Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!uunet!crdgw1!crdos1!davidsen
From: davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Unix on PC and AMI bios
Message-ID: <1663@crdos1.crd.ge.COM>
Date: 20 Nov 89 16:57:50 GMT
References: <776@prles2.UUCP> <1429@starfish.Convergent.COM> <694@disk.UUCP> <5641@holston.UUCP>
Reply-To: davidsen@crdos1.UUCP (bill davidsen)
Organization: GE Corp R&D Center, Schenectady NY
Lines: 24

In article <5641@holston.UUCP> barton@holston.UUCP (Barton A. Fisk) writes:

| I have heard of problems in multitasking environments relating 
| to the cache on the motherboard, but I think this can be defeated
| on most systems.
| 
| Perhaps someone with a little first hand experience would care
| to elaborate more. My info is second hand.

  Since you ask... I have run four versions of UNIX of 386 machines with
cache, from AMI, Dell, and Micronics. I have never seen a problem such
as you describe. Cache runs on physical addresses, so context switching
isn't a problem.

  I have heard rumors that certain controllers which write directly to
memory on the bus may cause problems if the cache is not marked
correctly, but that's not a multitasking problem since it could happen
under DOS as well. I've never had a chance to determine if the fault is
in the design of the cache or the controller.
-- 
bill davidsen	(davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen)
"The world is filled with fools. They blindly follow their so-called
'reason' in the face of the church and common sense. Any fool can see
that the world is flat!" - anon