Message-ID: <1403535525.10968.7.camel@lolumad> (sfid-20140623_165916_679557_1C150DAC)
Subject: Re: [PATCH 2/2] cfg80211: Use 5MHz bandwidth by default when
 checking usable channels
From: Rostislav Lisovy <lisovy@gmail.com>
To: Felix Fietkau <nbd@openwrt.org>
Cc: Johannes Berg <johannes@sipsolutions.net>,
	"John W. Linville" <linville@tuxdriver.com>,
	linux-wireless@vger.kernel.org, linux-kernel@vger.kernel.org,
	Michal Sojka <sojkam1@fel.cvut.cz>, s.sander@nordsys.de,
	jan-niklas.meier@volkswagen.de,
	Rostislav Lisovy <rostislav.lisovy@fel.cvut.cz>
Date: Mon, 23 Jun 2014 16:58:45 +0200
In-Reply-To: <53A83DE1.4050106@openwrt.org>
References: <1397565476-4348-1-git-send-email-rostislav.lisovy@fel.cvut.cz>
		 <1397565476-4348-3-git-send-email-rostislav.lisovy@fel.cvut.cz>
		 <53A6C0FF.9090104@openwrt.org>
		 <1403514536.4418.6.camel@jlt4.sipsolutions.net>
		 <53A7FBB2.2050703@openwrt.org>
		 <1403517938.4418.16.camel@jlt4.sipsolutions.net>
	 <1403534300.7341.6.camel@lolumad> <53A83DE1.4050106@openwrt.org>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Sender: linux-wireless-owner@vger.kernel.org

On Mon, 2014-06-23 at 16:46 +0200, Felix Fietkau wrote:
> On 2014-06-23 16:38, Rostislav Lisovy wrote:
> > On Mon, 2014-06-23 at 12:05 +0200, Johannes Berg wrote:
> >> On Mon, 2014-06-23 at 12:04 +0200, Felix Fietkau wrote:
> >> > On 2014-06-23 11:08, Johannes Berg wrote:
> >> > > On Sun, 2014-06-22 at 13:41 +0200, Felix Fietkau wrote:
> >> > >> On 2014-04-15 14:37, Rostislav Lisovy wrote:
> >> > >> > Current code checks if the 20MHz bandwidth is allowed for
> >> > >> > particular channel -- if it is not, the channel is disabled.
> >> > >> > Since we need to use 5/10 MHz channels, this code is modified in
> >> > >> > the way that the default bandwidth to check is 5MHz. If the
> >> > >> > maximum bandwidth allowed by the channel is smaller than 5MHz,
> >> > >> > the channel is disabled. Otherwise the channel is used and the
> >> > >> > flags are set according to the bandwidth allowed by the channel.
> >> > >> > 
> >> > >> > Signed-off-by: Rostislav Lisovy <rostislav.lisovy@fel.cvut.cz>
> >> > > 
> >> > >> This change causes a regression and needs to be reverted or fixed.
> >> > > 
> >> > > Noted, I've reverted it in mac80211.git.
> >> > > 
> >> > >> It leaves Channel 12 enabled for US regdomain and does not prevent
> >> > >> bringing up AP mode on it (IEEE80211_CHAN_NO_20MHZ does not get set).
> >> > > 
> >> > > I'm not sure this makes sense - CHAN_NO_20MHZ shouldn't get set on that
> >> > > channel? It should be disabled for other reasons for AP mode - e.g.
> >> > > NO_IR.
> >> > I was thinking it could be valid in AP mode for 5 MHz operation.
> >> 
> >> Huh, ok, maybe. I guess Rostislav can look into the details :)
> >> 
> > 
> > I am a bit confused right now. The regression is that the channel used
> > to be completely disabled and now it is not? The regulatory restrictions
> > prohibit the 20MHz BW for this channel but 5MHz BW is alright?
> Correct. The matching rule is this:
>         (2402 - 2472 @ 40), (N/A, 30), (N/A)
> 
> Channel 11 (2462 MHz) is allowed, because 2472 MHz is below the limit.
> Channel 12 (2467 MHz) is not allowed, because 2477 MHz is too high.
> 
> With your change, Channel 12 was enabled and usable for AP mode.

Thank you Felix, this makes sense.

I think the issue is the call
  bw_fits = reg_does_bw_fit(fr, center_freq, MHZ_TO_KHZ(5));
which checks if "2467 + 5" is equal or less than 2472 (which indeed is),
when the result is true, the maximum allowed BW (max_bandwidth_khz) is
wrongly used on that particular channel.

I hope I will send a fix on Friday.

Rostislav;