From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755691Ab0KHUlX (ORCPT ); Mon, 8 Nov 2010 15:41:23 -0500 Received: from canuck.infradead.org ([134.117.69.58]:46362 "EHLO canuck.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755369Ab0KHUlU convert rfc822-to-8bit (ORCPT ); Mon, 8 Nov 2010 15:41:20 -0500 Subject: Re: Q: perf_event && event->owner From: Peter Zijlstra To: Frederic Weisbecker Cc: Oleg Nesterov , Alan Stern , Arnaldo Carvalho de Melo , Ingo Molnar , Paul Mackerras , Prasad , Roland McGrath , linux-kernel@vger.kernel.org In-Reply-To: <20101108201108.GA6777@nowhere> References: <20101108145647.GA3426@redhat.com> <20101108145754.GB3434@redhat.com> <20101108201108.GA6777@nowhere> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8BIT Date: Mon, 08 Nov 2010 21:41:12 +0100 Message-ID: <1289248872.2191.27.camel@laptop> Mime-Version: 1.0 X-Mailer: Evolution 2.30.3 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 2010-11-08 at 21:11 +0100, Frederic Weisbecker wrote: > Perhaps the real owner should be the task on which we attach our > breakpoint. No the point of event->owner is to point to the task that creates the event, not the task we possibly attach it to (that should be reachable through event->ctx->task). As to removing event->owner as Oleg suggests, its a published ABI and there might be people using it. The use-case is a monitor thread wanting to stop all monitoring it initiated, for example because its wants to synchronize various counters attached to different tasks etc..