From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965074AbVIAHcL (ORCPT ); Thu, 1 Sep 2005 03:32:11 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S965076AbVIAHcL (ORCPT ); Thu, 1 Sep 2005 03:32:11 -0400 Received: from mx3.mail.elte.hu ([157.181.1.138]:61643 "EHLO mx3.mail.elte.hu") by vger.kernel.org with ESMTP id S965074AbVIAHcK (ORCPT ); Thu, 1 Sep 2005 03:32:10 -0400 Date: Thu, 1 Sep 2005 09:32:41 +0200 From: Ingo Molnar To: Florian Schmidt Cc: "Jack O'Quin" , Fernando Lopez-Lezcano , jackit-devel@lists.sourceforge.net, Lee Revell , linux-kernel@vger.kernel.org, cc@ccrma.Stanford.EDU, Steven Rostedt Subject: Re: [Jackit-devel] Re: jack, PREEMPT_DESKTOP, delayed interrupts? Message-ID: <20050901073241.GA6641@elte.hu> References: <1125453795.25823.121.camel@cmn37.stanford.edu> <20050831073518.GA7582@elte.hu> <7q64tmnwbb.fsf@io.com> <20050831175036.5640e221@mango.fruits.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20050831175036.5640e221@mango.fruits.de> User-Agent: Mutt/1.4.2.1i X-ELTE-SpamScore: 0.0 X-ELTE-SpamLevel: X-ELTE-SpamCheck: no X-ELTE-SpamVersion: ELTE 2.0 X-ELTE-SpamCheck-Details: score=0.0 required=5.9 tests=AWL autolearn=disabled SpamAssassin version=3.0.3 0.0 AWL AWL: From: address is in the auto white-list X-ELTE-VirusStatus: clean Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org * Florian Schmidt wrote: > > JACK sources already include a CHECK_PREEMPTION() macro which expands > > to Ingo's special gettimeofday() calls. The trace is turned on and > > then off automatically before and after the realtime critical section > > in the process thread (see libjack/client.c). > > Just for completeness sake: > > you need to build jackd with --enable-preemption-check this is another feature, unrelated to latency tracing. So (an adapted version of) the latency-tracing patch i sent should still be tried. --enable-preemption-check does the 'send SIGUSR2 if jackd gets scheduled unexpectedly'. That might unearth latencies, but it does not by itself measure latencies. Right now we are more interested in the latencies themselves. i suspect the confusion comes from the API hacks i'm using: user-space tracing is started/stopped via: gettimeofday(0,1); gettimeofday(0,0); while 'jackd does not want to be scheduled' flag is switched on/off via: gettimeofday(1,1); gettimeofday(1,0); (introducing extra syscalls for this is an overkill.) Ingo