ATAPI CD-ROMs on secondary controller

From OS2World.Com Wiki
Jump to navigation Jump to search

Last update: 7th February, 1996

Symptoms

ATAPI CD-ROM drive on secondary controller is not seen by the OS/2 Warp installation program.

Hardware

I believe this solution is not specific to this hardware, and may occur on any system with an ATAPI CD-ROM on a secondary controller.

   Gateway/Micronics 486 VLB motherboard with Enhanced IDE (ATAPI)
      controller built onto the motherboard.
   (2) Western Digital Caviar hard drives on PRIMARY controller.
   (1) Generic ATAPI-compatible double-speed CD-ROM drive set up
      as SLAVE on the SECONDARY controller.

Problem

OS/2 Warp's install program doesn't go out and try to find CD-ROMs on the secondary controller by default. When you add extra parameters to the IBM1S506.ADD driver, it will try to seek out drives on the secondary controller, as well.

Procedure

First make sure you have the updated ATAPI CD-ROM drivers. The file to look for is ATAPI.ZIP. You might have problems otherwise, and it eliminates one more variable in the troubleshooting process. Here's what's in the file:

Archive:  ATAPI.ZIP
 Length    Date    Time   Name 
 ------    ----    ----   ------- 
  28328  02-13-95  16:07  ibm1s506.add
  20490  02-13-95  17:33  ibmidecd.flt
  24344  02-13-95  15:54  os2cdrom.dmd
  33578  02-13-95  15:23  os2dasd.dmd
   6061  02-22-95  14:28  readme.txt
 ------                   -------
 112801                    5

You can get it via anon ftp at service.boulder.ibm.com/ps/products/os2/drivers/atapi/.

Second, you should check what's master and slave and on which controller. To repeat myself, I got Warp running with two hard drives on the primary controller, and a CD-ROM sitting alone as a slave on the secondary.

Finally, this is what you need to do if you really want Warp to wake up and see that drive: (this procedure, by the way, is stolen from [IBM's problem list].) It is titled "Configuring the NEC 260 IDE CD-ROM under OS/2 Warp". (but obviously it worked for my generic drive as well). Go into the config.sys file and add parameters to the DEVICE=IBM1S506.ADD line:

BASEDEV=IBM1S506.ADD /A:1 /U:0 /ATAPI

The /A: refers to the controller number, 0 or 1. The /U: refers to the device number. (I know, it's a slave and shouldn't be zero, but since it's the only device out there, this is what IBM said and what works). The /ATAPI enables ATAPI protocol.

Notes

I can't be sure about which CD-ROM drives require this sort of fix, but from talking to several people it would appear that all ATAPI CD-ROMs will need the /A:1 /U:0 /ATAPI to work on the secondary controller, while only some goofy clone drives, possibly only some SONY compatible drives will need the updated drivers (ATAPI.ZIP).

Relocating the CD-ROM drive to the primary controller is an easier solution, but I already have two IDE drives over there. Any hard drive I tried to put on the secondary controller was invisible to Boot Manager and was therefore not bootable. This is the only solution I could find that allowed me two bootable hard drives and a CD-ROM.

Johan W. Klüewer writes (14 October 1995):

I have just installed a NEC CDR-273 CD-ROM player in my machine, and would like to make a little comment to about Eric's note. The article helped me to install the player, but also caused me some trouble, for a silly reason that others might also like to be warned about.

My system is a 486 ASUSTek (I've never heard about it, either) PCI motherboard with two Western Digital Caviar hard drives on the primary on-board IDE controller and, now, the CD-ROM on the secondary, i.e. very similar to Seppanen's.

The problem occurred as follows: I jumpered the CD-ROM as a SLAVE, according to the recommendation in the article. This caused the system to boot up very slowly, spending a lot of time on trying to find something on the secondary controller, and it responded by saying (that is, IBMIDECD.FLT /V in the config.sys responded) that it had recognized a ?##/¤#! unit on the secondary controller. Booting with just Windows, the CD-ROM worked perfectly fine, so I started thinking bad thoughts about IBMIDECD.FLT working improperly.

Following advice from other sources (found through the Pharmacy and IBM's info service), I then tried turning off various features to see if the CD-ROM might be recognized. And it was! when I turned the external cache off and switched off "turbo" mode, reducing my computer to exactly the kind of thing I would like to be without. My suspicions against IBMIDECD.FLT rose when I discovered that the CD-ROM worked perfectly fine if turbo mode was turned on again after the filter had done its job during bootup: was it a weak link that couldn't cope with fast systems, etc.?

The solution turned out to be embarrassingly simple: Jumper the CD-ROM as master. This lets me use full processor speed and the external cache, and bootup time is about 1/3 of what the "slave" setting gave.

Eric Seppanen