> From: Joerg Czeranski <jc@joerch.org>

>	[ lots of ncr.c failures ]


Usually, those ncr errors go away, for me, when I do a >>> init or a power
cycle.

There appear to be two sets of problems. 1: problems sharing and initializing
the ncr controller between SRM and NetBSD. 2: general, pervasive lossage
in the driver.

The ncr driver uses a questionable technique for programming the ncr script
engine that is prone to race conditions and other errors.

One of these days, that driver should be rewritten, or at least revised
heavily enough to where it can be bus_dma(9)'ed. It's over 8,000 lines,
though, so don't hold your breath for this to happen. I once remerged our
ncr.c branch with the freebsd one, but it turned out they have made no real
changes either and the merged driver worked no better so I never checked
it in. My point is: I've tried the "cheap fix" for the ncr problems and it
didn't work...

Now, to get going: the driver hasn't really changed from 1.3.3, so I think
what you are seeing is problems reinitializing the chip for different
addresses and problems going back to SRM and then back again to NetBSD.

Try to boot directly from init or power cycle to 1.4, or boot the 1.4 kernel
from a floppy, or try other things. The ncr lossage almost always has a
workaround of some kind...

	ross.harvey@computer.org