From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail.candelatech.com ([208.74.158.172]:50629 "EHLO ns3.lanforge.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752472Ab2GIQCn (ORCPT ); Mon, 9 Jul 2012 12:02:43 -0400 Message-ID: <4FFB009D.7080400@candelatech.com> (sfid-20120709_180258_085752_5B90A744) Date: Mon, 09 Jul 2012 09:02:37 -0700 From: Ben Greear MIME-Version: 1.0 To: Johannes Berg CC: linux-wireless@vger.kernel.org Subject: Re: [PATCH] mac80211: use full power if off-channel References: <1341740130-6199-1-git-send-email-johannes@sipsolutions.net> <4FF99A80.9040700@candelatech.com> <1341822596.4455.6.camel@jlt3.sipsolutions.net> <4FFAEFCB.2080109@candelatech.com> <1341845721.4455.50.camel@jlt3.sipsolutions.net> In-Reply-To: <1341845721.4455.50.camel@jlt3.sipsolutions.net> Content-Type: text/plain; charset=UTF-8; format=flowed Sender: linux-wireless-owner@vger.kernel.org List-ID: On 07/09/2012 07:55 AM, Johannes Berg wrote: > On Mon, 2012-07-09 at 07:50 -0700, Ben Greear wrote: >> On 07/09/2012 01:29 AM, Johannes Berg wrote: >>> On Sun, 2012-07-08 at 07:34 -0700, Ben Greear wrote: >>>> On 07/08/2012 02:35 AM, Johannes Berg wrote: >>>>> From: Johannes Berg >>>>> >>>>> Instead of using full power when scanning, use >>>>> full power when off-channel. This is different >>>>> since off-channel can be remain-on-channel. >>>> >>>> What if we are scanning on our current channel? Shouldn't >>>> it be full-power then as well? >>> >>> Arguable. If we're on the channel that the AP explicitly told us to use >>> less power on, and we're not stopping normal traffic, then I'm not sure >>> we should. >> >> Well, if you are scanning, you might be wanting to find other APs that >> perhaps require full power to reach, so it seems to me it should scan >> at full power... > > Yes, but you're not supposed to send the data packets at the higher > power. So unless you have per-packet power control you'd be breaking > your TPC implementation (which is regulatory relevant) if you do > on-channel scan. > > I suppose the other option would be to skip the optimisation in that > case, but ... I guess I'd rather keep the scan-on-channel optimization and just tx at lower power. Another thing..with your change, we can skip a hardware config at the start of scan-on-channel since it was only there to configure the tx power anyway... In general, I'd like the ability to choose the tx power on a per-packet basis so that one virtual station could tx at a different power-setting than another..but I've no time to work on that now, and I'm not sure hardware would support that kind of thing anyway. Thanks, Ben > > johannes > -- Ben Greear Candela Technologies Inc http://www.candelatech.com