Newsgroup: alt.folklore.computers

In article <13167@scolex.sco.COM> erics@sco.COM (eric smith) writes:
>rom@freja.ericsson.se (Robert Malmgren TK/DG) writes:
>>I once read an article in the Bell Laboratories Technical Journal about
>>problems/experiences with porting the UNIX system. In the article they
>>talked about a porting to a 8088 (or was it a 8086?) that was done in
>>1979. [...]  So UNIX runned on an Intel-box long before Messy-DOS
>>even was invented....
>
>1979 seems a little early to me for an 8086 port. [...] I think the Altos
>port of Xenix to the 8086, of which I was a part, was probably the first
>8086-based flavor of Unix. That was in 1981 and used 512K memory and a 10MB
>disk. The 8086 is only capable of addressing 1MB of memory. As far as I know,
>there never was a version of Unix on the 8088.

If I remember correctly the BSTJ article is:
	%T UNIX Time-Sharing System: UNIX on a Microprocessor
	%A H. Lycklama
	%J Bell System Technical Journal
	%V 57
	%N 6
	%P 2087-2101
	%D 1978

>From the journal date we can conclude that the port was done before 
1978.  The 8088 and 8086 are almost functionaly identical from the 
software side, so an 8086 version of Unix could run without any changes 
on an 8088.  The machine used for Unix port to an 8086 had a custom-built 
MMU which provided adequate memory addressing and protection.  A problem
was that stack faults could not be restarted on the 8086 so the C compiler 
was changed to do expand the stack using idempotent instructions.

Diomidis
-- 
Diomidis Spinellis                  Internet:                 dds@doc.ic.ac.uk
Department of Computing             UUCP:                    ...!ukc!icdoc!dds
Imperial College, London SW7        #include "/dev/tty"

Unless otherwise expressly stated, all original material on this page created by Diomidis Spinellis is licensed under a Creative Commons Attribution-Share Alike 3.0 Greece License.