From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754422Ab1GHRfD (ORCPT ); Fri, 8 Jul 2011 13:35:03 -0400 Received: from cam-admin0.cambridge.arm.com ([217.140.96.50]:50869 "EHLO cam-admin0.cambridge.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751393Ab1GHRfB (ORCPT ); Fri, 8 Jul 2011 13:35:01 -0400 Date: Fri, 8 Jul 2011 18:34:42 +0100 From: Will Deacon To: "mingo@redhat.com" , "dengcheng.zhu@gmail.com" , "a.p.zijlstra@chello.nl" , "fweisbec@gmail.com" , "heiko.carstens@de.ibm.com" , "tglx@linutronix.de" , "dzickus@redhat.com" , "hpa@zytor.com" , "anton@samba.org" , "linux-kernel@vger.kernel.org" , "jason.wessel@windriver.com" , "lethal@linux-sh.org" , "emunson@mgebm.net" , "davem@davemloft.net" , "mcree@orcon.net.nz" , "mingo@elte.hu" Cc: "linux-tip-commits@vger.kernel.org" Subject: Re: [tip:perf/core] perf: Remove the nmi parameter from the swevent and overflow interface Message-ID: <20110708173442.GB31972@e102144-lin.cambridge.arm.com> References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.20 (2009-06-14) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Peter, On Fri, Jul 01, 2011 at 04:21:25PM +0100, tip-bot for Peter Zijlstra wrote: > Commit-ID: a8b0ca17b80e92faab46ee7179ba9e99ccb61233 > Gitweb: http://git.kernel.org/tip/a8b0ca17b80e92faab46ee7179ba9e99ccb61233 > Author: Peter Zijlstra > AuthorDate: Mon, 27 Jun 2011 14:41:57 +0200 > Committer: Ingo Molnar > CommitDate: Fri, 1 Jul 2011 11:06:35 +0200 > > perf: Remove the nmi parameter from the swevent and overflow interface > > The nmi parameter indicated if we could do wakeups from the current > context, if not, we would set some state and self-IPI and let the > resulting interrupt do the wakeup. > > For the various event classes: > > - hardware: nmi=0; PMI is in fact an NMI or we run irq_work_run from > the PMI-tail (ARM etc.) > - tracepoint: nmi=0; since tracepoint could be from NMI context. > - software: nmi=[0,1]; some, like the schedule thing cannot > perform wakeups, and hence need 0. > > As one can see, there is very little nmi=1 usage, and the down-side of > not using it is that on some platforms some software events can have a > jiffy delay in wakeup (when arch_irq_work_raise isn't implemented). > > The up-side however is that we can remove the nmi parameter and save a > bunch of conditionals in fast paths. > > Signed-off-by: Peter Zijlstra Whilst you updated the arch code in this series, you forgot to update the oprofile perf backend. I think you want something as simple as: diff --git a/drivers/oprofile/oprofile_perf.c b/drivers/oprofile/oprofile_perf.c index 59acf9e..94796f3 100644 --- a/drivers/oprofile/oprofile_perf.c +++ b/drivers/oprofile/oprofile_perf.c @@ -31,7 +31,7 @@ static int num_counters; /* * Overflow callback for oprofile. */ -static void op_overflow_handler(struct perf_event *event, int unused, +static void op_overflow_handler(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs) { int id; although I haven't tried it. Cheers, Will