Return-path: Received: from mail-iw0-f178.google.com ([209.85.223.178]:52013 "EHLO mail-iw0-f178.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752811AbZKLTqE convert rfc822-to-8bit (ORCPT ); Thu, 12 Nov 2009 14:46:04 -0500 Received: by iwn8 with SMTP id 8so2042209iwn.33 for ; Thu, 12 Nov 2009 11:46:10 -0800 (PST) MIME-Version: 1.0 In-Reply-To: <200911122037.39145.chunkeey@googlemail.com> References: <167ae39b0910071400n218d3d9ey7711b1011f290a73@mail.gmail.com> <200911121717.36906.chunkeey@googlemail.com> <4AFC5A5E.1050405@canonical.com> <200911122037.39145.chunkeey@googlemail.com> From: "Luis R. Rodriguez" Date: Thu, 12 Nov 2009 11:45:50 -0800 Message-ID: <43e72e890911121145i3d9f6891m362498fad2f99f79@mail.gmail.com> Subject: Re: Prism54/p54pci To: Christian Lamparter Cc: tim.gardner@canonical.com, James Grossmann , linux-wireless@vger.kernel.org Content-Type: text/plain; charset=UTF-8 Sender: linux-wireless-owner@vger.kernel.org List-ID: On Thu, Nov 12, 2009 at 11:37 AM, Christian Lamparter wrote: > On Thursday 12 November 2009 19:56:30 Tim Gardner wrote: >> Christian Lamparter wrote: >> > On Friday 06 November 2009 21:52:42 James Grossmann wrote: >> >> my caps lock & scroll lock leds are not blinking, and yet the computer >> >> is frozen.  (I'm not the best at terminology, I guess that might not >> >> be defined as a hard lock.)  I get no odd messages in my syslog.  I am >> >> using Ubuntu right now, so I'm guessing that they have all the >> >> prettyness going that they can, covering the important messages, but >> >> as it's not an actual oops, I'm not certain that's hurting us. >> > just a update: >> > >> > I can confirm the freeze with ubuntu 9.10 (2.6.31-14 & 2.6.31-15-generic) >> > + compat-wireless + p54pci (either as backport module or compiled from git). >> > The original modules (shipped with 2.6.31-14/15) are fine though..?! >> >> The Ubuntu Karmic linux-backports-modules package is based on >> compat-wireless-2.6.32-rc4. I'm wondering if its something specific to >> the p54pci driver? I'm using iwlagn (4965) from that package with no >> problems (so far). > > I put a ath5k card into the same machine and tried with the same > compat-wireless modules. Surprise, surprise: no problems... > > The affected device works fine with the current wireless-testing(pulled today) > and a vanilla 2.6.31.6 + compat-wireless (from today's git). > > I've no idea what voodoo is going on here, maybe there is > something wrong in compat-wireless pci glue code (if there's any?!) Should be easy to debug, the compat 2.6.31 code is in compat/compat-2.6.32.[ch] and since there is no 2.6.33 yet that is all. That and of course any ifdef which applies on the compat/patches/. >> Its not clear what version of stable compat-wireless you've tried. >> Does compat-wireless-2.6.32-rc6 exhibit the same problem? > "compiled from git", which should be compat-wireless-2009-11-12.tar.bz2 > and therefore -rc6. Well no, I believe what Tim means is the stable compat-wireless releases which are not based on master git on compat-wireless. These are based on a branch for the specific stable kernel. The stable compat-wireless releases use hpa's tree: git://git.kernel.org/pub/scm/linux/kernel/git/hpa/linux-2.6-allstable.git and I checkout specific 2.6.3x.y branches. So for 2.6.32 I just use master though which is based on Linus' tree and the latest *stable* 2.6.32-rcx. Luis