• Welcome to OS2World OLD-STATIC-BACKUP Forum.
 

News:

This is an old OS2World backup forum for reference only. IT IS READ ONLY!!!

If you need help with OS/2 - eComStation visit http://www.os2world.com/forum

Main Menu

ECS 2.0 RC 5 stops while loading OS2DASD.DMD

Started by Jan van Dijk, 2008.08.07, 17:09:15

Previous topic - Next topic

Jan van Dijk

Hello,

The following hardware setup:
IOMega Zip 250 attached to Jmicron JMB 363 with 80 pin flatcable. Jumper set to master on the zip drive.
Zipdrive works fine under Open Suse Linux 10.3.

If I now boot ECS, the system shows Loading OS2DASD.DMD in the lower left corner (altf2on.$$$ enabled).
In case I disable the JMicron controller the system boots normally, after loading OS2DASD.DMD, it says Installed OS2LVM.DMD.
In case I change the config.sys entry BASEDEV=DANIS506.ADD /!BIOS to BASEDEV=DANIS506.ADD /!BIOS /A:2 /I, the system boots normally.
I would like to use this zip drive under ECS for backup purposes.
Can some one give me a tip on how to enable my zipdrive.

Thanks in advance,

Jan

Pete

Hi Jan

I'm not that up on zip drives but wonder why you have the /!BIOS switch?

Try removing it, change the line to BASEDEV=DANIS506.ADD and reboot.

I also suspect the 80wire cable may be wasted on a zip drive but could be wrong...

Regards

Pete

DavidG

Jan

Is the zip drive the only drive attached to the cable?

Are you using the secondary or primary cable for the zip drive?


Jan van Dijk

Hi,

The zip drive is configured as master and the only drive attached to this IDE controller.

The /!BIOS option was added during default installation of ECS RC5.

The documentation of DANIS506 tells me that it might be the best to switch this option because information is now based upon the drive it self: (Quote of DANIS506.DOC)

     One user reported a boot failure in a particular situation which involved
     SCSI and ATA/ATAPI units. The solution was to avoid BIOS INT13 calls. The
     driver asks the BIOS about the geometry properties of the ATA drives. In
     general, this information is not required because the most authorative
     source of the geometry info is the drive itself (by inspecting the MBR and
     BPBs). This implies that no BIOS INT13 calls are required in almost any
     scenario. The only exception is if your disk is completely blank, and you
     want to partition this blank disk by OS/2 FDISK with a geometry compatible
     to the BIOS one.

     So, if you prefer to deny the driver calling the BIOS, add the option
     /!BIOS. When doing so, the "strange" VDM seen with some process management
     utilities is gone as well.


I don't think the cabling is the problem because I can use all functionality of the drive under Linux as well as Windows XP on this same system and setup. This is a multi boot configuration with those operating systems.
In the past I have used 80wire cables with zipdrives and this worked fine under ECS.

I rebooted with the /!BIOS removed, with BASEDEV=DANIS506.ADD.
The system booted and showed Loading OS2DASD.DMD and stayed there forever.

After switching back to the options /!BIOS /A:2 /I the system booted normally again.
I have attached dumpide output in case I boot with the /A:2 /I options.
Also attached current config.sys containing more information about the system.

Regards,

Jan

RobertM

Now, this may sound silly, but have you tried configuring the drive as slave?

Some older ATAPI drives I had would not work properly with OS/2 (or work at all) if configured as master... never stopped to determine if it was the drive or the mobo support... just re-jumpered as slave and let it go.

Also, though it probably doesnt matter, if set as master, is it on the master position on the 80pin cable? And if re-set to slave, is it on the slave position?


-R


|
|
Kirk's 5 Year Mission Continues at:
Star Trek New Voyages
|
|


DavidG

One other thing.  Does having the disk inside the enclosure on bootup make a difference or not?

The location on the cable can affect the drive.  All masters I always stick on the end of the cable.  Slaves I attach to the pin connection not located on the end.

Also, does the drive have (if I remember correctly) a CS setting.  I have found some drives not to work properly unless set to CS.

Jan van Dijk

Hi,

I tested the following combinations (I think these are all possible combinations):
Master while attached to Master: No Luck.
Master while attached to Slave: No Luck.
CS while attached to Master: No Luck.
CS while attached to Slave: No Luck.
Slave while attached to Master: No Luck.
Slave while attached to Slave: No Luck.
All above combinations also tested with a zip disk inserted while booting: No Luck.
It seems that if a zipdrive is connected to the JMicron IDE port booting keeps failing.

RobertM

Jan, is it connected to the primary or secondary IDE port?


|
|
Kirk's 5 Year Mission Continues at:
Star Trek New Voyages
|
|


Jan van Dijk

#8
There is only one flatcable connector available to which this drive is connected.
In the manual of the motherboard I see that the board has two ESATA ports connected to the JMicron chip also, I don't know if these are the primary or secondary ports, however these ports are not used at the moment.
Attached find the documentation of the JBM363 chip, it seems that the PATA connection is completely seperate from the SATA connection.

Jan van Dijk

Seems to be a configuration problem, I installed a PCI communication card with 2 serial and 1 parallel port (needed for some other project I am working on) and the system recognized the zip drive. Can it be that due to reshuffle of the interupt assignments the driver now loads correctly?

However LVM is not working anymore. I now get errormessage: unable to start lvm engine.
Can somebody tell me how this is possible and why now the driver loads correctly?

Thanks in advance,

Jan