From: Greg KH Subject: Re: [stable] [PATCH] stable - ext[234]: Avoid printk floods in the face of directory corruption (CVE-2008-3528) Date: Wed, 22 Oct 2008 10:37:52 -0700 Message-ID: <20081022173752.GA10587@kroah.com> References: <48FF42B8.3030606@redhat.com> <20081022161433.GB29489@kroah.com> <48FF52F4.10702@redhat.com> <20081022163421.GA28836@kroah.com> <48FF644A.50801@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: stable@kernel.org, ext4 development To: Eric Sandeen Return-path: Received: from kroah.org ([198.145.64.141]:49737 "EHLO coco.kroah.org" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1757888AbYJVRnh (ORCPT ); Wed, 22 Oct 2008 13:43:37 -0400 Content-Disposition: inline In-Reply-To: <48FF644A.50801@redhat.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, Oct 22, 2008 at 12:35:06PM -0500, Eric Sandeen wrote: > Greg KH wrote: > > On Wed, Oct 22, 2008 at 11:21:08AM -0500, Eric Sandeen wrote: > >> Greg KH wrote: > >>> On Wed, Oct 22, 2008 at 10:11:52AM -0500, Eric Sandeen wrote: > >>>> This is a trivial backport of the following upstream commits: > >>>> > >>>> - bd39597cbd42a784105a04010100e27267481c67 (ext2) > >>>> - cdbf6dba28e8e6268c8420857696309470009fd9 (ext3) > >>>> - 9d9f177572d9e4eba0f2e18523b44f90dd51fe74 (ext4) > >>>> > >>>> This addresses CVE-2008-3528 > >>>> > >>>> ext[234]: Avoid printk floods in the face of directory corruption > >>> For what kernel releases is this applicable? .27? .26? .25? Earlier? > >> Sorry.. it is applicable to pretty much any kernel in the past :) .27 > >> certainly (that's what the patch is against), .26, .25.... yes. > >> > >> It's not a particularly dangerous condition - you have to somehow get > >> the administrator to mount the filesystem before you can trigger the > >> "exploit" (which is a DoS, essentially) - so, I don't know if it's worth > >> porting back to the dawn of time... > > > > Well, I will not port it back to older kernels than .25, so that's not a > > big deal. > > > > As for the "admin mount a filesystem", you could put an ext2/3 fs on a > > usb stick and plug it into a box. It will be mounted automatically, no > > admin rights required, and the DoS would happen, right? > > If I wanted to DoS a box sitting in front of me, I'd just pull the plug. Yes, the fun "physical access" issue, right? But for some, who run Linux in a "kiosk" mode, or in semi-secured places like university labs, something like this would matter, so you might want to notify the distros of this issue through vendor-sec and let them make up their minds if they wish to backport the fixes to their supported releases. thanks, greg k-h