blob: 8ffad1eef07230af6a7a4697908cc994f536059c [file] [log] [blame]
A few virtual drivers (bonding, dummy) predate netlink (~2003), which means that
the kernel had no way to dynamically create network interfaces.
The solution was kernel module paramter which statically precreated a list of
these virtual interfaces. The number was generally set to 1 by default, meaning
that loading the module creates one of its interface.
For compatibility with legacy userspaces this is still kept around. We could
set the parameters to zero, but doing that everywhere is a pain. This just
patches the default values to zero.
diff --git a/drivers/net/dummy.c b/drivers/net/dummy.c
index c4b1b0aa438a..06b00f7a8eab 100644
--- a/drivers/net/dummy.c
+++ b/drivers/net/dummy.c
@@ -43,7 +43,7 @@
#define DRV_NAME "dummy"
-static int numdummies = 1;
+static int numdummies = 0;
/* fake multicast ability */
static void set_multicast_list(struct net_device *dev)
diff --git a/include/uapi/linux/if_bonding.h b/include/uapi/linux/if_bonding.h
index d174914a837d..bf8e2af101a3 100644
--- a/include/uapi/linux/if_bonding.h
+++ b/include/uapi/linux/if_bonding.h
@@ -82,7 +82,7 @@
#define BOND_STATE_ACTIVE 0 /* link is active */
#define BOND_STATE_BACKUP 1 /* link is backup */
-#define BOND_DEFAULT_MAX_BONDS 1 /* Default maximum number of devices to support */
+#define BOND_DEFAULT_MAX_BONDS 0 /* Default maximum number of devices to support */
#define BOND_DEFAULT_TX_QUEUES 16 /* Default number of tx queues per device */