SUMMARY: panic : hat_pmgreserve: invalid pmg & Maxtor LXT-340s

From: Alan G. Arndt (
Date: Mon Jun 10 1991 - 11:53:47 CDT

I got a few responses on the panic, most suggested things to try that were
in the right direction, only one had it entirely correct and is probably
close on the reason why.
>From what I have been told the 4Meg simms must come first as the hat,
hardware address translation, setup makes assumptions on tables. Meaning
that if you have 1Meg simms first you can't have more than 32Mb. I
have not been able to entirely test this theory as I do not have to correct
tool to remove these simm's and the 4Meg's don't have anything nice to grab.
So as I was affraid I did bust one pulling it out. So for a while I only have
24Meg. (aww)

Thanks To:
Jim Mattson <jmattson@UCSD.EDU>
Mike Maciolek <>
Hal Stern - Consultant <stern@sunne.East.Sun.COM>
David N. Blank <>

The Maxtor drive entry has yet to come in and I needed things up and going
so I made one up. I don't know that much about this task so I would still
appreciate getting the 'correct one'. I have also herd that for SCSI drives
that remap their own errors that you should commit and EMPTY back track list
instead of using the one provided by the manufacturer. I assume this is to
prevent the Sun from knowing anything about something that it really isn't
doing. argh. Is this correct? Also where does the cache entry come from
and how does one make that correctly? Here is the entry I am using that
APPEARS to work:

disk_type = "Maxtor LXT-340S" \
        : ctlr = MD21 : fmt_time = 4 \
        : trks_zone = 6 : atrks = 0 : asect = 1 \
        : ncyl = 1554 : acyl = 2 : pcyl = 1560 : nhead = 7 : nsect = 61 \
        : rpm = 3600 : bpt = 0

Alan Arndt, Comtech Labs, Inc. Palo Alto CA

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:15 CDT