Return-path: Received: from nbd.name ([46.4.11.11]:49489 "EHLO nbd.name" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754719Ab1KCRah (ORCPT ); Thu, 3 Nov 2011 13:30:37 -0400 Message-ID: <4EB2CFB3.2060802@openwrt.org> (sfid-20111103_183045_992727_CB439558) Date: Thu, 03 Nov 2011 18:30:27 +0100 From: Felix Fietkau MIME-Version: 1.0 To: Zefir Kurtisi CC: Christian Lamparter , linux-wireless@vger.kernel.org, ath9k-devel@lists.ath9k.org, rodrigue@qca.qualcomm.com Subject: Re: [RFC v2 2/2] ath9k: integrate initial DFS module References: <1320328553-28066-1-git-send-email-zefir.kurtisi@neratec.com> <1320328553-28066-3-git-send-email-zefir.kurtisi@neratec.com> <201111031651.46073.chunkeey@googlemail.com> <4EB2CE96.4060702@neratec.com> In-Reply-To: <4EB2CE96.4060702@neratec.com> Content-Type: text/plain; charset=ISO-8859-15; format=flowed Sender: linux-wireless-owner@vger.kernel.org List-ID: On 2011-11-03 6:25 PM, Zefir Kurtisi wrote: > On 11/03/2011 04:51 PM, Christian Lamparter wrote: >> On Thursday, November 03, 2011 02:55:53 PM Zefir Kurtisi wrote: >>> This patch integrates the DFS module into ath9k, including >>> * building the module into ath9k_hw >>> * setting up DFS debugfs >>> * defining HW capability flag for DFS support >>> * setting this flag by DFS supporting devices >>> (so far: AR_SREV_9280_20_OR_LATER, TBC) >>> * setting PHYRADAR rx filter flag to enable radar >>> pulse reporting >>> * forward radar PHY errors to dfs module >>> >>> This is WIP and at its current stage is limited to test ath9k >>> pulse detection capabilities. The DFS pattern matching is >>> TBD in the higher layers and is not part of this patch. >>> >>> CONFIG_ATH9K_DFS must be set to enable pulse detection. >> where? There's no kconfig option, yet it has the CONFIG_ perifx. >> So people will be looking for it. >> >> This is exactly why I made such a big fuss about CONFIG_XYZ issue last time. >> >> Just drop the ifdef. >> >> > Hi Christian, > > I just followed Luis' request to follow the regulatory statement and > make non regulatory certified code unusable by the 'common user'. > > Therefore, not (yet) selectable as a kconfig option, but for devs and > testers to be easily enabled. > > At this stage, removing the ifdef would not harm, because all the > code does is counting events. But for the long run I understand it is > a sensitive topic and requires to be disabled (at least as default > setting). I don't think it's a sensitive topic at all. Just leave out all the ifdef stuff. In the long run when we actually get to the part where we can test a full DFS implementation, the only thing that needs to be left out for normal users is the flag announcing support for it. - Felix