From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758388AbZCMNIJ (ORCPT ); Fri, 13 Mar 2009 09:08:09 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1759053AbZCMNHu (ORCPT ); Fri, 13 Mar 2009 09:07:50 -0400 Received: from mx2.mail.elte.hu ([157.181.151.9]:44886 "EHLO mx2.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759034AbZCMNHt (ORCPT ); Fri, 13 Mar 2009 09:07:49 -0400 Date: Fri, 13 Mar 2009 14:07:30 +0100 From: Ingo Molnar To: Peter Zijlstra , =?iso-8859-1?Q?Fr=E9d=E9ric?= Weisbecker , Steven Rostedt Cc: paulus@samba.org, tglx@linutronix.de, linux-kernel@vger.kernel.org Subject: Re: [PATCH 00/11] generic software counters -v2 Message-ID: <20090313130730.GA10117@elte.hu> References: <20090313112125.886730125@chello.nl> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090313112125.886730125@chello.nl> User-Agent: Mutt/1.5.18 (2008-05-17) X-ELTE-VirusStatus: clean X-ELTE-SpamScore: -1.5 X-ELTE-SpamLevel: X-ELTE-SpamCheck: no X-ELTE-SpamVersion: ELTE 2.0 X-ELTE-SpamCheck-Details: score=-1.5 required=5.9 tests=BAYES_00 autolearn=no SpamAssassin version=3.2.3 -1.5 BAYES_00 BODY: Bayesian spam probability is 0 to 1% [score: 0.0000] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org * Peter Zijlstra wrote: > This patch set improves the current software counters by > providing sampling events for most of them. > > The clock based software counters use hrtimers to generate > events, allowing sample based profiling without hardware > support -- obviously this has limited functionality for kernal > based profiling since we disable interrupts quite frequently, > but should be good for userspace. > > Other events like pagefaults insert simple hooks into the > actual pagefault code to generate these events. Very nice series! Applied to tip:core/locking and tip:perfcounters/core - thanks Peter! Fixed a trivial !PERF_COUNTERS build failure as well. > Future work (in-progress) will hook into the tracepoint > infrastructure to to provide profiling events. Cool! Especially for pagefaults we have a number of callbacks there: kmmio, kprobes, kmemcheck, tracing, perfcounters - they could indeed all be consolidated step by step onto a tracepoint. Ingo