From mboxrd@z Thu Jan 1 00:00:00 1970 From: ankita@in.ibm.com (Ankita Garg) Subject: Re: High 50us+ latencies in the process signal handling path Date: Fri, 19 Oct 2007 11:58:07 +0530 Message-ID: <20071019062807.GA10260@in.ibm.com> References: <20071018124827.GA22282@in.ibm.com> <20071018.055242.115933276.davem@davemloft.net> <20071018132614.GB22282@in.ibm.com> <20071018153618.GI12471@unix.sh> Reply-To: Ankita Garg Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Miller , linux-rt-users@vger.kernel.org, rostedt@goodmis.org, ghaskins@novell.com, mingo@elte.hu, dino@in.ibm.com, dvhltc@us.ibm.com, kravetz@us.ibm.com To: "Luis Claudio R. Goncalves" Return-path: Received: from E23SMTP03.au.ibm.com ([202.81.18.172]:54823 "EHLO e23smtp03.au.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754381AbXJSG2r (ORCPT ); Fri, 19 Oct 2007 02:28:47 -0400 Received: from d23relay03.au.ibm.com (d23relay03.au.ibm.com [202.81.18.234]) by e23smtp03.au.ibm.com (8.13.1/8.13.1) with ESMTP id l9J6Se3S011131 for ; Fri, 19 Oct 2007 16:28:40 +1000 Received: from d23av03.au.ibm.com (d23av03.au.ibm.com [9.190.234.97]) by d23relay03.au.ibm.com (8.13.8/8.13.8/NCO v8.5) with ESMTP id l9J6SfeT1294352 for ; Fri, 19 Oct 2007 16:28:41 +1000 Received: from d23av03.au.ibm.com (loopback [127.0.0.1]) by d23av03.au.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id l9J6SOiG008580 for ; Fri, 19 Oct 2007 16:28:26 +1000 Content-Disposition: inline In-Reply-To: <20071018153618.GI12471@unix.sh> Sender: linux-rt-users-owner@vger.kernel.org List-Id: linux-rt-users.vger.kernel.org Hi, On Thu, Oct 18, 2007 at 11:36:18AM -0400, Luis Claudio R. Goncalves wrote: > Ankita, > > On Thu, Oct 18, 2007 at 06:56:14PM +0530, Ankita Garg wrote: > | Hi David, > | > | On Thu, Oct 18, 2007 at 05:52:42AM -0700, David Miller wrote: > | > From: ankita@in.ibm.com (Ankita Garg) > | > Date: Thu, 18 Oct 2007 18:18:27 +0530 > | > > | > > Any thoughts on where in the code could these large latencies be > | > > attributed to? > | > > | > If the max latency is usually the first one, it's could be the page > | > fault the signal handler is taking the first time it is executed. > | > | Thanks for looking into this. Attaching the entire log from the testrun. > | It indicates that in each run of the 10000 iterations, there are about 100+ > | failures, with large iterations. Also, on intrumenting the testcode to > | find out the iterations which see large latencies, I see that as the > | number of iterations increase, the latencies increase. > > You might wanna check signaltest: > > git clone git://git.kernel.org/pub/scm/linux/kernel/git/tglx/rt-tests.git rt-tests > Are any previous logs available for signaltest? This is what I get on 2.6.23-rt1- # ./signaltest -q -p 85 -l100000 -t8 T: 0 ( 6791) P:85 C: 100000 Min: 109 Act: 148 Avg: 143 Max: 252 T: 1 ( 6792) P:85 C: 100000 Min: 109 Act: 148 Avg: 769 Max: 10240 T: 2 ( 6793) P:85 C: 100000 Min: 109 Act: 148 Avg: 769 Max: 10247 T: 3 ( 6794) P:85 C: 100000 Min: 109 Act: 161 Avg: 769 Max: 10238 T: 4 ( 6795) P:85 C: 100000 Min: 109 Act: 183 Avg: 769 Max: 10238 T: 5 ( 6796) P:85 C: 100000 Min: 109 Act: 214 Avg: 769 Max: 10239 T: 6 ( 6797) P:85 C: 100000 Min: 109 Act: 236 Avg: 769 Max: 10237 T: 7 ( 6798) P:85 C: 100000 Min: 109 Act: 329 Avg: 769 Max: 10238 And this on 2.6.21.4-rt10- # ./signaltest -q -p 85 -l100000 -t8 T: 0 ( 3767) P:85 C: 100000 Min: 56 Act: 101 Avg: 99 Max: 162 T: 1 ( 3768) P:85 C: 100000 Min: 56 Act: 101 Avg: 725 Max: 10164 T: 2 ( 3769) P:85 C: 100000 Min: 56 Act: 101 Avg: 725 Max: 10164 T: 3 ( 3770) P:85 C: 100000 Min: 56 Act: 102 Avg: 725 Max: 10164 T: 4 ( 3771) P:85 C: 100000 Min: 56 Act: 111 Avg: 725 Max: 10165 T: 5 ( 3772) P:85 C: 100000 Min: 56 Act: 120 Avg: 725 Max: 10163 T: 6 ( 3773) P:85 C: 100000 Min: 56 Act: 135 Avg: 725 Max: 10164 T: 7 ( 3774) P:85 C: 100000 Min: 56 Act: 178 Avg: 725 Max: 10165 The values seem rather high! > Regards, > Luis > -- > [ Luis Claudio R. Goncalves Bass - Gospel - RT ] > [ Fingerprint: 4FDD B8C4 3C59 34BD 8BE9 2696 7203 D980 A448 C8F8 ] > > - > To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- Regards, Ankita Garg (ankita@in.ibm.com) Linux Technology Center IBM India Systems & Technology Labs, Bangalore, India