Re: Recent instabilities?

[Havard Eidnes <he%NetBSD.org@localhost>]

> On Sat, Apr 02, 2022 at 10:58:28AM +0200, Havard Eidnes wrote:
>> This morning when I came to check on the status.  There were no
>> scary kernel messages, and I could enter another "ntpq -c pe"
>> command, but it failed to do anything when I pressed enter.  A
>> check with Control-T says tcsh is waiting on [uarea], and that
>> state is of course un-interruptible with control-C.
>
> That sounds like kernel memory allocation issues. Can you break into
> ddb, "show uvmexp" and ps, then continue, wait a minute and repeat?

As I tried explaining, "not easily".  This is a Mac Mini G4, with
a USB keyboard attached.  ddb(4) says

           macppc    Some models: <Command>-<Option>-<Power>

So only some unspecified models.  Also, despite me having a genuine
period-correct Apple keyboard (white keytops, blank plastic case), it
has no <Power> button (I beleive the upper right key is "eject"), and
I've tried most other <Command>-<Option>-<whatever> key combinations
to no effect other than echoing some escape sequences for those keys
which give that.

Also, as I hinted at: what are the chances of the USB keyboard working
at all when in DDB if I even manage to break into DDB?  Based on
earlier experience on i386/amd64, I am guessing "slim to none".

Maybe I have to dust off my old G4-upgraded PowerMac 7600 which *does*
have a serial console.  It's been stored for 10+ years...

Regards,

- Håvard