From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: [PATCH] x86 idle: repair large-server 50-watt idle-power regression Date: Thu, 19 Dec 2013 19:14:10 +0100 Message-ID: <20131219181410.GB32508@gmail.com> References: <20131219122257.GC11279@gmail.com> <52B316FF.50906@zytor.com> <20131219160210.GA28426@gmail.com> <52B31B21.6010901@zytor.com> <20131219162136.GM16438@laptop.programming.kicks-ass.net> <52B323BE.7090108@zytor.com> <20131219170741.GB30382@gmail.com> <20131219172535.GN16438@laptop.programming.kicks-ass.net> <20131219173629.GJ2480@laptop.programming.kicks-ass.net> <52B33567.9060205@zytor.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from mail-we0-f181.google.com ([74.125.82.181]:45087 "EHLO mail-we0-f181.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754830Ab3LSSOO (ORCPT ); Thu, 19 Dec 2013 13:14:14 -0500 Content-Disposition: inline In-Reply-To: <52B33567.9060205@zytor.com> Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: "H. Peter Anvin" Cc: Peter Zijlstra , Len Brown , x86@kernel.org, linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org, Len Brown , stable@vger.kernel.org, Linus Torvalds , Thomas Gleixner , Mike Galbraith , Borislav Petkov * H. Peter Anvin wrote: > On 12/19/2013 09:36 AM, Peter Zijlstra wrote: > > On Thu, Dec 19, 2013 at 06:25:35PM +0100, Peter Zijlstra wrote: > >> That said, I would find it very strange indeed if a CLFLUSH doesn't also > >> flush the store buffer. > > > > OK, it explicitly states it does not do that and you indeed need > > an mfence before the clflush. > > So, MONITOR is defined to be ordered as a load, which I think should > be adequate, but I really wonder if we should have mfence on both > sides of clflush. This now is up to 9 bytes, and perhaps pushing it > a bit with how much we would be willing to patch out. > > On the other hand - the CLFLUSH seems to have worked well enough by > itself, and this is all probabilistic anyway, so perhaps we should > just leave the naked CLFLUSH in and not worry about it unless > measurements say otherwise? So I think the window of breakage was rather large here, and since it seems to trigger on rare types of hardware I think we'd be better off by erring on the side of robustness this time around ... This is the 'go to idle' path, which isn't as time-critical as the 'get out of idle' code path. Thanks, Ingo