Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!rutgers!pyrnj!esquire!cmcl2!lanl!rbw
From: rbw@lanl.UUCP
Newsgroups: comp.sys.amiga
Subject: ASDG-RAM does not survive reboots
Message-ID: <12068@lanl.ARPA>
Date: Thu, 29-Jan-87 11:24:29 EST
Article-I.D.: lanl.12068
Posted: Thu Jan 29 11:24:29 1987
Date-Received: Sat, 31-Jan-87 02:22:35 EST
Distribution: na
Organization: Los Alamos National Laboratory
Lines: 40

Thanks to Perry for the ASDG-RAM disk software.  This has been on my
most-needed list for some time.  I think it downloaded here correctly,
but I am having trouble getting it to work consistently.  As soon as
I get my problem fixed, the $10 is yours.

I cannot get the ASDG-RAM to consistently survive voluntary reboots.
I do not know if it will survive a guru or a hung machine for me.  I
am running V1.2 KickStart and WB (33.47) with 512K chip and 2M fast
(Pacific Cypress) memory.  I have set HighCyl=(anything odd from 63
to 241).  All the proper files seem to be in the proper places, or it
would never work.  My Startup-Sequence file begins with
	   (a)             or             (b)
	MOUNT VD0:                     MOUNT VD0:
	CD VD0:                        CD VD0:
	INFO                           WAIT 4 SECS
				       INFO
The remainder of the startup-sequence creates vd0:c and copies c: to it,
followed by the usual kinds of Workbench things.  When the disk
survives, info shows data in the disk from the past;  when the disk
fails to survive, in case (a) above, info says the disk is being
validated, and in case (b) above, it says that the disk exists (empty)
with 2K of data stored in it.

The inconsistent survivability of the disk seems to be independent of
the exact command used on line 2 of the startup-sequence;  I have
tried CD VD0:, DIR > NIL: VD0:, and IF EXISTS VD0:C, in addition to
others.  Survivability seems to be independent of the size of the disk,
from 63 to 241 cylinders.  It seems to be independent of the size of
the Startup-Sequence file (< or > than 512 bytes).

As a test, I often reboot immediately after completing a boot sequence.
I have no reason to expect that the data in the ram disk has been
corrupted prior to rebooting, so I assume something in the startup
activity must be trashing some critical locations in fast ram.  Is this
possible?  If so, can I do anythingnything to prevent it?  Any advice
would be greatly appreciated.

Bob Walker (rbw%a@lanl.arpa)
Group T-12
Los Alamos National Laboratory