I just discovered today that the GENERIC_31 kernel from
ftp.eskimo.com (apparently built from the April 3rd sources) hangs on
my SE/30 with a Micron Xceed Color 30 card.  GENERIC_29 works fine, and
I never tried GENERIC_30, if there was one.

	It hangs immediately after printing (no newline):

grf0 at nubus0

	...which is when the working kernel prints:

grf0 at nubus0: 640 x 480 1 color MacVideo (Micron XCEED Color 30\M-*) display

	Actually, the "\M-*" isn't normally printed, but this is what
dmesg(8) is showing.  Why is it different?  Anyway, here's the
beggining of the trace, as transcribed by hand (there may be errors):

_Debugger(849ce,8a80e,fffffc78,16e6,0) + 6
_nmihand(0,d,1.4.0) + 28
_lev7intr(1) + 12
_via2_intr(fffffca4) + 50
_lev2intr(?)
_add_nubus_intr(f9000000,77290,0,a850a,fffffd28) + 14
_grfinit(a84fa,0) + 49
_grfprobe(a84fa,0,fffffdc4,14af8,c078c0) + 32
_grfprobe(c078c0,c078c0,984fa,a84fa,c078c0) + e2

	Unfortunately, I'm in over my head when it comes to device
drivers.  I took a look at nubus.c and grf.c at one point, at Allen's
suggestion, but it's Greek to me.  And no, I don't speak Greek...  :)

-- 
Tres Hofmeister                       Research Applications Program
tres@ncar.ucar.edu         National Center for Atmospheric Research