I was able to boot twice and the system ran fine for about 20 minutes
until it crashed with what looked like the same error as yours, but I
didn't write it down and I don't trust my memory too much. In any case, I
was not doing anything funny, just editing some files with vi, and I was
using NFS (not using the disk at all).

BTW I just tried it again, making sure I got to the rom prompt as soon as
possible, and again it crashed as before the "stray interrupt" error.

Is there any way of "disabling" SCSI altogether?

Cheers,
Matias.

On Tue, 16 Oct 2001, Goran Koruga wrote:

> On Tue, Oct 16 2001, Darrin B. Jewell wrote:
>
> Hi,
>
> I was also getting something similar when trying to access my SCSI disk
> (using the latest patches that added some sort of SCSI support) :
>
> NDMAP: *intrstat = 0x4000000<SCSI_DMA>
> NDMAP: *intrmask = 0xffffffffbc027648<NMI,TIMER,ENETX_DMA,ENETR_DMA,SCSI_DMA,SCC,DSP_4,DISK|C16_VIDEO,SCSI,ENETX,ENETR,DSP_3,KBD_MOUSE>
> NDMAP: nd->_nd_map->dm_mapsize = 8160
> NDMAP: nd->_nd_map->dm_nsegs = 3
> NDMAP: nd->_nd_map->dm_segs[2].ds_addr = 0x04f85000
> NDMAP: nd->_nd_map->dm_segs[2].ds_len = 128
> NDMAP: nd->_nd_map->dm_segs[2].ds_xfer_len = 305446639
> NDMAP: nd->_nd_map_cont->dm_mapsize = 80
> NDMAP: nd->_nd_map_cont->dm_nsegs = 1
> NDMAP: nd->_nd_map_cont->dm_segs[0].ds_addr = 0x0459b760
> NDMAP: nd->_nd_map_cont->dm_segs[0].ds_len = 80
> NDMAP: nd->_nd_map_cont->dm_segs[0].ds_xfer_len = 80
> NDMAP: dd->dd_csr          = 0xa000000<COMPLETE,SUPDATE>
> NDMAP: dd->dd_saved_next   = 0x04f420a0
> NDMAP: dd->dd_saved_limit  = 0x04f43000
> NDMAP: dd->dd_saved_start  = 0x0459b760
> NDMAP: dd->dd_saved_stop   = 0x0459b7b0
> NDMAP: dd->dd_next         = 0x04f85080
> NDMAP: dd->dd_next_initbuf = 0x04f85080
> NDMAP: dd->dd_limit        = 0x04f85080
> NDMAP: dd->dd_start        = 0x0459b760
> NDMAP: dd->dd_stop         = 0x0459b7b0
> NDMAP: interrupt ipl (6) intr(0x4000000<SCSI_DMA>)
> panic: DMA Unexpected dma state in interrupt
> (0xa000000<COMPLETE,SUPDATE>)
> Stopped at      _cpu_Debugger+0x6:      unlk    a6
>
> The rest of the details can be found in my mail in mail archives. But I
> only received this message when I was trying to play with the disk - if I
> used NFS root things worked smoothly.
>
> BR,
> Goran
>
> >
> > The stray interrupt is a SCSI interrupt that isn't handled because
> > there is no SCSI driver in the kernel.  I don't remember exactly
> > which conditions cause it, but I think it may be that you interrupted
> > a scsi boot sequence to get the rom prompt.
> >
> > I don't have access to a machine right now to experiment on how
> > to easily access the rom without interrupting an active scsi operation,
> > but it should be possible.
> >
> > Darrin
> >
> > Matias Atria <matias@sas.upenn.edu> writes:
> >
> > > Hi,
> > >
> > > I've been trying to boot NetBSD 1.5.2 on my color (non-turbo) slab, but I
> > > always get the same error (log below). I was running NEXTSTEP 4.2 without
> > > problems until I decided to try NetBSD. If you need more info just let me
> > > know (and if I'm off topic here, let me know too).
> > >
> > > Cheers,
> > > Matias.
> > >
> > > ...
> > > xe0 at intio0
> > > xe0 at MAX address 00:00:0f:00:ca:02
> > > nextkbd0 at intio0
> > > wskbd0 at nextkbd0: console keyboard
> > > nextdisplay0 at mainbus0: 1152 x 832, 16bpp
> > > wsdisplay0 at nextdisplay0: console (color, vt100 emulation), using wskbd0
> > > isrdispatch_autovec: stray level 3 interrupt
> > >   *intrstat = 0x1000<SCSI>
> > >   *intrmask = 0xffffffff98027648<NMI,ENETX_DMA,SCC,DSP_4,DISK|C16_VIDEO,
> > > SCSI,ENETX,ENETR,DSP_3,KYBD_MOUSE>
> > > ...
> > > (same thing over and over until I am thrown into the debugger).
> > >
> > > Any clues?
> >
>
> --
>
> Writing about music is like dancing about architecture.
>     -- Frank Zappa
>