Return-path: Received: from pne-smtpout2-sn1.fre.skanova.net ([81.228.11.159]:61808 "EHLO pne-smtpout2-sn1.fre.skanova.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753818AbYA0PCw (ORCPT ); Sun, 27 Jan 2008 10:02:52 -0500 From: "Lars Ericsson" To: "'Stefano Brivio'" Cc: , Subject: RE: [PATCH 2/2] rc80211-pid: add sanity check Date: Sun, 27 Jan 2008 14:53:00 +0100 Message-ID: <007a01c860eb$ee8231b0$0b3ca8c0@gotws1589> (sfid-20080127_150255_927409_8B226D07) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In-Reply-To: <20080127134252.776cebd6@morte> Sender: linux-wireless-owner@vger.kernel.org List-ID: Hi Stefano, What happens is that the while (newidx != sta->txrate) will continue until the newidx reaches 33. Reason is that BIT(33) is the same as BIT(1) In the rate_supported() function. New sta->txrate will be 33 which is far beyond the valid size (12 in my case). In my system that happens to be a 11g speed :o). I have a trace if you are intresting. Its about 260K data but it show How it happens. I think you need trafik (ping) in order for the relulator to operate. Regards Lars -----Original Message----- From: Stefano Brivio [mailto:stefano.brivio@polimi.it] Sent: den 27 januari 2008 13:43 To: Lars Cc: linux-wireless@vger.kernel.org Subject: Re: [PATCH 2/2] rc80211-pid: add sanity check On Sat, 26 Jan 2008 19:40:14 +0000 (UTC) Lars wrote: > Hi, > > I have spent some time debugging this problem from a different angle. > > If you use a 'b/g' STA connecting to a 'b' AP you will run into this > loop in a few seconds. What also happens is that the STA will try > using 'g' speeds. That was the way I located the problem. Thank you. I couldn't really figure out how to reproduce the bug. > In other words, I do not think a WARN_ON is good solution. Please could you try my 1/2 patch? In the meanwhile, I'll try to use your information for more debugging. -- Ciao Stefano