From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752655AbXA3BiF (ORCPT ); Mon, 29 Jan 2007 20:38:05 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752654AbXA3BiF (ORCPT ); Mon, 29 Jan 2007 20:38:05 -0500 Received: from ironport-out.pppoe.ca ([206.248.154.186]:2272 "EHLO ironport-out.pppoe.ca" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752655AbXA3BiD (ORCPT ); Mon, 29 Jan 2007 20:38:03 -0500 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AlwPAL4wvkVBJ8TuX2dsb2JhbACORRUPKQ X-IronPort-AV: i="4.13,254,1167627600"; d="scan'208"; a="70236410:sNHT16719888" Message-ID: <45BEA162.3030204@teksavvy.com> Date: Mon, 29 Jan 2007 20:37:38 -0500 From: Yavuz Onder Reply-To: yavuz@teksavvy.com Organization: Whatever ISP User-Agent: Thunderbird 1.5.0.9 (X11/20070104) MIME-Version: 1.0 To: linux-kernel@vger.kernel.org CC: yavuz@teksavvy.com Subject: Re: 2.6.19.2: bad reads/writes from/to CMOS clock References: <9e18e56a85195bf4a507163df8e2f62e@teksavvy.com> In-Reply-To: <9e18e56a85195bf4a507163df8e2f62e@teksavvy.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I have since found out that SMP kernels should(must?) have Enhanced RTC support built-in. It is stated in kernel config help for Enhanced RTC support, and I found a number of references on the net as well. Now I wonder why selecting SMP does not set Enhanced RTC support to "Y" automatically? I am building a kernel right now, and will monitor for a week or so, then come back and report if it removes the problem. Yavuz yavuz@teksavvy.com wrote: > Hi everyone, > > I am running 2.6.19.2 kernel from kernel.org. > > This is my first SMP kernel. > > The problem I describe below has not happend with non-SMP kernels ever... > > I have installed my new AMD64 x2 4800 CPU just a few days ago. My mobo is Asus A8N SLI > (Nvidia chipset). > > My problem with this new setup is that my CMOS clock is thrown off by varying > amounts of time. I have seen system times as long as ten months into future, and as > long as 25 days in the past. At the moment my system clock has correct > hh:mm:ss, but it shows the date Jan 1st, instead of correct 25th. > > In last few days have systematically checked CMOS clock after shutdown and before > boot-up. > > I have observed CMOS clock set to a wrong value, immediately after shutdown. > I also ended up with a wrong system time following a boot after verifying > correctness of the CMOS clock. > > Problem does not happen 100% of the time. But 6-7 times in a week is bad > enough. > > I searched mailing list archives, and found some NMI-RTC race condition discussion, > and some other discussions about timers not running, clock ticks being missed > etc. But they were all at least one year old. The fixes to those should have found > their way into 2.6.19 kernels. > > I would appreciate any ideas on how to follow-up on this problem. > > I am eager to try things, collect data as needed. Just tell me what to do. > > Thanks in advance. > > > > > Yavuz Onder > >