提交 5d261913 编写于 作者: V Veaceslav Falico 提交者: David S. Miller

net: add lower_dev_list to net_device and make a full mesh

This patch adds lower_dev_list list_head to net_device, which is the same
as upper_dev_list, only for lower devices, and begins to use it in the same
way as the upper list.

It also changes the way the whole adjacent device lists work - now they
contain *all* of upper/lower devices, not only the first level. The first
level devices are distinguished by the bool neighbour field in
netdev_adjacent, also added by this patch.

There are cases when a device can be added several times to the adjacent
list, the simplest would be:

     /---- eth0.10 ---\
eth0-		       --- bond0
     \---- eth0.20 ---/

where both bond0 and eth0 'see' each other in the adjacent lists two times.
To avoid duplication of netdev_adjacent structures ref_nr is being kept as
the number of times the device was added to the list.

The 'full view' is achieved by adding, on link creation, all of the
upper_dev's upper_dev_list devices as upper devices to all of the
lower_dev's lower_dev_list devices (and to the lower_dev itself), and vice
versa. On unlink they are removed using the same logic.

I've tested it with thousands vlans/bonds/bridges, everything works ok and
no observable lags even on a huge number of interfaces.

Memory footprint for 128 devices interconnected with each other via both
upper and lower (which is impossible, but for the comparison) lists would be:

128*128*2*sizeof(netdev_adjacent) = 1.5MB

but in the real world we usualy have at most several devices with slaves
and a lot of vlans, so the footprint will be much lower.

CC: "David S. Miller" <davem@davemloft.net>
CC: Eric Dumazet <edumazet@google.com>
CC: Jiri Pirko <jiri@resnulli.us>
CC: Alexander Duyck <alexander.h.duyck@intel.com>
CC: Cong Wang <amwang@redhat.com>
Signed-off-by: NVeaceslav Falico <vfalico@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
上级 aa9d8560
...@@ -1125,6 +1125,7 @@ struct net_device { ...@@ -1125,6 +1125,7 @@ struct net_device {
struct list_head napi_list; struct list_head napi_list;
struct list_head unreg_list; struct list_head unreg_list;
struct list_head upper_dev_list; /* List of upper devices */ struct list_head upper_dev_list; /* List of upper devices */
struct list_head lower_dev_list;
/* currently active device features */ /* currently active device features */
......
...@@ -4369,7 +4369,16 @@ static void net_rx_action(struct softirq_action *h) ...@@ -4369,7 +4369,16 @@ static void net_rx_action(struct softirq_action *h)
struct netdev_adjacent { struct netdev_adjacent {
struct net_device *dev; struct net_device *dev;
/* upper master flag, there can only be one master device per list */
bool master; bool master;
/* indicates that this dev is our first-level lower/upper device */
bool neighbour;
/* counter for the number of times this device was added to us */
u16 ref_nr;
struct list_head list; struct list_head list;
struct rcu_head rcu; struct rcu_head rcu;
struct list_head search_list; struct list_head search_list;
...@@ -4408,18 +4417,34 @@ static bool __netdev_search_upper_dev(struct net_device *dev, ...@@ -4408,18 +4417,34 @@ static bool __netdev_search_upper_dev(struct net_device *dev,
return ret; return ret;
} }
static struct netdev_adjacent *__netdev_find_upper(struct net_device *dev, static struct netdev_adjacent *__netdev_find_adj(struct net_device *dev,
struct net_device *upper_dev) struct net_device *adj_dev,
bool upper)
{ {
struct netdev_adjacent *upper; struct netdev_adjacent *adj;
struct list_head *dev_list;
list_for_each_entry(upper, &dev->upper_dev_list, list) { dev_list = upper ? &dev->upper_dev_list : &dev->lower_dev_list;
if (upper->dev == upper_dev)
return upper; list_for_each_entry(adj, dev_list, list) {
if (adj->dev == adj_dev)
return adj;
} }
return NULL; return NULL;
} }
static inline struct netdev_adjacent *__netdev_find_upper(struct net_device *dev,
struct net_device *udev)
{
return __netdev_find_adj(dev, udev, true);
}
static inline struct netdev_adjacent *__netdev_find_lower(struct net_device *dev,
struct net_device *ldev)
{
return __netdev_find_adj(dev, ldev, false);
}
/** /**
* netdev_has_upper_dev - Check if device is linked to an upper device * netdev_has_upper_dev - Check if device is linked to an upper device
* @dev: device * @dev: device
...@@ -4496,10 +4521,149 @@ struct net_device *netdev_master_upper_dev_get_rcu(struct net_device *dev) ...@@ -4496,10 +4521,149 @@ struct net_device *netdev_master_upper_dev_get_rcu(struct net_device *dev)
} }
EXPORT_SYMBOL(netdev_master_upper_dev_get_rcu); EXPORT_SYMBOL(netdev_master_upper_dev_get_rcu);
static int __netdev_adjacent_dev_insert(struct net_device *dev,
struct net_device *adj_dev,
bool neighbour, bool master,
bool upper)
{
struct netdev_adjacent *adj;
adj = __netdev_find_adj(dev, adj_dev, upper);
if (adj) {
BUG_ON(neighbour);
adj->ref_nr++;
return 0;
}
adj = kmalloc(sizeof(*adj), GFP_KERNEL);
if (!adj)
return -ENOMEM;
adj->dev = adj_dev;
adj->master = master;
adj->neighbour = neighbour;
adj->ref_nr = 1;
INIT_LIST_HEAD(&adj->search_list);
dev_hold(adj_dev);
pr_debug("dev_hold for %s, because of %s link added from %s to %s\n",
adj_dev->name, upper ? "upper" : "lower", dev->name,
adj_dev->name);
if (!upper) {
list_add_tail_rcu(&adj->list, &dev->lower_dev_list);
return 0;
}
/* Ensure that master upper link is always the first item in list. */
if (master)
list_add_rcu(&adj->list, &dev->upper_dev_list);
else
list_add_tail_rcu(&adj->list, &dev->upper_dev_list);
return 0;
}
static inline int __netdev_upper_dev_insert(struct net_device *dev,
struct net_device *udev,
bool master, bool neighbour)
{
return __netdev_adjacent_dev_insert(dev, udev, neighbour, master,
true);
}
static inline int __netdev_lower_dev_insert(struct net_device *dev,
struct net_device *ldev,
bool neighbour)
{
return __netdev_adjacent_dev_insert(dev, ldev, neighbour, false,
false);
}
void __netdev_adjacent_dev_remove(struct net_device *dev,
struct net_device *adj_dev, bool upper)
{
struct netdev_adjacent *adj;
if (upper)
adj = __netdev_find_upper(dev, adj_dev);
else
adj = __netdev_find_lower(dev, adj_dev);
if (!adj)
BUG();
if (adj->ref_nr > 1) {
adj->ref_nr--;
return;
}
list_del_rcu(&adj->list);
pr_debug("dev_put for %s, because of %s link removed from %s to %s\n",
adj_dev->name, upper ? "upper" : "lower", dev->name,
adj_dev->name);
dev_put(adj_dev);
kfree_rcu(adj, rcu);
}
static inline void __netdev_upper_dev_remove(struct net_device *dev,
struct net_device *udev)
{
return __netdev_adjacent_dev_remove(dev, udev, true);
}
static inline void __netdev_lower_dev_remove(struct net_device *dev,
struct net_device *ldev)
{
return __netdev_adjacent_dev_remove(dev, ldev, false);
}
int __netdev_adjacent_dev_insert_link(struct net_device *dev,
struct net_device *upper_dev,
bool master, bool neighbour)
{
int ret;
ret = __netdev_upper_dev_insert(dev, upper_dev, master, neighbour);
if (ret)
return ret;
ret = __netdev_lower_dev_insert(upper_dev, dev, neighbour);
if (ret) {
__netdev_upper_dev_remove(dev, upper_dev);
return ret;
}
return 0;
}
static inline int __netdev_adjacent_dev_link(struct net_device *dev,
struct net_device *udev)
{
return __netdev_adjacent_dev_insert_link(dev, udev, false, false);
}
static inline int __netdev_adjacent_dev_link_neighbour(struct net_device *dev,
struct net_device *udev,
bool master)
{
return __netdev_adjacent_dev_insert_link(dev, udev, master, true);
}
void __netdev_adjacent_dev_unlink(struct net_device *dev,
struct net_device *upper_dev)
{
__netdev_upper_dev_remove(dev, upper_dev);
__netdev_lower_dev_remove(upper_dev, dev);
}
static int __netdev_upper_dev_link(struct net_device *dev, static int __netdev_upper_dev_link(struct net_device *dev,
struct net_device *upper_dev, bool master) struct net_device *upper_dev, bool master)
{ {
struct netdev_adjacent *upper; struct netdev_adjacent *i, *j, *to_i, *to_j;
int ret = 0;
ASSERT_RTNL(); ASSERT_RTNL();
...@@ -4516,22 +4680,76 @@ static int __netdev_upper_dev_link(struct net_device *dev, ...@@ -4516,22 +4680,76 @@ static int __netdev_upper_dev_link(struct net_device *dev,
if (master && netdev_master_upper_dev_get(dev)) if (master && netdev_master_upper_dev_get(dev))
return -EBUSY; return -EBUSY;
upper = kmalloc(sizeof(*upper), GFP_KERNEL); ret = __netdev_adjacent_dev_link_neighbour(dev, upper_dev, master);
if (!upper) if (ret)
return -ENOMEM; return ret;
upper->dev = upper_dev; /* Now that we linked these devs, make all the upper_dev's
upper->master = master; * upper_dev_list visible to every dev's lower_dev_list and vice
INIT_LIST_HEAD(&upper->search_list); * versa, and don't forget the devices itself. All of these
* links are non-neighbours.
*/
list_for_each_entry(i, &upper_dev->upper_dev_list, list) {
list_for_each_entry(j, &dev->lower_dev_list, list) {
ret = __netdev_adjacent_dev_link(i->dev, j->dev);
if (ret)
goto rollback_mesh;
}
}
/* add dev to every upper_dev's upper device */
list_for_each_entry(i, &upper_dev->upper_dev_list, list) {
ret = __netdev_adjacent_dev_link(dev, i->dev);
if (ret)
goto rollback_upper_mesh;
}
/* add upper_dev to every dev's lower device */
list_for_each_entry(i, &dev->lower_dev_list, list) {
ret = __netdev_adjacent_dev_link(i->dev, upper_dev);
if (ret)
goto rollback_lower_mesh;
}
/* Ensure that master upper link is always the first item in list. */
if (master)
list_add_rcu(&upper->list, &dev->upper_dev_list);
else
list_add_tail_rcu(&upper->list, &dev->upper_dev_list);
dev_hold(upper_dev);
call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev); call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev);
return 0; return 0;
rollback_lower_mesh:
to_i = i;
list_for_each_entry(i, &dev->lower_dev_list, list) {
if (i == to_i)
break;
__netdev_adjacent_dev_unlink(i->dev, upper_dev);
}
i = NULL;
rollback_upper_mesh:
to_i = i;
list_for_each_entry(i, &upper_dev->upper_dev_list, list) {
if (i == to_i)
break;
__netdev_adjacent_dev_unlink(dev, i->dev);
}
i = j = NULL;
rollback_mesh:
to_i = i;
to_j = j;
list_for_each_entry(i, &dev->lower_dev_list, list) {
list_for_each_entry(j, &upper_dev->upper_dev_list, list) {
if (i == to_i && j == to_j)
break;
__netdev_adjacent_dev_unlink(i->dev, j->dev);
}
if (i == to_i)
break;
}
__netdev_adjacent_dev_unlink(dev, upper_dev);
return ret;
} }
/** /**
...@@ -4580,16 +4798,28 @@ EXPORT_SYMBOL(netdev_master_upper_dev_link); ...@@ -4580,16 +4798,28 @@ EXPORT_SYMBOL(netdev_master_upper_dev_link);
void netdev_upper_dev_unlink(struct net_device *dev, void netdev_upper_dev_unlink(struct net_device *dev,
struct net_device *upper_dev) struct net_device *upper_dev)
{ {
struct netdev_adjacent *upper; struct netdev_adjacent *i, *j;
ASSERT_RTNL(); ASSERT_RTNL();
upper = __netdev_find_upper(dev, upper_dev); __netdev_adjacent_dev_unlink(dev, upper_dev);
if (!upper)
return; /* Here is the tricky part. We must remove all dev's lower
list_del_rcu(&upper->list); * devices from all upper_dev's upper devices and vice
dev_put(upper_dev); * versa, to maintain the graph relationship.
kfree_rcu(upper, rcu); */
list_for_each_entry(i, &dev->lower_dev_list, list)
list_for_each_entry(j, &upper_dev->upper_dev_list, list)
__netdev_adjacent_dev_unlink(i->dev, j->dev);
/* remove also the devices itself from lower/upper device
* list
*/
list_for_each_entry(i, &dev->lower_dev_list, list)
__netdev_adjacent_dev_unlink(i->dev, upper_dev);
list_for_each_entry(i, &upper_dev->upper_dev_list, list)
__netdev_adjacent_dev_unlink(dev, i->dev);
call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev); call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev);
} }
EXPORT_SYMBOL(netdev_upper_dev_unlink); EXPORT_SYMBOL(netdev_upper_dev_unlink);
...@@ -5850,6 +6080,7 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, ...@@ -5850,6 +6080,7 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
INIT_LIST_HEAD(&dev->unreg_list); INIT_LIST_HEAD(&dev->unreg_list);
INIT_LIST_HEAD(&dev->link_watch_list); INIT_LIST_HEAD(&dev->link_watch_list);
INIT_LIST_HEAD(&dev->upper_dev_list); INIT_LIST_HEAD(&dev->upper_dev_list);
INIT_LIST_HEAD(&dev->lower_dev_list);
dev->priv_flags = IFF_XMIT_DST_RELEASE; dev->priv_flags = IFF_XMIT_DST_RELEASE;
setup(dev); setup(dev);
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册