*BSD News Article 7659


Return to BSD News archive

Newsgroups: comp.unix.bsd
Path: sserve!manuel.anu.edu.au!munnari.oz.au!news.hawaii.edu!ames!haven.umd.edu!uunet!mcsun!sunic!kth.se!love
From: love@stacken.kth.se (Love Feuer)
Subject: MicroVax II, Reno & Emulex
Message-ID: <1992Nov10.134525.6346@kth.se>
Sender: usenet@kth.se (Usenet)
Nntp-Posting-Host: sune.stacken.kth.se
Organization: Stacken Computer Club, Stockholm, Sweden
Date: Tue, 10 Nov 1992 13:45:25 GMT
Lines: 44

We have just laid our hand on an old MicroVax II (or actually a Vax Station 
GPX) equipped with an Emulex QD21 (I think...) ESDI controller together
with a 760 Mb Maxtor drive.

We plugged in an RD53 drive with 4.3BSD Reno with the ambitions to make
UNIX recognize the ESDI controller.

The VMS system that is on the Maxtor drive today is totally convinced that
the Emulex controller is an UDA50 with an RA81 attached to it.

 So we configured the controller as an UDA50 in our Reno system, defined 
some new drive parameters in uda.c that would conform to the Maxtor drive 
and recompiled the kernel together with the AVOID_EMULEX_BUG define that
we threw in just to be on the safe side.

When the new kernel initialized, it found the controller without any
problems, but when slave0 (the maxtor drive) was polled the kernel
panicked with an arithmetic error.

Some investigations gave that uda.c:uda_attach() asked the drive to go on 
line and when the drive interrupted that it was ready, the kernel went
down.

Some guy told us that Emulex has their controllers on a different interrupt 
level than the UDA50 that they are supposed to emulate.


And now for the big one:

What is wrong?
If it is the interrupt level, does anyone know how to set it to some better
value on the Emulex board?

Please mail any answers directly to me, since this problem is quite 
boring for the majority of this newsgroup.

Tnx in advance:
  /Love Feuer (love@stacken.kth.se)