Re: How to enter ddb on recent NetBSD/amd64 current

[Patrick Welche <prlw1%cam.ac.uk@localhost>]

On Tue, Jul 14, 2015 at 06:31:49PM +0000, Michael van Elst wrote:
> Sure. Only the console keyboard checks for magic input.
> 
> When wskbd is attached from a console keyboard, you see the "console keyboard"
> message. The first USB keyboard should be the 'console keyboard' unless
> pckbc_cnattach finds a PC keyboard.
> 
> The problem is, pckbc finds a PC keyboard even when you disable the driver
> because it just maps the hardware (and that a long time before autoconfig
> would attach the driver) and sends a simple enable command. Sending
> the command doesn't require any complex response from the hardware
> and it doesn't require a working keyboard.
> 
> There is a kernel option PCKBC_CNATTACH_SELFTEST which sends a selftest
> command to the PC keyboard and waits for a successful response. Maybe
> this helps in your case to really skip the PC keyboard for the console.
> 
> The alternative is of course to remove pckbd from the kernel configuration.

It's fighting back.

I quickly tried:

uc> find pckbc
[125] pckbc0 at isa? port ? size ? iomem ? iosiz ? irq ? drq ? drq2 ?
uc> find pckbd
[346] pckbd* at pckbc? slot ?
uc> disable pckbc
[125] pckbc0 disabled
uc> disable pckbd
[346] pckbd* disabled

and no joy:

ukbd0 at uhidev0: 8 modifier keys, 6 key codes
wskbd0 at ukbd0 mux 1
wskbd0: connecting to wsdisplay0

So then I tried the PCKBC_CNATTACH_SELFTEST option, leaving pckbc and
pckbd alone, and:

pckbc0 at isa0 port 0x60-0x64
pckbdprobe: reset error 5
pckbd0 at pckbc0 (kbd slot)
pckbc0: using irq 1 for kbd slot
wskbd0 at pckbd0: console keyboard, using wsdisplay0

The test seems to fail, but it grabs console anyway...

Thanks for the hints!

Patrick