Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!usc!apple!altos!altos86!rcollins
From: rcollins@altos86.Altos.COM (Robert Collins)
Newsgroups: comp.sys.ibm.pc.misc
Subject: Re: extended memory
Keywords: 286 extended memory
Message-ID: <4474@altos86.Altos.COM>
Date: 7 Dec 90 16:12:27 GMT
References: <18373@rasp.eng.cam.ac.uk> <4467@altos86.Altos.COM> <1990Dec6.222421.15840@agate.berkeley.edu>
Reply-To: rcollins@altos86.UUCP (Robert Collins)
Organization: Altos Computer Systems, San Jose, CA
Lines: 21

In article <1990Dec6.222421.15840@agate.berkeley.edu> raymond@math.berkeley.edu (Raymond Chen) writes:
>In article <4467@altos86.Altos.COM>, rcollins@altos86 (Robert Collins) writes:
>> [on using the memory from FFFF:0010 to FFFF:FFFF, known as the HMA]
>>Leave A20 always enabled.  
>
>This is dangerous. 

As a matter of fact, many modern BIOS implementations do exactly that...
leave A20 always on.  They do this for benchmarking purposes.

Granted some old software was so poorly designed that they use memory
wrapping, but we BIOS engineers shouldn't be constrained by such poor
designs.  Sure, if we are interested in compatibility, we should do what
we have always done, but the fact remains, 3 of the top 5 BIOS's on the
market now always leave A20 on.

-- 
"Worship the Lord your God, and serve him only."  Mat. 4:10
Robert Collins                 UUCP:  ...!sun!altos86!rcollins
HOME:  (408) 225-8002
WORK:  (408) 432-6200 x4356