On 4/6/06, James Carlson <[EMAIL PROTECTED]> wrote: > So, I don't see what would be added here. I'd like to be proven wrong > (do you have a concrete design to propose), but I don't see it.
Nothing to propose - just seems as though it is a class of problems that is creeping up that is not being addressed in all areas. It used to be that highly parallel system problems only came up on pretty high-end stuff but those rules seem to be changing rather quickly. As packets are received, presumably they could interrupt differnt cores, right? Does this imply that they would be processed by different kernel threads? If so, does the single-threaded userland code become a bottleneck? Mike -- Mike Gerdts http://mgerdts.blogspot.com/ _______________________________________________ networking-discuss mailing list [email protected]
