I've addressed the issue of starvation in an even simpler manner that is far more effective now. There are no known issues at this time. I'll be throwing a bootload of benchmarks at it again to see how it performs now.
http://ck.kolivas.org/patches/2.6/2.6.4/experimental/staircase/
even simpler manner
Any chance of us getting a high-level description of how this simpler, far more effective method works?
High level
In a nutshell the staircase is deadline by design. The only thing that caused starvation were tasks that would wakeup and go back to sleep in less than one tick. I now take that into consideration instead of bolting on a special deadline infrastructure. A high level description will follow soon.