Re: npxdna vector not initialized

[Erik Bertelsen <bertelsen.erik%gmail.com@localhost>]

2010/10/3 Erik Bertelsen <bertelsen.erik%gmail.com@localhost>:
> 2010/10/3 Erik Bertelsen <bertelsen.erik%gmail.com@localhost>:
>> On my Mac Mini, I have been unable to boot recent custom-built kernels.
>> My latest working kernel is from 6 Sep 2010, and my earliest saved
>> non-working kernel is from 10 Sep. Since then (and up to a few minutes
>> ago) I have made several unsuccessful attempts to boot new kernels.
>>
>> The newer kernels panic after the device probes as shown below:
>>
>> On-screen messages up to panic (retyped manually):
>>
>> (device probes)
>> boot device: wd0
>> root on wd0a dumps on wd0b
>> root file system tyep: ffs
>> panic: npxdna vector not initialized
>> fatal breakpoint trap in supervisor mode
>> trap type 1 code 0 eip c015ac24 cs 8 eflags 46 cr2 8066b80 ilevel 0
>> Stopped in pid 2.1 (sh) at           netbsd:breakpoint+0x4:  popl      %ebp
>
> Hello again,
>
> I was asked privately whether I had attempted to boot a GENERIC
> kernel, which I have to admit that I did not try.
>
> Doing this makes the machine progress further:
>
> It now continues to
>
> ...
> Mounting all filesystems...
> Oct  3 14:07:03 emini mdnsd: mDNSResponder (Engineering build) starting
> Oct  3 14:07:04 emini mdnsd: CheckNATMappings: Failed to allocate port
> multicast socket for NAT-PMP announcements
> uvm_fault(0xcc1e24f8, 0x2be98000, 1) -> 0xe
> fatal page fault in supervisor mode
> trap type 6 code 0 cip c07512b3 cs 8 cflags 10246 cr2 2be90556 ilevel ...
> panic: trap
> cpu0: Begin traceback...
> uvm_fault(0xcc1e24f8, 0, 1) -> 0xe
> fatal page fault in supervisor mode
> trap type 6 code 0 eip c0248c20 cs 8 eflags 10246 cr2 6 ilevel 0
> Skipping crash dump on recursive panic
> panic: trap
> Faulted in mid-traceback: aborting...rebooting...
>
>
> above is typed manually from a video recording of the screen done by
> my phone camera....
>
> - Erik
>

Hello again -- I have made some progress.

Based on the hint about using GENERIC and noticing that the original
panic was from npx code, I tried to add npx at isa0 and now the kernel
completed the initial boot but still paniced at the start-up of mdnsd.

Disabling mdnsd and also disabled mounts of nfs file systems, now made
system able to boot completely. Then rebuilding the system including
kernel modules then was the solution to eliminate the mdnsd and nfs
related panics.

In the end the net result seems to be: some change in september made
the npx attachment move from acpi to isa thus requiring the npx at
isa0 device declaration.

Another observation: up until this change, shutdown -p actually made
the machine power off, after the change shutdown -p stop with the
kernel message 'press any key to reboot'. Neither before or after
these September updates, the machine could complete a 'shutdown -r' --
it shuts down to the message 'rebooting...' and then after a moment
the screen goes black, but the machine does not restart.

Anyway I can continue using the machine :-)

- Erik