Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!agate!saturn!ken@gatech.edu
From: ken@gatech.edu (Ken Seefried iii)
Newsgroups: comp.os.research
Subject: Re: Inverted page tables
Message-ID: <6538@saturn.ucsc.edu>
Date: 1 Mar 89 14:20:25 GMT
Sender: usenet@saturn.ucsc.edu
Organization: School of Information and Computer Science, Georgia Tech, Atlanta
Lines: 24
Approved: comp-os-research@jupiter.ucsc.edu


In article <6527@saturn.ucsc.edu> mcvax!cs.aber.ac.uk!pcg@uunet.UU.NET writes:
>
>Inverted page tables are extremely good for operating systems, as they are
>very much more efficient than straight ones, especially for large sparse
>address spaces. It is somewhat laborious to implement shared memory with
>them, but then of course an operating system need not rely on or support
>shared memory (or messsage passing), like MUSS.  On the other hand the
>really good use for inverted page tables is for single address space
>machines with capability (and not multiple address space) based protection.
>

The MACH papers discuss some of the difficulties that they had porting
that operating system to the PC/RT because of the awkward
implemetation of shared memory on such an MMU.  I do not recall right
now how they solved the problem.  Perhaps someone at CMU could
elaborate.

Too bad the evil empire (IBM) slapped yet another copyright on this
idea.  It would, however, be interesting to hear how BigBlue does
System V shared memory on the ROMP MMU.

	...ken seefried iii
	   ken@gatech.edu