Johannes Berg wrote:
On Tue, 2012-06-26 at 14:37 +0200, Michal Kazior wrote:Hi, Here's another respin of channel context patches. Changes: - moved ieee80211_chan_mode - updated chanctx_conf comment - replaced vif_list with refcount - added lockdeps - channel type downgrading support (when a channel context is unassigned from a vif channel type is recalculated for the remaining vifs sharing the same channel context)Except for that last point which seems to have implementation issues, this looks really good too! How are you planning to proceed from here? I'm happy to pick these up even if you're not going to work on using it right away since I think in that case I would probably be working on using it pretty soon, but if you are planning to work on it that's even better :-)
The next step is to start using channel contexts in mac80211. I was thinking of using chanctx->channel instead of hw.conf.channel if .assign_vif_chanctx is defined. I expect some issues like using channel before it is set.
Hwsim requires .hw_scan and ROC to be implemented to test channel contexts. hw_scan should be simple (if we skip a few things..) but haven't looked into the ROC case yet. It also requires some changes to move channel logic per-vif.
Depending on how things turn out I may start to tackle on mac80211 a bit more and hw_scan for hwsim. I may send a few patches on monday.
I'll of course resend this patchset today once I fix the pointed out issues. -- Pozdrawiam / Best regards, Michal Kazior. -- 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
