Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753287Ab0AHRo0 (ORCPT ); Fri, 8 Jan 2010 12:44:26 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753142Ab0AHRoY (ORCPT ); Fri, 8 Jan 2010 12:44:24 -0500 Received: from nlpi129.sbcis.sbc.com ([207.115.36.143]:35724 "EHLO nlpi129.prodigy.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752975Ab0AHRoY (ORCPT ); Fri, 8 Jan 2010 12:44:24 -0500 Date: Fri, 8 Jan 2010 11:43:41 -0600 (CST) From: Christoph Lameter X-X-Sender: cl@router.home To: Linus Torvalds cc: Peter Zijlstra , KAMEZAWA Hiroyuki , Minchan Kim , "Paul E. McKenney" , "linux-kernel@vger.kernel.org" , "linux-mm@kvack.org" , "hugh.dickins" , Nick Piggin , Ingo Molnar Subject: Re: [RFC][PATCH 6/8] mm: handle_speculative_fault() In-Reply-To: Message-ID: References: <20100104182429.833180340@chello.nl> <20100104182813.753545361@chello.nl> <20100105092559.1de8b613.kamezawa.hiroyu@jp.fujitsu.com> <28c262361001042029w4b95f226lf54a3ed6a4291a3b@mail.gmail.com> <20100105134357.4bfb4951.kamezawa.hiroyu@jp.fujitsu.com> <20100105143046.73938ea2.kamezawa.hiroyu@jp.fujitsu.com> <20100105163939.a3f146fb.kamezawa.hiroyu@jp.fujitsu.com> <20100106092212.c8766aa8.kamezawa.hiroyu@jp.fujitsu.com> <20100106115233.5621bd5e.kamezawa.hiroyu@jp.fujitsu.com> <20100106125625.b02c1b3a.kamezawa.hiroyu@jp.fujitsu.com> <1262969610.4244.36.camel@laptop> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1336 Lines: 29 On Fri, 8 Jan 2010, Linus Torvalds wrote: > That's a huge jump. It's clear that the spinlock-based rwsem's simply > suck. The speculation gets rid of some additional mmap_sem contention, > but at least for two sockets it looks like the rwsem implementation was > the biggest problem by far. I'd say that the ticket lock sucks for short critical sections vs. a simple spinlock since it forces the cacheline into shared mode. > Of course, larger numbers of sockets will likely change the situation, but > at the same time I do suspect that workloads designed for hundreds of > cores will need to try to behave better than that benchmark anyway ;) Can we at least consider a typical standard business server, dual quad core hyperthreaded with 16 "cpus"? Cacheline contention will increase significantly there. > Because let's face it - if your workload does several million page faults > per second, you're just doing something fundamentally _wrong_. You may just want to get your app running and its trying to initialize its memory in parallel on all threads. Nothing wrong with that. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/