Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751122AbWCGRpV (ORCPT ); Tue, 7 Mar 2006 12:45:21 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751410AbWCGRpV (ORCPT ); Tue, 7 Mar 2006 12:45:21 -0500 Received: from tomts22.bellnexxia.net ([209.226.175.184]:64141 "EHLO tomts22-srv.bellnexxia.net") by vger.kernel.org with ESMTP id S1751122AbWCGRpT (ORCPT ); Tue, 7 Mar 2006 12:45:19 -0500 Date: Tue, 7 Mar 2006 12:45:18 -0500 From: Mathieu Desnoyers To: Steven Rostedt Cc: Serge Noiraud , linux-kernel@vger.kernel.org, ltt-dev@shafik.org, Ingo Molnar Subject: Re: RT patch and arch/i386/kernel/time.c question Message-ID: <20060307174517.GA24610@Krystal> References: <200603071742.42262.Serge.Noiraud@bull.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline In-Reply-To: X-Editor: vi X-Info: http://krystal.dyndns.org:8080 X-Operating-System: Linux/2.4.31-grsec (i686) X-Uptime: 12:41:47 up 4 days, 2:41, 3 users, load average: 0.58, 0.20, 0.12 User-Agent: Mutt/1.5.11+cvs20060126 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1586 Lines: 40 * Steven Rostedt (rostedt@goodmis.org) wrote: > > On Tue, 7 Mar 2006, Serge Noiraud wrote: > > > hi, > > > > I'm trying to port the LTTng patch over the rt20 and I got the following problem : > > The LTTng patch try to modify the arch/i386/kernel/time.c file in which the > > timer_interrupt function doesn't exist anymore. > > > > In which file / function could I try to patch the equivalent function ? > > > > The -rt patch uses the lastest stuff from Thomas Gleixner, John Stultz and > of course Ingo Molnar. The functions you are interested in, are in > kernel/time/ directory. Take a look at clockevents.c and perhaps > handle_tick(). I'm not sure what LTTng is doing there, but this will give > you a direction in which way to look. > > -- Steve > LTTng is updating its logical clock there. Note that this clock is only used on architectures missing a synchronised CPU timestamp counter (NUMA and old i586). If you have a fairly standard architecture (i686 with TSC available), this combination of jiffies counter and logical clock is not used. Look for where the xtime_lock seqlock is taken in the -rt patch : that's where the logical clock should be updated. Mathieu OpenPGP public key: http://krystal.dyndns.org:8080/key/compudj.gpg Key fingerprint: 8CD5 52C3 8E3C 4140 715F BA06 3F25 A8FE 3BAE 9A68 - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/