From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bhavesh Davda Subject: Re: [PATCH RT-TESTS] cyclictest: histogram overflow instance tracking Date: Tue, 13 Nov 2012 16:17:32 -0800 (PST) Message-ID: <186127385.49218418.1352852252623.JavaMail.root@vmware.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Cc: linux-rt-users@vger.kernel.org, frank rowand To: John Kacur Return-path: Received: from smtp-outbound-1.vmware.com ([208.91.2.12]:54971 "EHLO smtp-outbound-1.vmware.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755999Ab2KNARe (ORCPT ); Tue, 13 Nov 2012 19:17:34 -0500 In-Reply-To: Sender: linux-rt-users-owner@vger.kernel.org List-ID: No disrespect intended, sorry if it came across as such. And as I'm no cyclictest expert, I didn't quite catch the subtleties of Frank's important observation. I'll need to study the cyclictest code more carefully to really grok that, which I will. Stay tuned for a patch... -- Bhavesh Davda ----- Original Message ----- > From: "John Kacur" > To: "Bhavesh Davda" > Cc: linux-rt-users@vger.kernel.org, "frank rowand" > Sent: Tuesday, November 13, 2012 4:09:22 PM > Subject: Re: [PATCH RT-TESTS] cyclictest: histogram overflow instance tracking > > On Wed, Nov 14, 2012 at 1:04 AM, Bhavesh Davda > wrote: > > Okay, Frank's response raced with mine. Since Frank is not super > > motivated to submit a patch with his proposed changes to my patch, > > I'll do so. > > > > Ummm, ok, this is sounding a little condescending. When people review > your patch, this is a gift they gave you which they did not need to. > It means they found your patch interesting enough to suggest how to > make it better. Adding an option to turn it on, is trivial, I > believe, > the important point that Frank said is > > "One further thought... The histogram overflow cycle report shows > what cycle the overflow occurred in, not the actual time. Adding > the merged for all threads cycle times works because the histogram > turns off the "different intervals for different threads" option: > > if (!histogram) /* same interval on CPUs */ > interval += distance; > > but if that ever changes then cycle is not a useful value to be > reporting. > > So it seems like it would be useful to convert cycle to a time > in the report. This is something that would have to be done > anyway in post processing when trying to make use of the report." > > The motivation should belong to you. You should be proud that we > thought your ideas were important and good enough to incorporate into > the software and to review and comment on it. Careful there! > > Thanks > > John >