[96] in linux-security and linux-alert archive

home help back first fref pref prev next nref lref last post

Re: SvgaLib (was Re: Secure setup for file transfert)

daemon@ATHENA.MIT.EDU (Fabrizio Giudici)
Fri Mar 10 11:13:54 1995

Date: Fri, 10 Mar 1995 13:29:22 +0100
From: Fabrizio Giudici <fritz@dibe.unige.it>
To: linux-security@tarsier.cv.nrao.edu
In-Reply-To: <9503091410.AA12809@is1e.vub.ac.be>
Reply-To: linux-security@tarsier.cv.nrao.edu

On Thu, 9 Mar 1995, GLAUDE DAVID wrote:

> Mr Martin J Hargreaves said:
>
> Well, is there any way to secure program ussing svgalib.
> It seems that to access vga io port you need some priviledge wich is an 
>  increase of security (not anybody should be able to turn you screen upside
>  down). But because of the lack of security level in Unix (root or not root),
>  all program for Vga have to be run as root (I always log as root but don't
>  do as I do) or to be setuid root wich is a potential risk. (see above)
> Is there any other solution than setuid root thoses programs (like gs with
>  the vga console driver). Shouldn't a solution be search ?
> 
> -- 
> 	GLAUDE David	dglaude@is1.ulb.ac.be	[Glu]
> 	I speak French: "Linux est l'unique Unix de Linus." 
> 

I thought a possible solution could be to create a daemon (perhaps named
vgad?) that is able to process very-low-level queries like "write a bunch
of registers to vga" and so on. Such a daemon could also implement locking
mechanism to prevent from simultaneous VGA access more than one program 
(sometimes I unintentionally got X locked up by mistakenly running a program
based on a library of mine, similar to svgalib, while X was running).

I think this approach could be better than implementing special devices 
like /dev/vga or similar, because with a daemon no modifications to the 
kernel are required (and I think it's a good thing to keep the kernel as
"light" as possible).

This is my idea. I'm an relatively-experienced programmer, but I've still
*lots* of things to learn about UNIX/LINUX programming, so I really don't
know if there is some tech problem in my approach. So please don't flame
;)

Ciao.

.---------------------------------------------------------------------------.
|       Fabrizio Giudici (fritz@dibe.unige.it)        |                     |
|           WWW-PAGE: < work in progress >            | Style distinguishes |
|        PHONE: +39 10 3532163/3532780/3532781        | excellence between  |
|      Dept. of Biophys. and Elect. Eng.  (DIBE)      |   accomplishment.   |
|    University of Genoa - ITALY - EUROPE - EARTH     |                     |
|---------------------------------------------------------------------------|
|   "For a succesful technology, reality must take precedence over public   |
|   relations, for Nature cannot be fooled."         - Richard P. Feynman   |
`---------------------------------------------------------------------------'
 All expressed opinions are personal and not of the organization I work for. 

home help back first fref pref prev next nref lref last post