Here's a scheduler patch for 2.4.17. It looks like it changes a bunch of operations to O(1). I wasn't sure if this would be better for latency or throughput ( which are in conflict, right?). It also looks like it is being fixed to work with the pre-emp patch.
Here's the link to the summerary. http://kt.zork.net/kernel-traffic/latest.html#4 Can anyone tell me anything about this? Taybin
