From: jmorris@hunan.rastek.com (Jim Morris) Subject: Re: DPMI under dosemu (was Re: DOSEMU Success list!!) Date: Sun, 21 Mar 1993 05:45:50 GMT
In article <1993Mar20.200547.2864@kf8nh.wariat.org> bsa@kf8nh.wariat.org (Brandon S. Allbery) writes:
>In article <89285@hydra.gatech.EDU> gt8134b@prism.gatech.EDU (Howlin' Bob) writes:
>>In <1993Mar14.192241.18663@kf8nh.wariat.org> bsa@kf8nh.wariat.org writes:
>>>At great risk to myself, my computer, and my hard drives :-) I got bold,
>>>installed dosemu 0.48.1, loaded up BC++ 3.0 *under* dosemu, and ran DPMIRES.
>>
>>>"DPMI server initialization error->v86 task without vcpi"
>>
>>You betcha. And, from what I understand, VCPI is too free in its granting
>>of machine control power to be implemented under Linux (or Windows, or
>>OS/2).
>
>Okay, I guess I should have been REALLY verbose so folks wouldn't try to teach
>me what I already know.
>
>DPMIRES is supposed to be able to provide DPMI using XMS. I had hoped that it
>would be possible to have it ignore 386/EMM mode and just use XMS, which would
>make it compatible with dosemu, but it can't be done.
DPMIRES *DOES* provide DPMI services using just XMS. How else do you explain
the fact that DPMIRES runs on a lowly 80286? An 80286 certainly does not
provide V86 or VCPI support. I say all this based on my experiences with
Borland/Turbo C++ 3.xx, which utilize DPMI to run on both 80286 or 80386
CPU's, using Borland's DPMIRES program. I personally can vouch that DPMIRES
loads on an 80286 with 1MB XMS/Extended memory. Or course, you probably
don't WANT ro run it on a 286.... :-)
Since DOSEMU has almost a complete XMS implementation (I understand just
one or two things, like XMS UMB's, still need work), I do not really
understand why DPMI support is so totally out of the question.
Regards,
Jim Morris
(jmorris@hunan.rastek.com)