Brian Wildasinn wrote:
> Who out there in the mac68k universe is compiling the system files, such
> as kernels, on and for the Q700? What dated version are you using for
> system snapshot binaries and system source (sys.tar.gz) files?

I compiled a custom kernel last night for my new Q700.  Tonight I'm going
to try Allen's latest fast-scsi kernel and see how that works.  The
sources I used are from mid to late September.

> I've been trying unsuccessfully to get a REAL USEABLE kernel to compile.
> Oh yes, the netbsd kernel binary does indeed get compiled, but at
> boottime the system's db> throws out stack barf regarding kernal traps.
> Yes i've turned off all unneeded scsi drives and run "make clean" to
> start fresh in  /usr/src/sys/arch/mac68k/compile/GENERIC/ but each newly
> made kerenel results in the same stack dump into the debugger. I've kept
> a GENERIC 44 kernal backed up to keep the 970927 snapshot at least
> useful.
> 
> Any help is appreciated!

> notes: Q700; current snapshot 970927 for mac68k , GENERIC 44, and
> sys.tar.gz from
> netbsd.org/pub/NetBSD/NetBSD-current/tar_files/src/sys.tar.gz.
> 
> Here's what's been a happening:
> 
> > The kernel compiled, but turned out to be bogus since at boot time it dumped
> > its registers giving this explanation:
> >
> > trap type 1, code=0, v=16bccc
> > kernel: Address error trap
> > pid=0, pc=00E2416, p=2000, sfc=0001, dfc=0001...
> > panic: Address error
> >
> > Here's the dubugger stuff:
> >
> > db>t
> > _Debugger (1,16bc38, e6662, e636a, 8) +6
> > _panic (e636a, 8, 0, 0, 16bd 0f) +40
> > _trap (1, 0, 16bccc) + 1e6
> > faultstkadj (0, 8 347000, ffffff, 1) + 0

Hmmmm...what kind of nubus cards are you using?  Anything weird at all?
It appears to be dying whilst probing the NuBus slots (obviously :-), so I
would assume you have some strange card it's dying on.  Perhaps a video
card that's causing trouble?

> > _nubus_probe_slot (0, 8347000, a, 16bda6) +36
> > _nubus_attach (6b8f980, 6b8f740, 0, 6b8f980, 0) +88
> > _config_attach (6b8f980, f4206, 0, 0, 6b8f980) +28c
> > _mainbus_search (6b8f980, f4206, 0, f41c2, f4206) +2e
> > _mapply (16be7c, f4206) +3c
> > _config_search (e3874, 6b8f980, 0, e3846, 0) +56
> > _mainbus_attach (0, 6b8f980, dea79, dea79, dea79) +1e
> > _config_attach (0, f41ce, dea79, 0, 124) +28c
> > _config_rootfound (dea79, dea79) +3a
> > _configure (dfbae, cc, cc000, dfb9d, 1044000) + 24
> > _cpu_startup (f5004, 9f60, 9550, 80040a5a, f9001000) +492
> > _main () + 4a
> > _main () + 4a
> > db> show object
> > Object 0cbc2f2: size=0x32006131, res=6369280, ref=1630732385,
> > flags=0x3600, pip=24885, pager=0x360006435+ 0x643400,
> > shadow=(0x64330064) + 0x32006431 shadowers=(0x206e0008 trap type ,
> > code=0, v=16b9b4
> > Kernel: Address error trap
> > Caught exception in ddb

Later.

-- 
Colin Wood                                 cwood@ichips.intel.com
Component Design Engineer - MD6                 Intel Corporation
-----------------------------------------------------------------
I speak only on my own behalf, not for my employer.