From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932766AbWBUVAF (ORCPT ); Tue, 21 Feb 2006 16:00:05 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932767AbWBUVAF (ORCPT ); Tue, 21 Feb 2006 16:00:05 -0500 Received: from ns.suse.de ([195.135.220.2]:53933 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S932766AbWBUVAC (ORCPT ); Tue, 21 Feb 2006 16:00:02 -0500 From: Andi Kleen To: "David S. Miller" Subject: Re: softlockup interaction with slow consoles Date: Tue, 21 Feb 2006 21:43:46 +0100 User-Agent: KMail/1.8.2 Cc: linux-kernel@vger.kernel.org, mingo@elte.hu References: <200602212105.38075.ak@suse.de> <20060221.121948.60060362.davem@davemloft.net> In-Reply-To: <20060221.121948.60060362.davem@davemloft.net> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200602212143.46721.ak@suse.de> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tuesday 21 February 2006 21:19, David S. Miller wrote: > From: Andi Kleen > Date: Tue, 21 Feb 2006 21:05:37 +0100 > > > The classic way is to just use touch_nmi_watchdog() somewhere > > in the loop that does work. That touches the softwatchdog too > > these days. > > "jiffies" aren't advancing, since interrupts are disabled by > release_console_sem(), so that doesn't work. > > I tried that already :-) Ah I see the problem I guess. When you restart then your timer interrupt catches up the missing jiffies very quickly and that triggers the softwatchdog. Nasty. Perhaps it needs to look at xtime instead and let touch_softlockup_watchdog() update that. But that could break with NTP then. Perhaps using monotonic_clock() ? -Andi