From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tim Sander Subject: Re: Measuring timer interrupt latency Date: Mon, 13 Apr 2009 18:22:01 +0200 Message-ID: <200904131822.02125.tstone@vlsi.informatik.tu-darmstadt.de> References: <002901c9bc03$54aa23d0$3f52c70a@dlh.st.com> <20090413091405.380e42f8@torg> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE To: "'rt-users'" Return-path: Received: from lnx503.hrz.tu-darmstadt.de ([130.83.156.232]:53653 "EHLO lnx503.hrz.tu-darmstadt.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751863AbZDMQWJ convert rfc822-to-8bit (ORCPT ); Mon, 13 Apr 2009 12:22:09 -0400 Received: from mailserver3.hrz.tu-darmstadt.de (lnx504.hrz.tu-darmstadt.de [130.83.156.233]) by lnx503.hrz.tu-darmstadt.de (8.14.2/8.13.8/HRZ/PMX) with ESMTP id n3DGM7pq002641 for ; Mon, 13 Apr 2009 18:22:07 +0200 (envelope-from tstone@vlsi.informatik.tu-darmstadt.de) Received: from [213.157.27.17] (helo=IP-213157027017.dialin.heagmedianet.de) by mailserver3.hrz.tu-darmstadt.de with esmtpsa (TLS-1.0:DHE_RSA_AES_128_CBC_SHA1:16) (Exim 4.63) (envelope-from ) id 1LtOv9-00087B-Jn for linux-rt-users@vger.kernel.org; Mon, 13 Apr 2009 18:22:07 +0200 In-Reply-To: <20090413091405.380e42f8@torg> Content-Disposition: inline Sender: linux-rt-users-owner@vger.kernel.org List-ID: Hi > Have you tried cyclictest? It's part of the rt-tests package that > Thomas Gleixner wrote and that I now maintain. I think cyclictest is nice to get some first impressions and i really l= ike=20 it... but it only give the aggregate. > The thing is, I'm not sure how useful measuring the interval between > interrupt and ISR. It's such a small component of the total latency > value that it usually gets lost in the noise.=20 I don't think that you can neglect these numbers in a hard rt environme= nt. As far as i understand, the irq-tasklet has to be scheduled by the kern= el. > If you are wanting to decompose the latency into it's components on R= T, > I think you'd have: > 1. ISR latency > 2. Scheduler latency for IRQ thread > 3. IRQ handler time > 4. Scheduler latency for user-space thread > Cyclictest measures the aggregate of all these. One good aproach to minimize user mode latency are probably user mode d= rivers.=20 This approach has been presented at the Embedded in N=C3=BCrnberg this = year by=20 linutronix. So you can shave off #2 of the above list. Which means only= one=20 rescheduling runs instead of two. Usermode has of course to be run with= rt=20 priority. Tim --=20 ''=3D~('(?{'.('._@@[~'^'^-)./^').'"'.('}@@~-]@_^__'^'))-^~<.;;-}').',$/= })') -- To unsubscribe from this list: send the line "unsubscribe linux-rt-user= s" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html