Mail Archives: opendos/2000/12/04/15:20:59
----- Original Message -----
From: "Matthias Paul" <PAUL-MA AT reze-1 DOT rz DOT rwth-aachen DOT de>
To: <opendos AT delorie DOT com>
Sent: Monday, November 27, 2000 10:50 AM
Subject: Re: Optimizing CONFIG.SYS...
> On Sun, 26 Nov 2000 Patrick Moran wrote:
>
> > True, however, if I load NWCDEX in the CONFIG, it does not load
everything
> > high, there is still the 112 bytes env and 7248 bytes of it left in low
> > memory.
>
> The 7 KB footprint indicates it is using DPMS and that it had problems
> to load into UMBs because there was not enough free UMB memory
> available in one continous chunk (I recall something in the area
> 160 Kb that is required - this is why my INSTCDEX utility exists.)
Actually I discoved red it loads in two segments the largest of which seems
to be a little under 65000 bytes. Those people with a cramped UMB after EMS
page farme is installed may have trouble getting the 65000 bytes needed to
load it into extended memory. I know that QEMM 7.52 needed 128K to load it,
but I told QEMM to only use 65000 bytes and it would load okay then.
Otherwise some TSRs/drivers might not get loaded high at all. DR DOS 386EMM
should not have any problem with this. If it does, there is some fualty
coding in the EMM386
> (The 112 bytes environment footprint will probably vanish with
> future issues - not only with NWCDEX.)
I can live with the two that use 112 bytes each. I just load them last so
their env uses some of the last small protion of the B000 range.
> > QEMM cannot find the 1A handler because somthing moves it
> > during boot. I suspect it's the SCSI portion of the MB BIOS that is
doing
> > this.
>
> Try CONFIG.SYS YEAR2000=OFF. The DR-DOS 7.02+ IBMBIO.COM hooks INT 1Ah
> to correct the RTC date.
>
> This interferes with QEMM and many Symbios/LSI based SCSI drivers
> (from Symbios, ASUS, Nomai, etc.)
Well the Symbios (actually, Tyan) drivers do not get loaded until after the
QEMM line is put in. It hangs before it even gets to those drivers. However,
the MB BIOS has detection for this card and may be the culprit. This is what
I have always suspected anyway. The card has BIOS and the MB has Symbios. I
could probably turn off the card BIOS, but I don't believe that would do any
good. The MB BIOS turns over the control to the card BIOS bit they probably
both do about the same thing anyway.Then it seems that the Tyan driver then
overwrites or makes changes to the memory area that the BIOS used. It is
just above the video BIOS at C800, where it finally winds up. I think it
used 8k up there when everything is done. I think I also checked it without
loading any SCSI drivers and that is where it resides. I cannot turn off the
MB BIOS for this. Some MB can, but I got stuck with one that does not. It is
EPROM and not FLASH so I can't even remove the code from the MB BIOS.
Pat
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
- Raw text -