From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752951AbZKJIBM (ORCPT ); Tue, 10 Nov 2009 03:01:12 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752874AbZKJIBL (ORCPT ); Tue, 10 Nov 2009 03:01:11 -0500 Received: from out02.mta.xmission.com ([166.70.13.232]:47186 "EHLO out02.mta.xmission.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752294AbZKJIBK (ORCPT ); Tue, 10 Nov 2009 03:01:10 -0500 To: Andi Kleen Cc: Arnd Bergmann , Arjan van de Ven , linux-kernel@vger.kernel.org Subject: Re: [PATCH 22/23] sysctl arm: Remove binary sysctl support References: <20091109132830.GF26740@basil.fritz.box> <200911091628.47003.arnd@arndb.de> <20091109154621.GG26740@basil.fritz.box> From: ebiederm@xmission.com (Eric W. Biederman) Date: Tue, 10 Nov 2009 00:01:07 -0800 In-Reply-To: <20091109154621.GG26740@basil.fritz.box> (Andi Kleen's message of "Mon\, 9 Nov 2009 16\:46\:21 +0100") Message-ID: User-Agent: Gnus/5.11 (Gnus v5.11) Emacs/22.2 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-XM-SPF: eid=;;;mid=;;;hst=in02.mta.xmission.com;;;ip=76.21.114.89;;;frm=ebiederm@xmission.com;;;spf=neutral X-SA-Exim-Connect-IP: 76.21.114.89 X-SA-Exim-Mail-From: ebiederm@xmission.com X-SA-Exim-Version: 4.2.1 (built Thu, 25 Oct 2007 00:26:12 +0000) X-SA-Exim-Scanned: No (on in02.mta.xmission.com); Unknown failure Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Andi Kleen writes: >> can still use the config option for the compatibility code. >> There wouldn't even be a performance penalty over new glibc with >> new kernels which already use procfs. > > When he drops the sysctl(2) API completely the old userland will > be unhappy. When it comes to breaking user space the issue is not glibc or anything in a open source distribution. There are maybe 3 apps that are affected at all and as far as I can tell they were all fixed long ago. Last time I investigated this in 2007 none of the existing userlands would break if sysctl was disabled completely. The issue is all of the in-house and 3rd party software, that we don't have access to. We have no way of knowing what is used. So I do not favor a solution that kills some but not all sysctls. Either you might need them or you don't. So the decision must be made to you break the rare app from the writer who has given you no feed back after you have given him fair warning, or do you try and support him forever, with code that no cares about and no one tests. I think the solution in sysctl_binary.c strikes a good balance there. The code is simple and should survive a long time with little to no maintenance, so it is good if we need decide we are too chicken to disable the sysctl system call, and at the same time my changes remove the support overhead from everywhere else. Further with the existence of a single file for all of the binary sysctls we can modify it to strike any other balance between long term maintainability, and removal of unused cruft. I am happy to see additional patches from anyone else who chooses to seek a different balance. Eric