[comp.os.minix] Minix 1.1 and IBM portable hard disk

kab@juliet.cs.duke.edu (Kevin Anderson Brooks) (03/17/89)

	I'm a student who has been given MINIX 1.1 for one of my classes.
My problem is that I can't get minix to create a filesystem on my hard 
disk. I have a IBM portable computer with a 20Meg Segate ST225 hard disk.
I've followed the directions in the documentation and I not doing anything
wrong (I guess?).  I've tried both at_wini.c and xt_wini.c and I don't
know what to do next?  My best guess is that I don't have the correct 
driver for my hard disk.  The error message is as follows:

	I type -> # mkfs /dev/hd3 2176
	IT TYPES    Unrecoverable error device 3/3, block 2175
			Error: put_block couldn't write
			Line 1 being processed
                  Unrecoverable error device 3/3, block XX ( some numbers) 
                  Unrecoverable error device 3/3, block XX ( some numbers) 
							  .
							  .
							  .
                  Unrecoverable error device 3/3, block XX ( some numbers) 


	Any help would be deeply appreciated.

                             Kevin Brooks

  Internet:                                         Bitnet:
kab@northlab.cs.duke.edu    			         dboni@tucc
kab@dukeac.ac.duke.edu
kab@dukee.egr.duke.edu
			

ugkamins@sunybcs.uucp (John Kaminski) (03/21/89)

In article <13890@duke.cs.duke.edu> kab@juliet.cs.duke.edu (Kevin Anderson Brooks) writes:
>
>	I'm a student who has been given MINIX 1.1 for one of my classes.
>My problem is that I can't get minix to create a filesystem on my hard 
>disk. I have a IBM portable computer with a 20Meg Segate ST225 hard disk.
>I've followed the directions in the documentation and I not doing anything
>wrong (I guess?).  I've tried both at_wini.c and xt_wini.c and I don't
>know what to do next?  My best guess is that I don't have the correct 
>driver for my hard disk. 
[ ... rest of message deleted for brevity.  contained error msgs, primarily
         kernel msgs ... ]
On the 1.1 distribution I received for the AT, the file system was at most
17 blocks long (hmmm...number of sectors per track you say... why, of
course! :^)  Turns out the problem was in the driver (having the hard disk
controller programming specs proved very helpful) where a logial OR (with "||")
was used where a bitwise OR (with "|") should have been used in order to prop-
erly give the controller the commands for reading and writing past track
boundaries (i.e., going from sector 17 to sector 18, which is on the next
track).  I haven't played with that part of the system in quite some time, so
I do not even know precisely where that is.  I always thought that it would be
fixed in 1.2 (I ordered the system when 1.2 was just ready, and got 1.1, prob.
because P-H hadn't shipped all of their 1.1 copies and/or didn't have 1.2 quite
ready to ship).

Hope that might help.