Date: Fri, 21 Nov 1997 02:25:18 -0700 From: Dave Andersen <angio@angio.net> To: "Daniel Sobral" <Daniel_Sobral@voga.com.br> Cc: hackers@freebsd.org Subject: Re: Virtual Intel Machines? Message-ID: <199711210925.CAA01468@meowy.angio.net> In-Reply-To: Your message of "Thu, 20 Nov 1997 08:33:18 -0300." <03256555.003F5208.00@papagaio.voga.com.br>
next in thread | previous in thread | raw e-mail | index | archive | help
From: "Daniel Sobral"<Daniel_Sobral@voga.com.br>
> > Of course, given that Intel chips work the way they do, it sounds
> > like it'd be pretty much impossible to do what I was hoping for.
> > Sigh. It was such a cool idea, too.
>
> Err, maybe. That depends on _what_ you want. See the FLUX/FLUKE project,
> for instance. They run virtual machines, but not like the old vm machines.
> And work over Intel.
There are, however, some big differences between the "VM" provided
by Fluke and the the VMs provided by machines like the IBM VM/370 (and
between full virtual machine simulators). The major difference is
that traditional virtual machine simulators of either flavor
provide the operating system with what appears to it to be a direct
hardware interface. In contrast, Fluke provides a set of low-level
access routines that differ widely from the underlying hardware. The
idea behind the recursive virtual machine support in Fluke is to
provide an efficient mechanism for supporting relatively deep
nested virtual machine hierarchies, _not_ to provide an environment
where an operating system can run, unmodified, alongside another
complete and unmodified operating system. (This is what the original
poster was inquiring about).
-Dave
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199711210925.CAA01468>