Return-path: Received: from mail.atheros.com ([12.36.123.2]:39145 "EHLO mail.atheros.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752270AbYIHFoQ (ORCPT ); Mon, 8 Sep 2008 01:44:16 -0400 Date: Mon, 8 Sep 2008 11:08:51 +0530 From: Vasanthakumar Thiagarajan To: Marcel Holtmann CC: Vasanth Thiagarajan , "linville@tuxdriver.com" , "linux-wireless@vger.kernel.org" , Jouni Malinen , Luis Rodriguez Subject: Re: [PATCH] Ath9k: Add RF kill support. Message-ID: <20080908053851.GC28149@vasanth-lnx.users.atheros.com> (sfid-20080908_074424_415864_35998CB6) References: <20080906123842.GA27504@vasanth-lnx.users.atheros.com> <1220721347.6714.114.camel@californication> <20080908045926.GA28149@vasanth-lnx.users.atheros.com> <1220851340.6714.132.camel@californication> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In-Reply-To: <1220851340.6714.132.camel@californication> Sender: linux-wireless-owner@vger.kernel.org List-ID: On Mon, Sep 08, 2008 at 10:52:20AM +0530, Marcel Holtmann wrote: > Hi Vasanthakumar, > > > > > CONFIG_ATH9K_RFKILL needs to be set > > > > to enable this support. > > > > > > why do we have to introduce yet another config option for this? What is > > > the advantage of disabling RFKILL support? I really don't see it and all > > > these extra config options are rather confusing than useful. > > Normally home users may not hit a situation where they have to > > disable the radio, so they can compile out the rf kill support > > and essentially disable the periodic run of the work queue > > (rfkill_poll ,runs for every 2 secs).If this does not make > > any sense, I will remove this config option. > > I don't know about the actual implementation, but if you use a polling > mechanism then it is wrong. Can't you make this interrupt driven. There is some issue in GPIO intrrupt genenration, we can't make it interrupt driven. > > The desktop use case needs RFKILL support, because that will be used for > switching off the radio. Just think about the flight mode case. So the > only case where disabling RFKILL support makes sense would be an > embedded system with specific constraints. However even in the embedded > case, I don't see that usefulness. > Yeah, I understand. > Regards > > Marcel > Regards Vasanth >