The kernel will not use the serial console unless it's been
told to in the config file.

So the fact that just having pckbc is the config file makes
the USB keyboard not be used console is really a bug, and I will
fix it.  The pckbc_cnattach() function has an error code that
tells if it managed to attach a console keyboard or not.  I'll
try using that to determine if there is a PS/2 keyboard or not.

	-- Lennart

Chris Tribo wrote:
> I'm building a kernel without pckbc to test with now.
> 
> Wouldn't that mean that serial console/debug console machines would  not 
> have the console on the serial port be the system console as well?
> 
> On Apr 29, 2005, at 3:37 PM, Lennart Augustsson wrote:
> 
>> Then I'm pretty sure your USB keyboard is not considered the console.
>> There must be no mention of pckbc in the kernel config file for the
>> USB keyboard to be the console.  This is flaw, admittedly.
>>
>>     -- Lennart
>>
>> Chris Tribo wrote:
>>
>>> I tried control-alt-escape as per the ddb man page and it didn't  work.
>>> hw.cnmagic = \x27\x02
>>> ddb.fromconsole = 1
>>> On Apr 29, 2005, at 3:21 PM, Lennart Augustsson wrote:
>>>
>>>> Can you get into DDB after the system has started up?
>>>> With the USB keyboard, I mean.
>>>>
>>>>     -- Lennart
>>>>
>>>>
>>>> Chris Tribo wrote:
>>>>
>>>>
>>>>> Only:
>>>>> Apr 29 09:10:26 atlantis /netbsd: wsdisplay0 at vga0 kbdmux 1:    
>>>>> console (80x25, vt100 emulation)
>>>>> which seems to be the case as far back as my logs go.
>>>>> There's no mention of pckbdc or pckbd anything in the boot  message...
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>>
>>
>>
>> !DSPAM:42728d0f83771344317433!
>>
>