vlan: Fix lockdep warning when vlan dev handle notification

When I open the LOCKDEP config and run these steps:

modprobe 8021q
vconfig add eth2 20
vconfig add eth2.20 30
ifconfig eth2 xx.xx.xx.xx

then the Call Trace happened:

[32524.386288] =============================================
[32524.386293] [ INFO: possible recursive locking detected ]
[32524.386298] 3.14.0-rc2-0.7-default+ #35 Tainted: G           O
[32524.386302] ---------------------------------------------
[32524.386306] ifconfig/3103 is trying to acquire lock:
[32524.386310]  (&vlan_netdev_addr_lock_key/1){+.....}, at: [<ffffffff814275f4>] dev_mc_sync+0x64/0xb0
[32524.386326]
[32524.386326] but task is already holding lock:
[32524.386330]  (&vlan_netdev_addr_lock_key/1){+.....}, at: [<ffffffff8141af83>] dev_set_rx_mode+0x23/0x40
[32524.386341]
[32524.386341] other info that might help us debug this:
[32524.386345]  Possible unsafe locking scenario:
[32524.386345]
[32524.386350]        CPU0
[32524.386352]        ----
[32524.386354]   lock(&vlan_netdev_addr_lock_key/1);
[32524.386359]   lock(&vlan_netdev_addr_lock_key/1);
[32524.386364]
[32524.386364]  *** DEADLOCK ***
[32524.386364]
[32524.386368]  May be due to missing lock nesting notation
[32524.386368]
[32524.386373] 2 locks held by ifconfig/3103:
[32524.386376]  #0:  (rtnl_mutex){+.+.+.}, at: [<ffffffff81431d42>] rtnl_lock+0x12/0x20
[32524.386387]  #1:  (&vlan_netdev_addr_lock_key/1){+.....}, at: [<ffffffff8141af83>] dev_set_rx_mode+0x23/0x40
[32524.386398]
[32524.386398] stack backtrace:
[32524.386403] CPU: 1 PID: 3103 Comm: ifconfig Tainted: G           O 3.14.0-rc2-0.7-default+ #35
[32524.386409] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
[32524.386414]  ffffffff81ffae40 ffff8800d9625ae8 ffffffff814f68a2 ffff8800d9625bc8
[32524.386421]  ffffffff810a35fb ffff8800d8a8d9d0 00000000d9625b28 ffff8800d8a8e5d0
[32524.386428]  000003cc00000000 0000000000000002 ffff8800d8a8e5f8 0000000000000000
[32524.386435] Call Trace:
[32524.386441]  [<ffffffff814f68a2>] dump_stack+0x6a/0x78
[32524.386448]  [<ffffffff810a35fb>] __lock_acquire+0x7ab/0x1940
[32524.386454]  [<ffffffff810a323a>] ? __lock_acquire+0x3ea/0x1940
[32524.386459]  [<ffffffff810a4874>] lock_acquire+0xe4/0x110
[32524.386464]  [<ffffffff814275f4>] ? dev_mc_sync+0x64/0xb0
[32524.386471]  [<ffffffff814fc07a>] _raw_spin_lock_nested+0x2a/0x40
[32524.386476]  [<ffffffff814275f4>] ? dev_mc_sync+0x64/0xb0
[32524.386481]  [<ffffffff814275f4>] dev_mc_sync+0x64/0xb0
[32524.386489]  [<ffffffffa0500cab>] vlan_dev_set_rx_mode+0x2b/0x50 [8021q]
[32524.386495]  [<ffffffff8141addf>] __dev_set_rx_mode+0x5f/0xb0
[32524.386500]  [<ffffffff8141af8b>] dev_set_rx_mode+0x2b/0x40
[32524.386506]  [<ffffffff8141b3cf>] __dev_open+0xef/0x150
[32524.386511]  [<ffffffff8141b177>] __dev_change_flags+0xa7/0x190
[32524.386516]  [<ffffffff8141b292>] dev_change_flags+0x32/0x80
[32524.386524]  [<ffffffff8149ca56>] devinet_ioctl+0x7d6/0x830
[32524.386532]  [<ffffffff81437b0b>] ? dev_ioctl+0x34b/0x660
[32524.386540]  [<ffffffff814a05b0>] inet_ioctl+0x80/0xa0
[32524.386550]  [<ffffffff8140199d>] sock_do_ioctl+0x2d/0x60
[32524.386558]  [<ffffffff81401a52>] sock_ioctl+0x82/0x2a0
[32524.386568]  [<ffffffff811a7123>] do_vfs_ioctl+0x93/0x590
[32524.386578]  [<ffffffff811b2705>] ? rcu_read_lock_held+0x45/0x50
[32524.386586]  [<ffffffff811b39e5>] ? __fget_light+0x105/0x110
[32524.386594]  [<ffffffff811a76b1>] SyS_ioctl+0x91/0xb0
[32524.386604]  [<ffffffff815057e2>] system_call_fastpath+0x16/0x1b

========================================================================

The reason is that all of the vlan dev have the same class key for dev_lock_list, if we up or down the real dev, the notification will change the state for every vlan dev in the vlan group, then the vlan dev will hold netif_addr_lock and the real dev also hold its own netif_addr_lock together, so the warning happened.

The best way to fix the problem is add a new class key for the vlan dev if its real dev is a vlan dev too.

---
 net/8021q/vlan_dev.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

index 566adbf..4632ec8 100644
@@ -506,6 +506,7 @@ static void vlan_dev_set_rx_mode(struct net_device *vlan_dev)
  */
 static struct lock_class_key vlan_netdev_xmit_lock_key;
 static struct lock_class_key vlan_netdev_addr_lock_key;
+static struct lock_class_key vlan_netdev_addr_lock_subkey;
 
 static void vlan_dev_set_lockdep_one(struct net_device *dev,
 				     struct netdev_queue *txq,

@@ -518,9 +519,15 @@ static void vlan_dev_set_lockdep_one(struct net_device *dev,
 
 static void vlan_dev_set_lockdep_class(struct net_device *dev, int subclass)
 {
-	lockdep_set_class_and_subclass(&dev->addr_list_lock,
-				       &vlan_netdev_addr_lock_key,
-				       subclass);
+	if (subclass)
+		lockdep_set_class_and_subclass(&dev->addr_list_lock,
+					       &vlan_netdev_addr_lock_subkey,
+					       subclass);
+	else
+		lockdep_set_class_and_subclass(&dev->addr_list_lock,
+					       &vlan_netdev_addr_lock_key,
+					       subclass);
+
 	netdev_for_each_tx_queue(dev, vlan_dev_set_lockdep_one, &subclass);
 }
 

1.8.0

From: Ding Tianhong <[email protected]>
Date: Thu, 6 Mar 2014 22:13:33 +0800

...

So what happens when one does:

vconfig add eth2 20
vconfig add eth2.20 30
vconfig add eth2.30 40
ifconfig eth2 xx.xx.xx.xx

the old vlan list:

eth2---------------------->eth2.20-------------------->eth2.20.30 (xxx_addr_lock_key)

			(xxx_addr_lock_key)|
					    ---------->eth2.20.50 (xxx_addr_lock_key)

after this patch:

eth2---------------------->eth2.20-------------------->eth2.20.30 (xxx_addr_lock_subkey)

			(xxx_addr_lock_key)|
					    ---------->eth2.20.50 (xxx_addr_lock_subkey)

The eth2.20 will only sync mc ddress with eth2.20.30 or eth2.20.50, and the eth2.20.30 would never sync mc with eth2.20.50, so the xxx_addr_lock_subkey will not meet each other.

But I think if I vconfig add eth2.20.30 50, the warning would happen again, but till now I could not find any scene that use 3 vlan id on a virtual device, so I think it is not a problem.

From: Ding Tianhong <[email protected]>
Date: Fri, 7 Mar 2014 09:55:25 +0800

My example was to show that you aren't really fixing the bug, but rather are papering over it.

In that sense, it is a problem.

I also happen to disagree with your assertion that a 3 layered VLAN is invalid, it is valid. And if we allow the configuration it shouldn't throw locking assertions.

OK, thanks for your advise. I would think and solve this problem once-again.

Regards
Ding