Subject: Re: hp425 kernel panic with DCM ports
To: Chris Rupnik <port-hp300@netbsd.org>
From: Jarkko Teppo <jarkko.teppo@nfp.fi>
List: port-hp300
Date: 07/06/1999 10:55:46
Chris Rupnik wrote..
>Hi
> Ok, made a lot of progress with the serial ports. I >am now able to run a
>getty off the tty0 port and that port seems to work >perfectly, but not the
>tty00 ports (DCM)
>
> I can connect using tty00 via cu to my alpha >running a getty however, and
>i get this
>
>ptrap: bad kernel read access at 0xaa
>trap type 8,code 0x525, v = 0xaa
>kernel program counter = 0xca3ca
>kernel: MMU fault trap
>pid = -1 , pc = 000CA3CA, ps = 2304, sfc = 1, dfc = >1
>
>lots of numbers deleted


<snip>
>dcm0 at dio0 scode 26 ipl 3
>dcm1 at dio0 scode 27 ipl 3
>dcm2 at dio0 scode 28 ipl 3
>dcm3 at dio0 scode 29 ipl 3


Try changing the ipl from three to five, worked for
me. I got the same crash as you did and the
backtrace looked similar. Changing the ipl
made them work.

btw, the dcm:s *seem* to work on distribution
generic kernels with ipl3 but not if you
try to compile a generic kernel.

>interrupt levels: bio = 4, net = 5, tty = 5
                                     =======?
>Chris
>

--
Jarkko Teppo jarkko.teppo@nfp.fi