linux-kernel - Re: [PATCH 5/5] x86/vsyscall: Change the default vsyscall mode to xonly

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] [day] [month] [year] [list]

Message-ID: <201906132218.E923F38F@keescook>
Date:   Thu, 13 Jun 2019 22:19:23 -0700
From:   Kees Cook <keescook@...omium.org>
To:     Andy Lutomirski <luto@...nel.org>
Cc:     X86 ML <x86@...nel.org>, LKML <linux-kernel@...r.kernel.org>,
        Borislav Petkov <bp@...en8.de>,
        Kernel Hardening <kernel-hardening@...ts.openwall.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH 5/5] x86/vsyscall: Change the default vsyscall mode to
 xonly

On Thu, Jun 13, 2019 at 12:14:50PM -0700, Andy Lutomirski wrote:
> On Mon, Jun 10, 2019 at 1:44 PM Kees Cook <keescook@...omium.org> wrote:
> >
> > On Mon, Jun 10, 2019 at 01:25:31PM -0700, Andy Lutomirski wrote:
> > > The use case for full emulation over xonly is very esoteric.  Let's
> > > change the default to the safer xonly mode.
> >
> > Perhaps describe the esoteric cases here (and maybe in the Kconfig help
> > text)? That should a user determine if they actually need it. (What
> > would the failure under xonly look like for someone needing emulate?)
> 
> I added it to the Kconfig text.
> 
> Right now, the failure will just be a segfault.  I could add some
> logic so that it would log "invalid read to vsyscall page -- fix your
> userspace or boot with vsyscall=emulate".  Do you think that's
> important?

I think it would be a friendly way to help anyone wondering why
something suddenly started segfaulting, yeah. Just a pr_warn_once() or
something (not a WARN() since it's "intentionally" reachable by
userspace).

-- 
Kees Cook