Re: CVE-2016-4484: - Cryptsetup Initrd root Shell

[John Haxby <john.haxby () oracle com>]

On 17/11/16 17:50, Jason Cooper wrote:
> Hi John,
>
> On Thu, Nov 17, 2016 at 04:56:06PM +0000, John Haxby wrote:
> > > On 17/11/16 16:39, Jason Cooper wrote:
> > > > > However, the golden rule still applies.  Physical access trumps all
> > > > > defensive measures.  The absolute best you can do is detect that
> > > > > physical access occurred.  From there, you're hoping there are no
> > > > > hardware implants or other devices outside the scope of software
> > > > > security.
> > >
> > > I agree.  However, it ought be to be harder than leaning on the enter
> > > key to break into a system.  You lock your doors even though it doesn't
> > > stop a determined burglar?
> Yes, as I said before, non-deterministic failure modes are bad.  This
> CVE is a bug in the initrd script and needs to be fixed.  What I
> disagree with, and still do, is the "sky is falling!" nature of the
> alert.

Yup.  I agree there, but that's down to the publicity its received.  We
can't do much but grin and bear that.   Red Hat have given this a CVSS2
score of 7.2 which reflects the _potential_ severity but marked it as
"moderate" which reflects the actual effect.   It's most serious for
people like me who have an encrypted root but no grub password (and no
rd.shell=0) but, no, the sky is not falling.

jch