From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+willy=40w.ods.org-S261561AbULIRFu@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S261561AbULIRFu (ORCPT <rfc822;willy@w.ods.org>);
	Thu, 9 Dec 2004 12:05:50 -0500
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261566AbULIRFs
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Thu, 9 Dec 2004 12:05:48 -0500
Received: from omx1-ext.sgi.com ([192.48.179.11]:14247 "EHLO
	omx1.americas.sgi.com") by vger.kernel.org with ESMTP
	id S261561AbULIRFd (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Thu, 9 Dec 2004 12:05:33 -0500
Date: Thu, 9 Dec 2004 09:05:25 -0800 (PST)
From: Christoph Lameter <clameter@sgi.com>
X-X-Sender: clameter@schroedinger.engr.sgi.com
To: Pavel Machek <pavel@ucw.cz>
cc: nickpiggin@yahoo.com.au, Jeff Garzik <jgarzik@pobox.com>,
       torvalds@osdl.org, hugh@veritas.com, benh@kernel.crashing.org,
       linux-mm@kvack.org, linux-ia64@vger.kernel.org,
       linux-kernel@vger.kernel.org
Subject: Re: Anticipatory prefaulting in the page fault handler V1
In-Reply-To: <20041209105753.GB1131@elf.ucw.cz>
Message-ID: <Pine.LNX.4.58.0412090904100.10400@schroedinger.engr.sgi.com>
References: <Pine.LNX.4.58.0412011539170.5721@schroedinger.engr.sgi.com>
 <Pine.LNX.4.58.0412011608500.22796@ppc970.osdl.org> <41AEB44D.2040805@pobox.com>
 <20041201223441.3820fbc0.akpm@osdl.org> <41AEBAB9.3050705@pobox.com>
 <20041201230217.1d2071a8.akpm@osdl.org> <179540000.1101972418@[10.10.2.4]>
 <41AEC4D7.4060507@pobox.com> <20041202101029.7fe8b303.cliffw@osdl.org>
 <Pine.LNX.4.58.0412080920240.27156@schroedinger.engr.sgi.com>
 <20041209105753.GB1131@elf.ucw.cz>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org

On Thu, 9 Dec 2004, Pavel Machek wrote:

> Hi!
>
> > Standard Kernel on a 512 Cpu machine allocating 32GB with an increasing
> > number of threads (and thus increasing parallellism of page faults):
> >
> >  Gb Rep Threads   User      System     Wall flt/cpu/s fault/wsec
> >  32   3    1    1.416s    138.165s 139.050s 45073.831  45097.498
> ...
> > Patched kernel:
> >
> > Gb Rep Threads   User      System     Wall flt/cpu/s fault/wsec
> >  32   3    1    1.098s    138.544s 139.063s 45053.657  45057.920
> ...
> > These number are roughly equal to what can be accomplished with the
> > page fault scalability patches.
> >
> > Kernel patches with both the page fault scalability patches and
> > prefaulting:
> >
> >  Gb Rep Threads   User      System     Wall flt/cpu/s fault/wsec
> >  32  10    1    4.103s    456.384s 460.046s 45541.992  45544.369
> ...
> >
> > The fault rate doubles when both patches are applied.
> ...
> > We are getting into an almost linear scalability in the high end with
> > both patches and end up with a fault rate > 3 mio faults per second.
>
> Well, with both patches you also slow single-threaded case more than
> twice. What are the effects of this patch on UP system?

The faults per second are slightly increased, so its faster. The last
numbers are 10  repetitions and not 3. Do not look at the wall time.