Excerpts from netbsd: 28-Sep-97 port-i386/4177: pccons.c sc..
dholland@eecs.harvard.ed (2006)

> 	The tsleep() in sput() that pccons.c uses to hold output 
> 	when scroll-lock is on can sometimes get called from an
> 	interrupt handler, leading to a total system lockup.

My system even crashes (NULL proc pointer dereference, with
KTRACE).

As I understand it, the original SCROLL LOCK behaviour is
hard to implement cleanly, especially when it comes to the
separation of keyboard part and output part one day.

I'd remove this function, or perhaps, replace it by a simple
XON/XOFF equivalent (as F1 does on real keyboards).

So, current users:
Would anybody miss it?

best regards
Matthias Drochner