From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754482AbYISWIK (ORCPT ); Fri, 19 Sep 2008 18:08:10 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751869AbYISWH5 (ORCPT ); Fri, 19 Sep 2008 18:07:57 -0400 Received: from wa-out-1112.google.com ([209.85.146.180]:43986 "EHLO wa-out-1112.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751694AbYISWH4 (ORCPT ); Fri, 19 Sep 2008 18:07:56 -0400 DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:sender:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references:x-google-sender-auth; b=tNSqW32REeZ3uUaAfkrehy1r32s0rMrbhbMdJmm3SmV4wSjrJvL121ZipT03/jcB06 R1+LFDiQi1kyYuLg7ViRLpmGB/sjcaXPDtJd/oza1Xk4qW4F6Q9z6248U1KNAE3JkluU Xp+Cnl3vTFxP2h8K0BffCzFSkGNQUeg9gZQT0= Message-ID: <84144f020809191507w64ff1c47m40c2c500da3fdfa6@mail.gmail.com> Date: Sat, 20 Sep 2008 01:07:55 +0300 From: "Pekka Enberg" To: "Christoph Lameter" Subject: Re: [patch 3/3] Increase default reserve percpu area Cc: akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, jeremy@goop.org, ebiederm@xmission.com, travis@sgi.com, herbert@gondor.apana.org.au, xemul@openvz.org In-Reply-To: <84144f020809191502i35805980n97dc3073ab8a52bc@mail.gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline References: <20080919203703.312007962@quilx.com> <20080919203724.474751340@quilx.com> <84144f020809191502i35805980n97dc3073ab8a52bc@mail.gmail.com> X-Google-Sender-Auth: eec10f270d661d22 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Sep 20, 2008 at 1:02 AM, Pekka Enberg wrote: > On Fri, Sep 19, 2008 at 11:37 PM, Christoph Lameter > wrote: >> SLUB now requires a portion of the per cpu reserve. There are on average >> about 70 real slabs on a system (aliases do not count) and each needs 12 bytes >> of per cpu space. Thats 840 bytes. In debug mode all slabs will be real slabs >> which will make us end up with 150 -> 1800. Give it some slack and add 2000 >> bytes to the default size. >> >> Things work fine without this patch but then slub will reduce the percpu reserve >> for modules. > > Hmm, shouldn't this be dynamically configured at runtime by > multiplying the number of possible CPUs with some constant? Oh, never mind, setup_per_cpu_areas() already does that, obviously.