> What can we do to improve the situation without getting new HW?
> One way would be to measure spl* protected regions starting at
> level 5 and up, to see where long stalls occur. 

how might one go about this measuring?.. reading code, or is there a
more efficient way (and more precice) .. I want to know more about
observing kernel performance in general.. any recommened reading?
unfortuntely my O/S class isnt until next year :(

thanks..

btw.. over the summer someone posted the name of a book about the
porting and general writing of BSD unices.. does anyone remember the
name? (or have another favorite)


later

bart niswonger

				   
----		    -------------------------------			-------
finger me |  One day he found, topped with a smutty grin,   |	from:
for the	  |  The small corpse of a monkey, partly eaten.    |	 Anthony HHecht
rest.  	  |  Force of the sun had split the bluish skin,    |	
	  |  Which, by their questioning and entering in,   |	 Alceste in the
	  |  A swarm of bees had been concerned to sweeten. |	  Wilderness

		      My PGP public key opinion.