Return-path: Received: from mail-fx0-f158.google.com ([209.85.220.158]:48782 "EHLO mail-fx0-f158.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753555AbZDPPp6 convert rfc822-to-8bit (ORCPT ); Thu, 16 Apr 2009 11:45:58 -0400 Received: by fxm2 with SMTP id 2so497011fxm.37 for ; Thu, 16 Apr 2009 08:45:56 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <49E74D46.6030707@nets.rwth-aachen.de> References: <49E71982.5050204@bitnet.be> <40f31dec0904160611l58500309wa1ba92fe3aeff843@mail.gmail.com> <69e28c910904160637j5769be3au94821d48bfecf6f2@mail.gmail.com> <49E74D46.6030707@nets.rwth-aachen.de> Date: Thu, 16 Apr 2009 18:45:56 +0300 Message-ID: <40f31dec0904160845w4a6a55a8w27888c701f4b3340@mail.gmail.com> (sfid-20090416_174600_717859_99F0F7B1) Subject: Re: how to set antenna iw + ath5k From: Nick Kossifidis To: Arnd Hannemann Cc: =?UTF-8?Q?G=C3=A1bor_Stefanik?= , Hans Maes , "linux-wireless@vger.kernel.org" Content-Type: text/plain; charset=UTF-8 Sender: linux-wireless-owner@vger.kernel.org List-ID: 2009/4/16 Arnd Hannemann : > G=C3=A1bor Stefanik wrote: >> On Thu, Apr 16, 2009 at 3:11 PM, Nick Kossifidis wrote: >> >>> 2009/4/16 Hans Maes : >>> >>>> Hello, >>>> >>>> I'm not sure this is the correct place to ask my question, but I'v= e been >>>> searching on google for ages now and can't seem to find an answer = so I'm >>>> asking it here. >>>> If not, please yell at me and tell me where to go whine instead ;-= ) >>>> >>>> I'm using CM11 a/b/g cards with the Atheros AR5414A Chip, which ha= ve 2 >>>> antenna connectors. >>>> I only have 1 antenna connected to the MAIN connector. >>>> Running a self compiled 2.6.29 kernel on a few debian lenny boxes = with ath5k >>>> module for the wireless cards, set up in mesh mode. >>>> Seems to work fine on the test bench, but behaves rather poorly in >>>> production environment. >>>> Everything related to the mesh network setup seems to work, but I = have about >>>> 40% - 60% packet loss on the entire netwerk (about 15 nodes) >>>> >>>> I'm guessing this has something to do with the antenna diversity s= etting. >>>> >>>> Can anybody tell me how to tell the ath5k module to disable antenn= a >>>> diversity and only use the MAIN antenna ? >>>> >>>> If this is a RTFM thingie, then I apologize, I just can't seem to = find it >>>> anywhere. >>>> >>>> Please reply to my personal e-mail address as well, I'm not subscr= ibed to >>>> the list. >>>> >>>> Thanks for your help. >>>> >>>> Regards, >>>> >>>> Hans >>>> >>> After this patch... >>> https://lists.ath5k.org/pipermail/ath5k-devel/2009-April/002264.htm= l >>> >>> It's possible to tell the driver to use one fixed antenna (in your >>> case its fixed antenna A). Just change... >>> ath5k_hw_set_antenna_mode(ah, AR5K_ANTMODE_DEFAULT); >>> >>> with >>> ath5k_hw_set_antenna_mode(ah, AR5K_ANTMODE_FIXED_A); >>> >>> we will have a better way to change this in the future ;-) >>> >> >> I wonder if just exporting that as a modparam is a good-enough >> choice... though perhaps the best solution would be something like "= iw >> phy phy0 antenna 1 up/down". Or perhaps we should make antenna >> diversity depend on the PCI ID, defaulting to 1 antenna when diversi= ty >> support cannot be determined. >> >> > modparam seems somewhat broken, thinking of multiple wireless cards..= =2E > My feature request would be to allow setup fixed txantenna only for > broadcast/multicast frames. Well such frames should be sent on all antennas i believe (i did some experiments and it seems we can force this by writing 0xf on antenna field of tx descriptor). > I think it would work for almost all setups, even if only > one antenna connector is used, as diversity logic > will naturally handle unicast frames right. Or did I miss something? > > BTW: is it theoretically possible to turn off RX antenna diversity, a= nd > what would be the potential gain? > Yes, check out the patch. The only problem i see with rx diversity in some cases is when you get a frame on the wrong antenna, this can lead on updating tx antenna (based on acks etc) which can lead to missed frames. --=20 GPG ID: 0xD21DB2DB As you read this post global entropy rises. Have Fun ;-) Nick -- To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html