drm/amdkfd: Allocate gws on device initialization

On device initialization, KFD allocates all (64) gws which
is shared by all KFD processes.

Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>