massar@think.ARPA (JP Massar) (10/18/85)
A few weeks ago I posted a request for information on why no one
had yet ported 4.[23] to the MicroVax II.
Here are the responses I've gotten. Note that responses 3 and 4
seem (to me at least) completely contradictory!
Thanks to everyone who responded.
-----------------------------------------------------------------------------
From: Rick Adams <rick@seismo.CSS.GOV>
I'll tell you why I'm not. I have no intentions of buying a microvaxII!
The Sun3 runs circles around the microvaxII in all benchmarks I have seen
and costs substantially less. It is delivered with 4.3bsd (OK, Sun3.0).
The only possible reason to buy a microvax is if you require the
ability to run vax binaries on it. Must people don't, thats
one of the nice things about unix, you aren't tied to a particular
hardware line (like VMS).
------------------------------------------------------------------------------
From: Jason Venner <jason@ucbopal.Berkeley.EDU>
Perhaps the biggest stumbling block is that it is nesessary to write
the instruction simulation code for the missing vax instructions.
This is likeley to happen soon as Ultrix on the uvaxen is complete
and utter trash. We have ~20 here, and not one will run ultrix
dependably. The network flakes out, or the machines crash, or there
is misterious lossage of characters from files (like /etc/rc.local
losing characters from the ifconfig line!!!!).
The other reason is that the CSRG people are too busy with the recent
4.3 and the port to the 8600 (which is also a bit flaky).
------------------------------------------------------------------------------
From: Sebastian Schmitz <mcvax!unido!ecrcvax!snoopy@seismo.CSS.GOV>
The answer you want is simple.
"Big" VAXen are Massbuss/Unibus machines.
"Little" VAXen (i.e. MV1&2) are Q-Bus machines.
This requires a different set of drivers altogether because the
I/O address spaces are different. You also have a totally
different structure: on big VAXen the interfaces are all
addressed as an offset from a MASSBUS or UNIBUS- ADAPTER.
The Q-Bus does not have this.
Hence no simple port is possible. Its an obvious port but
not as easy, as one might hope.
------------------------------------------------------------------------------
From: Craig Partridge <craig@LOKI.ARPA>
According to the guy at DEC who ported ULTRIX to the uVAX II
it is trivial -- the MicroVAX hardware is designed to look just like
bigger VAXen (it makes the Qbus look like a Unibus).
DEC is supporting 4.2 on a uVAX II in the form of Ultrix. As
for everyone else -- I suspect that Berkeley just hasn't gotten around
to bring up 4.X on the machine. I know they have some.
------------------------------------------------------------------------------
--
-- JP Massar, Thinking Machines Corporation, Cambridge, MA
-- ihnp4!godot!massar, massar@think.com.arpa
-- 617-876-1111