From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ove Karlsen Subject: Re: [ANNOUNCE] 3.6.2-rt4 Date: Mon, 29 Oct 2012 21:39:49 +0100 Message-ID: <508EE995.3030109@paradoxuncreated.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE To: tglx@linutronix.de, linux-rt-users@vger.kernel.org Return-path: Received: from smtp.domeneshop.no ([194.63.252.54]:48121 "EHLO smtp.domeneshop.no" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932224Ab2J2U4Q (ORCPT ); Mon, 29 Oct 2012 16:56:16 -0400 Sender: linux-rt-users-owner@vger.kernel.org List-ID: Hiya, I am switching email-client currently, so excuse a bit of chaos. I am reading=20 https://rt.wiki.kernel.org/index.php/Frequently_Asked_Questions#What_so= rt_of_real-time_performance_should_I_expect.3F I think it sounds really interesting that you have 10-30uS latencies=20 back in 2007. The concept sounds thorough and tried then. So I shouldn`= t=20 expect any big SMI messing things up, on my core2duo right? I think I really wanna get this to work. No solutions? :) I am reading more here: https://rt.wiki.kernel.org/index.php/CONFIG_PREEMPT_RT_Patch "Works ok, but frequency scaling has to be turned off for decent=20 results. A task run at 1 kHz repetition rate with low load was found to= =20 be t-distributed with 4 degrees of freedom, with 1% value 0.9842 ms, 99= %=20 value 1.0103 ms (ideal realtime performance would get exactly 1 ms for=20 both values), with the frequency governor set to "performance"." Quite low. So he has got it running with low jitter, on a core 2 quad. SMI sounds not much of a problem then. I want this guys. Altough as sai= d=20 I am buying a computer, particulary for low-jitter, we can always try=20 and make this work on my older box in the mean time. Remember the prob is with threadirqs on mainline aswell. Could it be=20 some clockproblem? A task for the thirsty bughunter, and patch-glory to= =20 mainline. --=20 =46red v=E6re med deg / Peace Be with You, Ove Karlsen http://www.paradoxuncreated.com -- 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