Return-path: Received: from rv-out-0910.google.com ([209.85.198.185]:1734 "EHLO rv-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933484AbYCEL7W (ORCPT ); Wed, 5 Mar 2008 06:59:22 -0500 Received: by rv-out-0910.google.com with SMTP id k20so870289rvb.1 for ; Wed, 05 Mar 2008 03:59:19 -0800 (PST) Message-ID: (sfid-20080305_115928_724010_432923AD) Date: Wed, 5 Mar 2008 13:59:19 +0200 From: "Ron Rindjunsky" To: "Johannes Berg" Subject: Re: [PATCH 05/11] mac80211: adding mac80211_tx_control_flags andHT flags Cc: "Winkler, Tomas" , "Chatre, Reinette" , linville@tuxdriver.com, linux-wireless@vger.kernel.org, ipw3945-devel@lists.sourceforge.net In-Reply-To: <1204715685.25502.66.camel@johannes.berg> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 References: <1204682972-26004-1-git-send-email-reinette.chatre@intel.com> <1204682972-26004-5-git-send-email-reinette.chatre@intel.com> <1204682972-26004-6-git-send-email-reinette.chatre@intel.com> <1204710116.25502.48.camel@johannes.berg> <1204711417.25502.59.camel@johannes.berg> <1204713755.25502.62.camel@johannes.berg> <1204715685.25502.66.camel@johannes.berg> Sender: linux-wireless-owner@vger.kernel.org List-ID: > > > It seems to me that these flags aren't entirely orthogonal to the > > > bitrates? > > > > no, they are all orthogonal to each other. the example you gave will > > give us 108MBit. all flags can be used, as long they match the spec, > > and that's the algorithm responsibility to check and determine. > > I guess that the flag that is more confusing is OFDM_HT. since 11n > > supports (theoretically...) n streams of data, using flags like SISO / > > MIMO / MIMO_3_ANT / MIMO_... seemed the wrong way. in stead, once you > > know OFDM_HT is up, a check to antenna_sel_tx will determine the Tx > > expected method. > > Ahh! Ok, thanks for the explanation. Is antenna_sel_tx really the right > thing though? Tomas mentioned that you can have three antennas and only > two chains so should that rather be a chain selection? > yes, this is correct from low level driver's perspective, but i think that from rate scaling perspective keeping the antenna term is clearer then a chain. > > It may be good to add this to the file's documentation. > > Indeed. Or to a DOC: section somewhere so it can be added to an 11n > chapter of the mac80211 book (since the patch is merged I'm > automatically generating it nightly at > http://linuxwireless.org/mac80211book/) > I'll make a patch with the explanation