From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757484AbZBTRHM (ORCPT ); Fri, 20 Feb 2009 12:07:12 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753384AbZBTRGz (ORCPT ); Fri, 20 Feb 2009 12:06:55 -0500 Received: from cdptpa-omtalb.mail.rr.com ([75.180.132.123]:46420 "EHLO cdptpa-omtalb.mail.rr.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753065AbZBTRGy (ORCPT ); Fri, 20 Feb 2009 12:06:54 -0500 Message-ID: <499EE32B.1040800@cfl.rr.com> Date: Fri, 20 Feb 2009 12:06:51 -0500 From: Mark Hounschell User-Agent: Thunderbird 2.0.0.19 (X11/20081227) MIME-Version: 1.0 To: Andreas Herrmann CC: Borislav Petkov , Mark Hounschell , john stultz , linux-kernel , Thomas Gleixner Subject: Re: PROBLEM: Can't boot a (HZ = 1000) kernel using an AMD Phenom-II processor References: <499C0913.9050905@compro.net> <20090218141843.GB13304@aftab> <499C3DED.1090807@cfl.rr.com> <20090218174243.GD13304@aftab> <499C7AC8.4080609@cfl.rr.com> <499D364C.3010704@cfl.rr.com> <20090219141131.GM4301@alberich.amd.com> <20090220115357.GE4834@alberich.amd.com> <499ECF87.2070204@cfl.rr.com> <20090220160843.GF4834@alberich.amd.com> In-Reply-To: <20090220160843.GF4834@alberich.amd.com> X-Enigmail-Version: 0.95.7 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Andreas Herrmann wrote: > On Fri, Feb 20, 2009 at 10:43:03AM -0500, Mark Hounschell wrote: >> Andreas Herrmann wrote: >>> On Thu, Feb 19, 2009 at 10:54:19AM -0500, Mark Hounschell wrote: >>>> Borislav Petkov wrote: >>>>> On Thu, Feb 19, 2009 at 10:16:40AM -0500, Mark Hounschell wrote: >>>>>> Andreas Herrmann wrote: >>>>>>> Hi Mark, >>>>>>> >>>>>>> Can you please provide us output of lspci -vvxxx. >>>>>>> (Please run this as root.) >>>>>>> >>>> Attached... >>>> >>>> >>>>>>> Furthermore you can try booting your 1000HZ kernel >>>>>>> with following kernel command lines: >>>>>>> >>> Thanks for trying this out. >>> But nothing suspicious found. I.e. timer/hpet interrupt is enabled >>> and its routing seems to be ok. >>> >>> Here's one more check that's worth trying: In one of your logs I've >>> seen that HPET id reports 3 timers but later on 4 timers were detected >>> for HPET. >>> >>> To check what's going on I'd like to see the HPET configuration in the >>> course of booting your system. Please apply attached patch and boot >>> with hpet=debug -- both with 250HZ and with 1000HZ -- and provide dmesg >>> output. >>> >>> >> Here is the short 1000Hz bootlog. I'm afraid non of your debug output made it >> out however. > > Sigh. > >> I thought it was because it just wasn't getting that far but >> looking at the 250Hz log I don't see any there either??? > > >> The patch applied (with fuz) to both kernel sources and I built/installed them >> from clean. I don't find any messages starting with "hpet:" in either trace??? > > Your 250HZ kernel was built on Feb, 2nd, right? > So the patch wasn't applied I guess, no? > The patch was applied. These 2 kernels were built from "clean" at the same time. >> Linux version 2.6.28.3-1000Hz (root@harley) (gcc version 4.3.1 20080507 >> (prerelease) [gcc-4_3-branch revision 135036] (SUSE Linux) ) #7 SMP PREEMPT Fri Feb >> 20 09:01:29 EST 2009 > > I assume the patch was aplied. > Yes it was. >> Linux version 2.6.28.3-250Hz (root@harley) (gcc version 4.3.1 >> 20080507 (prerelease) [gcc-4_3-branch revision 135036] (SUSE Linux) >> ) #7 SMP PREEMPT Fri Feb 2 0 09:01:54 EST 2009 > ^ | Look closer it was built Feb 20 09:01:53. Not Feb 2 0 09:01:53. There is a space between the 2 and 0 because of a minicom linewrap I guess. > Here I doubt that's the right kernel -- at least it was not built today. > Can you please double check this. > My double check is correct. both these kernels were built this morning from "clean" after your patch was applied. I will do it all again from scratch if you wish. Not a problem. I am at your disposal. I'm not ignorant when it comes to building and installing kernels. Standing by. Mark