Re: [PATCH] cfg80211: 80MHz (11ac) regulatory change

[Johannes Berg <johannes@sipsolutions.net>]

It would be really good if you could convince your email client to put >
in front of each quoted line ... this is really confusing as is.

> Also, I believe there are many more possibilities, since we count from
> the control channel -- ie. for HT HT40+ means secondary channel is above
> the control channel. For VHT 80, you're going to have 4 possibilities:
> 
> |-1-|-2-|-3-|-4-|
> 
> the control channel can be any one of these four I believe? So you'd
> have configurations like
> 
> VHT_CHAN_LAYOUT_0_3
> VHT_CHAN_LAYOUT_1_2
> VHT_CHAN_LAYOUT_2_1
> VHT_CHAN_LAYOUT_3_0
> 
> indicating the number of channels below/above control (for control
> channel 1,2,3,4 respectively). Similarly, for VHT160 you'd have 8
> possibilities:
> 
> |-1-|-2-|-3-|-4-|-5-|-6-|-7-|-8-|
> 
> (which one could again capture as VHT_CHAN_LAYOUT_0_7 etc.)
> 
> [MP] I see your point. But according to 11ac spec, AP will use primary
> chan as specified in HT operation IE chan num. Secondary channel is
> center freq specified in VHT Operation IE. So I am thinking secondary
> channel is not relative offset to primary channel. Hope I am not
> mistaken here.

Ok so HT has primary channel and secondary, and VHT has secondary VHT
which can again be above/below? That would make sense, but you wouldn't
be covering it.

> > +     /* This would happen when regulatory rules disallow VHT80 completely */
> > +     if (IEEE80211_CHAN_NO_VHT80 == (chan->flags & (IEEE80211_CHAN_NO_VHT80)))
> > +             return true;
> 
> Is that really right? Need to document what the return value of this
> function should be, I guess?
> 
> [MP] I guess, it's possible for a channel not allowed for 80Mhz operation.

Yeah but should it really check *all* the bits rather than any one of
them?

> > +     /*
> > +      * Please note that this assumes target bandwidth is 40 MHz,
> > +      * if that ever changes we also need to change the below logic
> > +      * to include that as well.
> > +      */
> 
> ???
> 
> [MP] Can you explain? This function doesn't make any sense?

The comment about 40 MHz doesn't make any sense.

johannes