Date: Fri, 11 Jun 1999 22:00:02 +0200
From: "GWDVMS::MOELLER" moeller at gwdvms.dnet.gwdg.de
Subject: Vs3100 & Vs2000 news
Dear all,
I found the e-mail addresses to whom I'm sending this mail on some web pages dealing with VAX models, or the Vs2000 in particular.
I see that none of the web pages yet seem to have learned of the news about the SCSI on the VAXstation/uVAX 2000, formerly known as "tape controller" (TPC) - I've been practicing it since 1997, triggered by B.Barth's "discovery" of it working just fine with NetBSD.
In particular, it's time for doing away with the "almost SCSI" urban legends perpetuated at some web sites ... it's all the same hardware as the Vs3100, only equipped with a somewhat scarce "DMA buffer" of 16 kB (Vs3100 has 128 kB) which _may_ mean that in general, you can't do real large SCSI "DMA" transfers (the last word isn't out on this, so far my VMS driver assumes this restriction to be valid, but there is a chance that I learn of a work-around when I get one original DEC developer to remember well enough how come he put some obscure, possibly never exercised, code into the original VMS 'tape' driver, that at least _pretends_ being able to do 'segmented' DMA transfers, without the tape disconnecting/reconnecting every 12kB, as that old TK50Z apparently did). This DMA limit doesn't even come into play with the BOOT stuff mentioned below, as it's conservatively using "programmed SCSI I/O" throughout (sure that's slooow!).
I lately updated the VMS "PK2K" driver & supporting patches to accomodate several different VMS releases, and a settable SCSI host id, and added BOOT images plus a BOOT ROM upgrade that might even be useful with operating systems other than VMS - 'boot block' boot support would be identical to what the Vs2000 can do with 'old' ST506 disks, and for that matter, the Vs3100 (which - down to the ROMs - is just the Vs2000's big brother).
The posting of mine that I'll include below, mainly deals with my 'updates' to the Vs3100 and Vs2000 ROMs, but further below proceeds to mention the "PK2K" URL as well (all of the "ftp://" URLs also ought to work with an "http://" prefix).
I may still be the only one to have tried out all of this and while I'm pretty sure (and I'm not _that_ clueless :-) about it all working just fine & dandy, could use some more 'early adopters'. Since on the other hand, I don't like writing (web pages, in particular) - maybe some of you could help me out a bit, or at least spread the word ...
I'm mainly a 'VMS person', so "patches", where mentioned below, are ASCII files that can be input to the VAX/VMS PATCH utility, which allows for "editing" binaries at the level of (symbolic!) machine-instructions. If there's interest in non-VMS circles, let me know what tool would be best for doing these things without VMS - while I'm not particularly eager to run some UNIX on these 'small VAXen' _myself_ (and I don't), seriously I'd be happy to help those who prefer doing so.
Best regards,
Wolfgang J. Moeller, Tel. +49 551 2011516 or -510, moeller at gwdvms.dnet.gwdg.de
GWDG, D-37077 Goettingen, F.R.Germany | Disclaimer: No claim intended!
moeller at decus.decus.de - moeller at gwdg.de - w.moeller at ieee.org
On INFO-VAX (comp.os.vms), I posted on 21-MAY-1999 ...
The Vs3100 boot EPROMs (all models, also found in early uVAX 3100 models) have the infamous 1GB limitation for the (SCSI) system disk (i.e. the boot/dump device).
I have patches for some popular boot EPROM versions (plus VMS tools which read the existing EPROMs & create "binary" files suitable for EPROM writers) in
ftp://ftp.gwdg.de/pub/vms/ka420/README.txt ftp://ftp.gwdg.de/pub/vms/ka420/ka420-rom-patch_010.zip
The EPROMs (2 chips, 64k x 16) are standard parts, and are easy to come by (they cost a few $). All you need besides VMS and the chips is someone with an EPROM writer (no soldering involved!) ...
All `standard disclaimers' for free software apply, however my home Vs3100 works real fine with patched boot ROMs, and 2GB Seagate & IBM disks (4 GB IBM DDRS also tried once, there is no reason to expect surprises). If your ROM version is not among the ones listed in the README, please let me know.
-----
Lately I also patched my Vs3100's "storage controller" EPROM (the one that sits on the disk controller board with a Vs3100/* and on the main board with a uVAX 3100/10) to eliminate the _bug_ that would indicate a bogus self-test error (000000D4) with some newer SCSI disks, and thus inhibit automatic boot at power up.
I know that there are at least 3 versions of this ROM in the field. Version numbers are displayed by the ">>> T 50" console command, under the "SCSI-A", "SCSI-B", or "STRG-1" (i.e. ST506) headings. ROMs also differ among the dual SCSI vs. SCSI+ST506 controller boards. The one I'm using is "V1.30" of the SCSI-A/STRG-1 variety...
Depending on interest, I guess I'd see fit to adapt this patch (quite simple) to the other ROM versions - please let me know (too lazy to put the 1st one up for FTP, as long as no-one asks for it, or even complains about the bug :-).
-----
A huge patch that converts the "V2.3" Vs2000 ROMs (4 of 64k x 8) into ones that can boot from SCSI disks (*instead* of from TK50) is part of my "PK2K" kit (this SCSI port driver for VMS is obviously required to be present on the media to be booted).
ftp://ftp.gwdg.de/pub/vms/pk2k/preview/
I know that it works fine for all of my Vs2000s, and for all SCSI disks (up to 4GB) and the two CD-ROM drives that I happen to have around, but _really_ ought to be more widely tested ... (no danger of either a 1GB limitation, or the "000000D4" bug!).
Again, I'd know how to also patch the older ROM versions, if someone really cared for them ...
----
OBTW, EPROM 'burning' is rather fast, compared to CD-R 'burning'. Writers cost somewhat more, and tend to require a PC's parallel port to drive (plus proprietary software) ...
Wolfgang J. Moeller, Tel. +49 551 2011516 or -510, moeller at gwdvms.dnet.gwdg.de
GWDG, D-37077 Goettingen, F.R.Germany | Disclaimer: No claim intended!