lkml.org 
[lkml]   [2000]   [Jul]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: disk-destroyer.c
Andre Hedrick writes:
> On Fri, 21 Jul 2000, Karen Shaeffer wrote:
>
> > Thank you for the nice summary. And, certainly the patch should be applied.
> >
> > But you have described defective hardware. And despite the patch, it is still
> > possible for a malicious root to exploit this defective hardware. So, this
> > is ultimately a hardware manufacturer issue. A list of boards with the defective
> > disk controller chip should be released to the public, so consumers can be
> > aware of the problem and *NOT* purchase these controllers in the future.
>
> Assume no hardware has defects, it is an exploit against the ATA SPECS.

Fine, then the ATA specification is broken instead of the disk hardware.
If the ATA specification mandates that compliant drives contain the
equivalent of a self-destruct command, then this isn't a problem with
the Linux kernel. Even if the IDE driver in the kernel has checks to
prevent the self-destruct command from being sent to the drive via
normal IDE driver operations, this can't guarantee that the
self-destruct command can never be sent to the drive; malicious kernel
code or root-run programs will still be able to deliver the command
directly to the drive without going through the IDE driver.

Can we just be done with this now? This is by no means the first or the
worst case where bad kernel code or incompetence or malice on the part
of a root user can damage hardware devices beyond repair. If there's an
earth-shattering revelation here, it's that the ATA specification is
fundamentally flawed and needs to be fixed for there to be a real
solution to this problem.

And please, Andre, count to 10 and relax a bit before sending your
replies. Practically all I've seen from you on this is hysterical and
sometimes incomprehensible ranting, and gratuitous insults directed
towards even the most reasoned respondents. If you just calmly
explained the exact nature of the problem in the ATA specification, it
would help us all figure out what solution, if any, could be applied to
the problem in the Linux kernel.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:57    [W:0.677 / U:0.092 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site