On Mon, 2011-05-09 at 09:53 -0700, Ben Greear wrote: > >> I think there is no problem with having different beacon > >> intervals, as long as they are all a multiple of > >> the smallest interval and the driver does things properly. > >> > >> I'm not sure ath9k or ath5k currently supports this properly, > >> but there was a patch floating around for a while that did > >> this for ath9k I think... > > > > Yes, in theory that's possible, but apparently no driver actually did > > this correctly. Also, it didn't seem like anyone really cares, and we > > need to enforce some restrictions because otherwise drivers will end up > > doing it wrong, and you'll end up having a beacon interval of 200 while > > advertising 150 for example, which will totally throw off powersaving > > clients. > > > > If you really care greatly about having different beacon intervals (and > > I don't see why you would?) then maybe you can think how we can enforce > > and advertise that to userspace. For now, I'm more comfortable just > > restricting it. > > I guess we could add a flag to the driver when it supports it properly > and modify your logic to check for even multiples instead of just == > if that flag is set? Yes, but it's not just a flag, you also need the smallest divisor that the driver needs, for example at least 50 TU. Also, when multiple channels are used (not right now of course) that might factor into the calculations of when which beacon is sent. 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
