tpc@leibniz.UUCP (Tom Chmara) (09/21/88)
Does anyone know where in the literature (USENIX proceedings, etc) I could find information on the pitfalls and difficulties in layering UNIX on a "Virtual machine"? This VM would handle the native hardware, giving UNIX a sanitized version of reality. No, it's not my idea of a performance- oriented machine, but there is interest in building something like that here (small, fast nucleus offering services to UNIX and other OSes) Thanks for your time... ---tpc--- -- I am sole owner of the above opinions. Licensing inquiries welcome. ------------------------------------------------------------------------ Tom Chmara UUCP: ..utgpu!bnr-vpa!bnr-di!leibniz!tpc BNR Ltd. BITNET: TPC@BNR.CA
dempsey@handel.colostate.edu. (Steve Dempsey) (09/23/88)
In article <164@leibniz.UUCP> tpc@leibniz.UUCP (Tom Chmara) writes: >Does anyone know where in the literature (USENIX proceedings, etc) I could >find information on the pitfalls and difficulties in layering UNIX on a >"Virtual machine"? This VM would handle the native hardware, giving UNIX ^^^^^^^ ^^^^^^^ ^^ >a sanitized version of reality. No, it's not my idea of a performance- >oriented machine, but there is interest in building something like that >here (small, fast nucleus offering services to UNIX and other OSes) ^^^^^^^ > Thanks for your time... > ---tpc--- >-- >Tom Chmara UUCP: ..utgpu!bnr-vpa!bnr-di!leibniz!tpc Hmmm... sounds like someone wants to run UN*X on IBM hardware. This is certainly an admirable goal. I'm working on something along these lines, but we all know IBM+UNIX=Vaporware :-), at least on a reasonably capable cpu. My experience in searching for such literature as Tom is seeking has been disappointing at best; perhaps I'm looking in the wrong places. I would also be interested in anything on the subject. Steve -- /\ \Steve Dempsey, Center For \steved@longs.LANCE.ColoState.Edu \/ _|/ _ _\Computer Asisted Engineering\dempsey@handel.CS.ColoState.Edu /\ | (_) | |_(_)\Colorado State University \...!ccncsu!handel!dempsey /_/_/(_/\_/ V \_ \Fort Collins, CO 80523 \(303)-491-7585
sauer@auschs.UUCP (Charlie Sauer) (09/23/88)
I'll be a little presumptuous and assume the subject line should have read
"AIX VRM". The most recent readily available paper is
L.K. Loucks and C.H. Sauer, "Advanced Interactive Executive (AIX) Operating
System Overview," IBM Systems Journal 26, 4 (1987).
Older, but still largely accurate, is
T.G. Lang, M.S. Greenberg and C.H. Sauer, "The Virtual Resource Manager,"
IBM RT Personal Computer Technology, SA23-1057 (January 1986).
I have a fair number of reprints of the first article and some copies of the
book including the second one. I'm willing to mail copies upon e-mail request.
--
Charlie Sauer IBM AES/ESD, D75/802 uucp: cs.utexas.edu!ibmaus!sauer
11400 Burnet Road csnet: ibmaus!sauer@CS.UTEXAS.EDU
Austin, Texas 78758 aesnet: sauer@auschs
(512) 823-3692 vnet: SAUER at AUSVM6tsp@ncsc1.AT&T.NCSC (Tom Poindexter x8862) (09/23/88)
In article <164@leibniz.UUCP>, tpc@leibniz.UUCP (Tom Chmara) writes: > Does anyone know where in the literature (USENIX proceedings, etc) I could > find information on the pitfalls and difficulties in layering UNIX on a > "Virtual machine"? This VM would handle the native hardware, giving UNIX If you are interested in the IBM RT implementation, there was an "IBM Systems Journal" devoted to the RT, including its virtual machine, VRM (?), and how AIX was ported on top of it. I don't remember the date of publication, but it was very near the date of the RT's announcement. Call the IBM office nearest you for info. -- Tom Poindexter {where ever}!att!ncsc5!tsp
jfh@rpp386.Dallas.TX.US (The Beach Bum) (10/01/88)
In article <371@ccncsu.ColoState.EDU> dempsey@handel.colostate.edu..UUCP (Steve Dempsey) writes: >Hmmm... sounds like someone wants to run UN*X on IBM hardware. This is >certainly an admirable goal. I'm working on something along these lines, >but we all know IBM+UNIX=Vaporware :-), at least on a reasonably capable cpu. >My experience in searching for such literature as Tom is seeking has been >disappointing at best; perhaps I'm looking in the wrong places. I would >also be interested in anything on the subject. Various UNIX ports exist which run on top of a different operating system. IBM's 370/ix port runs under VM on their 370 family of CPU's, which do include some very "reasonable capable cpu"s. There are also native mode UNIX ports [ AT&T did these some years ago ] for IBM 370's and plug compatible machines. I'd suggest starting with the obvious sources for literature, IBM and AT&T should have written something. You may want to try the ``UNIX Edition'' of the BSTJ for starters. -- John F. Haugh II (jfh@rpp386.Dallas.TX.US) HASA, "S" Division "Why waste negative entropy on comments, when you could use the same entropy to create bugs instead?" -- Steve Elias
egs@killer.DALLAS.TX.US (Eric Schnoebelen) (10/02/88)
In article <7376@rpp386.Dallas.TX.US> jfh@rpp386.Dallas.TX.US (The Beach Bum) writes: > >I'd suggest starting with the obvious sources for literature, IBM and AT&T >should have written something. You may want to try the ``UNIX Edition'' >of the BSTJ for starters. Anybody tried talking to Amdahl recently? They had a big booth at the Feburary UniForum promoting their UTS ( System V based ) port to IBM plug compatible machines, both native and guest under VM. >-- >John F. Haugh II (jfh@rpp386.Dallas.TX.US) HASA, "S" Division > > "Why waste negative entropy on comments, when you could use the same > entropy to create bugs instead?" -- Steve Elias -- Eric Schnoebelen ( just nobody in particular... )
mkhaw@teknowledge-vaxc.ARPA (Mike Khaw) (10/03/88)
I was sure someone else would mention it, but since I haven't seen it:
SRI published some papers on Eunice (Unix workalike running
on top of VMS) many years ago. You can probably get reprints
from the Wollongong Group, Palo Alto.
I believe the early incarnations of HP-UX were Unix on top of some proprietary
HP kernel OS. I don't know if HP ever published anything about it.
Mike Khaw
--
internet: mkhaw@teknowledge.arpa
uucp: {uunet|sun|ucbvax|decwrl|uw-beaver}!mkhaw%teknowledge.arpa
hardcopy: Teknowledge Inc, 1850 Embarcadero Rd, POB 10119, Palo Alto, CA 94303peter@ficc.uu.net (Peter da Silva) (10/03/88)
In article <371@ccncsu.ColoState.EDU>, dempsey@handel.colostate.edu. (Steve Dempsey) writes: > In article <164@leibniz.UUCP> tpc@leibniz.UUCP (Tom Chmara) writes: > >here (small, fast nucleus offering services to UNIX and other OSes) > Hmmm... sounds like someone wants to run UN*X on IBM hardware. Not necessarily. He could be talking about MERT or Mach or (my fantasy) UNIX under Amiga Exec (with UNIX and Amiga DOS being the O/Ses supported). -- Peter da Silva `-_-' Ferranti International Controls Corporation. "Have you hugged U your wolf today?" peter@ficc.uu.net
craig@amdahl.uts.amdahl.com (Craig Harmer) (10/04/88)
In article <5694@killer.DALLAS.TX.US> u-word!egs@killer.Dallas.TX.US (Eric Schnoebelen) writes: > > Anybody tried talking to Amdahl recently? They had a big booth >at the Feburary UniForum promoting their UTS ( System V based ) port to >IBM plug compatible machines, both native and guest under VM. > ... >-- >Eric Schnoebelen >( just nobody in particular... ) Amdahl has a System V based Unix which runs on Amdahl (and IBM-compatible) hardware. It's the "Most powerful Unix machine in the world" native, and it also runs under VM (though not as fast). Amdahl's hardware is faster than IBM's, so it beats an Unix IBM has (yet) to deliver. Sorry, i don't think we've written any papers on the subject of running under VM. There's really not alot to say, aside from paying attention to all the idiosyncracies of whatever VM you're running under, in order to get decent performance. It's particularly hard to get decent performance using virtual memory as a "Virtual Operating System". -- George Bush snorts Contra coke. craig@uts.amdahl.com {uunet,sun,decwrl}!amdahl!craig [views above shouldn't be viewed as Amdahl views, or as views from Amdahl, or as Amdahl views views, or as views by Mr. Amdahl, or as views from his house]
pugs%whitsun@Sun.COM (Tom Lyon) (10/05/88)
In article <bcHsHc5qnd1010RzzbI@amdahl.uts.amdahl.com>, craig@amdahl.uts.amdahl.com (Craig Harmer) writes: > Sorry, i don't think we've written any papers on the subject of running > under VM. There's really not alot to say, aside from paying attention to > all the idiosyncracies of whatever VM you're running under, in order to > get decent performance. It's particularly hard to get decent performance > using virtual memory as a "Virtual Operating System". Actually, I wrote a paper at Amdahl 9 years ago, when UTS was still called "Au",called "How to get the Pb out of Au". The single thing which increased performance most under VM was to make the kernel run with interrupts disabled except in "sleep", so that all the calls to spl() could be removed! The calls to spl() required traps to VM to implement them, so the ran several orders of magnitude slower in a virtual machine than on a real machine. Lest anyone think this is a terrible idea, it turns out that CP (the VM kernel) does exactly the same thing. IBM I/O devices are built for high latency. The other neat thing under VM was virtual page faults; VM would notify the kernel when a user process took a page fault so the UNIX kernel could schedule some other process; another notification would arrive after VM got the page in. This level of page faulting was independent of the paging eventually implemented by the UTS kernel.
pingel@wang7.UUCP (pingel) (10/07/88)
We already do at Wang. The same kernel is used both native and on top
of VM. Our VM provides a 'real machine' interface for the virtual machine.
Our machines are IBM 360 clones, so experiences should similiar.
--
Lee Pingel {frog,masscomp,ima}!wang7!pingel
137 Aiken Ave "Better to die on one's feet, than to live on one's knees"
Lowell, MA 01850 - Zapata ka@june.cs.washington.edu (Kenneth Almquist) (10/07/88)
dempsey@handel.colostate.edu. (Steve Dempsey) writes: > Hmmm... sounds like someone wants to run UN*X on IBM hardware. This is > certainly an admirable goal. I'm working on something along these lines, > but we all know IBM+UNIX=Vaporware :-), at least on a reasonably capable cpu. Both AT&T and Amdahl have ported UN*X to the IBM 370 architecture. The AT&T port runs on top of a kernel provided by IBM, which makes it possible to run TSO concurrently on the same machine, should you want to. (This does not mean that TSO users can communicate with UN*X users on the same machine; file formats and everything else are different.) The Amdahl port is a native implementation. I don't know anything about purchasing these versions of UN*X. It used to be that the AT&T port could be purchased for $100,000, but since the kernel provided by IBM was not for sale to anyone besides AT&T, IBM was the only possible customer. The Amdahl port might be a better bet. Kenneth Almquist