On Sat, 2021-12-18 at 00:36 +0100, Greg KH wrote: > > Yeah, I don't think it's needed, but as it was marked to go back really > far, I wanted to give you the option to do so or not. > Yeah I marked it because I'm pretty sure that problem _was_ introduced that far back with the TXQ code, but OTOH I'm not even sure initially it had users there, and then users that would run into the restart issue ... probably not, otherwise I'd think we'd have heard about it earlier? Backporting the IEEE80211_TXQ_STOP_NETIF_TX stuff doesn't seem that problematic either, but I'm not convinced it's worth it, and I suppose somebody whose driver actually uses it on those kernels would better take a look - I'm not familiar with all the drivers in detail, and iwlwifi doesn't use it yet there. johannes