Return-path: Received: from mfe1.polimi.it ([131.175.12.23]:45672 "EHLO polimi.it" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1756076AbXLPJ4u (ORCPT ); Sun, 16 Dec 2007 04:56:50 -0500 Date: Sun, 16 Dec 2007 10:40:31 +0100 From: Stefano Brivio To: Stefano Brivio Cc: linux-wireless , Mattias Nissler , "John W. Linville" , Johannes Berg Subject: Re: [RFC/T][PATCH 3/3] rc80211-pid: allow for parameters to be set through sysfs Message-ID: <20071216104031.7e70cf56@morte> (sfid-20071216_095654_741506_4FA0CEA1) In-Reply-To: <20071209212842.1515704c@morte> References: <20071209211547.2d7fca32@morte> <20071209212842.1515704c@morte> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Sender: linux-wireless-owner@vger.kernel.org List-ID: On Sun, 9 Dec 2007 21:28:42 +0100 Stefano Brivio wrote: > This patch allows for tuning parameters to be set through sysfs. Note that > this lacks locking, as another copy of the whole parameters data would have > been needed and this won't be the final approach anyway, so don't bother > too much. > > > Signed-off-by: Stefano Brivio > > --- > > Mattias: oops. Just noticed that my previous approach to locking here was > bogus due to misconceptions about sysfs parameters. It's possible to get > locking right here, but quite troublesome. > Everybody: this lacks proper locking. Never apply this to any sane tree, > please. Actually, it looks like we can assume atomic operation by sysfs here. So I'd go with this approach until we don't add the needed hooks to cfg80211. -- Ciao Stefano