From mboxrd@z Thu Jan 1 00:00:00 1970 From: Carsten Emde Subject: Re: Latest PreemptRT patch error on imx35 Date: Mon, 31 Aug 2009 13:48:23 +0200 Message-ID: <4A9BB887.1040407@osadl.org> References: <200908311143.08952.tim01@vlsi.informatik.tu-darmstadt.de> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: linux-rt-users To: Tim Sander Return-path: Received: from toro.web-alm.net ([62.245.132.31]:42942 "EHLO toro.web-alm.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751753AbZHaMnZ (ORCPT ); Mon, 31 Aug 2009 08:43:25 -0400 In-Reply-To: <200908311143.08952.tim01@vlsi.informatik.tu-darmstadt.de> Sender: linux-rt-users-owner@vger.kernel.org List-ID: Tim, > I just bootet 2.6.31-rc8-rt9 on the phycore-imx35 development board,=20 > without any additional patches. The FEC Ethernet Controller went bell= y=20 > up, but the system booted nevertheless. I tried cyclictest: > root@pmx:~ cyclictest -t1 -p 80 -n -i 10000 -l 10000 As a role of thumb, the -i option is normally set to twice the expected latency. If, for example you expect 100 us, set it to -i200. Your loop end is then reached 50 times faster, or - what is much better - you may specify 50 times more loops for a test of the same duration which considerably increases the quality of your assessment. Normally, however, you would probably wait for 100 million or even a billion loop= s. > 0.05 0.10 0.05 1/39 570 > T: 0 ( 570) P:80 I:10000 C: 10000 Min: 14 Act: 20 Avg: 19 M= ax: 27 > These are Numbers from an Idle Machine (no network), but they look r= eally good. With an average of 19 and a maximum of 27, the result is probably not very relevant. This indicates that the CPU was idle most of the time when the wakeup occurred - every OS can do this. The challenge of an RTOS is to wakeup in time also when the kernel is not idle, e.g. from kernel or driver code while a lock is held. Therefore, adequate stress tools must be running during a latency measurement. > They are much better than my last tests on the imx31 with an older pr= eemptrt-kernel=20 > (2.6.29) which had idle latencys up to 80=B5s.=20 > Given the fact, that ftrace is enabled, this is really impressive pro= gress! Please keep in mind that ftrace configured is not ftrace enabled. To enable it, use cyclictest's -f option or execute echo function >/sys/kernel/debug/tracing/current_tracer echo 1 >/sys/kernel/debug/tracing/tracing_enabled from command line. I hope that this hasn't tempered your joy with your new hardware. Carsten. -- 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