From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753510AbXDJWtf (ORCPT ); Tue, 10 Apr 2007 18:49:35 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753514AbXDJWte (ORCPT ); Tue, 10 Apr 2007 18:49:34 -0400 Received: from gw.goop.org ([64.81.55.164]:33853 "EHLO mail.goop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753510AbXDJWte (ORCPT ); Tue, 10 Apr 2007 18:49:34 -0400 Message-ID: <461C1470.60308@goop.org> Date: Tue, 10 Apr 2007 15:49:20 -0700 From: Jeremy Fitzhardinge User-Agent: Thunderbird 1.5.0.10 (X11/20070302) MIME-Version: 1.0 To: "Eric W. Biederman" CC: Andrew Morton , Andi Kleen , Linux Kernel Mailing List , Rusty Russell Subject: Re: [PATCH 2/2] Allow percpu variables to be page-aligned References: <461BC4B8.40902@goop.org> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Eric W. Biederman wrote: > Jeremy Fitzhardinge writes: > > >> Let's allow page-alignment in general for per-cpu data (wanted by Xen, and >> Ingo suggested KVM as well). >> >> Because larger alignments can use more room, we increase the max per-cpu >> memory to 64k rather than 32k: it's getting a little tight. >> > > The second paragraph of the comment is dated. You are not changing the per-cpu > memory size at all now. > Yup. > Although we should probably fail the module load if the requested alignment > is to great, but that is just picking nits, and is no worse than what we > do today. > Yes. And we barely support a module's use of page-aligned data, since it would potentially chew through the percpu space very quickly. J