Re: Linux, UDI and SCO.

Ernest ter Kuile (TERKUILE@KVI.nl)
Sat, 19 Sep 1998 18:08:18 +0100 (MET)


I still don't quiet accept that UDI is a bad idea. Linux still would have
complete control of the UDI interface itself (well, at least whatever
is behind it) as implemented in the kernel.

As the UDI driver is not allowed to use anything else, how hard would it
be to build enough safegards into the interface to stop any malignant UDI
drivers using this (or doing it by accident, bugs *can* be legitimate) to
crash the kernel ?

Isn't it because of speed of execution that all drivers are so deep
into the kernel ? Let all UDI drivers be removed from the inner kernel
(I don't know.. an other privilege ring ? is that possible ? (does that
even exists ?) maybe even user space), and be heavely checked if a
problem is suspected.

At least it would point a finger to whatever is trying to crash the
system. This should be very acceptable for closed drivers (it might
even help debug open drivers).

Just applying this traitement to UDI drivers not recompiled with the
current kernel for example, (that is, one without sources) might even force
hw makers to give there sources away.

Sorry, I not a kernel programmer. If this make you laugh your head off,
don't blame me ;^).

Cheers,

Ernest.

----------------------------------------------------------------------
E-mail : terkuile@kvi.nl ing. Ernest J.W. ter Kuile
Tel. : (NL)(0)50-3633545 Kernfysisch Versneller Instituut
ICQ : 10055163 Groningen, The Netherlands

-
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/