>Number:         5372
>Category:       kern
>Synopsis:       pcbkd driver "finds" non-existent keyboards
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    kern-bug-people (Kernel Bug People)
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Tue Apr 28 10:35:00 1998
>Last-Modified:
>Originator:     
>Organization:
Numerical Aerospace Simulation Facility - NASA Ames
>Release:        NetBSD 1.3E, April 28, 1998
>Environment:
	
System: NetBSD bishop 1.3E NetBSD 1.3E (BISHOP) #338: Mon Apr 27 10:37:02 PDT 1998 thorpej@bishop:/tmp_mnt/dracul/u5/netbsd/src/sys/arch/alpha/compile/BISHOP alpha


>Description:
	The pckbd driver "finds" (i.e. successfully probes) keyboards that
	do not exist.  A wskbd instance is then attached to the bogus
	pckbd instance.

	It appears as if something similar happens when the PS/2 mouse
	port is probed, but the psm driver correctly fails to find
	the mouse.

	As a small nit, the probes of these devices are too verbose.  See
	below.

>How-To-Repeat:
	Boot a GENERIC kernel on an AlphaStation with no keyboard or mouse
	attached (i.e. uses serial console).  In my case, this is an
	AlphaStation 500/500 used as a build server, which shares a common
	kernel configuration with my AlphaStation 600, which has a keyboard
	and mouse connected.

	Note the autoconfiguration output:

pckbc0 at isa0 port 0x60-0x64
pckbc: cmd failed
pcprobe: reset error 1
pckbd0 at pckbc0 (kbd slot)
pckbc: cmd failed
pckbd: error setting scanset 2
wskbd0 at pckbd0
pckbc: cmd failed
pmsprobe: command error
pcppi0 at isa0 port 0x61
spkr0 at pcppi0

>Fix:
	None supplied.

	I would suggest making the keyboard probe fail if the command fails,
	as is done with the PS/2 mouse driver.  Also, it would be nice if
	the autoconfiguration messages were quiet by default, unless perhaps
	the kernel was build with DEBUG, or something.

>Audit-Trail:
>Unformatted: