[rescue] More IBM emulation fun

Geoffrey S. Mendelson gsm at mendelson.com
Tue Feb 10 01:12:32 CST 2009


On Mon, Feb 09, 2009 at 11:08:00PM -0700, Scott M wrote:
>On Mon, 9 Feb 2009 20:04:36 -0600, Bill Bradford wrote:
>> ... VM/370 ... documentation ... 
>http://www.bitsavers.org/pdf/ibm/370/vm370/
>
>At this URL, this doc:  Introduction_To_VM370_Course_Jul75.pdf
>has this chapter heading:  "VIRTUAL MACHINE DESCRIPTION". 
>Virtualization in 1975?  Really? 

Cut me a break. What makes you think it is something new? BTW, 
VM existed before 1975, it was originally released around 1967, and ran
on the system 360 model 67, which was a model 65 with VM hardware. 

>From what I remember at the time, VM was an implemention of an older
idea, a commercialization of something that IBM used internally, and
adapted from something even earlier.



>However with so little memory available, it seems strange that 
>anyone would want to put more than one OS on a machine of that era. 
>Page 8 of this doc shows a "256K DOS BATCH" virtual machine running 
>on a S/370 Model 145 along with five "320K CMS" virtual machines. 
>But DOS didn't exist in 1975!   -Turns out this isn't MS-DOS, but 
>IBM DOS/360 or some variant of it: 
>http://en.wikipedia.org/wiki/DOS/360
>Maybe DOS/360 and CMS were single-user and virtualization was a way 
>to make the machine multi-user?  Or maybe IBM had the same problems 
>as today where code needs to be isolated because it tries to write 
>to memory that belongs to someone else? 

You could run single user DOS systems, circa 1965 in 16k. Memory protection
hardware existed in the 1960s. Not only did IBM have it, but as far as I
remember, so did CDC, Univac, HP, DEC, and a bunch of others.

Geoff.


-- 
Geoffrey S. Mendelson, Jerusalem, Israel gsm at mendelson.com  N3OWJ/4X1GM



More information about the rescue mailing list