* 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
#include <linux/module.h>
#include <linux/init.h>
#include <linux/rtnetlink.h>
+#include <linux/if_ether.h>
#include <net/sock.h>
#include "br_private.h"
* 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;
-
- 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;
- case SPEED_10000:
- return 2;
- case SPEED_10:
- return 100;
+ if (dev->ethtool_ops && dev->ethtool_ops->get_settings) {
+ struct ethtool_cmd ecmd = { .cmd = ETHTOOL_GSET, };
+
+ if (!dev->ethtool_ops->get_settings(dev, &ecmd)) {
+ switch(ecmd.speed) {
+ case SPEED_10000:
+ return 2;
+ case SPEED_1000:
+ return 4;
+ case SPEED_100:
+ return 19;
+ case SPEED_10:
+ return 100;
+ }
}
}
* 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)
+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 (netif_carrier_ok(dev))
+ p->path_cost = port_cost(dev);
- spin_lock_bh(&p->br->lock);
- if (p->state == BR_STATE_DISABLED) {
- p->path_cost = cost;
- br_stp_enable_port(p);
+ if (netif_running(br->dev)) {
+ spin_lock_bh(&br->lock);
+ if (netif_carrier_ok(dev)) {
+ if (p->state == BR_STATE_DISABLED)
+ br_stp_enable_port(p);
+ } else {
+ if (p->state != BR_STATE_DISABLED)
+ br_stp_disable_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);
+ spin_unlock_bh(&br->lock);
}
- rtnl_unlock();
}
+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;
p->dev = NULL;
dev_put(dev);
- br_sysfs_freeif(p);
+ kobject_put(&p->kobj);
}
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, dev->name);
- cancel_delayed_work(&p->carrier_check);
- flush_scheduled_work();
+ dev_set_promiscuity(dev, -1);
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);
+
+ br_fdb_delete_by_port(br, p, 1);
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);
-
+ rcu_assign_pointer(dev->br_port, NULL);
+
+ kobject_uevent(&p->kobj, KOBJ_REMOVE);
+ kobject_del(&p->kobj);
+
call_rcu(&p->rcu, destroy_nbp_rcu);
}
/* called with RTNL */
-static void del_br(struct net_bridge *br)
+static void del_br(struct net_bridge *br, struct list_head *head)
{
struct net_bridge_port *p, *n;
list_for_each_entry_safe(p, n, &br->port_list, list) {
- br_sysfs_removeif(p);
del_nbp(p);
}
del_timer_sync(&br->gc_timer);
br_sysfs_delbr(br->dev);
- unregister_netdevice(br->dev);
+ unregister_netdevice_queue(br->dev, head);
}
-static struct net_device *new_bridge_dev(const char *name)
+static struct net_device *new_bridge_dev(struct net *net, 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;
+ dev_net_set(dev, net);
br = netdev_priv(dev);
br->dev = dev;
br->bridge_id.prio[0] = 0x80;
br->bridge_id.prio[1] = 0x00;
- memset(br->bridge_id.addr, 0, ETH_ALEN);
+
+ memcpy(br->group_addr, br_group_address, ETH_ALEN);
br->feature_mask = dev->features;
- br->stp_enabled = 0;
+ br->stp_enabled = BR_NO_STP;
br->designated_root = br->bridge_id;
br->root_path_cost = 0;
br->root_port = 0;
br->topology_change = 0;
br->topology_change_detected = 0;
br->ageing_time = 300 * HZ;
- INIT_LIST_HEAD(&br->age_list);
+
+ br_netfilter_rtable_init(br);
br_stp_timer_init(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);
}
/* 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 = 0;
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);
return p;
}
-int br_add_bridge(const char *name)
+static struct device_type br_type = {
+ .name = "bridge",
+};
+
+int br_add_bridge(struct net *net, const char *name)
{
struct net_device *dev;
int ret;
- dev = new_bridge_dev(name);
- if (!dev)
+ dev = new_bridge_dev(net, name);
+ if (!dev)
return -ENOMEM;
rtnl_lock();
if (strchr(dev->name, '%')) {
ret = dev_alloc_name(dev, dev->name);
if (ret < 0)
- goto err1;
+ goto out_free;
}
+ SET_NETDEV_DEVTYPE(dev, &br_type);
+
ret = register_netdevice(dev);
if (ret)
- goto err2;
-
- /* 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();
+ goto out_free;
ret = br_sysfs_addbr(dev);
- dev_put(dev);
-
- if (ret)
- unregister_netdev(dev);
+ if (ret)
+ unregister_netdevice(dev);
out:
+ rtnl_unlock();
return ret;
- err2:
+out_free:
free_netdev(dev);
- err1:
- rtnl_unlock();
goto out;
}
-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)) {
else if (dev->flags & IFF_UP) {
/* Not shutdown yet. */
ret = -EBUSY;
- }
+ }
- else
- del_br(netdev_priv(dev));
+ else
+ del_br(netdev_priv(dev), NULL);
rtnl_unlock();
return ret;
}
-/* Mtu of the bridge pseudo-device 1500 or the minimum of the ports */
+/* MTU of the bridge pseudo-device: ETH_DATA_LEN or the minimum of the ports */
int br_min_mtu(const struct net_bridge *br)
{
const struct net_bridge_port *p;
ASSERT_RTNL();
if (list_empty(&br->port_list))
- mtu = 1500;
+ mtu = ETH_DATA_LEN;
else {
list_for_each_entry(p, &br->port_list, list) {
if (!mtu || p->dev->mtu < mtu)
void br_features_recompute(struct net_bridge *br)
{
struct net_bridge_port *p;
- unsigned long features, checksum;
+ unsigned long features, mask;
+
+ features = mask = br->feature_mask;
+ if (list_empty(&br->port_list))
+ goto done;
- features = br->feature_mask &~ NETIF_F_IP_CSUM;
- checksum = br->feature_mask & NETIF_F_IP_CSUM;
+ 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;
+done:
+ br->dev->features = netdev_fix_features(features, NULL);
}
/* called with RTNL */
struct net_bridge_port *p;
int err = 0;
- 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)
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;
+ /* Device is already being bridged */
if (dev->br_port != NULL)
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);
+ err = dev_set_promiscuity(dev, 1);
+ if (err)
+ goto put_back;
- list_add_rcu(&p->list, &br->port_list);
+ err = kobject_init_and_add(&p->kobj, &brport_ktype, &(dev->dev.kobj),
+ SYSFS_BRIDGE_PORT_ATTR);
+ if (err)
+ goto err0;
- 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 = br_fdb_insert(br, p, dev->dev_addr);
+ if (err)
+ goto err1;
- dev_set_mtu(br->dev, br_min_mtu(br));
- }
+ err = br_sysfs_addif(p);
+ if (err)
+ goto err2;
+
+ rcu_assign_pointer(dev->br_port, p);
+ dev_disable_lro(dev);
+
+ list_add_rcu(&p->list, &br->port_list);
+ spin_lock_bh(&br->lock);
+ br_stp_recalculate_bridge_id(br);
+ br_features_recompute(br);
+
+ if ((dev->flags & IFF_UP) && netif_carrier_ok(dev) &&
+ (br->dev->flags & IFF_UP))
+ br_stp_enable_port(p);
+ spin_unlock_bh(&br->lock);
+
+ br_ifinfo_notify(RTM_NEWLINK, p);
+
+ dev_set_mtu(br->dev, br_min_mtu(br));
+
+ kobject_uevent(&p->kobj, KOBJ_ADD);
+
+ return 0;
+err2:
+ br_fdb_delete_by_port(br, p, 1);
+err1:
+ kobject_put(&p->kobj);
+ p = NULL; /* kobject_put frees */
+err0:
+ dev_set_promiscuity(dev, -1);
+put_back:
+ dev_put(dev);
+ kfree(p);
return err;
}
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)
+
+ if (!p || p->br != br)
return -EINVAL;
- br_sysfs_removeif(p);
del_nbp(p);
spin_lock_bh(&br->lock);
return 0;
}
-void __exit br_cleanup_bridges(void)
+void __net_exit br_net_exit(struct net *net)
{
- struct net_device *dev, *nxt;
+ struct net_device *dev;
+ LIST_HEAD(list);
rtnl_lock();
- for (dev = dev_base; dev; dev = nxt) {
- nxt = dev->next;
+ for_each_netdev(net, dev)
if (dev->priv_flags & IFF_EBRIDGE)
- del_br(dev->priv);
- }
+ del_br(netdev_priv(dev), &list);
+
+ unregister_netdevice_many(&list);
rtnl_unlock();
}