Primary IDE controller <!>d out.

Hi,

Any advice gratefully received as this is an old ongoing thorn in my side.

System is fairly old, P2 with 8.4Gb and 3.2Gb ATA33 HDDs 440LX chipset mobo running Win98SE, but performs well for my 12yr old son alongside my own newer P3 machine. It has the 2 HDDs on the primary set to master/slave correctly and a CDROM alone on the secondary. It constantly shows a compromised primary IDE controller in Device Manager and consequently has no HDDs showing up. After a complete reformat/reinstall it showed correct, but within an hour had reverted back to compromised. The master controller, Intel 82371 AB/EB, shows as working correctly and has Dual IDE Channel Settings set to "Default". Even the compromised primary controller shows "no conflicts" under resources. The BIOS is set to Autodetect all three units and finds and identifies them correctly at startup. That NOIDE setting in the registry is correctly not present. Autoexec.bat and Config.sys have nothing I can spot as unusual:

Autoexec. bat:
mode con codepage prepare=((850) C:\WINDOWS\COMMAND\ega.cpi)
mode con codepage select=850
keyb uk,,C:\WINDOWS\COMMAND\keyboard.sys

Config.sys:
device=C:\WINDOWS\COMMAND\display.sys con=(ega,,1)
Country=044,850,C:\WINDOWS\COMMAND\country.sys

so I don't know where to go from here. IOS.log is listed below but I am not at a level of understanding where it speaks to me. It looks flaky but I don't really understand why.

IOS.log:
ESDI data miscompare
Unit number 02 going through real mode drivers.
Unit number 03 going through real mode drivers.

Driver Name: io.sys
Block Driver controlling 04 unit(s)
Driver Info:
Driver Address: 0070005E
IOSYS INT 13: F0009794
INT 13 Hook: 007003EE
INT 4B Hook: F000E7C6
INT 4F Hook: 007004FC
Config.sys line number: 00000000

Driver Name: display.sys
Character Driver
Driver Info:
Driver Address: 02500000
Config.sys line number: 00000001
Config.sys date/time stamp: 24210312
Hardware interrupt hook map: 00000001

Driver Name: himem.sys
Character Driver
Driver Info:
Driver Address: 06B90000
Config.sys line number: 00000000

Driver Name: dblbuff.sys
Character Driver
Driver Info:
Driver Address: 06FF0000
INT 13 Hook: 06FF0148
Config.sys line number: 00000000

Driver Name: ifshlp.sys
Character Driver
Driver Info:
Driver Address: 07B90000
Config.sys line number: 00000000

Autoexec Date/Time Stamp: 2E2E6AE0
TSR Name: KEYB
Autoexec Position: 0000005C
Hardware Interrupt Hook Bit Map: 00000002

Other than that there seems to be no conflict with anything else, all other hardware shows as working correctly in Device Manager.

Anyone got any other ideas as to where I can go from here to sort out this problem? I will be honest it would be easier to give up on the hardware but I just WANT to know what is going on.

 

So are you saying the machine won't even start up then?

When you say no hard drives show up, you mean in the bios? Otherwise, I guess I'm not understanding how you would get to the desktop, if both hard drives aren't appearing.

 

No, you're quite right I didn't make that clear, sorry.

Everything seems fine BIOSwise and the machine is completely usable. Windows comes up without problems, but once it does if you open Device Manager the Primary IDE controller is shown with that dreaded yellow PLING on it. This means of course the disc drives cannot be detected and reported in their own slot in Device Manager. The "devices working in MS-DOS compatability mode" message is given and disc access is slowed.

I understand that this is related to devices using real mode drivers, but have no idea which ones or why!

 

Albert
Here's a little "light" reading for you...obviously all do not apply to your situation but may be worth a glance anyway..

W95/98 Troubleshooting MS-DOS Compatibility Mode on Hard Disks [Q130179]
http://support.microsoft.com/support/kb/articles/q130/1/79.asp

W95/98 Large IDE Hard Disk May Run In Compatibility Mode [Q175629]
http://support.microsoft.com/support/kb/articles/q175/6/29.asp

W95 MS-DOS Compatibility Mode Problems with PCI-IDE Controllers [Q151911]
http://support.microsoft.com/support/kb/articles/q151/9/11.asp


W95 MS-DOS Compatibility Mode When Booting from Removable Disk [Q163317]
http://support.microsoft.com/default.aspx?scid=kb;EN-US;Q163317



W95 MS-DOS Compatibility Mode on National Semiconductor Controller [Q153164]
http://support.microsoft.com/default.aspx?scid=kb;EN-US;Q153164

W95 Drives Use MS-DOS Compatibility Mode with Iomega Zip Drive [Q136742]
http://support.microsoft.com/default.aspx?scid=kb;EN-US;Q136742


W95/98 FAT32 Using Compatibility Mode with OnTrack Disk Manager [Q152701]
http://support.microsoft.com/support/kb/articles/q152/7/01.asp

W95/98 Mh32bit.386 Device Driver Causes MS-DOS Compatibility Mode [Q137405]
http://support.microsoft.com/default.aspx?scid=kb;EN-US;Q137405

W98 Performance/MS-DOS Compatibility Mode Issues with VIA Busmaster [Q242607]
http://support.microsoft.com/default.aspx?scid=kb;EN-US;Q242607

 

I was afraid of that. This sometimes is impossible to solve.

Can you uncheck the config.sys and autoexec.bat files from loading....start button, run, then type msconfig and hit ok.....then restart. (Although, I don't think that is going to help, we may as well try to at least rule it out).

 

Ok AcaCandy I have disabled them and restarted but as you suspected there is no difference. It would have been too easy wouldn't it?

Now, jwbirdsong, that is some reading list, and believe it or not I recognise some of those numbers. I actually have copies of a few of them on disc from the last time I tried to grapple with this problem. One thing I did learn from before is that I was up against a coven of mobo/BIOS/chipset/HDD manufacturers, where it was acknowledged that there existed rogue combinations of hardware, (I found one on I think Maxtor or Seagate's website for example), which just plain caused problems. And of course it was always the OTHER manufacturers' responsibility to sort them out. So AcaCandy I understand where the tone of resignation in your posting is coming from!

I will bone up on the papers JW has offered and repost in a day or so, (or maybe a bit longer if they are written for the priesthood by the priesthood).

Don't worry, in this case I will be very pragmatic and not expect too much time and effort to be poured by any of us into a problem with perhaps a very low return value. Thanks for now and watch this space.

 

I'll dig around and try to find some similar threads I've worked on prior (and by the way, I don't give up easily as you'll see once you start reading the threads ).........although, as you indicate too, it is a VERY frustrating problem.......


http://forums.techguy.org/showthread.php?postid=252688#post252688


There's one for starters....I know there's more around here...just gotta find them.......

 

JW, one of those papers has just jogged my memory. 151911, MS-DOS Compatibility Mode Problems with PCI IDE Controllers, has at the bottom a list of IDE controllers known to cause this problem. Guess what? Intel 82371 AB/EB is on plain view. And that is part of my 440LX setup of course.

Now I remembered that the problem first appeared when I downloaded an infamous update (!) from Micro$oft's site which among other things upgraded the IDE controller drivers. Things immediately went pear shaped! I then found out from Intel's site that in fact the Windows drivers installed were absolutely the wrong ones for that chipset under Win98SE. So I needed to download the removal tool that suddenly appeared on the Intel site, but I could never get the thing to behave correctly again. The Micro$oft forum at the time was ablaze with people whose system's had been screwed by this download, and without a word it simply disappeared overnight. I hope my memory has got things the right way around and I'm not slagging off the one I should be praising but the gist of it is correct I think. The situation wasn't helped by my combination of a Maxtor and a Seagate drive on that primary.

Now here's the strange thing, I am numerous reformat/reinstalls down the line and I STILL can't get the thing to work properly. I wonder if there isn't some Intel update to the mobo setup that I am missing which would sort it out. Intel of course no longer actively support the chipset on their site. Down to me now I guess.

 

Another thing gets me thinking, when you say that it appears to hold for awhile....what happens? Do you turn the machine off, then the next restart is when the problem occurs? It's not able to change midstream.......at least, I wouldn't think so.........

Here's another one that may help:

http://forums.techguy.org/showthread.php?s=&threadid=87827&highlight=ide+controller

 

Got it! I've just had a quick look at Intel's site in the legacy dept and found the right info regarding this MS/Intel screw up. Intel brought out their "The Intel Bus Master IDE driver" and Microsoft incorporated it into a download from their site which their advisory automatic update system recommended to everyone per se. Intel of course were adamant that it should NOT be installed on Win98 systems as it did nasties to the IDE setup. I'm rereading the info there at the moment to see if it sheds any light.

 

Real progress at last!

On an impulse I booted up into Safe mode with the intention of maybe removing the IDE controller and reinstalling. I know how impossible this is in Normal mode as it locks up immediately the controller is removed. According to Safe Mode I had 3 HDDs not 2! It showed type 46 HDDs as C: & D: and an additional type 47 as C:. Yes C: was somehow duplicated as both a type 46 and a type 47. I have deleted the C: type 47 drive leaving only 2 type 46, chosen only because the type 47 is the odd man out, and on rebooting - no difference. So I went back into Safe Mode to see if it was the same, and now I still had only the 2 type 46 HDDs. I then took my courage in both hands and deleted the IDE master controller and quickly closed down before it froze. Then a bootup into Win98 and hey presto "I have found an IDE controller". That installed and another requested restart and - voila - I now have a correctly functioning IDE subsystem and 2 type 47 HDDs. Yes, now they are type 47.

They are both currently set with DMA unchecked and have been reported by the BIOS as UDMA 4 and UDMA 2, told you they were old! Should I leave the DMA setting unchecked?

I'm not going to fiddle with this for a while but I still do not know what is going on so if anyone could please give me any thoughts you have on the whys and wherefores of the affair I would be grateful. I HATE solving a problem without learning how I solved it. It all seems so pointless then. And of course there is every possibility that tomorrow I will be back where I was five minutes ago!

Thanks all.

 

Well things have held stable since I last posted through three or four startups so I guess we could say that they are going to stay that way until I decide to do something else "adventurous". The answer was just to remove and reinstall all along. I do find the "ghost" drive curious though.

Huge thanks to jwbirdsong and AcaCandy for your support and time, as always it is truly greatly appreciated, (and I am sure is doing no harm to your karma). Incidentally it is a great pleasure to be assisted by someone who I can have a legitimate reason to believe must look like Linda Carter, (drool runs from side of mouth). Now if only I could borrow those wristlets...

 

You'd have to visit me in the Random section at the TSG Tavern........

stop by......we have fun there