From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755865AbYB2CkY (ORCPT ); Thu, 28 Feb 2008 21:40:24 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751667AbYB2CkI (ORCPT ); Thu, 28 Feb 2008 21:40:08 -0500 Received: from mx1.redhat.com ([66.187.233.31]:54744 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753433AbYB2CkG (ORCPT ); Thu, 28 Feb 2008 21:40:06 -0500 Date: Thu, 28 Feb 2008 21:39:09 -0500 From: Jason Baron To: Mathieu Desnoyers Cc: Max Krasnyanskiy , akpm@linux-foundation.org, Ingo Molnar , linux-kernel@vger.kernel.org, Rusty Russell Subject: Re: [patch 1/2] add ALL_CPUS option to stop_machine_run() Message-ID: <20080229023909.GA3440@redhat.com> References: <20080202210828.840735763@polymtl.ca> <20080202211204.268876860@polymtl.ca> <20080226225242.GA15926@redhat.com> <20080227190519.GA14335@Krystal> <20080228163303.GA6195@redhat.com> <47C73121.2000109@qualcomm.com> <20080228221430.GA22958@Krystal> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080228221430.GA22958@Krystal> User-Agent: Mutt/1.5.14 (2007-02-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Feb 28, 2008 at 05:14:30PM -0500, Mathieu Desnoyers wrote: > * Max Krasnyanskiy (maxk@qualcomm.com) wrote: > > Jason Baron wrote: > >> -allow stop_mahcine_run() to call a function on all cpus. Calling > >> stop_machine_run() with a 'ALL_CPUS' invokes this new behavior. > >> stop_machine_run() proceeds as normal until the calling cpu has > >> invoked 'fn'. Then, we tell all the other cpus to call 'fn'. > > > > Jason, we're actually trying to reduce the usage of the stop_machine in > > general. It's a very big hammer that kills latencies and stuff. It'd be > > nice if we did not introduce any more dependencies on it. I guess in some > > case there is simply no other way to handle what need to do. But please > > think twice > > (or more :)). > > > > Max > > > > > > I have a "more complex" immediate value implementation that does not > depend on such heavy lock. I made this simplified version because Rusty > preferred it, although I say from the beginning that it kills interrupt > latency. I could propose the atomic, nmi-safe version directly if enough > people are in favor of it. > > Mathieu > to me the updating of the immdiate values isn't the critical path, but obviously i'd be in favor of a more efficient implementation. -Jason