Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!ncar!ames!henry.jpl.nasa.gov!elroy.jpl.nasa.gov!ucla-cs!michael
From: michael@maui.cs.ucla.edu (michael gersten)
Newsgroups: comp.sys.amiga
Subject: Weird bug in 1.3 or Cltd
Message-ID: <26765@shemp.CS.UCLA.EDU>
Date: 29 Aug 89 19:57:16 GMT
Sender: news@CS.UCLA.EDU
Reply-To: michael@maui.cs.ucla.edu (michael gersten)
Organization: UCLA Computer Science Department
Lines: 32

( You'll love this one :-)

Hardware: 2000, one meg chip, 2 meg fast, CLtd SCSI board.
	2 SCSI drives, one a Konika 10 meg removable (unit 0 in mountlist),
	one a 380 fixed (unit 1 in mountlist).
	One external amiga floppy, one internal IBM floppy that
	does not supply diskchange info (system thinks there is
	always a floppy in the drive, diskchange command used manually)
Software: 1.3 ROM, 1.3 workbench, booting from RAD, setpatch r active.
	Cltd's scsi driver version 2.03

The problem? If I have a floppy in the internal drive, everything works
fine. Otherwise, attempting to mount the Konika results in a "Not a dos
disk in unit 0" error. (My mountlist specifies mount=1, BTW). Hitting
retry has no effect (it comes right back). Hitting cancel will reboot
and give me a recoverable alert (Ha!. This is with setpatch running).

Someone is checking the state of the floppy drive unit 0, which is not
a dos disk, and using that to report an error on the Konika. I am using the
Fast file system on the konika; the error is comming while FFS is in
validation stage (at priority -120)

My question: Is this the FFS in 1.3, or is this an error on Cltd's software?

My second question: Can you please raise the priority of the FFS's initial
set up? Or make the validation a different task? Right now I'm constantly
having to raise it up from -120 to 0 so it can resume execution (I generally
have PerfMon running, and sometimes another "soak all cycles at -99" program.)
The mount command doesn't return until the system is happy (3 inch floppy
in drive 0).

			Michael