microcoding a PC into a PDP-11 (was: RE: Classic Computers vs. Classic Computing)

From: Jonathan Engdahl <engdahl_at_cle.ab.com>
Date: Tue Sep 18 14:14:42 2001

The idea is to write PDP-11 microcode for the PC platform, rather running an
"emulator" under Windows or whatever. The Pentium would be viewed as the
micro-architecture, the PDP-11 as the real machine. It would be table driven
and fully expanded, using the PC memory rather extravagantly. You should be
able to emulate simple instructions at the rate of about 4~8 Pentium opcodes
for every PDP-11 opcode. If you rely on the Pentium MMU to trap accesses to
the I/O page, you don't have to check for non-memory accesses from within
the CPU model. The trap routines would emulate PDP-11 I/O, mapping it onto
the PC hardware, rather than onto file I/O as in an emulator. The Pentium
MMU can also be used to emulate the PDP-11 MMU. Map the PDP-11 registers
onto Pentium registers, and never save them in memory except on a trap. This
gives you a very, very fast PDP-11, IBM 1130, or whatever. If you can figure
out a way to cause the machine to boot this "microcode" at powerup instead
of Microsoft Wincrash, I argue that you could legitimately call this a
PDP-11.

I think I remember hearing that the IBM 360 VM OS did this.

It seems a clean way of preserving classic architecture without having to
mess with decayed disk drives, and without the compromises imposed by
emulation.

Sort of like rebuilding the Parthenon with injection-molded faux-marble
columns and friezes. ;-)

--
Jonathan Engdahl???????????????? Rockwell Automation
Principal Research Engineer????? 24800 Tungsten Road
Advanced Technology????????????? Euclid, OH 44117, USA
Euclid Labs????????????????????? engdahl_at_cle.ab.com  216-266-6409
> -----Original Message-----
> From: owner-classiccmp_at_classiccmp.org
> [mailto:owner-classiccmp_at_classiccmp.org]On Behalf Of emanuel stiebler
> Sent: Tuesday, September 18, 2001 1:01 PM
> To: classiccmp
> Subject: Re: microcoding a PC into a PDP-11 (was: RE: Classic Computers
> vs. Classic Computing)
>
>
> Jonathan Engdahl wrote:
> >
> > > -----Original Message-----
> > > [mailto:owner-classiccmp_at_classiccmp.org]On Behalf Of emanuel stiebler
> > > Jonathan Engdahl wrote:
> > >
> > > sizes. Prices are
> > > > about $100 per meg. Something called "flash DIMM" shows up on
> > > >
> > > Looking at this prices, what happened to the old 3.5" floppy drive ?
> > > If you can start a complete/compressed linux/firewall from
> there, should
> > > be enough to start an emulator.
> > > And, BTW, not all of the pc motherboard chip sets supports FLASH DIMM
> > > (any ?),
> > > so you're stuck then with some motherboards.
> >
> > Very good point. I was thinking that it would be nice to have a
> machine that
> > was a PDP-11 as soon as you flipped on the power, but probably
> not worth the
> > cost and hassle of the flash.
>
> And, what I forgot to write is that the flash is slower anyway, so you
> copy the
> software from flash to *RAM anyway.
>
> > Didn't some of the VAXen boot their microcode
> > from a floppy?
>
> yes
>
> > Also, that way you could have one microcode floppy for each classic
> > architecture.
>
> works only if you have the OS on the floppy too.
>
> > I think that my emulator idea can be made to work under Win32.
>
> Sorry, I missed that. What is so special about your idea ?
> (No offense, just missed you posting I guess ;-))
>
> > It appears there are facilities allowing an application
> > program to catch access violations.
>
> Sure. But it is easier to check this yourself.
>
> cheers
>

Received on Tue Sep 18 2001 - 14:14:42 BST