home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.os.os2.misc
- Path: sparky!uunet!nih-csl!lhc!dabl2
- From: dabl2@nlm.nih.gov (Don A.B. Lindbergh)
- Subject: Re: OS/2 and SCSI
- Message-ID: <1992Jul31.144338.4905@nlm.nih.gov>
- Organization: National Library of Medicine
- References: <rick.19.712447261@sjsumcs.sjsu.edu> <1992Jul30.140855.219272@cs.cmu.edu> <1992Jul30.174240.13771@mksol.dseg.ti.com>
- Date: Fri, 31 Jul 92 14:43:38 GMT
- Lines: 44
-
- >The following is from memory. It may be incorrect in detail, but I
- >believe that it is essential correct in broad.
- >
- >I believe the problem is with how the 1542 is set up. 'Out of the
- >box' behaviour has the timing between the hard and floppy disks set in
- >such a way that the OS/2 installation fails. This leads to all the
- >disk thrashing, etc. I think you can change the timing on the 1542 by
- >changing some jumpers on the controller, but I can't state that with
- >certainty because I don't have this kind of card.
- >
-
- I have a 1542B and my understanding is that the initial problems most people
- were having (including me) with the first release aha154x.add driver was
- floppy access WHILE the hardisk was doing stuff caused errors. It was
- apparently because the bus on/off time needed to be re-programmed in this
- situation and that code needed to be put in the driver, hence the updated
- aha154x.add driver. Since then there have been at least one more I think,
- addressing different issues like > 1 Gig support. The bus on/off time
- parameters are software only adjustable, so it's unlikely that hardware
- switches would accomplish what you suggest, but I am no 1542B 'expert'.
- The simple way to get a successful installation is to let the hardisk
- continue 'doing it's thing' around disk 9 (3.5") until it's finished
- THEN insert the next disk. After the installation is all over, replace
- the aha54x.add driver with the new one. I don't think you can easily
- aha154x.add driver on the original diskettes because it seems to be 'packed'
- with other stuff. If you want to get tricky, at the step between disks 5 and
- 6 (I think) there is a 'break" in the installation, where you reboot and things
- go 'graphical' at this step, you could boot from an OS/2 floppy (or really
- even DOS if you're doing FAT) and put the 'good' aha154x.add driver in place
- in the /OS2 directory. (I think it shouldn't have put the 'bad' one there
- at this point.) Then reboot from the hardisk and continue with the
- installation. This worked for me, I'm guessing that the install sees that
- the aha154x.add is already there and doesn't add it, maybe I was just lucky.
- Actually now that I think about it, the way I understand the config.sys to work
- is that it is parsed in it's entirety and only relevent stuff is used, which is
- why it really doesn't matter if the INT13 driver is in the config.sys. As long
- as the DEVICE=aha154x.add is ALSO in there, THAT is the one that's used.
- Therefore, it might be wise to make sure that line is in the config.sys during
- the 'break'. But this is really for thrill seekers, the easiest way is to just
- let the drive settle down before continuing at disk 9. Whew, sorry for the
- lengthy explanation.
-
- --Don
- dabl2@lhc.nlm.nih.gov
-