net: linkwatch: add check for netdevice being present to linkwatch_do_dev
authorHeiner Kallweit <hkallweit1@gmail.com>
Tue, 18 Sep 2018 19:55:36 +0000 (21:55 +0200)
committerDavid S. Miller <davem@davemloft.net>
Thu, 20 Sep 2018 04:06:46 +0000 (21:06 -0700)
When bringing down the netdevice (incl. detaching it) and calling
netif_carrier_off directly or indirectly the latter triggers an
asynchronous linkwatch event.
This linkwatch event eventually may fail to access chip registers in
the ndo_get_stats/ndo_get_stats64 callback because the device isn't
accessible any longer, see call trace in [0].

To prevent this scenario don't check for IFF_UP only, but also make
sure that the netdevice is present.

[0] https://lists.openwall.net/netdev/2018/03/15/62

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/core/link_watch.c

index e38e641e98d5b5c5ec92ed07aa3ce3d2af6f50e3..7f51efb2b3ab9715cff64f61072e1b9b242a3a5b 100644 (file)
@@ -155,7 +155,7 @@ static void linkwatch_do_dev(struct net_device *dev)
        clear_bit(__LINK_STATE_LINKWATCH_PENDING, &dev->state);
 
        rfc2863_policy(dev);
-       if (dev->flags & IFF_UP) {
+       if (dev->flags & IFF_UP && netif_device_present(dev)) {
                if (netif_carrier_ok(dev))
                        dev_activate(dev);
                else