(In reply to Takashi Iwai from comment #1) > Hm, are you sure that this is the real fix commit? It looks merely like a > conversion of the container_of() cast with a helper function (plus a > different way to retrieve dev pointer). Ineed, this is an unintentional fix for a real issue. Please refer to the discussion under original patch before it was accepted for stable backport: https://www.spinics.net/lists/netdev/msg690904.html