>No kmem-grovelers should be broken by this; I didn't change the size
>or layout of struct proc in any way, merely added a new state.  Obviously
>old programs will not know about this state, but currently, only one
>process at any given time will be in this state, so it's not too
>horrible.

wouldn't that be one per cpu (eventually)?  and only the process
that's actually doing the looking will find itself in that state,
right?  or might it context switch to a run queue while the data's
being copied to its userspace from kmem?

>I'm going to commit these just before I commit the scheduler variable
>rearragement, and then bump the NetBSD version after the variable
>rearrangement.

1.4Z?!

-- 
|-----< "CODE WARRIOR" >-----|
codewarrior@daemon.org             * "ah!  i see you have the internet
twofsonet@graffiti.com (Andrew Brown)                that goes *ping*!"
andrew@crossbar.com       * "information is power -- share the wealth."