[1427] in linux-scsi channel archive
(fwd) place-holder: adaptec UW scsi + SMP woes?
daemon@ATHENA.MIT.EDU (Peter T. Breuer)
Sat Feb 15 18:43:47 1997
Date: Sun, 16 Feb 1997 00:40:27 +0100
From: "Peter T. Breuer" <ptb@it.uc3m.es>
To: linux-scsi@vger.rutgers.edu, ptb@oboe.it.uc3m.es
Path: ordago.gsyc.inf.uc3m.es!oboe.it.uc3m.es!ptb
From: ptb@it.uc3m.es (Peter T. Breuer)
Newsgroups: linux.dev.scsi
Subject: place-holder: adaptec UW scsi + SMP woes?
Date: 15 Feb 1997 23:35:19 GMT
Organization: Area Telematica, Universidad Carlos III, Madrid, Spain
Lines: 64
Message-ID: <5e5h7n$8a4@ordago.gsyc.inf.uc3m.es>
NNTP-Posting-Host: oboe.it.uc3m.es
X-Newsreader: TIN [version 1.2 PL2]
Hi - this is a placeholder while I go away and do some research on the
problem. I have the impression that adaptec are probably going to have a
frimware upgrade somewhere ... :-)
I have just come back from a frustrating evening with a dual PP200
with an adaptec uw on the MB. You won't get the version numbers from me
until I get back to the machine in a few days, but I can tell you that it
boots just fine with the aic78xx controller in the 2.0.25 smp kernel.
There is a wide quantum (?) on the wide cable and a jaz and a scsi cdrom
on the other side on the narrow cable, in that order. There is a standard
ami bios on board. Yecch. WinNT and Win95 were on the machine. I have
two ide drives in it as well. The ide controller is on a pci card.
I just repartitioned the extended partition on the scsi drive from linux.
Bang. I cannot reboot in dos FROM A DOS SYSTEM DISKETTE IN A:. It looks to me
as though the bios is out to lunch. In contrast, I can boot winnt from the
ide drive and I can boot linux from a boot diskette. Probably both replace
the bios. I cannot boot win95 from the IDE. Taking the scsi disk out
removes the problem. I can boot win95/dos just fine from A: or C: again.
This is the second time this has happened to this machine. It just came back
from two weeks at the shop. This time I took it apart. The adaptec was
TERMINATED both low and high and was in the middle of the chain. That's wrong.
But I left it because at least winnt and win95 were working that way. At worst
we are going to lose commands on the bus, not invent crazy ones out of
nothing. And if the controller goes bang I can take it back and show them the
smoke and tell them it is their fault.
The termination was via the cmos, not by jumper. The crunch is that there
is no ctl-A entrance to the "scsiSelect" utility. You get a dos utility AND I
CANNOT NOW BOOT DOS UNLESS I TAKE THE DISK OUT. If I do take the disk out
then I can remove the termination in cmos - not that I have yet - but
it will leave the chain unterminated for a moment. I'll try it in a few days,
but I don't think it'll help.
The problem looks to me as though the adaptec cmos has lost something of its
bios. As soon as control transfers to it it starts tring to do something with
the scsi disk, possibly with incorrect geometry (its a 4G). Then dos hangs.
I know it shouldn't be the case that a boot from A: depends on something on
F:G:H:, but it appears to be so.
My question: is there anybody out there who has seen the same crazy behaviour?
Is there something amiss with:
adaptec UW on the MB
dual PPro200
ami bios
quantum scsi 4G drive
as a combination? My bet would be yes. Details later ... meanwhile, please
mail me if you have a good idea. Do the ami and adaptec bisoses communicate
right?
Thanks!
--
---------------------------------------------------------------------
Peter T. Breuer Phd. Ing.,
Area de Ingenieria Telematica E-mail: ptb@it.uc3m.es
Dpto. Ingenieria Tel: +34 1 624 99 47
Universidad Carlos III de Madrid Fax: +34 1 624 94 30/65
Butarque 15, E-28911 Leganes URL: http://www.it.uc3m.es/~ptb
Spain
--
---------------------------------------------------------------------
Peter T. Breuer Phd. Ing.,
Area de Ingenieria Telematica E-mail: ptb@it.uc3m.es
Dpto. Ingenieria Tel: +34 1 624 99 47
Universidad Carlos III de Madrid Fax: +34 1 624 94 30/65
Butarque 15, E-28911 Leganes URL: http://www.it.uc3m.es/~ptb
Spain
