[Intel-gfx] Trivial scheduler, take 2

[Chris Wilson]

Not much to report here. Biggest change is converting the recursive DFS
for PI into an iterative algorithm. The biggest challenge remains trying
to get a consistent set of names so we avoid confusing ourselves going
forwards into preemption. Biggest TODO task is check for perf regressions
from the deferred submission mechanism and for the impact of using a
2-slot dispatch for the GuC workqueue (and the extra tasklet wakeups for
execlists, which should not be so bad as the tasklet is due to be woken
up within a few tens of nanoseconds after the user-interrupt is sent!)
-Chris