Hard Disk Compatibility-Mode Woes
E Smith
95. It is connected to the primary IDE controller of an Opti-chipset based
motherboard. The CD-ROM, which is attached to the secondary controller,
works fine.
1) In the Device Manager panel of the system control panel, the primary
controller has a yellow exclaimation point. Under properties it says "This
device is not present, not working properly or does not have all the
drivers installed. No conflicts." The secondary controller used to have a
yellow exclaimation point until I applied a service diskette (OPTCDFIX)
from IBM (the vendor). Thinking that I could manually apply IBM's voodoo
to the primary controller as well, I examined the service disk with an
editor. As far as I can tell, it deletes two registry keys
(hkey_local_machine\enum\mf and
hkey_local_machine\enum\pci\ven_1045&dev_c621), replaces a third
(hkey_local_machine\system\currentcontrolset\services\vxd\ios) and
overwrites MSHDF.INF with a modified version. There is nothing on the disk
to suggest that it is controller specific. Furthermore, "NOIDE" does not
appear anywhere in my registry. Trying to uncheck "use automatic setting"
for the parent device has no effect.
2) I tried flashing the BIOS with the latest version. No change. When I
temporarily disabled LBA translation, my system would not boot at all.
When I temporarily disabled Read Prefetch, the computer took 12 minutes to
boot and still ran in compatibility mode. The BIOS cannot disable the
controller.
3) I applied Win95 Service Pack 1 and most of the recommended Microsoft
patches to my original version (preloaded) of Windows 95. I even applied
the Y2K patch, however I did not apply the MS Exchange or Imaging patch.
No change.
4) Search Microsoft's Knowledge Base, I found article Q130179 -
Troubleshooting MS-DOS Compatibility Mode on Hard Disks. Section 3c
suggests that the default protected mode driver is ESDI_506.PDR. A search
of the registry revealed that this is the driver assigned to both the
primary and secondary controller. Furthermore, it states that a file
called IOS.LOG is created whenever any device uses compatibility mode.
Sure enough, I found ios.log in my Windows directory. Among other things
it states "Unit number 02 going through real mode driver." Duh. After
reading Knowledge Base article Q138899 - Description of the ios.log file,
I decided to temporarily remove my CONFIG.SYS and AUTOEXEC.BAT files
simulateously. The hard drive continued to run in compatibility mode,
although my ios.log file became much smaller.
5) Returning to the original KB article, I tried a logged boot (F8). The
bootlog revealed load failures for ndis2sup.vxd and vshare, an init
failure for the first instance of esdi_506.pdf and an init success for a
second instance of esdi_506.pdr. To eliminate the possibility of a
corrupted file (section 3d of the KB article), I extracted esdi_506.pdr
and ndis2sup.vxd from the original CAB files. No change. Obviously, the
Windows default IDE driver is not loading on the primary controller, but
why? KB article Q151911 - "MS-DOS Compatibility Mode Problems with PCI-IDE
Controllers" gives one theory, but no solutions.
Can anyone help?
Thanks.
Dog
E Smith <ern...@worldnet.att.net> wrote in message
news:ernsts-2508...@107.white-plains-08-09rs16rt.ny.dial-access.att
.net...
sco...@my-deja.com
Spent more than a few days trying to patch this type of problem on an
AST system which I added a CDROM to. A royal pain. It seems that when
installing Windoze 95, it didn't give the option for 32 bit file access
mode when formating...then all the yellow flags after loading. It
cannot support an IDE controller with HD at 32 bit and CDROM at 16 bit.
This was with a store bought version of W95. The system could not
recognize all the hardware. A buddy gave me a different version of W95,
it asked the question about formating for 32 bit, installed and ran.
THAT SUCKS BIG TIME! too bad we can't sue M$FT for such stupidity on
their part.
Before getting Windoze 98, try getting a different WIN95 CD.
Stew Corman
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
Larry S. & Karen L. Moore
Large IDE Hard Disk May Run In Compatibility Mode: http://support.microsoft.com/support/kb/articles/Q175/6/29.ASP
Q154435 - Windows 95 IDE Support for SMART, DMA, and ATAPI Tape Drives: http://support.microsoft.com/support/kb/articles/q154/4/35.asp
Hope this helps, Good day!
Larry
<sco...@my-deja.com> wrote in message news:7q3jfs$prd$1...@nnrp1.deja.com...
: Been there, done that.
E Smith
Unfortunately, I do not have that option at this time. Upgrading to Win98
will open a whole can of worms, because I need to retain the software on
the machine (i.e. no clean install) For what it's worth, I do have Win 98
Second Edition on order, and I have begun reading the Win 98 Resource Kit.
> A buddy gave me a different version of W95, it asked the question
> about formating for 32 bit, installed and ran.
I believe that Mr. Corman refers to the Win 95 OSR 2.x release of Windows
95 which is only available to OEMs. This release allowed OEMs to configure
pre-loaded computers with FAT32 support for large (>2 GB) hard drives.
There is a FAQ at www.compuclinic.com that describes the OSR releases in
more detail. It is my understanding that FAT32 is not necessary for my 1
GB hard disk. In fact, other computers with the same version of Windows
(4.1.950) work fine.
> Here's some MSKB's about this...[snip]
I appreciate the effort, but I mentioned these KB articles in my original
post at the beginning of this thread. The MS KB articles were helpful for
diagnosis, but are sorely lacking in solutions.
I guess what I'm looking for is someone who was stuck in compatibility
mode with similar hardware (Opti Viper-M 82C558M/82C557M) who remembers
how they solved the problem.
- E
Lindsey Ken McMillen
wrote:
>Problem: My hard disk operates in MS-DOS compatibility mode under Windows
>95. It is connected to the primary IDE controller of an Opti-chipset based
>motherboard. The CD-ROM, which is attached to the secondary controller,
>works fine.
>
After ensuring that you have already installed the most recent BIOS
for you system from IBM, retrieve a copy of Microsoft's Knowledge Base
article Q189518, and follow the instructions it contains. I've used
this trick with IBM Model 2144s and 2176s, and several ACER systems,
running Windows 95, Windows 98 and Windows 98 Second Edition with
great success. Not only has it cured the problem that you described,
but each and every system I've applied it to runs significantly faster
afterwards. It can't hurt.
Lindsey "Ken" McMillen
lkmcmillen
earthlink
net
Jim Sweeney
working beyond his intellectual limit.
sco...@my-deja.com wrote:
> Been there, done that.
> Spent more than a few days trying to patch this type of problem on an
> AST system which I added a CDROM to. A royal pain. It seems that when
> installing Windoze 95, it didn't give the option for 32 bit file access
> mode when formating...then all the yellow flags after loading. It
> cannot support an IDE controller with HD at 32 bit and CDROM at 16 bit.
> This was with a store bought version of W95. The system could not
> recognize all the hardware. A buddy gave me a different version of W95,
> it asked the question about formating for 32 bit, installed and ran.
E Smith
(Lindsey "Ken" McMillen) wrote:
> On Wed, 25 Aug 1999 23:55:24 -0500, ern...@worldnet.att.net (E Smith)
> wrote:
> After ensuring that you have already installed the most recent BIOS
> for you system from IBM, retrieve a copy of Microsoft's Knowledge Base
> article Q189518, and follow the instructions it contains. I've used
Having nothing to lose, I went ahead and tried the technique listed in
this article. On my system, IDEATAPI.INF did not exist. When I rebooted,
Windows told me that it had detected "new" hardware: a PCI IDE controller.
Then, it rebuilt its driver database, identified the hardware as an "Opti
Dual IDE Controller" and prompted me to reboot. When I entered the Device
Manager I saw that nothing had changed. The primary controller was still
running in compatibility mode.
With the belief that Q189518 was half of a solution, I went to the OPTI
ftp site. (I had previously been to their website) In
/pub/chipset/download I saw ESDI_506.PDR and MSHDC.INF without any
documentation. On a hunch, I placed these files in the appropriate Windows
directories and tried Q189518 again. No change.
When I examined the properties tab of the ESDI_506.PDR from the Opti site,
I realized that it was older than the one I had obtained in the
REMIDEUP.EXE update from Microsoft (KB Q154435). I restored the newer
ESDI_506.PDR and tried Q189518 again. No change.
Mr. McMillen, this is an IBM 2144 (Aptiva M series). Does it matter if I
used IBM's OPTCDFIX diskette earlier? Is there anything else you can
suggest?