On Mon, 2025-06-23 at 01:24 +0300, Andrei Otcheretianski wrote: > Implement and export a new API that will be used by drivers to notify > user space about upcoming NAN Discovery Window (DW). Seems there's not that much point in splitting a 67 line patch and a 7 line patch (#3) that really belong together? > + if (!wdev->owner_nlportid) > + genlmsg_multicast_netns(&nl80211_fam, wiphy_net(&rdev->wiphy), > + msg, 0, NL80211_MCGRP_NAN, gfp); > + else > + genlmsg_unicast(wiphy_net(&rdev->wiphy), msg, > + wdev->owner_nlportid); I feel like perhaps the new design should just enforce the owner? Otherwise how would all of this possibly work ... any application starts listening and sending frames? Seems like a recipe for confusion. johannes
