	42	struct net_device *dev;
	43	struct list_head list;
	44	struct hlist_node hlist;
	45	struct macvlan_port *port;
	46	struct net_device *lowerdev;
	47	};
	48
	49
	50	static struct macvlan_dev macvlan_hash_lookup(const struct macvlan_port port,
	51	const unsigned char *addr)
	52	{
	53	struct macvlan_dev *vlan;
	54	struct hlist_node *n;
	55
	56	hlist_for_each_entry_rcu(vlan, n, &port->vlan_hash[addr[5]], hlist) {
ac06713d	57	if (!compare_ether_addr_64bits(vlan->dev->dev_addr, addr))
b863ceb7 PM	58	return vlan;
	59	}
	60	return NULL;
	61	}
	62
f9ac30f0 EB	63	static void macvlan_hash_add(struct macvlan_dev *vlan)
	64	{
	65	struct macvlan_port *port = vlan->port;
	66	const unsigned char *addr = vlan->dev->dev_addr;
	67
	68	hlist_add_head_rcu(&vlan->hlist, &port->vlan_hash[addr[5]]);
	69	}
	70
	71	static void macvlan_hash_del(struct macvlan_dev *vlan)
	72	{
	73	hlist_del_rcu(&vlan->hlist);
	74	synchronize_rcu();
	75	}
	76
	77	static void macvlan_hash_change_addr(struct macvlan_dev *vlan,
	78	const unsigned char *addr)
	79	{
	80	macvlan_hash_del(vlan);
	81	/* Now that we are unhashed it is safe to change the device
	82	* address without confusing packet delivery.
	83	*/
	84	memcpy(vlan->dev->dev_addr, addr, ETH_ALEN);
	85	macvlan_hash_add(vlan);
	86	}
	87
	88	static int macvlan_addr_busy(const struct macvlan_port *port,
	89	const unsigned char *addr)
	90	{
	91	/* Test to see if the specified multicast address is
	92	* currently in use by the underlying device or
	93	* another macvlan.
	94	*/
ac06713d	95	if (!compare_ether_addr_64bits(port->dev->dev_addr, addr))
f9ac30f0 EB	96	return 1;
	97
	98	if (macvlan_hash_lookup(port, addr))
	99	return 1;
	100
	101	return 0;
	102	}
	103
b863ceb7 PM	104	static void macvlan_broadcast(struct sk_buff *skb,
	105	const struct macvlan_port *port)
	106	{
	107	const struct ethhdr *eth = eth_hdr(skb);
	108	const struct macvlan_dev *vlan;
	109	struct hlist_node *n;
	110	struct net_device *dev;
	111	struct sk_buff *nskb;
	112	unsigned int i;
	113
efbbced3 PM	114	if (skb->protocol == htons(ETH_P_PAUSE))
	115	return;
	116
b863ceb7 PM	117	for (i = 0; i < MACVLAN_HASH_SIZE; i++) {
	118	hlist_for_each_entry_rcu(vlan, n, &port->vlan_hash[i], hlist) {
	119	dev = vlan->dev;
b863ceb7 PM	120
	121	nskb = skb_clone(skb, GFP_ATOMIC);
	122	if (nskb == NULL) {
	123	dev->stats.rx_errors++;
	124	dev->stats.rx_dropped++;
	125	continue;
	126	}
	127
	128	dev->stats.rx_bytes += skb->len + ETH_HLEN;
	129	dev->stats.rx_packets++;
	130	dev->stats.multicast++;
b863ceb7 PM	131
b863ceb7 PM	132	nskb->dev = dev;
ac06713d	133	if (!compare_ether_addr_64bits(eth->h_dest, dev->broadcast))
b863ceb7 PM	134	nskb->pkt_type = PACKET_BROADCAST;
	135	else
	136	nskb->pkt_type = PACKET_MULTICAST;
	137
	138	netif_rx(nskb);
	139	}
	140	}
	141	}
	142
	143	/* called under rcu_read_lock() from netif_receive_skb */
	144	static struct sk_buff macvlan_handle_frame(struct sk_buff skb)
	145	{
	146	const struct ethhdr *eth = eth_hdr(skb);
	147	const struct macvlan_port *port;
	148	const struct macvlan_dev *vlan;
	149	struct net_device *dev;
	150
	151	port = rcu_dereference(skb->dev->macvlan_port);
	152	if (port == NULL)
	153	return skb;
	154
	155	if (is_multicast_ether_addr(eth->h_dest)) {
	156	macvlan_broadcast(skb, port);
	157	return skb;
	158	}
	159
	160	vlan = macvlan_hash_lookup(port, eth->h_dest);
	161	if (vlan == NULL)
	162	return skb;
	163
	164	dev = vlan->dev;
	165	if (unlikely(!(dev->flags & IFF_UP))) {
	166	kfree_skb(skb);
	167	return NULL;
	168	}
	169
	170	skb = skb_share_check(skb, GFP_ATOMIC);
	171	if (skb == NULL) {
	172	dev->stats.rx_errors++;
	173	dev->stats.rx_dropped++;
	174	return NULL;
	175	}
	176
	177	dev->stats.rx_bytes += skb->len + ETH_HLEN;
	178	dev->stats.rx_packets++;
b863ceb7 PM	179
	180	skb->dev = dev;
	181	skb->pkt_type = PACKET_HOST;
	182
	183	netif_rx(skb);
	184	return NULL;
	185	}
	186
424efe9c SH	187	static netdev_tx_t macvlan_start_xmit(struct sk_buff *skb,
424efe9c SH	188	struct net_device *dev)
b863ceb7	189	{
2c114553 ED	190	int i = skb_get_queue_mapping(skb);
2c114553 ED	191	struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
b863ceb7 PM	192	const struct macvlan_dev *vlan = netdev_priv(dev);
	193	unsigned int len = skb->len;
	194	int ret;
	195
	196	skb->dev = vlan->lowerdev;
	197	ret = dev_queue_xmit(skb);
	198
	199	if (likely(ret == NET_XMIT_SUCCESS)) {
2c114553 ED	200	txq->tx_packets++;
	201	txq->tx_bytes += len;
	202	} else
	203	txq->tx_dropped++;
	204
b863ceb7 PM	205	return NETDEV_TX_OK;
	206	}
	207
	208	static int macvlan_hard_header(struct sk_buff skb, struct net_device dev,
3b04ddde SH	209	unsigned short type, const void *daddr,
3b04ddde SH	210	const void *saddr, unsigned len)
b863ceb7 PM	211	{
	212	const struct macvlan_dev *vlan = netdev_priv(dev);
	213	struct net_device *lowerdev = vlan->lowerdev;
	214
0c4e8581 SH	215	return dev_hard_header(skb, lowerdev, type, daddr,
0c4e8581 SH	216	saddr ? : dev->dev_addr, len);
b863ceb7 PM	217	}
b863ceb7 PM	218
3b04ddde SH	219	static const struct header_ops macvlan_hard_header_ops = {
	220	.create = macvlan_hard_header,
	221	.rebuild = eth_rebuild_header,
	222	.parse = eth_header_parse,
3b04ddde SH	223	.cache = eth_header_cache,
	224	.cache_update = eth_header_cache_update,
	225	};
	226
b863ceb7 PM	227	static int macvlan_open(struct net_device *dev)
	228	{
	229	struct macvlan_dev *vlan = netdev_priv(dev);
b863ceb7 PM	230	struct net_device *lowerdev = vlan->lowerdev;
	231	int err;
	232
f9ac30f0 EB	233	err = -EBUSY;
	234	if (macvlan_addr_busy(vlan->port, dev->dev_addr))
	235	goto out;
	236
ccffad25	237	err = dev_unicast_add(lowerdev, dev->dev_addr);
b863ceb7	238	if (err < 0)
b89fb7da WC	239	goto out;
	240	if (dev->flags & IFF_ALLMULTI) {
	241	err = dev_set_allmulti(lowerdev, 1);
	242	if (err < 0)
	243	goto del_unicast;
	244	}
f9ac30f0	245	macvlan_hash_add(vlan);
b863ceb7	246	return 0;
b89fb7da WC	247
b89fb7da WC	248	del_unicast:
ccffad25	249	dev_unicast_delete(lowerdev, dev->dev_addr);
b89fb7da WC	250	out:
b89fb7da WC	251	return err;
b863ceb7 PM	252	}
	253
	254	static int macvlan_stop(struct net_device *dev)
	255	{
	256	struct macvlan_dev *vlan = netdev_priv(dev);
	257	struct net_device *lowerdev = vlan->lowerdev;
	258
	259	dev_mc_unsync(lowerdev, dev);
	260	if (dev->flags & IFF_ALLMULTI)
	261	dev_set_allmulti(lowerdev, -1);
	262
ccffad25	263	dev_unicast_delete(lowerdev, dev->dev_addr);
b863ceb7	264
f9ac30f0	265	macvlan_hash_del(vlan);
b863ceb7 PM	266	return 0;
	267	}
	268
ad5d20a6 PM	269	static int macvlan_set_mac_address(struct net_device dev, void p)
	270	{
	271	struct macvlan_dev *vlan = netdev_priv(dev);
	272	struct net_device *lowerdev = vlan->lowerdev;
	273	struct sockaddr *addr = p;
	274	int err;
	275
	276	if (!is_valid_ether_addr(addr->sa_data))
	277	return -EADDRNOTAVAIL;
	278
f9ac30f0 EB	279	if (!(dev->flags & IFF_UP)) {
	280	/* Just copy in the new address */
	281	memcpy(dev->dev_addr, addr->sa_data, ETH_ALEN);
	282	} else {
	283	/* Rehash and update the device filters */
	284	if (macvlan_addr_busy(vlan->port, addr->sa_data))
	285	return -EBUSY;
ad5d20a6	286
ccffad25 JP	287	err = dev_unicast_add(lowerdev, addr->sa_data);
ccffad25 JP	288	if (err)
f9ac30f0	289	return err;
ad5d20a6	290
ccffad25	291	dev_unicast_delete(lowerdev, dev->dev_addr);
f9ac30f0 EB	292
	293	macvlan_hash_change_addr(vlan, addr->sa_data);
	294	}
ad5d20a6 PM	295	return 0;
	296	}
	297
b863ceb7 PM	298	static void macvlan_change_rx_flags(struct net_device *dev, int change)
	299	{
	300	struct macvlan_dev *vlan = netdev_priv(dev);
	301	struct net_device *lowerdev = vlan->lowerdev;
	302
	303	if (change & IFF_ALLMULTI)
	304	dev_set_allmulti(lowerdev, dev->flags & IFF_ALLMULTI ? 1 : -1);
	305	}
	306
	307	static void macvlan_set_multicast_list(struct net_device *dev)
	308	{
	309	struct macvlan_dev *vlan = netdev_priv(dev);
	310
	311	dev_mc_sync(vlan->lowerdev, dev);
	312	}
	313
	314	static int macvlan_change_mtu(struct net_device *dev, int new_mtu)
	315	{
	316	struct macvlan_dev *vlan = netdev_priv(dev);
	317
	318	if (new_mtu < 68 \|\| vlan->lowerdev->mtu < new_mtu)
	319	return -EINVAL;
	320	dev->mtu = new_mtu;
	321	return 0;
	322	}
	323
	324	/*
	325	* macvlan network devices have devices nesting below it and are a special
	326	* "super class" of normal network devices; split their locks off into a
	327	* separate class since they always nest.
	328	*/
	329	static struct lock_class_key macvlan_netdev_xmit_lock_key;
cf508b12	330	static struct lock_class_key macvlan_netdev_addr_lock_key;
b863ceb7 PM	331
	332	#define MACVLAN_FEATURES \
	333	(NETIF_F_SG \| NETIF_F_ALL_CSUM \| NETIF_F_HIGHDMA \| NETIF_F_FRAGLIST \| \
	334	NETIF_F_GSO \| NETIF_F_TSO \| NETIF_F_UFO \| NETIF_F_GSO_ROBUST \| \
	335	NETIF_F_TSO_ECN \| NETIF_F_TSO6)
	336
	337	#define MACVLAN_STATE_MASK \
	338	((1<<__LINK_STATE_NOCARRIER) \| (1<<__LINK_STATE_DORMANT))
	339
e8a0464c DM	340	static void macvlan_set_lockdep_class_one(struct net_device *dev,
	341	struct netdev_queue *txq,
	342	void *_unused)
c773e847 DM	343	{
	344	lockdep_set_class(&txq->_xmit_lock,
	345	&macvlan_netdev_xmit_lock_key);
	346	}
	347
	348	static void macvlan_set_lockdep_class(struct net_device *dev)
	349	{
cf508b12 DM	350	lockdep_set_class(&dev->addr_list_lock,
cf508b12 DM	351	&macvlan_netdev_addr_lock_key);
e8a0464c	352	netdev_for_each_tx_queue(dev, macvlan_set_lockdep_class_one, NULL);
c773e847 DM	353	}
c773e847 DM	354
b863ceb7 PM	355	static int macvlan_init(struct net_device *dev)
	356	{
	357	struct macvlan_dev *vlan = netdev_priv(dev);
	358	const struct net_device *lowerdev = vlan->lowerdev;
	359
	360	dev->state = (dev->state & ~MACVLAN_STATE_MASK) \|
	361	(lowerdev->state & MACVLAN_STATE_MASK);
	362	dev->features = lowerdev->features & MACVLAN_FEATURES;
8c2acc53	363	dev->gso_max_size = lowerdev->gso_max_size;
b863ceb7	364	dev->iflink = lowerdev->ifindex;
ef5c8996	365	dev->hard_header_len = lowerdev->hard_header_len;
b863ceb7	366
c773e847 DM	367	macvlan_set_lockdep_class(dev);
c773e847 DM	368
b863ceb7 PM	369	return 0;
	370	}
	371
	372	static void macvlan_ethtool_get_drvinfo(struct net_device *dev,
	373	struct ethtool_drvinfo *drvinfo)
	374	{
	375	snprintf(drvinfo->driver, 32, "macvlan");
	376	snprintf(drvinfo->version, 32, "0.1");
	377	}
	378
	379	static u32 macvlan_ethtool_get_rx_csum(struct net_device *dev)
	380	{
	381	const struct macvlan_dev *vlan = netdev_priv(dev);
b1b67dd4	382	return dev_ethtool_get_rx_csum(vlan->lowerdev);
b863ceb7 PM	383	}
b863ceb7 PM	384
9edb8bb6 SH	385	static int macvlan_ethtool_get_settings(struct net_device *dev,
	386	struct ethtool_cmd *cmd)
	387	{
	388	const struct macvlan_dev *vlan = netdev_priv(dev);
b1b67dd4	389	return dev_ethtool_get_settings(vlan->lowerdev, cmd);
9edb8bb6 SH	390	}
	391
	392	static u32 macvlan_ethtool_get_flags(struct net_device *dev)
	393	{
	394	const struct macvlan_dev *vlan = netdev_priv(dev);
b1b67dd4	395	return dev_ethtool_get_flags(vlan->lowerdev);
9edb8bb6 SH	396	}
9edb8bb6 SH	397
b863ceb7 PM	398	static const struct ethtool_ops macvlan_ethtool_ops = {
b863ceb7 PM	399	.get_link = ethtool_op_get_link,
9edb8bb6	400	.get_settings = macvlan_ethtool_get_settings,
b863ceb7	401	.get_rx_csum = macvlan_ethtool_get_rx_csum,
b863ceb7	402	.get_drvinfo = macvlan_ethtool_get_drvinfo,
9edb8bb6	403	.get_flags = macvlan_ethtool_get_flags,
b863ceb7 PM	404	};
b863ceb7 PM	405
54a30c97 SH	406	static const struct net_device_ops macvlan_netdev_ops = {
	407	.ndo_init = macvlan_init,
	408	.ndo_open = macvlan_open,
	409	.ndo_stop = macvlan_stop,
00829823	410	.ndo_start_xmit = macvlan_start_xmit,
54a30c97 SH	411	.ndo_change_mtu = macvlan_change_mtu,
	412	.ndo_change_rx_flags = macvlan_change_rx_flags,
	413	.ndo_set_mac_address = macvlan_set_mac_address,
	414	.ndo_set_multicast_list = macvlan_set_multicast_list,
	415	.ndo_validate_addr = eth_validate_addr,
	416	};
	417
b863ceb7 PM	418	static void macvlan_setup(struct net_device *dev)
	419	{
	420	ether_setup(dev);
	421
93f154b5	422	dev->priv_flags &= ~IFF_XMIT_DST_RELEASE;
54a30c97	423	dev->netdev_ops = &macvlan_netdev_ops;
b863ceb7	424	dev->destructor = free_netdev;
3b04ddde	425	dev->header_ops = &macvlan_hard_header_ops,
b863ceb7 PM	426	dev->ethtool_ops = &macvlan_ethtool_ops;
	427	dev->tx_queue_len = 0;
	428	}
	429
	430	static int macvlan_port_create(struct net_device *dev)
	431	{
	432	struct macvlan_port *port;
	433	unsigned int i;
	434
	435	if (dev->type != ARPHRD_ETHER \|\| dev->flags & IFF_LOOPBACK)
	436	return -EINVAL;
	437
	438	port = kzalloc(sizeof(*port), GFP_KERNEL);
	439	if (port == NULL)
	440	return -ENOMEM;
	441
	442	port->dev = dev;
	443	INIT_LIST_HEAD(&port->vlans);
	444	for (i = 0; i < MACVLAN_HASH_SIZE; i++)
	445	INIT_HLIST_HEAD(&port->vlan_hash[i]);
	446	rcu_assign_pointer(dev->macvlan_port, port);
	447	return 0;
	448	}
	449
	450	static void macvlan_port_destroy(struct net_device *dev)
	451	{
	452	struct macvlan_port *port = dev->macvlan_port;
	453
	454	rcu_assign_pointer(dev->macvlan_port, NULL);
	455	synchronize_rcu();
	456	kfree(port);
	457	}
	458
	459	static void macvlan_transfer_operstate(struct net_device *dev)
	460	{
	461	struct macvlan_dev *vlan = netdev_priv(dev);
	462	const struct net_device *lowerdev = vlan->lowerdev;
	463
	464	if (lowerdev->operstate == IF_OPER_DORMANT)
	465	netif_dormant_on(dev);
	466	else
	467	netif_dormant_off(dev);
	468
	469	if (netif_carrier_ok(lowerdev)) {
	470	if (!netif_carrier_ok(dev))
	471	netif_carrier_on(dev);
	472	} else {
f12ca5f9	473	if (netif_carrier_ok(dev))
b863ceb7 PM	474	netif_carrier_off(dev);
	475	}
	476	}
	477
	478	static int macvlan_validate(struct nlattr tb[], struct nlattr data[])
	479	{
	480	if (tb[IFLA_ADDRESS]) {
	481	if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
	482	return -EINVAL;
	483	if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
	484	return -EADDRNOTAVAIL;
	485	}
	486	return 0;
	487	}
	488
2c114553 ED	489	static int macvlan_get_tx_queues(struct net *net,
	490	struct nlattr *tb[],
	491	unsigned int *num_tx_queues,
	492	unsigned int *real_num_tx_queues)
	493	{
	494	struct net_device *real_dev;
	495
	496	if (!tb[IFLA_LINK])
	497	return -EINVAL;
	498
	499	real_dev = __dev_get_by_index(net, nla_get_u32(tb[IFLA_LINK]));
	500	if (!real_dev)
	501	return -ENODEV;
	502
	503	*num_tx_queues = real_dev->num_tx_queues;
	504	*real_num_tx_queues = real_dev->real_num_tx_queues;
	505	return 0;
	506	}
	507
b863ceb7 PM	508	static int macvlan_newlink(struct net_device *dev,
	509	struct nlattr tb[], struct nlattr data[])
	510	{
	511	struct macvlan_dev *vlan = netdev_priv(dev);
	512	struct macvlan_port *port;
	513	struct net_device *lowerdev;
	514	int err;
	515
	516	if (!tb[IFLA_LINK])
	517	return -EINVAL;
	518
c346dca1	519	lowerdev = __dev_get_by_index(dev_net(dev), nla_get_u32(tb[IFLA_LINK]));
b863ceb7 PM	520	if (lowerdev == NULL)
	521	return -ENODEV;
	522
b0832a29 EB	523	/* When creating macvlans on top of other macvlans - use
b0832a29 EB	524	* the real device as the lowerdev.
a6ca5f1d	525	*/
b0832a29 EB	526	if (lowerdev->rtnl_link_ops == dev->rtnl_link_ops) {
	527	struct macvlan_dev *lowervlan = netdev_priv(lowerdev);
	528	lowerdev = lowervlan->lowerdev;
	529	}
a6ca5f1d	530
b863ceb7 PM	531	if (!tb[IFLA_MTU])
	532	dev->mtu = lowerdev->mtu;
	533	else if (dev->mtu > lowerdev->mtu)
	534	return -EINVAL;
	535
	536	if (!tb[IFLA_ADDRESS])
	537	random_ether_addr(dev->dev_addr);
	538
	539	if (lowerdev->macvlan_port == NULL) {
	540	err = macvlan_port_create(lowerdev);
	541	if (err < 0)
	542	return err;
	543	}
	544	port = lowerdev->macvlan_port;
	545
	546	vlan->lowerdev = lowerdev;
	547	vlan->dev = dev;
	548	vlan->port = port;
	549
	550	err = register_netdevice(dev);
	551	if (err < 0)
	552	return err;
	553
	554	list_add_tail(&vlan->list, &port->vlans);
	555	macvlan_transfer_operstate(dev);
	556	return 0;
	557	}
	558
	559	static void macvlan_dellink(struct net_device *dev)
	560	{
	561	struct macvlan_dev *vlan = netdev_priv(dev);
	562	struct macvlan_port *port = vlan->port;
	563
	564	list_del(&vlan->list);
	565	unregister_netdevice(dev);
	566
	567	if (list_empty(&port->vlans))
73120964	568	macvlan_port_destroy(port->dev);
b863ceb7 PM	569	}
	570
	571	static struct rtnl_link_ops macvlan_link_ops __read_mostly = {
	572	.kind = "macvlan",
	573	.priv_size = sizeof(struct macvlan_dev),
2c114553	574	.get_tx_queues = macvlan_get_tx_queues,
b863ceb7 PM	575	.setup = macvlan_setup,
	576	.validate = macvlan_validate,
	577	.newlink = macvlan_newlink,
	578	.dellink = macvlan_dellink,
	579	};
	580
	581	static int macvlan_device_event(struct notifier_block *unused,
	582	unsigned long event, void *ptr)
	583	{
	584	struct net_device *dev = ptr;
	585	struct macvlan_dev vlan, next;
	586	struct macvlan_port *port;
	587
	588	port = dev->macvlan_port;
	589	if (port == NULL)
	590	return NOTIFY_DONE;
	591
	592	switch (event) {
	593	case NETDEV_CHANGE:
	594	list_for_each_entry(vlan, &port->vlans, list)
	595	macvlan_transfer_operstate(vlan->dev);
	596	break;
	597	case NETDEV_FEAT_CHANGE:
	598	list_for_each_entry(vlan, &port->vlans, list) {
	599	vlan->dev->features = dev->features & MACVLAN_FEATURES;
8c2acc53	600	vlan->dev->gso_max_size = dev->gso_max_size;
b863ceb7 PM	601	netdev_features_change(vlan->dev);
	602	}
	603	break;
	604	case NETDEV_UNREGISTER:
	605	list_for_each_entry_safe(vlan, next, &port->vlans, list)
	606	macvlan_dellink(vlan->dev);
	607	break;
	608	}
	609	return NOTIFY_DONE;
	610	}
	611
	612	static struct notifier_block macvlan_notifier_block __read_mostly = {
	613	.notifier_call = macvlan_device_event,
	614	};
	615
	616	static int __init macvlan_init_module(void)
	617	{
	618	int err;
	619
	620	register_netdevice_notifier(&macvlan_notifier_block);
	621	macvlan_handle_frame_hook = macvlan_handle_frame;
	622
	623	err = rtnl_link_register(&macvlan_link_ops);
	624	if (err < 0)
	625	goto err1;
	626	return 0;
	627	err1:
52913246	628	macvlan_handle_frame_hook = NULL;
b863ceb7 PM	629	unregister_netdevice_notifier(&macvlan_notifier_block);
	630	return err;
	631	}
	632
	633	static void __exit macvlan_cleanup_module(void)
	634	{
	635	rtnl_link_unregister(&macvlan_link_ops);
	636	macvlan_handle_frame_hook = NULL;
	637	unregister_netdevice_notifier(&macvlan_notifier_block);
	638	}
	639
	640	module_init(macvlan_init_module);
	641	module_exit(macvlan_cleanup_module);
	642
	643	MODULE_LICENSE("GPL");
	644	MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");
	645	MODULE_DESCRIPTION("Driver for MAC address based VLANs");
	646	MODULE_ALIAS_RTNL_LINK("macvlan");