linux-kernel - Re: [PATCH] proc: fix NULL dereference when reading /proc/<pid>/auxv

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20161019165958.GM24393@dhcp22.suse.cz>
Date:   Wed, 19 Oct 2016 19:00:00 +0200
From:   Michal Hocko <mhocko@...nel.org>
To:     Oleg Nesterov <oleg@...hat.com>
Cc:     Leon Yu <chianglungyu@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Al Viro <viro@...iv.linux.org.uk>,
        Kees Cook <keescook@...omium.org>,
        John Stultz <john.stultz@...aro.org>,
        Mateusz Guzik <mguzik@...hat.com>,
        Janis Danisevskis <jdanis@...gle.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] proc: fix NULL dereference when reading /proc/<pid>/auxv

On Wed 19-10-16 18:44:29, Oleg Nesterov wrote:
> On 10/19, Michal Hocko wrote:
> >
> > On Wed 19-10-16 18:12:08, Oleg Nesterov wrote:
> > > On 10/19, Michal Hocko wrote:
> > > >
> > > > Why would we even want to open that file if there is no mm struct?
> > >
> > > Agreed this is strange, but I am not sure we can change this old
> > > behaviour. Say, cat /proc/$pid-of-kthread/mem doesn't fail, it shows
> > > the "empty mm".
> >
> > What kind of application would break?
> 
> I have no idea, probably nobody will ever notice this change, just
> I am always nervous when it comes to user-visible changes.
> 
> But why do we want this change? Not that I am going to argue if you
> submit a patch, but personally I see no real point.

I find it much easier to not allow open than a) have a weird side
effects like seeing a hijacked mm and b) risk that similar NULL ptr will
repeat again because all the code paths have to check for it explicitly.
-- 
Michal Hocko
SUSE Labs