> 
> We have a number of different Dilog MSCP controllers, for both SCSI and
> SMD and while playing with an SMD version (DQ256) I was reminded again
> of some local stuff we did to handle certain "features" of these
> controllers.
> 
> The bottom line is that they just don't work when polled too fast (I
> suspect their processor cannot do its thing and respond to constant
> register polling at the same time).
> 
> We also found that our SCSI versions (SQ706A and SU726A) while going
> through the initialization steps on start up needed to be queried an
> extra time or so to get a successful response.
> 
> Here is a massaged copy of our changes to give the flavor of what we did,
> if for some reason anyone has found they've got a 3rd party controller
> that just doesn't respond as I have on several occasions over the years.
> 
Thanx, I'll incorporate these patches (almost) immediately.
Have you tried these cards with NetBSD? In that case I'll put
them on the known-to-work-hardware list.
-- Ragge