Return-path: Received: from 128-177-27-249.ip.openhosting.com ([128.177.27.249]:41635 "EHLO jmalinen.user.openhosting.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754748AbZFPPQz (ORCPT ); Tue, 16 Jun 2009 11:16:55 -0400 Date: Tue, 16 Jun 2009 18:12:58 +0300 From: Jouni Malinen To: Bob Copeland Cc: Jouni Malinen , "John W. Linville" , linux-wireless@vger.kernel.org, stable@kernel.org Subject: Re: [PATCH] ath9k: Fix PCI FATAL interrupts by restoring RETRY_TIMEOUT disabling Message-ID: <20090616151258.GA22849@jm.kir.nu> References: <20090616085922.GA10176@jm.kir.nu> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: Sender: linux-wireless-owner@vger.kernel.org List-ID: On Tue, Jun 16, 2009 at 10:33:29AM -0400, Bob Copeland wrote: > > An earlier commit, b572b24c578ab1be9d1fcb11d2d8244878757a66, removed > > code that was documented to disable RETRY_TIMEOUT register (PCI reg > > 0x41) since it was claimed to be a no-op. > > Jouni, can you confirm that e23a9014... also needs reverting? > (In this case only affects resume path). I wish I could, but I can only guess that it probably should be reverted. > I seem to remember the provenance of this code was copy-paste from > an intel driver, so while it does "something," the comment may not > match the code, 0x41 being vendor-defined. The exact story behind this has been a bit more than trivial task to figure out ;-). I would assume this comment is referring to a madwifi changeset: http://madwifi-project.org/changeset/584 I've yet to find documentation that would explain the exact meaning of PCI register 0x41 in case of ath9k/ath5k-supported devices, but maybe I will some day ;-). Or well, I've seen documentation for said register, but that does not match with the comment (but seems to indicate that 0 is a reasonable value to write there). At this point, I can really only say that I can now easily reproduce this with ath9k and AR5416 and zeroing the 0x41 value does make the problem disappear. Since that code has been in use for five years, I do not expect problems with it and we can just eventually fix the comment if somebody actually find out what that register is supposed to control ;-). -- Jouni Malinen PGP id EFC895FA