From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: [patch] Performance Counters for Linux, v2 Date: Tue, 9 Dec 2008 20:02:54 +0100 Message-ID: <20081209190254.GA27833@elte.hu> References: <20081208012211.GA23106@elte.hu> <7c86c4470812082237ne58c814s7218cc663f3b49e9@mail.gmail.com> <20081209134636.GA1926@elte.hu> <493E9F3E.3020902@nortel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mx3.mail.elte.hu ([157.181.1.138]:35713 "EHLO mx3.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753841AbYLITDg (ORCPT ); Tue, 9 Dec 2008 14:03:36 -0500 Content-Disposition: inline In-Reply-To: <493E9F3E.3020902@nortel.com> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Chris Friesen Cc: eranian@gmail.com, linux-kernel@vger.kernel.org, Thomas Gleixner , linux-arch@vger.kernel.org, Andrew Morton , Eric Dumazet , Robert Richter , Arjan van de Veen , Peter Anvin , Peter Zijlstra , Steven Rostedt , David Miller , Paul Mackerras , Paolo Ciarrocchi * Chris Friesen wrote: > Ingo Molnar wrote: > >> When there are two (or more) hw metrics to profile, the ideally best >> (i.e. the statistically most stable and most relevant) sampling for >> the two statistical variables (say of l2_misses versus l2_accesses) is >> to sample them independently, via their own metric. Not via a static >> 1khz rate - or via picking one of the variables to generate samples. > > Regardless of sampling method, don't you still want some way to > enable/disable the various counters as close to simultaneously as > possible? If it's about counter control for the monitored task, then we sure could do something about that. (apps/libraries could thus select a subset of functions to profile/measure, runtime, etc.) If it's about counter control for the profiler/debugger, i'm not sure how useful that is - do you have a good usecase for it? Ingo