[TUHS] Re: "Webster's Second on the Head of a Pin"?

[Douglas McIlroy <douglas.mcilroy@dartmouth.edu>]

I think $IFS ranks at the top of Jim's cabinet of horrors. For our
multilevel-secure Unix (https://www.cs.dartmouth.edu/~doug/IX), he
took almost everything out of the administrative shell, nosh (the
no-shell shell), The classic /bin/sh was not endowed with any of the
capabilities necessary to override Bell-LaPadula-Biba secrecy or
integrity constraints.

On Wed, Jan 1, 2025 at 1:11 PM Rik Farrow <rik@rikfarrow.com> wrote:
>
> I wonder what Reeds meant. I know there are issues. For example, the 3B2 I administered for a while in the late 80s had multiple accounts with rsh, the restricted shell, as the login shell. That was okay, unless you used su and then had access to a root shell.
>
> HP/UX was way worse, with over 120 SUID shell scripts in the 90s. A much more interesting example of insecurity. But somehow, I'm guessing that's not what Reeds wrote about.
>
> Rik
>
>
> On Wed, Jan 1, 2025 at 8:02 AM Douglas McIlroy <douglas.mcilroy@dartmouth.edu> wrote:
>>
>> I have it and will try to scan it in the next few days. Bug me if it
>> doesn't appear.
>>
>> Doug
>>
>> On Tue, Dec 31, 2024 at 11:37 AM Chet Ramey <chet.ramey@case.edu> wrote:
>> >
>> > On 12/29/24 8:44 AM, Douglas McIlroy wrote:
>> > > I can supply a copy if no one else has beaten me to it.
>> > >
>> > > Ron Hardin subsequently pushed the limit even further. Unfortunately,
>> > > I do not have a record of that work.
>> >
>> > Along these same lines, does anyone on the list have a copy of
>> >
>> > "J. A. Reeds, /bin/sh: The biggest UNIX security Loophole,
>> > 11217-840302-04TM, AT&T Bell Laboratories, Murray Hill, NJ (1984)"?
>> >
>> > Years ago, in another lifetime, I wrote and asked him for a copy, but
>> > never got a reply.
>> >
>> > --
>> > ``The lyf so short, the craft so long to lerne.'' - Chaucer
>> >                  ``Ars longa, vita brevis'' - Hippocrates
>> > Chet Ramey, UTech, CWRU    chet@case.edu    http://tiswww.cwru.edu/~chet/