1 files changed, 325 insertions, 192 deletions

diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c
index ba442883e87..3eca3fdf8fe 100644
--- a/net/bridge/br_if.c
+++ b/net/bridge/br_if.c
@@ -5,8 +5,6 @@
  *	Authors:
  *	Lennert Buytenhek		<buytenh@gnu.org>
  *
- *	$Id: br_if.c,v 1.7 2001/12/24 00:59:55 davem Exp $
- *
  *	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
@@ -15,13 +13,17 @@
 
 #include <linux/kernel.h>
 #include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/netpoll.h>
 #include <linux/ethtool.h>
 #include <linux/if_arp.h>
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/rtnetlink.h>
 #include <linux/if_ether.h>
+#include <linux/slab.h>
 #include <net/sock.h>
+#include <linux/if_vlan.h>
 
 #include "br_private.h"
 
@@ -29,33 +31,20 @@
  * Determine initial path cost based on speed.
  * using recommendations from 802.1d standard
  *
- * Need to simulate user ioctl because not all device's that support
- * ethtool, use ethtool_ops.  Also, since driver might sleep need to
- * not be holding any locks.
+ * Since driver might sleep need to not be holding any locks.
  */
 static int port_cost(struct net_device *dev)
 {
-	struct ethtool_cmd ecmd = { ETHTOOL_GSET };
-	struct ifreq ifr;
-	mm_segment_t old_fs;
-	int err;
-
-	strncpy(ifr.ifr_name, dev->name, IFNAMSIZ);
-	ifr.ifr_data = (void __user *) &ecmd;
+	struct ethtool_cmd ecmd;
 
-	old_fs = get_fs();
-	set_fs(KERNEL_DS);
-	err = dev_ethtool(&ifr);
-	set_fs(old_fs);
-	
-	if (!err) {
-		switch(ecmd.speed) {
-		case SPEED_100:
-			return 19;
-		case SPEED_1000:
-			return 4;
+	if (!__ethtool_get_settings(dev, &ecmd)) {
+		switch (ethtool_cmd_speed(&ecmd)) {
 		case SPEED_10000:
 			return 2;
+		case SPEED_1000:
+			return 4;
+		case SPEED_100:
+			return 19;
 		case SPEED_10:
 			return 100;
 		}
@@ -72,34 +61,149 @@ static int port_cost(struct net_device *dev)
 }
 
 
-/*
- * Check for port carrier transistions.
- * Called from work queue to allow for calling functions that
- * might sleep (such as speed check), and to debounce.
- */
-static void port_carrier_check(void *arg)
+/* Check for port carrier transitions. */
+void br_port_carrier_check(struct net_bridge_port *p)
 {
-	struct net_bridge_port *p = arg;
+	struct net_device *dev = p->dev;
+	struct net_bridge *br = p->br;
 
-	rtnl_lock();
-	if (netif_carrier_ok(p->dev)) {
-		u32 cost = port_cost(p->dev);
+	if (!(p->flags & BR_ADMIN_COST) &&
+	    netif_running(dev) && netif_oper_up(dev))
+		p->path_cost = port_cost(dev);
 
-		spin_lock_bh(&p->br->lock);
-		if (p->state == BR_STATE_DISABLED) {
-			p->path_cost = cost;
+	if (!netif_running(br->dev))
+		return;
+
+	spin_lock_bh(&br->lock);
+	if (netif_running(dev) && netif_oper_up(dev)) {
+		if (p->state == BR_STATE_DISABLED)
 			br_stp_enable_port(p);
-		}
-		spin_unlock_bh(&p->br->lock);
 	} else {
-		spin_lock_bh(&p->br->lock);
 		if (p->state != BR_STATE_DISABLED)
 			br_stp_disable_port(p);
-		spin_unlock_bh(&p->br->lock);
 	}
-	rtnl_unlock();
+	spin_unlock_bh(&br->lock);
+}
+
+static void br_port_set_promisc(struct net_bridge_port *p)
+{
+	int err = 0;
+
+	if (br_promisc_port(p))
+		return;
+
+	err = dev_set_promiscuity(p->dev, 1);
+	if (err)
+		return;
+
+	br_fdb_unsync_static(p->br, p);
+	p->flags |= BR_PROMISC;
+}
+
+static void br_port_clear_promisc(struct net_bridge_port *p)
+{
+	int err;
+
+	/* Check if the port is already non-promisc or if it doesn't
+	 * support UNICAST filtering.  Without unicast filtering support
+	 * we'll end up re-enabling promisc mode anyway, so just check for
+	 * it here.
+	 */
+	if (!br_promisc_port(p) || !(p->dev->priv_flags & IFF_UNICAST_FLT))
+		return;
+
+	/* Since we'll be clearing the promisc mode, program the port
+	 * first so that we don't have interruption in traffic.
+	 */
+	err = br_fdb_sync_static(p->br, p);
+	if (err)
+		return;
+
+	dev_set_promiscuity(p->dev, -1);
+	p->flags &= ~BR_PROMISC;
+}
+
+/* When a port is added or removed or when certain port flags
+ * change, this function is called to automatically manage
+ * promiscuity setting of all the bridge ports.  We are always called
+ * under RTNL so can skip using rcu primitives.
+ */
+void br_manage_promisc(struct net_bridge *br)
+{
+	struct net_bridge_port *p;
+	bool set_all = false;
+
+	/* If vlan filtering is disabled or bridge interface is placed
+	 * into promiscuous mode, place all ports in promiscuous mode.
+	 */
+	if ((br->dev->flags & IFF_PROMISC) || !br_vlan_enabled(br))
+		set_all = true;
+
+	list_for_each_entry(p, &br->port_list, list) {
+		if (set_all) {
+			br_port_set_promisc(p);
+		} else {
+			/* If the number of auto-ports is <= 1, then all other
+			 * ports will have their output configuration
+			 * statically specified through fdbs.  Since ingress
+			 * on the auto-port becomes forwarding/egress to other
+			 * ports and egress configuration is statically known,
+			 * we can say that ingress configuration of the
+			 * auto-port is also statically known.
+			 * This lets us disable promiscuous mode and write
+			 * this config to hw.
+			 */
+			if (br->auto_cnt == 0 ||
+			    (br->auto_cnt == 1 && br_auto_port(p)))
+				br_port_clear_promisc(p);
+			else
+				br_port_set_promisc(p);
+		}
+	}
+}
+
+static void nbp_update_port_count(struct net_bridge *br)
+{
+	struct net_bridge_port *p;
+	u32 cnt = 0;
+
+	list_for_each_entry(p, &br->port_list, list) {
+		if (br_auto_port(p))
+			cnt++;
+	}
+	if (br->auto_cnt != cnt) {
+		br->auto_cnt = cnt;
+		br_manage_promisc(br);
+	}
+}
+
+static void nbp_delete_promisc(struct net_bridge_port *p)
+{
+	/* If port is currently promiscuous, unset promiscuity.
+	 * Otherwise, it is a static port so remove all addresses
+	 * from it.
+	 */
+	dev_set_allmulti(p->dev, -1);
+	if (br_promisc_port(p))
+		dev_set_promiscuity(p->dev, -1);
+	else
+		br_fdb_unsync_static(p->br, p);
+}
+
+static void release_nbp(struct kobject *kobj)
+{
+	struct net_bridge_port *p
+		= container_of(kobj, struct net_bridge_port, kobj);
+	kfree(p);
 }
 
+static struct kobj_type brport_ktype = {
+#ifdef CONFIG_SYSFS
+	.sysfs_ops = &brport_sysfs_ops,
+#endif
+	.release = release_nbp,
+};
+
 static void destroy_nbp(struct net_bridge_port *p)
 {
 	struct net_device *dev = p->dev;
@@ -108,7 +212,7 @@ static void destroy_nbp(struct net_bridge_port *p)
 	p->dev = NULL;
 	dev_put(dev);
 
-	br_sysfs_freeif(p);
+	kobject_put(&p->kobj);
 }
 
 static void destroy_nbp_rcu(struct rcu_head *head)
@@ -118,87 +222,69 @@ static void destroy_nbp_rcu(struct rcu_head *head)
 	destroy_nbp(p);
 }
 
-/* called with RTNL */
+/* Delete port(interface) from bridge is done in two steps.
+ * via RCU. First step, marks device as down. That deletes
+ * all the timers and stops new packets from flowing through.
+ *
+ * Final cleanup doesn't occur until after all CPU's finished
+ * processing packets.
+ *
+ * Protected from multiple admin operations by RTNL mutex
+ */
 static void del_nbp(struct net_bridge_port *p)
 {
 	struct net_bridge *br = p->br;
 	struct net_device *dev = p->dev;
 
-	dev->br_port = NULL;
-	dev_set_promiscuity(dev, -1);
+	sysfs_remove_link(br->ifobj, p->dev->name);
 
-	cancel_delayed_work(&p->carrier_check);
-	flush_scheduled_work();
+	nbp_delete_promisc(p);
 
 	spin_lock_bh(&br->lock);
 	br_stp_disable_port(p);
 	spin_unlock_bh(&br->lock);
 
-	br_fdb_delete_by_port(br, p);
+	br_ifinfo_notify(RTM_DELLINK, p);
 
 	list_del_rcu(&p->list);
 
-	del_timer_sync(&p->message_age_timer);
-	del_timer_sync(&p->forward_delay_timer);
-	del_timer_sync(&p->hold_timer);
-	
+	nbp_vlan_flush(p);
+	br_fdb_delete_by_port(br, p, 1);
+	nbp_update_port_count(br);
+
+	dev->priv_flags &= ~IFF_BRIDGE_PORT;
+
+	netdev_rx_handler_unregister(dev);
+
+	netdev_upper_dev_unlink(dev, br->dev);
+
+	br_multicast_del_port(p);
+
+	kobject_uevent(&p->kobj, KOBJ_REMOVE);
+	kobject_del(&p->kobj);
+
+	br_netpoll_disable(p);
+
 	call_rcu(&p->rcu, destroy_nbp_rcu);
 }
 
-/* called with RTNL */
-static void del_br(struct net_bridge *br)
+/* Delete bridge device */
+void br_dev_delete(struct net_device *dev, struct list_head *head)
 {
+	struct net_bridge *br = netdev_priv(dev);
 	struct net_bridge_port *p, *n;
 
 	list_for_each_entry_safe(p, n, &br->port_list, list) {
-		br_sysfs_removeif(p);
 		del_nbp(p);
 	}
 
+	br_fdb_delete_by_port(br, NULL, 1);
+
+	br_vlan_flush(br);
 	del_timer_sync(&br->gc_timer);
 
 	br_sysfs_delbr(br->dev);
- 	unregister_netdevice(br->dev);
-}
-
-static struct net_device *new_bridge_dev(const char *name)
-{
-	struct net_bridge *br;
-	struct net_device *dev;
-
-	dev = alloc_netdev(sizeof(struct net_bridge), name,
-			   br_dev_setup);
-	
-	if (!dev)
-		return NULL;
-
-	br = netdev_priv(dev);
-	br->dev = dev;
-
-	spin_lock_init(&br->lock);
-	INIT_LIST_HEAD(&br->port_list);
-	spin_lock_init(&br->hash_lock);
-
-	br->bridge_id.prio[0] = 0x80;
-	br->bridge_id.prio[1] = 0x00;
-	memset(br->bridge_id.addr, 0, ETH_ALEN);
-
-	br->feature_mask = dev->features;
-	br->stp_enabled = 0;
-	br->designated_root = br->bridge_id;
-	br->root_path_cost = 0;
-	br->root_port = 0;
-	br->bridge_max_age = br->max_age = 20 * HZ;
-	br->bridge_hello_time = br->hello_time = 2 * HZ;
-	br->bridge_forward_delay = br->forward_delay = 15 * HZ;
-	br->topology_change = 0;
-	br->topology_change_detected = 0;
-	br->ageing_time = 300 * HZ;
-	INIT_LIST_HEAD(&br->age_list);
-
-	br_stp_timer_init(br);
-
-	return dev;
+	unregister_netdevice_queue(br->dev, head);
 }
 
 /* find an available port number */
@@ -208,12 +294,11 @@ static int find_portno(struct net_bridge *br)
 	struct net_bridge_port *p;
 	unsigned long *inuse;
 
-	inuse = kmalloc(BITS_TO_LONGS(BR_MAX_PORTS)*sizeof(unsigned long),
+	inuse = kcalloc(BITS_TO_LONGS(BR_MAX_PORTS), sizeof(unsigned long),
 			GFP_KERNEL);
 	if (!inuse)
 		return -ENOMEM;
 
-	memset(inuse, 0, BITS_TO_LONGS(BR_MAX_PORTS)*sizeof(unsigned long));
 	set_bit(0, inuse);	/* zero is reserved */
 	list_for_each_entry(p, &br->port_list, list) {
 		set_bit(p->port_no, inuse);
@@ -225,86 +310,63 @@ static int find_portno(struct net_bridge *br)
 }
 
 /* called with RTNL but without bridge lock */
-static struct net_bridge_port *new_nbp(struct net_bridge *br, 
+static struct net_bridge_port *new_nbp(struct net_bridge *br,
 				       struct net_device *dev)
 {
 	int index;
 	struct net_bridge_port *p;
-	
+
 	index = find_portno(br);
 	if (index < 0)
 		return ERR_PTR(index);
 
-	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	p = kzalloc(sizeof(*p), GFP_KERNEL);
 	if (p == NULL)
 		return ERR_PTR(-ENOMEM);
 
-	memset(p, 0, sizeof(*p));
 	p->br = br;
 	dev_hold(dev);
 	p->dev = dev;
 	p->path_cost = port_cost(dev);
- 	p->priority = 0x8000 >> BR_PORT_BITS;
-	dev->br_port = p;
+	p->priority = 0x8000 >> BR_PORT_BITS;
 	p->port_no = index;
+	p->flags = BR_LEARNING | BR_FLOOD;
 	br_init_port(p);
 	p->state = BR_STATE_DISABLED;
-	INIT_WORK(&p->carrier_check, port_carrier_check, p);
-	kobject_init(&p->kobj);
+	br_stp_port_timer_init(p);
+	br_multicast_add_port(p);
 
 	return p;
 }
 
-int br_add_bridge(const char *name)
+int br_add_bridge(struct net *net, const char *name)
 {
 	struct net_device *dev;
-	int ret;
-
-	dev = new_bridge_dev(name);
-	if (!dev) 
-		return -ENOMEM;
-
-	rtnl_lock();
-	if (strchr(dev->name, '%')) {
-		ret = dev_alloc_name(dev, dev->name);
-		if (ret < 0)
-			goto err1;
-	}
-
-	ret = register_netdevice(dev);
-	if (ret)
-		goto err2;
+	int res;
 
-	/* network device kobject is not setup until
-	 * after rtnl_unlock does it's hotplug magic.
-	 * so hold reference to avoid race.
-	 */
-	dev_hold(dev);
-	rtnl_unlock();
+	dev = alloc_netdev(sizeof(struct net_bridge), name,
+			   br_dev_setup);
 
-	ret = br_sysfs_addbr(dev);
-	dev_put(dev);
+	if (!dev)
+		return -ENOMEM;
 
-	if (ret) 
-		unregister_netdev(dev);
- out:
-	return ret;
+	dev_net_set(dev, net);
+	dev->rtnl_link_ops = &br_link_ops;
 
- err2:
-	free_netdev(dev);
- err1:
-	rtnl_unlock();
-	goto out;
+	res = register_netdev(dev);
+	if (res)
+		free_netdev(dev);
+	return res;
 }
 
-int br_del_bridge(const char *name)
+int br_del_bridge(struct net *net, const char *name)
 {
 	struct net_device *dev;
 	int ret = 0;
 
 	rtnl_lock();
-	dev = __dev_get_by_name(name);
-	if (dev == NULL) 
+	dev = __dev_get_by_name(net, name);
+	if (dev == NULL)
 		ret =  -ENXIO; 	/* Could not find device */
 
 	else if (!(dev->priv_flags & IFF_EBRIDGE)) {
@@ -315,10 +377,10 @@ int br_del_bridge(const char *name)
 	else if (dev->flags & IFF_UP) {
 		/* Not shutdown yet. */
 		ret = -EBUSY;
-	} 
+	}
 
-	else 
-		del_br(netdev_priv(dev));
+	else
+		br_dev_delete(dev, NULL);
 
 	rtnl_unlock();
 	return ret;
@@ -346,22 +408,24 @@ int br_min_mtu(const struct net_bridge *br)
 /*
  * Recomputes features using slave's features
  */
-void br_features_recompute(struct net_bridge *br)
+netdev_features_t br_features_recompute(struct net_bridge *br,
+	netdev_features_t features)
 {
 	struct net_bridge_port *p;
-	unsigned long features, checksum;
+	netdev_features_t mask;
 
-	features = br->feature_mask &~ NETIF_F_IP_CSUM;
-	checksum = br->feature_mask & NETIF_F_IP_CSUM;
+	if (list_empty(&br->port_list))
+		return features;
+
+	mask = features;
+	features &= ~NETIF_F_ONE_FOR_ALL;
 
 	list_for_each_entry(p, &br->port_list, list) {
-		if (!(p->dev->features 
-		      & (NETIF_F_IP_CSUM|NETIF_F_NO_CSUM|NETIF_F_HW_CSUM)))
-			checksum = 0;
-		features &= p->dev->features;
+		features = netdev_increment_features(features,
+						     p->dev->features, mask);
 	}
 
-	br->dev->features = features | checksum | NETIF_F_LLTX;
+	return features;
 }
 
 /* called with RTNL */
@@ -369,72 +433,141 @@ int br_add_if(struct net_bridge *br, struct net_device *dev)
 {
 	struct net_bridge_port *p;
 	int err = 0;
+	bool changed_addr;
 
-	if (dev->flags & IFF_LOOPBACK || dev->type != ARPHRD_ETHER)
+	/* Don't allow bridging non-ethernet like devices */
+	if ((dev->flags & IFF_LOOPBACK) ||
+	    dev->type != ARPHRD_ETHER || dev->addr_len != ETH_ALEN ||
+	    !is_valid_ether_addr(dev->dev_addr))
 		return -EINVAL;
 
-	if (dev->hard_start_xmit == br_dev_xmit)
+	/* No bridging of bridges */
+	if (dev->netdev_ops->ndo_start_xmit == br_dev_xmit)
 		return -ELOOP;
 
-	if (dev->br_port != NULL)
+	/* Device is already being bridged */
+	if (br_port_exists(dev))
 		return -EBUSY;
 
-	if (IS_ERR(p = new_nbp(br, dev)))
+	/* No bridging devices that dislike that (e.g. wireless) */
+	if (dev->priv_flags & IFF_DONT_BRIDGE)
+		return -EOPNOTSUPP;
+
+	p = new_nbp(br, dev);
+	if (IS_ERR(p))
 		return PTR_ERR(p);
 
- 	if ((err = br_fdb_insert(br, p, dev->dev_addr)))
-		destroy_nbp(p);
- 
-	else if ((err = br_sysfs_addif(p)))
-		del_nbp(p);
-	else {
-		dev_set_promiscuity(dev, 1);
+	call_netdevice_notifiers(NETDEV_JOIN, dev);
 
-		list_add_rcu(&p->list, &br->port_list);
+	err = dev_set_allmulti(dev, 1);
+	if (err)
+		goto put_back;
 
-		spin_lock_bh(&br->lock);
-		br_stp_recalculate_bridge_id(br);
-		br_features_recompute(br);
-		if ((br->dev->flags & IFF_UP) 
-		    && (dev->flags & IFF_UP) && netif_carrier_ok(dev))
-			br_stp_enable_port(p);
-		spin_unlock_bh(&br->lock);
+	err = kobject_init_and_add(&p->kobj, &brport_ktype, &(dev->dev.kobj),
+				   SYSFS_BRIDGE_PORT_ATTR);
+	if (err)
+		goto err1;
 
-		dev_set_mtu(br->dev, br_min_mtu(br));
-	}
+	err = br_sysfs_addif(p);
+	if (err)
+		goto err2;
+
+	err = br_netpoll_enable(p);
+	if (err)
+		goto err3;
+
+	err = netdev_master_upper_dev_link(dev, br->dev);
+	if (err)
+		goto err4;
+
+	err = netdev_rx_handler_register(dev, br_handle_frame, p);
+	if (err)
+		goto err5;
+
+	dev->priv_flags |= IFF_BRIDGE_PORT;
+
+	dev_disable_lro(dev);
+
+	list_add_rcu(&p->list, &br->port_list);
+
+	nbp_update_port_count(br);
+
+	netdev_update_features(br->dev);
+
+	if (br->dev->needed_headroom < dev->needed_headroom)
+		br->dev->needed_headroom = dev->needed_headroom;
+
+	if (br_fdb_insert(br, p, dev->dev_addr, 0))
+		netdev_err(dev, "failed insert local address bridge forwarding table\n");
+
+	spin_lock_bh(&br->lock);
+	changed_addr = br_stp_recalculate_bridge_id(br);
+
+	if (netif_running(dev) && netif_oper_up(dev) &&
+	    (br->dev->flags & IFF_UP))
+		br_stp_enable_port(p);
+	spin_unlock_bh(&br->lock);
 
+	br_ifinfo_notify(RTM_NEWLINK, p);
+
+	if (changed_addr)
+		call_netdevice_notifiers(NETDEV_CHANGEADDR, br->dev);
+
+	dev_set_mtu(br->dev, br_min_mtu(br));
+
+	kobject_uevent(&p->kobj, KOBJ_ADD);
+
+	return 0;
+
+err5:
+	netdev_upper_dev_unlink(dev, br->dev);
+err4:
+	br_netpoll_disable(p);
+err3:
+	sysfs_remove_link(br->ifobj, p->dev->name);
+err2:
+	kobject_put(&p->kobj);
+	p = NULL; /* kobject_put frees */
+err1:
+	dev_set_allmulti(dev, -1);
+put_back:
+	dev_put(dev);
+	kfree(p);
 	return err;
 }
 
 /* called with RTNL */
 int br_del_if(struct net_bridge *br, struct net_device *dev)
 {
-	struct net_bridge_port *p = dev->br_port;
-	
-	if (!p || p->br != br) 
+	struct net_bridge_port *p;
+	bool changed_addr;
+
+	p = br_port_get_rtnl(dev);
+	if (!p || p->br != br)
 		return -EINVAL;
 
-	br_sysfs_removeif(p);
+	/* Since more than one interface can be attached to a bridge,
+	 * there still maybe an alternate path for netconsole to use;
+	 * therefore there is no reason for a NETDEV_RELEASE event.
+	 */
 	del_nbp(p);
 
 	spin_lock_bh(&br->lock);
-	br_stp_recalculate_bridge_id(br);
-	br_features_recompute(br);
+	changed_addr = br_stp_recalculate_bridge_id(br);
 	spin_unlock_bh(&br->lock);
 
+	if (changed_addr)
+		call_netdevice_notifiers(NETDEV_CHANGEADDR, br->dev);
+
+	netdev_update_features(br->dev);
+
 	return 0;
 }
 
-void __exit br_cleanup_bridges(void)
+void br_port_flags_change(struct net_bridge_port *p, unsigned long mask)
 {
-	struct net_device *dev, *nxt;
-
-	rtnl_lock();
-	for (dev = dev_base; dev; dev = nxt) {
-		nxt = dev->next;
-		if (dev->priv_flags & IFF_EBRIDGE)
-			del_br(dev->priv);
-	}
-	rtnl_unlock();
+	struct net_bridge *br = p->br;
 
+	if (mask & BR_AUTO_MASK)
+		nbp_update_port_count(br);
 }