From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pa0-f42.google.com (mail-pa0-f42.google.com [209.85.220.42]) by kanga.kvack.org (Postfix) with ESMTP id 2704A6B0038 for ; Wed, 17 Dec 2014 13:52:10 -0500 (EST) Received: by mail-pa0-f42.google.com with SMTP id et14so16964012pad.1 for ; Wed, 17 Dec 2014 10:52:09 -0800 (PST) Received: from blackbird.sr71.net ([2001:19d0:2:6:209:6bff:fe9a:902]) by mx.google.com with ESMTP id ou2si6670813pbb.214.2014.12.17.10.52.04 for ; Wed, 17 Dec 2014 10:52:05 -0800 (PST) Message-ID: <5491D0D2.5070103@sr71.net> Date: Wed, 17 Dec 2014 10:52:02 -0800 From: Dave Hansen MIME-Version: 1.0 Subject: Re: post-3.18 performance regression in TLB flushing code References: <5490A5F8.6050504@sr71.net> <20141217100810.GA3461@arm.com> <20141217165310.GJ870@arm.com> In-Reply-To: <20141217165310.GJ870@arm.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Will Deacon , Linus Torvalds Cc: Benjamin Herrenschmidt , Peter Zijlstra , Russell King - ARM Linux , Michal Simek , LKML , Linux-MM On 12/17/2014 08:53 AM, Will Deacon wrote: > On Wed, Dec 17, 2014 at 04:28:23PM +0000, Linus Torvalds wrote: >> On Wed, Dec 17, 2014 at 2:08 AM, Will Deacon wrote: >> So why not just this trivial patch, to make the logic be the same it >> used to be (just using "end > 0" instead of the old "need_flush")? > > Looks fine to me... Dave? First of all, this is quite observable when testing single-threaded on a desktop. This is a mildly crusty Sandybridge CPU from 2011. I made 3 runs with a single thread: ./brk1_processes -s 30 -t 1 fb7332a9fed : 4323385 fb7332a9fed^: 4503736 fb7332a9fed+Linus's fix: 4516761 These things are also a little bit noisy, so we're well within the margin of error with Linus's fix. This also holds up on the large system. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org