[Intel-gfx] Make execbuf fast[er]

[Chris Wilson]

Many eons ago, we add ppgtt support. Among the rejoicing, was a bitter
pill, it was slow, much slower due to driver overhead in looking up the
vma. In part this was due to obj_to_vma being a linear walk, but there
was also the effect in execbuf of going from handle to obj to vma, and
we have had talked for a long time of finding a faster method, so why
not use a ht to go from handle to vma. That's the headline, the majority
of the code deals with little fixes to avoid relocations and to reduce
the cost in processing them (and reduce the cost in avoiding them).
-Chris