On Jun 02 04:33:23, kolip...@exoticsilicon.com wrote:
> On Mon, Jun 02, 2025 at 09:17:47AM +0200, Jan Stary wrote:
> > On Jun 02 03:16:13, kolip...@exoticsilicon.com wrote:
> > > On Mon, Jun 02, 2025 at 08:04:22AM +0200, Jan Stary wrote:
> > > > In retrospect, why did this USB keyboard ever attach via pckbc?
> > >
> > > It didn't. At least, not in the dmesgs that you attached.
> > > The pckbc and pckbd drivers are strictly PS/2.
> > > The uhidev and ukbd drivers are USB.
> > > Both pckbd and ukbd can be used as wskbd devices, I.E. a keyboard
> > > that is recognised and used by wscons.
> > > Your USB keyboard is using ukbd in both of those dmesgs.
> >
> > Maybe I am misreading it:
> >
> > pckbd0 at pckbc0 (kbd slot)
> > wskbd0 at pckbd0: console keyboard
> > ukbd0 at uhidev0: 8 variable keys, 6 key codes
> > wskbd1 at ukbd0 mux 1
> > wskbd2 at ucc0 mux 1
> > wsdisplay0 at inteldrm0 mux 1: console (std, vt100 emulation), using wskbd0
> > wskbd1: connecting to wsdisplay0
> > wskbd2: connecting to wsdisplay0
> >
> > So that's wskbd0 at pckbd0 at pckbc0, right?
>
> Yes, but wskbd0 never actually did anything on your system. All your input to
> wscons was coming from wskbd1.
>
> You have, (had), three wskbd devices:
>
> wskbd0 - PS/2 keyboard which does not physically exist on your system.
> wskbd1 - USB keyboard which does exist.
> wskbd2 - The 'extra' keys on the USB keyboard which attach as a separate
> device.
>
> It's possible to use several keyboards at the same time with wscons.
>
> In your case, the driver for PS/2 was active and detecting something that was
> not there. That did not cause a problem, except for the timeout messages you
> reported seeing in the logs.
Thanks, this is what I've being misunderstanding.
> When 7.8 is released, I expect we'll see a few people complaining that their
> PS/2 devices no longer work because they didn't test this code :-/.
Right, you robbed of the timeouts of the nonexistent kbd :-)
Thanks again
Jan
