[TUHS] Unix on Zilog Z8000?

[TUHS] Unix on Zilog Z8000?

[Jon Forrest]

There's been a lot of discussion about early Unix on Intel, National
Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
the Z8000, and if not, why not.

As I remember the Z8000 was going to be the great white hope that
would continue Zilog's success with the Z80 into modern times.
But, it obviously didn't happen.

Why?

Jon

Re: [TUHS] Unix on Zilog Z8000?

[Steve Nickolas]

On Tue, 21 Jan 2020, Jon Forrest wrote:

> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>
> As I remember the Z8000 was going to be the great white hope that
> would continue Zilog's success with the Z80 into modern times.
> But, it obviously didn't happen.
>
> Why?
>
> Jon
>

Didn't Coherent run on the Z8K?  Though that's not really Unix.

-uso.

Re: [TUHS] Unix on Zilog Z8000?

[William Pechter]

[-- Attachment #1: Type: text/plain, Size: 1427 bytes --]

IIRC... Zilog had Zeus (a SYSIII version) running.  Exxon Office Systems swallowed and regurgitated Zilog. They were a Vax customer in Princeton in the late '80s.  I installed one of the Vax 11/780s with the first RM80 I (and my office) ever saw as a brand new Field Service grunt. 

The US Internal Revenue Service was looking at replacing (Z8000 systems) them with AT&T sold Pyramid boxes when I was there.  Then they swallowed NCR and it all fell apart. This was around '94 or so. 

AT&T and Siemens were both Pyramid OEMs and were about 50% of Pyramid's business (I was told). Anyone have further info?   When Exxon pulled the plug on EOS I thing things were up in the air... 

Bill

Sent from pechter@gmail.com

-----Original Message-----
From: Steve Nickolas <usotsuki@buric.co>
To: Jon Forrest <nobozo@gmail.com>
Cc: The Unix Heritage Society mailing list <tuhs@tuhs.org>
Sent: Tue, 21 Jan 2020 13:14
Subject: Re: [TUHS] Unix on Zilog Z8000?

On Tue, 21 Jan 2020, Jon Forrest wrote:

> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>
> As I remember the Z8000 was going to be the great white hope that
> would continue Zilog's success with the Z80 into modern times.
> But, it obviously didn't happen.
>
> Why?
>
> Jon
>

Didn't Coherent run on the Z8K?  Though that's not really Unix.

-uso.

[-- Attachment #2: Type: text/html, Size: 1912 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Henry Bent]

[-- Attachment #1: Type: text/plain, Size: 387 bytes --]

On Tue, 21 Jan 2020 at 12:53, Jon Forrest <nobozo@gmail.com> wrote:

> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>

A tiny bit of research would have answered this question for you:
https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems

-Henry

[-- Attachment #2: Type: text/html, Size: 837 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Larry McVoy]

On Tue, Jan 21, 2020 at 01:15:53PM -0500, Henry Bent wrote:
> On Tue, 21 Jan 2020 at 12:53, Jon Forrest <nobozo@gmail.com> wrote:
> 
> > There's been a lot of discussion about early Unix on Intel, National
> > Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> > the Z8000, and if not, why not.
> >
> 
> A tiny bit of research would have answered this question for you:
> https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems

Yeah, it ran on the 16 bit one but I looked and couldn't find if they
got Unix on the z80000 (which I suspect is what Jon meant).

Re: [TUHS] Unix on Zilog Z8000?

[Nigel Williams]

On Wed, Jan 22, 2020 at 5:22 AM Larry McVoy <lm@mcvoy.com> wrote:
> Yeah, it ran on the 16 bit one but I looked and couldn't find if they
> got Unix on the z80000 (which I suspect is what Jon meant).

I spent some time attempting to confirm whether the Z80000 (the 32-bit
flat address space development of the Z8000) ever went to silicon and
was not able to confirm it did. My suspicion remains that the Z80000
was only ever a paper specification and did not even get to a first
spin. If anyone knows and has some evidence I would be glad to update
the wikipedia page.

Re: [TUHS] Unix on Zilog Z8000?

[Henry Bent]

[-- Attachment #1: Type: text/plain, Size: 964 bytes --]

On Tue, 21 Jan 2020 at 16:49, Nigel Williams <nw@retrocomputingtasmania.com>
wrote:

> On Wed, Jan 22, 2020 at 5:22 AM Larry McVoy <lm@mcvoy.com> wrote:
> > Yeah, it ran on the 16 bit one but I looked and couldn't find if they
> > got Unix on the z80000 (which I suspect is what Jon meant).
>
> I spent some time attempting to confirm whether the Z80000 (the 32-bit
> flat address space development of the Z8000) ever went to silicon and
> was not able to confirm it did. My suspicion remains that the Z80000
> was only ever a paper specification and did not even get to a first
> spin. If anyone knows and has some evidence I would be glad to update
> the wikipedia page.
>

This thread reminded me that I own a copy of the Zilog "1983/84 Components
Data Book," which I am talking to Al Kossow about scanning.  It fully
details the Z80,000 on paper but does not say anything about hardware,
everything is labeled preliminary.  It is dated September 1983.

-Henry

[-- Attachment #2: Type: text/html, Size: 1408 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Derek Fawcus]

On Tue, Jan 21, 2020 at 05:00:50PM -0500, Henry Bent wrote:
> 
> This thread reminded me that I own a copy of the Zilog "1983/84 Components
> Data Book," which I am talking to Al Kossow about scanning.  It fully
> details the Z80,000 on paper but does not say anything about hardware,
> everything is labeled preliminary.  It is dated September 1983.

http://bitsavers.trailing-edge.com/components/zilog/_dataBooks/1983_84_Components_Data_Book.pdf

DF

Re: [TUHS] Unix on Zilog Z8000?

[Jason Stevens]

[-- Attachment #1: Type: text/plain, Size: 999 bytes --]

The only one I know is Coherent.  Disk images recently surfaced 




https://www.autometer.de/unix4fun/coherent/ftp/distrib/Coherent-0.7/ 




This is for the Commodore B 900 prototype. 




http://www.floodgap.com/retrobits/ckb/secret/900.html 




Get Outlook for Android







On Wed, Jan 22, 2020 at 2:22 AM +0800, "Larry McVoy" <lm@mcvoy.com> wrote:










On Tue, Jan 21, 2020 at 01:15:53PM -0500, Henry Bent wrote:
> On Tue, 21 Jan 2020 at 12:53, Jon Forrest  wrote:
> 
> > There's been a lot of discussion about early Unix on Intel, National
> > Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> > the Z8000, and if not, why not.
> >
> 
> A tiny bit of research would have answered this question for you:
> https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems

Yeah, it ran on the 16 bit one but I looked and couldn't find if they
got Unix on the z80000 (which I suspect is what Jon meant).






[-- Attachment #2: Type: text/html, Size: 2284 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Cornelius Keck]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: TEXT/PLAIN; charset=X-UNKNOWN, Size: 1869 bytes --]

That would be Cohere 0.7.3. Snappy system, loved it. Commodore Germany in
Braunschweig got told to ditch the project after successfully presenting
it at CeBit in Hannover. Many of these systems made it into hobbyists'
hands. It's pretty much a version 7 lookalike, but internally different
enough to encur AT&T's wrath. No network. Loved that thing. It's in
storage in Germany. Gotta wonder if it still works. Other stuff I've
brought back here did, so....



On Tue, 21 Jan 2020, Jason Stevens wrote:

> Date: Tue, 21 Jan 2020 22:35:59 +0000 (UTC)
> From: Jason Stevens <jsteve@superglobalmegacorp.com>
> To: Henry Bent <henry.r.bent@gmail.com>, Larry McVoy <lm@mcvoy.com>
> Cc: The Unix Heritage Society mailing list <tuhs@tuhs.org>
> Subject: Re: [TUHS] Unix on Zilog Z8000?
>
> The only one I know is Coherent.  Disk images recently surfaced
>
>
>
>
> https://www.autometer.de/unix4fun/coherent/ftp/distrib/Coherent-0.7/
>
>
>
>
> This is for the Commodore B 900 prototype.
>
>
>
>
> http://www.floodgap.com/retrobits/ckb/secret/900.html
>
>
>
>
> Get Outlook for Android
>
>
>
>
>
>
>
> On Wed, Jan 22, 2020 at 2:22 AM +0800, "Larry McVoy" <lm@mcvoy.com> wrote:
>
>
>
>
>
>
>
>
>
>
> On Tue, Jan 21, 2020 at 01:15:53PM -0500, Henry Bent wrote:
> > On Tue, 21 Jan 2020 at 12:53, Jon Forrest  wrote:
> >
> > > There's been a lot of discussion about early Unix on Intel, National
> > > Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> > > the Z8000, and if not, why not.
> > >
> >
> > A tiny bit of research would have answered this question for you:
> > https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems
>
> Yeah, it ran on the 16 bit one but I looked and couldn't find if they
> got Unix on the z80000 (which I suspect is what Jon meant).
>
>
>
>
>
>

Re: [TUHS] Unix on Zilog Z8000?

[Wesley Parish]

Is it still in existence? In the form of floppies? Any source?

Thanks

Wesley Parish

On 3/16/20, Cornelius Keck <tuhs@keck.us> wrote:
> That would be Cohere 0.7.3. Snappy system, loved it. Commodore Germany in
> Braunschweig got told to ditch the project after successfully presenting
> it at CeBit in Hannover. Many of these systems made it into hobbyists'
> hands. It's pretty much a version 7 lookalike, but internally different
> enough to encur AT&T's wrath. No network. Loved that thing. It's in
> storage in Germany. Gotta wonder if it still works. Other stuff I've
> brought back here did, so....
>
>
>
> On Tue, 21 Jan 2020, Jason Stevens wrote:
>
>> Date: Tue, 21 Jan 2020 22:35:59 +0000 (UTC)
>> From: Jason Stevens <jsteve@superglobalmegacorp.com>
>> To: Henry Bent <henry.r.bent@gmail.com>, Larry McVoy <lm@mcvoy.com>
>> Cc: The Unix Heritage Society mailing list <tuhs@tuhs.org>
>> Subject: Re: [TUHS] Unix on Zilog Z8000?
>>
>> The only one I know is Coherent.  Disk images recently surfaced
>>
>>
>>
>>
>> https://www.autometer.de/unix4fun/coherent/ftp/distrib/Coherent-0.7/
>>
>>
>>
>>
>> This is for the Commodore B 900 prototype.
>>
>>
>>
>>
>> http://www.floodgap.com/retrobits/ckb/secret/900.html
>>
>>
>>
>>
>> Get Outlook for Android
>>
>>
>>
>>
>>
>>
>>
>> On Wed, Jan 22, 2020 at 2:22 AM +0800, "Larry McVoy" <lm@mcvoy.com>
>> wrote:
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On Tue, Jan 21, 2020 at 01:15:53PM -0500, Henry Bent wrote:
>> > On Tue, 21 Jan 2020 at 12:53, Jon Forrest  wrote:
>> >
>> > > There's been a lot of discussion about early Unix on Intel, National
>> > > Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
>> > > the Z8000, and if not, why not.
>> > >
>> >
>> > A tiny bit of research would have answered this question for you:
>> > https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems
>>
>> Yeah, it ran on the 16 bit one but I looked and couldn't find if they
>> got Unix on the z80000 (which I suspect is what Jon meant).
>>
>>
>>
>>
>>
>>
>

Re: [TUHS] Unix on Zilog Z8000?

[Derek Fawcus]

On Mon, Mar 16, 2020 at 01:28:26PM +1300, Wesley Parish wrote:
> Is it still in existence?

> In the form of floppies?

https://www.autometer.de/unix4fun/coherent/ftp/distrib/Coherent-0.7/

Certainly disk2 looks like it has a filesystem on it, since there are
blocks which appear to be 16 byte directory entries, presumably
inode number and 14 char name.

> Any source?

I don't know about historical sources of the time.
The availale Coherent source has a few files with z8001 in their name:

$ gzip -dc ~/Downloads/mwc-COHERENT-Source.tgz|tar tvf - |fgrep -i z800
-rw-r--r--  0 steve  staff     6521  6 Feb  1991 romana/relic/d/lib/libc/sys/mch/z8001.tar.Z
drwxr-xr-x  0 steve  staff        0 14 Jun  2003 romana/relic/d/lib/libc/stdio/z8001/
-rw-r--r--  0 steve  staff      268  9 Nov  1984 romana/relic/d/lib/libc/stdio/z8001/fgetw.c
-rw-r--r--  0 steve  staff      169  9 Nov  1984 romana/relic/d/lib/libc/stdio/z8001/fputw.c
-rw-r--r--  0 steve  staff      340  9 Nov  1984 romana/relic/d/lib/libc/stdio/z8001/getw.c
-rw-r--r--  0 steve  staff      231  9 Nov  1984 romana/relic/d/lib/libc/stdio/z8001/putw.c
-rw-r--r--  0 steve  staff     4906  6 Feb  1991 romana/relic/d/lib/libc/gen/mch/z8001.tar.Z
drwxr-xr-x  0 steve  staff        0 14 Jun  2003 romana/relic/d/bin/diff/z8001/
-rw-r--r--  0 steve  staff     8445  2 Nov  1984 romana/relic/d/bin/diff/z8001/diff1.c
-rw-r--r--  0 steve  staff     7767 23 Mar  1990 romana/relic/d/bin/diff/z8001/diff2.c
-rw-r--r--  0 steve  staff       92  8 Jul  1985 romana/relic/d/bin/diff/z8001/Readme
-rw-r--r--  0 steve  staff    17512  7 Mar  1991 romana/relic/d/bin/db/mch/z8001.tar.Z
-rw-r--r--  0 steve  staff    13929  7 Mar  1991 romana/relic/d/bin/db/mch/z8002.tar.Z
-rw-r--r--  0 steve  staff    21823  6 Feb  1991 romana/relic/d/bin/as/mch/z8001.tar.Z
-rw-r--r--  0 steve  staff    19275  6 Feb  1991 romana/relic/d/bin/as/mch/z8002.tar.Z
-rw-r--r--  0 steve  staff     6521  6 Feb  1991 romana/relic/b/lib/libc/sys/mch/z8001.tar.Z
drwxr-xr-x  0 steve  staff        0 14 Jun  2003 romana/relic/b/lib/libc/stdio/z8001/
-rw-r--r--  0 steve  staff      268  9 Nov  1984 romana/relic/b/lib/libc/stdio/z8001/fgetw.c
-rw-r--r--  0 steve  staff      169  9 Nov  1984 romana/relic/b/lib/libc/stdio/z8001/fputw.c
-rw-r--r--  0 steve  staff      340  9 Nov  1984 romana/relic/b/lib/libc/stdio/z8001/getw.c
-rw-r--r--  0 steve  staff      231  9 Nov  1984 romana/relic/b/lib/libc/stdio/z8001/putw.c
-rw-r--r--  0 steve  staff     4906  6 Feb  1991 romana/relic/b/lib/libc/gen/mch/z8001.tar.Z

I looked inside the first one, and it gives this set of files, being the library for the syscalls:

$ ls z8001/
_brk.s		chroot.s	getegid.s	link.s		pause.s		signal.s	umount.s
access.s	close.s		geteuid.s	lock.s		pipe.s		sload.s		unique.s
acct.s		creat.s		getgid.s	lseek.s		profil.s	stat.s		unlink.s
alarm.s		dup.s		getpid.s	mknod.s		ptrace.s	stime.s		utime.s
bpt.s		execve.s	getuid.s	mount.s		read.s		suload.s	wait.s
chdir.s		fork.s		halt.s		mpx.s		sbrk.c		sync.s		write.s
chmod.s		fstat.s		ioctl.s		nice.s		setgid.s	times.s
chown.s		ftime.s		kill.s		open.s		setuid.s	umask.s

I've not explored any further, but if the kernel source isn't there, I rather suspect
the 286 kernel source could be beaten in to some sort of shape to work on the z8000.

There some RCS files in there if you wish to explore further.

That still leaves you with having to come up with drivers, but there do seem to be
(some) technical docs available for the machine.

DF

Re: [TUHS] Unix on Zilog Z8000?

[William Pechter]

[-- Attachment #1: Type: text/plain, Size: 968 bytes --]

IRS Pyramid buy and 3B2-Mips on Google. 

https://books.google.com/books?id=a4Jp5vySB-UC&pg=PA100&lpg=PA100&dq=Irs+to+buy+pyramid+at%26t&source=bl&ots=G_oVO6qpOx&sig=ACfU3U3EckPvlhXgj1NGrwhO1lsaBp3iWQ&hl=en&sa=X&ved=2ahUKEwjF1uy2qZXnAhVxmuAKHXhCCqEQ6AEwAHoECAEQAQ#v=onepage&q=Irs%20to%20buy%20pyramid%20at%26t&f=false
Sent from pechter@gmail.com

-----Original Message-----
From: Henry Bent <henry.r.bent@gmail.com>
To: Jon Forrest <nobozo@gmail.com>
Cc: The Unix Heritage Society mailing list <tuhs@tuhs.org>
Sent: Tue, 21 Jan 2020 13:17
Subject: Re: [TUHS] Unix on Zilog Z8000?

On Tue, 21 Jan 2020 at 12:53, Jon Forrest <nobozo@gmail.com> wrote:

> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>

A tiny bit of research would have answered this question for you:
https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems

-Henry

[-- Attachment #2: Type: text/html, Size: 1958 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Jon Forrest]

On 1/21/2020 10:15 AM, Henry Bent wrote:
>     There's been a lot of discussion about early Unix on Intel, National
>     Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
>     the Z8000, and if not, why not.
> 
> 
> A tiny bit of research would have answered this question for you: 
> https://en.wikipedia.org/wiki/Zilog_Z8000#Z8000_CPU_based_systems

True, but sometimes oldtimers like those of us who hang out here
remember additional, or different, things than what Wikipedia pages
say.

Jon

Re: [TUHS] Unix on Zilog Z8000?

[Clem Cole]

[-- Attachment #1: Type: text/plain, Size: 1625 bytes --]

The Onyx box redated all the 68K and later Intel or other systems.   John
Bass brought one to USENIX to demo in early 1980 ru a V7 port and everyone
was blown away. Playing with it. It was a desktop (19" rack) system that
worked like a PDP-11.   I don't remember the bus, but I would guess it was
either custom or Multibus-I.

Besides being one of the first non-PDP-11 'ports', the original lockf(2)
system call was defined for the database that they had built.  John would
release the specs to /usr/group later.  I remember at one meeting in the
early 1980s discussing if file locking needed to be in the original
specification (Heinz probably remembers also as the chair of that
meeting).  I'm not at home, so I don't have the document to see if it was
picked up.  The argument was that serious computers like VMS or the like
ran real databases and without file locking UNIX would never be considered
a real OS that people could use.

BTW: Joy would later use Bass's call as a model for the 4.2 call, but Joy
made the locks advisory, Bass's version was full / mandatory locks.

FWIW: I think a search will pick up a number of other Z8000 based systems,
but Onyx was the first UNIX box.

On Tue, Jan 21, 2020 at 12:53 PM Jon Forrest <nobozo@gmail.com> wrote:

> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>
> As I remember the Z8000 was going to be the great white hope that
> would continue Zilog's success with the Z80 into modern times.
> But, it obviously didn't happen.
>
> Why?
>
> Jon
>

[-- Attachment #2: Type: text/html, Size: 2560 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Heinz Lycklama]

[-- Attachment #1: Type: text/plain, Size: 1923 bytes --]

That is correct. John Bass did contribute the lockf(2) system call
to the /usr/group standard. It is in the 1984 document.

Heinz

On 1/21/2020 10:28 AM, Clem Cole wrote:
> The Onyx box redated all the 68K and later Intel or other systems.  
>  John Bass brought one to USENIX to demo in early 1980 ru a V7 port 
> and everyone was blown away. Playing with it. It was a desktop (19" 
> rack) system that worked like a PDP-11.   I don't remember the bus, 
> but I would guess it was either custom or Multibus-I.
>
> Besides being one of the first non-PDP-11 'ports', the original 
> lockf(2) system call was defined for the database that they had built. 
> John would release the specs to /usr/group later.  I remember at one 
> meeting in the early 1980s discussing if file locking needed to be in 
> the original specification (Heinz probably remembers also as the chair 
> of that meeting).  I'm not at home, so I don't have the document to 
> see if it was picked up.  The argument was that serious computers like 
> VMS or the like ran real databases and without file locking UNIX would 
> never be considered a real OS that people could use.
>
> BTW: Joy would later use Bass's call as a model for the 4.2 call, but 
> Joy made the locks advisory, Bass's version was full / mandatory locks.
>
> FWIW: I think a search will pick up a number of other Z8000 based 
> systems, but Onyx was the first UNIX box.
>
> On Tue, Jan 21, 2020 at 12:53 PM Jon Forrest <nobozo@gmail.com 
> <mailto:nobozo@gmail.com>> wrote:
>
>     There's been a lot of discussion about early Unix on Intel, National
>     Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
>     the Z8000, and if not, why not.
>
>     As I remember the Z8000 was going to be the great white hope that
>     would continue Zilog's success with the Z80 into modern times.
>     But, it obviously didn't happen.
>
>     Why?
>
>     Jon
>


[-- Attachment #2: Type: text/html, Size: 3715 bytes --]

[TUHS] Onyx (was Re: Unix on Zilog Z8000?)

[Derek Fawcus]

On Tue, Jan 21, 2020 at 01:28:14PM -0500, Clem Cole wrote:
> The Onyx box redated all the 68K and later Intel or other systems.

That was a fun bit of grubbing around courtesy of a bitsavers mirror
(https://www.mirrorservice.org/sites/www.bitsavers.org/pdf/onyx/).

It seems they started with a board based upon the non-segmented Z8002
and only later switched to using the segmented Z8001.  In the initial
board, they created their own MMU:

  Page 6 of: https://www.mirrorservice.org/sites/www.bitsavers.org/pdf/onyx/c8002/Onyx_C8002_Brochure.pdf

  Memory Management Controller:

  The Memory Management Controller (MMC) enables the C8002 to perform
  address translation, memory block protection, and separation of
  instruction and data spaces. Sixteen independent map sets are
  implemented, with each map set consisting of an instruction map and
  a data map. Within each map there are 32 page registers. Each page
  register relocates and validates a 2K byte page. The MMC generates
  a 20 bit address allowing the C8002 to access up to one Mbyte of
  physical memory.

So I'd guess the MMC was actually programed through I/O instuctions
to io space, and hence preserved the necessary protection domains.

Cute.  I've had a background interest in the Z8000 (triggered by reading
a Z80000 datasheet around 87/88), and always though about using
the segmented rather than unsegmented device.

The following has a bit more info about the version of System III
ported to their boxes:

  https://www.mirrorservice.org/sites/www.bitsavers.org/pdf/onyx/c8002/UNIX_3.0.3_Software_Release_Notice_May83.pdf

DF

Re: [TUHS] Unix on Zilog Z8000?

[Arrigo Triulzi]

[-- Attachment #1: Type: text/plain, Size: 680 bytes --]

On 21 Jan 2020, at 19:30, Clem Cole <clemc@ccc.com> wrote:
> FWIW: I think a search will pick up a number of other Z8000 based systems, but Onyx was the first UNIX box.

The very first “Unix at home” system for me was an Onyx 8002 “pedestal” which was connected to three serial terminals: dad, uncle, me :) It was definitely 1980 as it was delivered at home the day of my (unforgettable) 7th birthday. Previously I had only “played” with dad’s TTY into the University systems. Ended up being taught C instead of riding a bicycle… 

We still have the manuals somewhere but the machine was gone when the Avioon came in. No idea where it ended, sadly.

Arrigo 

[-- Attachment #2: Type: text/html, Size: 1057 bytes --]

Re: [TUHS] Unix on Zilog Z8000?

[Mary Ann Horton]

Absolutely. When I was an impoverished grad student at Berkeley, Zilog 
hired me as a consultant to port vi and the other Berkeley tools to 
their Z8000 UNIX system. It was a treasured paying gig.

As I recall, it was a 16 bit system (with some addressing enhancements 
ala the 11/70). By then, the VAX was popular and everybody wanted 32 bit 
systems. People were pinning their micro-UNIX hopes on the Motorola 68K.

Even before Zilog's ZEUS, Onyx came out with a microwave oven-sized box 
based on the Z8000. They loaned one to Berkeley, and it was my first 
home computer when I took it home to port the tools. Everything had to 
be copied over by serial port.

     Mary Ann

On 1/21/20 9:52 AM, Jon Forrest wrote:
> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>
> As I remember the Z8000 was going to be the great white hope that
> would continue Zilog's success with the Z80 into modern times.
> But, it obviously didn't happen.
>
> Why?
>
> Jon

[TUHS] Life at UC Berkeley (was Unix on Zilog Z8000?)

[Jon Forrest]

On 1/22/2020 9:00 AM, Mary Ann Horton wrote:

> Even before Zilog's ZEUS, Onyx came out with a microwave oven-sized box 
> based on the Z8000. They loaned one to Berkeley, and it was my first 
> home computer when I took it home to port the tools. 

As someone who played a very minor role at UC Berkeley, I thought
I'd amplify this a little. It was amazing how the various vendors
would happily lend (or give) equipment to UC Berkeley. I don't
know how things were at the other top CS schools, but at Berkeley
I found that if I needed something, telling my faculty boss/mentor
would often make the item appear.

For example, I did an early port of Postgres to Windows NT using
a DEC PC with 16MB of RAM. With such a small amount of RAM this
was excruciatingly slow but I made enough progress to show Mike
Stonebraker. He then talked to his friends in industry and all of
a sudden a MIPS(!) box with 64MB of RAM showed up. I was then able
to get the port to the point where it could run the Wisconsin benchmark.

Something similar happened when the group wanted to port an early
version of Postgres95 to Solaris (it had been developed on DEC Alphas 
running OSF/1). Mike said to write up a proposal to Sun that he'd sign
and then send in, which I did. Soon after a couple of Sparcstation 10s
(or 20, I don't remember) showed up and the port was done.

There other systems research groups at UCB in the 90s also were
the happy recipients of all kinds of great hardware. The limiting factor
for such donations was usually finding space to house them. The end
result was that the RAID, Risc, and Postgres groups weren't slowed down
by lack of hardware.

I have no idea what it's like today. Eric Allman, would you care to
comment?

Jon Forrest

Re: [TUHS] Unix on Zilog Z8000?

[Erik Fair]

… and after Mary Ann was done with it (sometime 1981?), Professor Fabry gave the UCB Onyx to a bunch of undergraduates which they used as the first computer of the Undergraduate Computing Facility (UCF, predecessor to the UCB XCF which had a pile of Suns to play with) in room B50 (basement) of Evans Hall. I was one of those undergrads.

That machine became the “x” host on the UCB BerkNet, and it was where B version netnews software was written by Matt Glickman (he was a high school student at the time and while he had Mary Ann’s patronage, my memory is that none of the more official computer services in CS or the Computer Center would give him an account - but we did). At the time, every other computer had restrictions on who could use ‘em, or that you had to pay hefty fees for use (the computer center PDP-11/70s were in that class) - our policy was: if you have a current UCB student ID, we’ll give you an account. We didn’t care whether you were a CS or Engineering student, or not.

I remember fixing bugs in various bits of BSD stuff we added to the userland and kernel - we had full source to play with, which was nice. Adding job control was a top priority, and one of our hacks was to change the tty line discipline to restore previous “cooked” character processing state if the program in foreground which had last changed it exited non-zero - that way, programs that modified tty state didn’t need to be recoded for job control too, which, in a system with PDP-11-like memory restrictions (64k text, 64k data maximum per process), was useful.

We trained an awful lot of students in the ways of Unix, and many of them became system administrators of the explosion of Unix systems which came to UCB later: the workstation clusters, other microcomputer based Unixes, etc. Kernel hackers (systems programmers) too.

Mary Ann also allowed me to photocopy a samizdat copy of the Lions Book, too. I still have that … somewhere, though I bought a copy of the Peter Salus 1996 republication, too.

	Erik Fair

> On Jan 22, 2020, at 09:00, Mary Ann Horton <mah@mhorton.net> wrote:
> 
> Absolutely. When I was an impoverished grad student at Berkeley, Zilog hired me as a consultant to port vi and the other Berkeley tools to their Z8000 UNIX system. It was a treasured paying gig.
> 
> As I recall, it was a 16 bit system (with some addressing enhancements ala the 11/70). By then, the VAX was popular and everybody wanted 32 bit systems. People were pinning their micro-UNIX hopes on the Motorola 68K.
> 
> Even before Zilog's ZEUS, Onyx came out with a microwave oven-sized box based on the Z8000. They loaned one to Berkeley, and it was my first home computer when I took it home to port the tools. Everything had to be copied over by serial port.
> 
>     Mary Ann
> 
> On 1/21/20 9:52 AM, Jon Forrest wrote:
>> There's been a lot of discussion about early Unix on Intel, National
>> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
>> the Z8000, and if not, why not.
>> 
>> As I remember the Z8000 was going to be the great white hope that
>> would continue Zilog's success with the Z80 into modern times.
>> But, it obviously didn't happen.
>> 
>> Why?
>> 
>> Jon

Re: [TUHS] Unix on Zilog Z8000?

[Andreas Hein]

As far as i know, there was a U8001 (Z8001 CPU Clone) in the System 
P8000 which run a System III based Unix System named "Wega"
It was designed and produced by Company EAW from East-Berlin GDR (German 
Democratic Republic)

Details with Docs, Schematics and Sourcecode at: http://www.pofo.de/P8000/

KR,

Andreas

Am 21.01.20 um 18:52 schrieb Jon Forrest:
> There's been a lot of discussion about early Unix on Intel, National
> Semi, Motorola, and Sparc processors. I don't recall if Unix ran on
> the Z8000, and if not, why not.
>
> As I remember the Z8000 was going to be the great white hope that
> would continue Zilog's success with the Z80 into modern times.
> But, it obviously didn't happen.
>
> Why?
>
> Jon

Re: [TUHS] Unix on Zilog Z8000?

[Oliver Lehmann]

Hi,

let me just reply to this single message :)

> Details with Docs, Schematics and Sourcecode at: http://www.pofo.de/P8000/

Wow - my page gets quoted on TUHS-ML :)

The 16 Bit board runs WEGA (renamed copy of ZEUS). The board is a copy  
of the 16 Bit Board from Zilogs System 8000. Some parts where added  
because the subsystem design is different (serial ports, harddisk,  
floppy disks instead of tape,...).
WEGA only runs on a Z8001 (no Z8002) and the OS was able to execute  
(and compile) segmented (full memory access) and unsegmented (only 64K  
access iirc) binaries.

The OS of the P8000 is as I said a copy of ZEUS. The name ZEUS was  
"replaced" with WEGA in the executables (hex editor).
Some parts of the system where reverse engineered and altered. I had  
contact to a kernel developer after the system collapsed who told me  
some stories ;)
While the userland ist nearly a 100% copy of ZEUS, main parts of the  
kernel (dev) where changed to handle the different system layout.

I own a complete and running P8000 as well as a S8000 which could run  
if I would have
a) a copy of the Zilog ZEUS Installation Tape (+ system diagnostic tape)
b) a working tape drive (non-standard tape not compatible to something else)
c) a SA1100 harddisk, or c) a SMD Controller + SMD disks.
d) time

After I designed, built and programmed a harddisk subsystem emulator  
for the P8000, I also planed to emulate the whole disk subsystem of  
the S8000. Additionally I planned to use a partial backup of a ZEUS  
installation I have + WEGA to generate a running ZEUS-a-like but then  
I became a father and time was gone ;)

So right now my S8000 sits in my basement togehter with some circuit  
ideas and logical analysation results and waits for more time ;)

Some links...

Kernel Sources of WEGA I could get my hands on, or are disassembled  
and "make-it-C-code-again"ed by me
https://github.com/OlliL/P8000/tree/master/WEGA/src/uts
Beside that there is also other stuff like C compiler sources and so  
on - just navigate around :)

Zilog S8000 pics
http://pics.pofo.de/index.php?/category/129

EAW P8000 pics
http://pics.pofo.de/index.php?/category/21

Some S8000 sources (mostly firmware, rebuilt with having access to the  
P8000 firmware)
https://github.com/OlliL/S8000

Any questions are welcome....

Best regards,
Oliver

Re: [TUHS] Dhrystone Benchmark (was: Unix on Zilog Z8000?)

[Oliver Lehmann]

By the way.... I found some dhry-c output I ran on WEGA

#83 cc -O -DHZ=60  -DREG=register -c dhry_1.c
#84 cc -O -DREG=register -c dhry_2.c
#85 cc -o dhry dhry_1.o dhry_2.o
#86 echo 30000 | ./dhry | awk '/Lang/ || /^Micro/ {print} \
/per Sec/ {mips=$4/1757;print;printf("Dhrystone MIPS:%38.04f\n",mips)}'
Dhrystone Benchmark, Version 2.1 (Language: C)
Microseconds for one run through Dhrystone: 1652.8
Dhrystones per Second:                       605.0
Dhrystone MIPS:                                0.3443
#87 echo 30000 | ./dhry | awk '/Lang/ || /^Micro/ {print} \
/per Sec/ {mips=$4/1757;print;printf("Dhrystone MIPS:%38.04f\n",mips)}'
Dhrystone Benchmark, Version 2.1 (Language: C)
Microseconds for one run through Dhrystone: 1652.2
Dhrystones per Second:                       605.2
Dhrystone MIPS:                                0.3445
Oliver Lehmann <lehmann@ans-netz.de> wrote:

> Hi,
>
> let me just reply to this single message :)
>
>> Details with Docs, Schematics and Sourcecode at: http://www.pofo.de/P8000/
>
> Wow - my page gets quoted on TUHS-ML :)
>
> The 16 Bit board runs WEGA (renamed copy of ZEUS). The board is a  
> copy of the 16 Bit Board from Zilogs System 8000. Some parts where  
> added because the subsystem design is different (serial ports,  
> harddisk, floppy disks instead of tape,...).
> WEGA only runs on a Z8001 (no Z8002) and the OS was able to execute  
> (and compile) segmented (full memory access) and unsegmented (only  
> 64K access iirc) binaries.
>
> The OS of the P8000 is as I said a copy of ZEUS. The name ZEUS was  
> "replaced" with WEGA in the executables (hex editor).
> Some parts of the system where reverse engineered and altered. I had  
> contact to a kernel developer after the system collapsed who told me  
> some stories ;)
> While the userland ist nearly a 100% copy of ZEUS, main parts of the  
> kernel (dev) where changed to handle the different system layout.
>
> I own a complete and running P8000 as well as a S8000 which could  
> run if I would have
> a) a copy of the Zilog ZEUS Installation Tape (+ system diagnostic tape)
> b) a working tape drive (non-standard tape not compatible to something else)
> c) a SA1100 harddisk, or c) a SMD Controller + SMD disks.
> d) time
>
> After I designed, built and programmed a harddisk subsystem emulator  
> for the P8000, I also planed to emulate the whole disk subsystem of  
> the S8000. Additionally I planned to use a partial backup of a ZEUS  
> installation I have + WEGA to generate a running ZEUS-a-like but  
> then I became a father and time was gone ;)
>
> So right now my S8000 sits in my basement togehter with some circuit  
> ideas and logical analysation results and waits for more time ;)
>
> Some links...
>
> Kernel Sources of WEGA I could get my hands on, or are disassembled  
> and "make-it-C-code-again"ed by me
> https://github.com/OlliL/P8000/tree/master/WEGA/src/uts
> Beside that there is also other stuff like C compiler sources and so  
> on - just navigate around :)
>
> Zilog S8000 pics
> http://pics.pofo.de/index.php?/category/129
>
> EAW P8000 pics
> http://pics.pofo.de/index.php?/category/21
>
> Some S8000 sources (mostly firmware, rebuilt with having access to  
> the P8000 firmware)
> https://github.com/OlliL/S8000
>
> Any questions are welcome....
>
> Best regards,
> Oliver

Re: [TUHS] Dhrystone Benchmark (was: Unix on Zilog Z8000?)

[Oliver Lehmann]

Sorry... I hit send to early by mistake... I wanted to include a  
comparison chart...

CPU   MIPS   MIPS
     System                      OS          CPU     (MHz)  V1.1   V2.1  REF
### ---------------------- ------------ ----------- ----- ------ ------ ---
...
316 IBM PC/AT              PCDOS 3.0    80286         6.0    0.39 -----   0
317 ATT PC6300             MSDOS 2.11   8086          8.0    0.39 -----   0
318 Onyx C8002             IS/1 1.1(V7) Z8000         4.0    0.29 -----   0
319 PDP 11/34              UNIX V7M     ----------- -----    0.25 -----   0
...

So WEGA@P8000 is between 317 and 318 - and 318 is the the Onyx  
mentioned as well.... but what is "IS/1 1.1"?

Best Regards,
Oliver

Oliver Lehmann <lehmann@ans-netz.de> wrote:

> By the way.... I found some dhry-c output I ran on WEGA
>
> #83 cc -O -DHZ=60  -DREG=register -c dhry_1.c
> #84 cc -O -DREG=register -c dhry_2.c
> #85 cc -o dhry dhry_1.o dhry_2.o
> #86 echo 30000 | ./dhry | awk '/Lang/ || /^Micro/ {print} \
> /per Sec/ {mips=$4/1757;print;printf("Dhrystone MIPS:%38.04f\n",mips)}'
> Dhrystone Benchmark, Version 2.1 (Language: C)
> Microseconds for one run through Dhrystone: 1652.8
> Dhrystones per Second:                       605.0
> Dhrystone MIPS:                                0.3443
> #87 echo 30000 | ./dhry | awk '/Lang/ || /^Micro/ {print} \
> /per Sec/ {mips=$4/1757;print;printf("Dhrystone MIPS:%38.04f\n",mips)}'
> Dhrystone Benchmark, Version 2.1 (Language: C)
> Microseconds for one run through Dhrystone: 1652.2
> Dhrystones per Second:                       605.2
> Dhrystone MIPS:                                0.3445

Re: [TUHS] Dhrystone Benchmark (was: Unix on Zilog Z8000?)

[Warner Losh]

[-- Attachment #1: Type: text/plain, Size: 2148 bytes --]

On Fri, Jan 24, 2020, 6:50 PM Oliver Lehmann <lehmann@ans-netz.de> wrote:

> Sorry... I hit send to early by mistake... I wanted to include a
> comparison chart...
>
> CPU   MIPS   MIPS
>      System                      OS          CPU     (MHz)  V1.1   V2.1
> REF
> ### ---------------------- ------------ ----------- ----- ------ ------ ---
> ...
> 316 IBM PC/AT              PCDOS 3.0    80286         6.0    0.39 -----   0
> 317 ATT PC6300             MSDOS 2.11   8086          8.0    0.39 -----   0
> 318 Onyx C8002             IS/1 1.1(V7) Z8000         4.0    0.29 -----   0
> 319 PDP 11/34              UNIX V7M     ----------- -----    0.25 -----   0
> ...
>
> So WEGA@P8000 is between 317 and 318 - and 318 is the the Onyx
> mentioned as well.... but what is "IS/1 1.1"?
>

This would be the Interactive Systems V6 port, I think. At least that is
what wikipedia says:

"ISC's 1977 offering, IS/1, was a Version 6 Unix
<https://en.m.wikipedia.org/wiki/Version_6_Unix> variant enhanced for
office automation running on the PDP-11
<https://en.m.wikipedia.org/wiki/PDP-11>."

But there is a V7 after it, so maybe it's a Version 7 port instead...

Warner

Best Regards,
> Oliver
>
> Oliver Lehmann <lehmann@ans-netz.de> wrote:
>
> > By the way.... I found some dhry-c output I ran on WEGA
> >
> > #83 cc -O -DHZ=60  -DREG=register -c dhry_1.c
> > #84 cc -O -DREG=register -c dhry_2.c
> > #85 cc -o dhry dhry_1.o dhry_2.o
> > #86 echo 30000 | ./dhry | awk '/Lang/ || /^Micro/ {print} \
> > /per Sec/ {mips=$4/1757;print;printf("Dhrystone MIPS:%38.04f\n",mips)}'
> > Dhrystone Benchmark, Version 2.1 (Language: C)
> > Microseconds for one run through Dhrystone: 1652.8
> > Dhrystones per Second:                       605.0
> > Dhrystone MIPS:                                0.3443
> > #87 echo 30000 | ./dhry | awk '/Lang/ || /^Micro/ {print} \
> > /per Sec/ {mips=$4/1757;print;printf("Dhrystone MIPS:%38.04f\n",mips)}'
> > Dhrystone Benchmark, Version 2.1 (Language: C)
> > Microseconds for one run through Dhrystone: 1652.2
> > Dhrystones per Second:                       605.2
> > Dhrystone MIPS:                                0.3445
>

[-- Attachment #2: Type: text/html, Size: 4889 bytes --]