Does WINDOWS run under VP/ix?

[Glenn Geers]

>From article <1991May17.010926.338 at gorgon.uucp>, by dag at gorgon.uucp (Daniel A. Glasser):
> In article <8466 at uceng.UC.EDU> dmocsny at minerva.che.uc.edu (Daniel Mocsny) writes:
>>In article <1991May14.164301.15364 at aai.com> leo at aai.com writes:
>>>An update, available from ISC, is necessary to run Windows 3.0 under VPIX
>>>on ISC 2.2.
>>Wow! Will this update run any of the other troublesome vpix-buster
>>software such as the Borland Turbo Debugger? And can it run 386-
>>protected-mode DOS software such as the Borland TD386 version of
>>the Turbo Debugger?
> 
> Sorry, no.  The update from ISC for VPIX lets MS-Windows 3.0 run in
> real mode only.  No standard ('286) mode, no Extended ('386) mode.
> It would be amusing, but Intel didn't make the '386 (or '486) such
> that it is easy for one set of '386 protected mode code to simulate
> the iron for another set of '386 protected mode code which is simulaing
> the iron for a bunch of 8086 (real-mode) virtual machines.  It might
> be possible to do, just not easy (both Unix and Windows want to directly
> bang on the MMU...  This is hard to trap/simulate transparently.)
> Maybe someone ought to solve the DOS under Unix problem for '386 systems
> the way I've seen DOS and CP/M done for non-ISA machines for years...
> A hardware solution!  A real '386sx on a bus slot with some shared RAM or
> DMA and some custom ICs that look to the 'sx as if they are standard
> ISA peripherals (floppy controller, HD controller, VGA registers, keyboard
> port, serial, etc.) and allow the main CPU to control access to the
> real devices or simulate the hardware and do wholesale swaps of
> machine context to allow multiple users to use the same board without
> the 'protected mode' software knowing about it at all.  You should even
> be able to run some '386 version of Unix (though this might be pushing
> it) on the inner box and run VPIX on that :)

Don't forget UNIX runs in one 4 Gb segment. I think it is possible to have
a supervisor running in another segment that could switch between UNIX and
windows 3.0 in protected mode. The segment selector is 16 bits wide so you
get 65536 segments (i.e. the 386 can address up to 2^16*2^32 bytes of (very)
virtual memory). What I'm saying is that you could run a lot of different
protected mode operating systems if you had a suitable supervisor program
to manage segment swapping - I'd hate to wait for a 4Gb swap! - and memory
fragmentation and compaction would be a pain!

This idea is explained more fully in "Programming the 80386" by [name forgotten]
and Gellsinger (sp?) - a friend has my copy at present...

> 
> 					Just a few thoughts from
> 					the endless stream of
> 					meaningless drivel from
> 					incoherent mind of
> 
> 						Daniel A. Glasser
> 
> Disclaimer:  The opinions expressed in this posting are not necessarily
>              those of WOOF FM, it's management or sponsers.  They belong
>              to the poster, and he is welcome to them.
> -- 
> Daniel A. Glasser                       One of those things that goes
> dag%gorgon at persoft.com                  "BUMP! (ouch!)" in the night.

Even more mindless drivel,

Glenn
--
___________________________________________________________________________
Glenn Geers                       | "So when it's over, 
                                  |  we're back to people.
Department of Theoretical Physics |  Just to prove that human touch
The University of Sydney          |  can have no equal."
Sydney NSW 2006 Australia         |  - Basia Trzetrzelewska, 'Prime Time TV'
                                  |
Phone: +61 2 692-3241 (voice)     |_________________________________________
       +61 2 660-2903 (fax)       |
                                  |
glenn at qed.physics.su.oz.au        | #include <standard_disclaimer.h>
                                  |
----------------------------------------------------------------------------