i915/gem_tiled_fence_blits: Trim workload

Similar to gem_tiled_blits and gem_linear_blits, we only need to just
force the system to be thrashing the GTT for the test to be effective,
so trim the working set to just a be one element larger than could fit,
and parallelise the checking across multiple cpus.

Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1586


Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>