From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 4 Dec 2002 01:44:22 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 4 Dec 2002 01:44:22 -0500 Received: from dilbert.sider.net ([200.196.245.9]:60089 "EHLO dilbert.sider.net") by vger.kernel.org with ESMTP id ; Wed, 4 Dec 2002 01:44:21 -0500 Date: Wed, 4 Dec 2002 04:52:58 -0200 From: Felipe Massia To: linux-kernel@vger.kernel.org Subject: timer glitch (solved?) Message-Id: <20021204045258.590d7732.felmasper@yahoo.com.br> X-Mailer: Sylpheed version 0.8.6claws9 (GTK+ 1.2.10; ) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org My clock, some times, gives wrong results. When I call the folowing command bash$ while : ; do date ; done the time shown is not correct. When the seconds are changing, for a small period of time, the time is offset by about 1h12min. E.g. when the clock is changing from 1h30min30sec to 1h30min31sec, some 2 or 3 lines shown are 2h42min. I don't know how to reproduce the problem. It seems the timer enters this wrong state after using mplayer (maybe because it's "timer-hungry"). I've read the "timer glitch on 2.4.18: solved" thread in this list and I think it's the same problem. Still not clear to me: should I experience this problem w/ 2.4.19? Is there a way so I can have the clock working properly? BTW, I don't know if it's important to mention: I'm using rtc. But even when I remove the rtc module, the problem persists. Some data on my system (maybe useful): $ uname -a Linux felipe 2.4.19-k6 #1 Sun Oct 6 19:53:19 EST 2002 i586 AMD-K6(tm) 3D processor AuthenticAMD GNU/Linux # kernel-image for k6 from Debian unstable $ dmesg | grep -i via PCI: Using IRQ router VIA [1106/0686] at 00:07.0 VP_IDE: VIA vt82c686a (rev 1b) IDE UDMA66 controller on pci00:07.1 parport_pc: Via 686A parallel port: io=0x378 # I read something about VIA686 mb in the thread $ cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 5 model : 8 model name : AMD-K6(tm) 3D processor stepping : 12 cpu MHz : 501.157 tia, Felipe