From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755387AbZBILtb (ORCPT ); Mon, 9 Feb 2009 06:49:31 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754057AbZBILtG (ORCPT ); Mon, 9 Feb 2009 06:49:06 -0500 Received: from mx3.mail.elte.hu ([157.181.1.138]:34719 "EHLO mx3.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753976AbZBILtE (ORCPT ); Mon, 9 Feb 2009 06:49:04 -0500 Date: Mon, 9 Feb 2009 12:48:52 +0100 From: Ingo Molnar To: Paul Mackerras Cc: "Zhang, Yanmin" , linux-kernel@vger.kernel.org Subject: Re: [PATCH] perf_counter: Make software counters work as per-cpu counters Message-ID: <20090209114852.GF18757@elte.hu> References: <18832.5815.706605.434454@cargo.ozlabs.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <18832.5815.706605.434454@cargo.ozlabs.ibm.com> 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 * Paul Mackerras wrote: > Impact: kernel crash fix > > Yanmin Zhang reported that using a PERF_COUNT_TASK_CLOCK software > counter as a per-cpu counter would reliably crash the system, because > it calls __task_delta_exec with a null pointer. The page fault, > context switch and cpu migration counters also won't function > correctly as per-cpu counters since they reference the current task. > > This fixes the problem by redirecting the task_clock counter to the > cpu_clock counter when used as a per-cpu counter, and by implementing > per-cpu page fault, context switch and cpu migration counters. > > Along the way, this: > > - Initializes counter->ctx earlier, in perf_counter_alloc, so that > sw_perf_counter_init can use it > - Adds code to kernel/sched.c to count task migrations into each > cpu, in rq->nr_migrations_in > - Exports the per-cpu context switch and task migration counts > via new functions added to kernel/sched.c > - Makes sure that if sw_perf_counter_init fails, we don't try to > initialize the counter as a hardware counter. Since the user has > passed a negative, non-raw event type, they clearly don't intend > for it to be interpreted as a hardware event. > > Signed-off-by: Paul Mackerras Very nice, thanks Paul! > I'm a little concerned about the use of u64 for the existing > rq->nr_switches and the new rq->nr_migrations_in. On 32-bit machines > this will get updated in two halves, so there is a very small but > non-zero probability that reading it will give a bogus value. It's > not clear to me what the best way to fix it is, since not all 32-bit > platforms have atomic64_t. Maybe make nr_switches and > nr_migrations_in unsigned long, or atomic_t? It used to be 'just stats' so we dont really mind. But now that we expose it in a bit more systematic way i agree that it should be fixed. Updating to 'unsigned long' sounds good to me. > I have some sort of git problem with my perfcounters.git repository. > I'll let you know when I have that sorted out. ok - i've applied your patch from email to allow Yanmin to test tip:master. Ingo