iris: Use copy_region and staging resources to avoid transfer stalls

This is similar to intel_miptree_map_blit and intel_buffer_object.c's
temporary blits in i965.

Improves performance of DiRT Rally by 20-25% by eliminating stalls.

Breaks piglit's spec/arb_shader_image_load_store/host-mem-barrier,
by using the GPU to do uploads, exposing a st/mesa issue where it
doesn't give us memory_barrier() calls.  This is a pre-existing issue
and will be fixed by a later patch (currently out for review).

src/gallium/drivers/iris/iris_resource.h

@@ -180,6 +180,11 @@ struct iris_transfer {
    void *buffer;
    void *ptr;
 
+   /** A linear staging resource for GPU-based copy_region transfers. */
+   struct pipe_resource *staging;
+   struct blorp_context *blorp;
+   struct iris_batch *batch;
+
    void (*unmap)(struct iris_transfer *);
 };