[34755] in bugtraq
Re: http://www.smashguard.org
daemon@ATHENA.MIT.EDU (Theo de Raadt)
Sat May 1 12:25:52 2004
Message-Id: <200405010045.i410jg9M010605@cvs.openbsd.org>
To: Pavel Machek <pavel@ucw.cz>
Cc: Crispin Cowan <crispin@immunix.com>,
Hilmi Ozdoganoglu <cyprian@purdue.edu>,
Dave Paris <dparis@w3works.com>, bugtraq@securityfocus.com
In-reply-to: Your message of "Thu, 29 Apr 2004 23:55:07 +0200."
<20040429215507.GA22284@elf.ucw.cz>
Date: Fri, 30 Apr 2004 18:45:42 -0600
From: Theo de Raadt <deraadt@cvs.openbsd.org>
> > >The idea is not to create "custom CPUs" but to have our modification
> > >picked up by major vendors. Clearly there is interest in applying
> > >hardware to solve security issues based on the latest press releases
> > >from AMD that AMD chips include buffer-overflow protection (see
> > >Computer World, January 15, 2004).
> > >
> > As Theo said, the AMD buffer overflow "protection" is nothing more than
> > sensible separation of R and X bits per page, fixing a glaring and
>
> Actually it is not "sensible", and it is not separation.
>
> You can have r--, r-x, but you can't have --x.
Oh for the record. A few chips make it possible to have --x
permissions.
alpha (I am not positive)
sparc64 (I am not positive)
ia64
hppa
amd29k
m88k
The first two have software tlb refillers with a split tlb architecture,
but I am not sure if there is tlb "leak"
The next three have specific page table bits for kernel (r w x) and
user (r w x).
The last has a harvard-style split mmu (entirely different mmu for
code and data), and it should be possible to play games to do it...
