Hi all, Continuing my quest to make things better for drivers that are not on the extreme end of the soft/hard-MAC spectrum like ath9k, like the station state stuff, I'm now thinking about per-mode callbacks. For example, we have per-vif idle today. It's fairly obvious what this means, but the sequences aren't always completely obvious (or even correct). Similarly, calls to bss_info_changed() could appear to happen at any time, even though they really shouldn't. I'm also continuing to think about multi-channel scenarios, and those can be even more complex in the driver/device setup that they need. So I'm thinking that maybe we should have per-interface & per-type calls, modelled more like how cfg80211 works, maybe like this: managed/p2p-client ------------------ - prepare_connect(vif, bssid, channel, ?) - connected(vif, ...) [optional -- if/after assoc success] - disconnect(vif) AP mode ------- - start_ap(vif, bssid, channel, ...) - stop_ap(vif) ... Some of this would not interact well with software-based scanning & remain-on-channel, but I'm beginning to think about pushing that implementation closer to (some) drivers by shared code/callbacks. Thoughts? johannes -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
