From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1161202AbXDKFXv (ORCPT ); Wed, 11 Apr 2007 01:23:51 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1161209AbXDKFXv (ORCPT ); Wed, 11 Apr 2007 01:23:51 -0400 Received: from madara.hpl.hp.com ([192.6.19.124]:55451 "EHLO madara.hpl.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1161202AbXDKFXu (ORCPT ); Wed, 11 Apr 2007 01:23:50 -0400 Date: Tue, 10 Apr 2007 22:20:07 -0700 From: Stephane Eranian To: Venki Pallipadi Cc: Adam Belay , Andi Kleen , "Rafael J. Wysocki" , Andrew Morton , linux-kernel@vger.kernel.org, randy.dunlap@oracle.com, lenb@kernel.org Subject: Re: 2.6.21-rc6-mm1 Message-ID: <20070411052007.GA24056@frankl.hpl.hp.com> Reply-To: eranian@hpl.hp.com References: <653FFBB4508B9042B5D43DC9E18836F539ADFB@scsmsx415.amr.corp.intel.com> <200704091940.53439.rjw@sisk.pl> <1176249569.32554.229.camel@localhost.localdomain> <20070411001513.GA2975@linux-os.sc.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20070411001513.GA2975@linux-os.sc.intel.com> User-Agent: Mutt/1.4.1i Organisation: HP Labs Palo Alto Address: HP Labs, 1U-17, 1501 Page Mill road, Palo Alto, CA 94304, USA. E-mail: eranian@hpl.hp.com X-HPL-MailScanner: Found to be clean X-HPL-MailScanner-From: eranian@hpl.hp.com Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Venki, On Tue, Apr 10, 2007 at 05:15:14PM -0700, Venki Pallipadi wrote: > > > x86-64 expects all idle handlers to enable interrupts before returning from > > > idle handler. This is due to enter_idle(), exit_idle() races. Make > > > cpuidle_idle_call() confirm to this when there is no pm_idle_old. > > > > > > Also, cpuidle look at the return values of attch_driver() and set > > > current_driver to NULL if attach fails on all CPUs. > > > > My vote would be to instead remove enter_idle() and exit_idle() from > > x86-64, just as was done with i386. Performance monitoring > > infrastructure shouldn't be interfering with the idle interrupt > > delivery, as that could only hurt performance... Besides, there's > > probably a better way of doing this than an idle notifier anyway. > > > > Agreed. I did not like local_irq_enable() in cpuidle either, but added it > anyway as it was a corner case when cpuidle is active and no driver is > active and not a common case. I thought we will have it as a bandaid solution > until enter_idle, exit_idle is around. > > Andi/Stephane: What are the plans around enter_idle exit_idle in x86-64. > Is it still being used by perfmon for x86-64 arch? > The next kernel patch for Perfmon will not make use of the idle notification anymore on any platform. -- -Stephane