Return-path: Received: from static-ip-62-75-166-246.inaddr.intergenia.de ([62.75.166.246]:41498 "EHLO vs166246.vserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1762026AbXFAUmA (ORCPT ); Fri, 1 Jun 2007 16:42:00 -0400 From: Michael Buesch To: James Ketrenos Subject: Re: [PATCH] mac80211: Update stop_queues kdoc Date: Fri, 1 Jun 2007 22:41:04 +0200 Cc: Michael Wu , Jiri Benc , John Linville , linux-wireless@vger.kernel.org References: <200706011129.12432.mb@bu3sch.de> <200706011204.23666.flamingice@sourmilk.net> <46605814.1090806@linux.intel.com> In-Reply-To: <46605814.1090806@linux.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Message-Id: <200706012241.05103.mb@bu3sch.de> Sender: linux-wireless-owner@vger.kernel.org List-ID: On Friday 01 June 2007 19:32:04 James Ketrenos wrote: > iwlwifi currently calls ieee80211_stop_queue from the Tx handler and calls ieee80211_stop_queues in the event the adapter is reset (due to HW error, RF kill transition, hw tear down, etc.) as it asynchronously brings re-initializes the hardware. I think we do that in bcm43xx, too, and I think it is indeed racy, too. So it may trigger the freeze, too. A workaround might be to aquire the netdev_tx_lock around the stop_queues call. But I'm not sure. -- Greetings Michael.