From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751775AbaJAEyT (ORCPT ); Wed, 1 Oct 2014 00:54:19 -0400 Received: from mail-la0-f48.google.com ([209.85.215.48]:36249 "EHLO mail-la0-f48.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750925AbaJAEyS (ORCPT ); Wed, 1 Oct 2014 00:54:18 -0400 MIME-Version: 1.0 In-Reply-To: <542B7200.6030902@landley.net> References: <5c6381879bea68aebb13530442f1cf8a052be97f.1411958379.git.luto@amacapital.net> <542B4DA3.5080105@gmail.com> <542B519B.6010001@landley.net> <542B5E44.40303@gmail.com> <542B7200.6030902@landley.net> From: Andy Lutomirski Date: Tue, 30 Sep 2014 21:53:56 -0700 Message-ID: Subject: Re: [PATCH v5] init: Disable defaults if init= fails To: Rob Landley Cc: frowand.list@gmail.com, Andrew Morton , "linux-kernel@vger.kernel.org" , Chuck Ebbert , Randy Dunlap , Shuah Khan , Rusty Russell Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Sep 30, 2014 at 8:16 PM, Rob Landley wrote: > On 09/30/14 20:52, Frank Rowand wrote: >> On 9/30/2014 5:58 PM, Rob Landley wrote: >>> If you're going to argue that it should "default y", that's a defensible >>> choice. But please don't argue for kernel config symbols with a negative >>> meaning or we'll start having allyesconfig_n brain damage too... >> >> Yes, "default y" is a valid answer to my request. > > Works for me. > >>>> Instead of using a config option, would adding another kernel >>>> command line option, such as 'init_fail_is_fatal', work for >>>> your needs? >>> >>> That was the previous series of patches you ignored, which added code so >>> you can provide _extra_ kernel commands to tell it _not_ to do stuff. >>> The patches did not generate noticeable enthusiasm. >> >> But there also was not a strong push back either. Just Chuck's suggestion >> of an alternate syntax, and your suggestion of instead using a config >> option (and possibly immediately deprecating the config option). >> >> You could as easily frame the argument that the added code was to >> tell the kernel to "_do_ stuff" (panic) instead of "_not_ do stuff". >> But that is just semantics on my part; whatever. >> >> I thought the general trend was to try to avoid adding config options. >> The strictinit method seems fine to me. > > Embedded guys care: > > http://elinux.org/Linux_Tiny > > http://lkml.iu.edu//hypermail/linux/kernel/1409.2/03763.html > >>>> I have a feeling this has already been proposed, >>>> as the 'strictinit' option mentioned in the changes from v3 >>>> below might be the same concept? >>> >>> That was it, yes. >>> >>> Having to get your kernel config right (and your kernel command line >>> right) in order for your system to boot is not really a new concept, is >>> it? You can still specify "init=/bin/sh" if you want that. (I do it all >>> the time when I need to edit a system I haven't bothered to look up the >>> root password to.) >> >> Yes, of course I can. So it falls back to personal preference (as I said, >> I like that some failed boots will drop into a shell without having to >> change the kernel command line). > > The config option lets it do that. Default Y preserves the old behavior. I significantly prefer default N. Scripts that play with init= really don't want the fallback, and I can imagine contexts in which it could be a security problem. --Andy