MBECK@AI.AI.MIT.EDU ("Mark E. Becker") (06/25/88)
Hello Netland (spoken in a Glum Voice) -
Not too long ago I picked up one of these Mylex 386-based AT
style machines. Very nice. Runs everything like crazy. I've only
seen two pieces of software that wouldn't run on it (apparently a
victim of 8088/8086/80286-only type instructions).
Now for the Fun Part.
It used to be that, if someone had a disk drive whose geometry
wasn't in tables, it was easy enough (for the hardware types) to patch
the ROM and re-calculate checksums so as to get the most out of the
drive without having to mess around with those programs designed to
"fix this problem".
The Mylex board uses the AMI-386 BIOS rom set. This set consists
of four ROM's, a total of 128 KBytes. Bar none, this thing has about
the most sophisticated built-in SETUP program ever devised.. drop
down menus, software to test everything in the box, software to help
determine the optimum interleave for a hard drive.. test floppy
drives.. it'll even format floppies for you. Everything including the
kitchen sink.
The checksum algorithm is different than in "standard" ROMs.
Telephone calls to AMI have resulted in a quiet silence to my
questions; the checksum algorithm is "proprietary".
I am quietly peeved.. In every flavor of computer I've ever
worked with in the past, requesting information about machine insides
after purchase one has either (a) been given away freely or (b)
required me to sign a non-disclosure agreement. The thing isn't going
anywhere.. it sits on my desk at home, letting me have fun with the
MS-DOS world. I even have a small second drive wired up to it so I
can tinker with MINIX.
AMI's technical people have a good product (int 19 doesn't seem
to work "right".. but I'm told that is a common problem) and the
supplied documentation it is more than I've seen from anywhere else.
Sigh. Just another hacker wishing for a return to the "old
days".
Suggestions?
Mark Becker Disclaimer: These opinions are _mine_ and do not
necessarily reflect those of the people
running this computer.syd@dsinc.UUCP (Syd Weinstein) (06/29/88)
In his note, KPETERSEN complains about the AMI bios as not handling
odd geometries. Please note that this bios has a special entry for
odd geometries that handles and non standard one. Note that one
entry, the last one, is empty. If you select this one it prompts you
for all the parameters and I have used it to play with my disk to
see if it had any extra tracks. It also handles my RLL controller this
way to handle the extra sectors. It works fine. You don't need to
mess with the checksum or change roms at all.
(I have the same Mylex board)
--
=====================================================================
Sydney S. Weinstein, CDP, CCP
Datacomp Systems, Inc. Voice: (215) 947-9900
{allegra,bellcore,bpa,vu-vlsi}!dsinc!syd FAX: (215) 938-0235tneff@dasys1.UUCP (Tom Neff) (07/06/88)
Mark Becker complains that the AMI BIOS ROMS use a proprietary checksum algorithm that prevents him from burning patched copies with a "fudged" checksum. May I suggest that in this case you patch the code to no-op the checksum test itself. It shouldn't be too hard to track down with a debugger. -- Tom Neff UUCP: ...!cmcl2!phri!dasys1!tneff "None of your toys CIS: 76556,2536 MCI: TNEFF will function..." GEnie: TOMNEFF BIX: t.neff (no kidding)
zeeff@b-tech.UUCP (Jon Zeeff) (07/06/88)
I've found that the AMI bios does have a different checksum method, but if you simply change the number of sectors in one entry and make up for it in the next one, it doesn't create a problem. I needed RLL support, so I changed the number of sectors for one entry from 17 to 26. I changed the next entry from 17 to 8. Note that this also means you only have to modify/burn one chip. Completely changing an entry (numbers of head, cyls, etc) doesn't work. Please let me know if you find the location (or a byte pattern) of the checksum code. -- Jon Zeeff Branch Technology, uunet!umix!b-tech!zeeff zeeff%b-tech.uucp@umix.cc.umich.edu