Carl Lowenstein wrote in
news:200505021807.j42I7CH11861@opihi.ucsd.edu:

>> From: "Meat Loaf"
>> Subject: Re: First boot on PDP-11/23!
>> Date: Sun, 01 May 2005 22:35:10 -0500
>> To: info-pdp11@village.org
>> > bill@cs.uofs.edu (Bill Gunshannon) wrote in
>> > news:3dbu28F6sp4bsU1@individual.net:
>> >
>> > > In article ,
>> > > William Korvine writes:
>> > >> Powered 'er up and got "28/START?"!!!
>> > >> Yay, I made it to the console debugger!
>> > >>
>> > >> Help: The "28" means 28K detected. I have 256K installed. I have
>> > >> a memory problem. How can I proceed?
>> > >
>> > > I think 28KW is all the startup ever detects.
>> > >
>> > > bill
>> > >
>> >
>> > Ok, I used ODT and poked values into n00004 where 0<=n<=7 and read
>> > the correct values back. I conclude that despite the value at the
>> > START prompt, I actually have 256K of working RAM.
>> >
>> > I'm going to end my works here for a little while. For those who
>> > have been following my travails, thanks for your attention and aid.
>> >
>> > I now have a working CPU with 256kb of RAM and a working 4-port
>> > serial card. This is everything I planned to accomplish at this
>> > stage.
>> >
>> > I have no storage devices, so there's not much more I can do until
>> > I get them. Since I have no room for RL or RX drives, I must wait
>> > for a SCSI interface to become available or build an emulator
>> > device. I've seen the plans for the emulator boards such as at
>> > http://www.geocities.com/saipan59/dec/rxv11_emu.html
>> > and doubt I will attempt it. Besides, if I wanted an emulated PDP,
>> > I would have downloaded one.
>> >
>> > I still have the RLV11 boards, although the capacitors on one of
>> > them are physically damaged and need replacing. I will probably
>> > attempt that eventually just to preserve the boards. And who knows,
>> > I might have room for RLs some day.
>> >
>> > Last post for a little while. Thanks, and take care.
>> >
>> > .wk.
>>
>> What you will find is that the PDP-11's basically use 28K for memory
>> and 4K to address what they call the I/O page. Unlike the Intel type
>> processor that has an input and output instruction the PDP's (and
>> VAX's) used memory mapped I/O. Also to use memory greater than 28K
>> it is done with memory management hardware that rearranges 4k pages
>> to point to different areas of physical memory. The PDP-11/23 also
>> has a memory management seperation for kernel and user mode
>> operation. This allowed some added protection, such as a user mode
>> "HALT" instruction would not execute instead it would cause an
>> illegal instruction TRAP to occur.
>
> Something that can be very confusing on an 11/23: Console ODT does
> not have the capability of addressing the full 22-bit memory address
> space.
>
> Top of 16-bit memory space is 160000 - 2
> Top of 18-bit memory space is 760000 - 2
> Top of 22-bit memory space is 17760000 - 2
>
> If you type addresses greater than 777776 to Console ODT, they will be
> wrapped in the 18-bit space.
>
> carl

I didn't know that. I'm not too worried because I only have an 18-bit
address space anyway (I'm pretty sure).

..wk.