From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mail.atheros.com ([12.36.123.2]:20588 "EHLO mail.atheros.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752697AbZAPVHY (ORCPT ); Fri, 16 Jan 2009 16:07:24 -0500 Received: from mail.atheros.com ([10.10.20.108]) by sidewinder.atheros.com for ; Fri, 16 Jan 2009 13:07:24 -0800 Date: Fri, 16 Jan 2009 13:06:55 -0800 From: "Luis R. Rodriguez" To: Johannes Berg CC: Luis Rodriguez , "linville@tuxdriver.com" , "linux-wireless@vger.kernel.org" Subject: Re: [PATCH 08/13] cfg80211: save original values on regulatory hints Message-ID: <20090116210655.GK29609@tesla> (sfid-20090116_220729_241466_FDBA49F1) References: <1232064746-17134-3-git-send-email-lrodriguez@atheros.com> <1232064746-17134-4-git-send-email-lrodriguez@atheros.com> <1232064746-17134-5-git-send-email-lrodriguez@atheros.com> <1232064746-17134-6-git-send-email-lrodriguez@atheros.com> <1232064746-17134-7-git-send-email-lrodriguez@atheros.com> <1232064746-17134-8-git-send-email-lrodriguez@atheros.com> <1232064746-17134-9-git-send-email-lrodriguez@atheros.com> <1232097921.3854.28.camel@johannes> <20090116163100.GE29609@tesla> <1232138062.3745.6.camel@johannes> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In-Reply-To: <1232138062.3745.6.camel@johannes> Sender: linux-wireless-owner@vger.kernel.org List-ID: On Fri, Jan 16, 2009 at 12:34:22PM -0800, Johannes Berg wrote: > On Fri, 2009-01-16 at 08:31 -0800, Luis R. Rodriguez wrote: > > On Fri, Jan 16, 2009 at 01:25:21AM -0800, Johannes Berg wrote: > > > On Thu, 2009-01-15 at 16:12 -0800, Luis R. Rodriguez wrote: > > > > When a driver issues a regulatory_hint() lets save the received > > > > values as original channel settings. This allows users to change > > > > regulatory domains multiple times while always respecting the driver's > > > > own regulatory setings. > > > > > > This definitely isn't the right way to do things here. This means that > > > my card that's programmed to US can never do channel 13 here, something > > > which on b43 we definitely want to allow. > > > > Ah -- are you sure you want to allow for that? > > Yes. > > > It seems I was misunderstanding a bit > > how things would be done for 11d in ath9k and the fact is that we *don't* allow > > for channels beyond what the programmed regulatory domain allows. This is because > > calibration stuff has only been tested/ensured/certified/call-it-what you want > > for the channels in that regulatory domain SKU. > > Yes, I know why you want this, but I think you should do it slightly > differently. > > > Are you certain that b43 can operate > > well on channels not in their regulatory domain SKU? If so then how about a wiphy flag > > to let drivers pick this. > > I don't see a need for a flag. Well in ath9k we want this, and orig_* stuff is used only for custom regulatory domains. How would we install information on orig_* from a regulatory_hint() on drivers if we have drivres which may or may not want to save that information on orig_* parameters? > We just need to have the driver set > orig_flags rather than cfg80211, no? Oh you mean in the reg_notifier() on REGDOM_SET_BY_DRIVER? Sure, that works as well, if so desired. Luis