From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751225Ab1ITRvc (ORCPT ); Tue, 20 Sep 2011 13:51:32 -0400 Received: from mga02.intel.com ([134.134.136.20]:46355 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750778Ab1ITRvb (ORCPT ); Tue, 20 Sep 2011 13:51:31 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.67,352,1309762800"; d="scan'208";a="50390951" Message-ID: <4E78D2A2.6010106@linux.intel.com> Date: Tue, 20 Sep 2011 10:51:30 -0700 From: Andi Kleen User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2.15) Gecko/20110303 Thunderbird/3.1.9 MIME-Version: 1.0 To: Peter Zijlstra CC: Andi Kleen , acme@ghostprotocols.net, linux-kernel@vger.kernel.org Subject: Re: [PATCH 3/4] perf_events: Support a lock_parent event flag References: <1316127412-1912-1-git-send-email-andi@firstfloor.org> <1316127412-1912-3-git-send-email-andi@firstfloor.org> <1316510160.11841.12.camel@twins> In-Reply-To: <1316510160.11841.12.camel@twins> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > This all just sucks horridly. So this name tells us we need inherited > counters so that whenever we encounter a lock it will be accounted in > the parent process, or whatever -- which doesn't make any sense. Thanks for the kind words. So what name do you suggest? account_in_caller_of_lock ? i_write_a_novel_in_every_variable_name? > Furthermore, the sole reason you want this is because you don't want > callchains, supposedly because they're too expensive, but then you don't > say that. perf top doesn't support callchains. And yes it's a lot cheaper too, but that wasn't the main motivation. perf top is just useless to diagnose any lock problems currently. I say that in the changelog for the user space. > How about you provide means of limiting the callchain depth, and then > frob the in_lock_function() and unwind 1 crap in userspace? Is that equivalent? The normal call chains don't work without frame pointers. profile_pc does. -Andi