CCTK_CUDA_KERNEL calc_rhs_2 TYPE=gpu_cuda/3dblock TILE="8,8,8" SHARECODE=yes STENCIL="1,1,1,1,1,1"
{
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=in {phi} "phi"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=out {phirhs} "phirhs"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=in {pi} "pi"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=out {pirhs} "pirhs"
}

CCTK_CUDA_KERNEL calc_rhs_4 TYPE=gpu_cuda/3dblock TILE="8,8,8" SHARECODE=yes STENCIL="2,2,2,2,2,2"
{
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=in {phi} "phi"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=out {phirhs} "phirhs"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=in {pi} "pi"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=out {pirhs} "pirhs"
}

CCTK_CUDA_KERNEL calc_bound_rhs TYPE=gpu_cuda/boundary_s TILE="8,8,8" SHARECODE=yes
{
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=out {phirhs} "phirhs"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=out {pirhs} "pirhs"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=in {xCopy} "xCopy"
}

CCTK_CUDA_KERNEL copy_to_device TYPE=gpu_cuda/3dblock TILE="8,8,8" SHARECODE=yes STENCIL="0,0,0,0,0,0"
{
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=inout {phi} "phi"
  CCTK_CUDA_KERNEL_VARIABLE  cached=no intent=inout {pi} "pi"
}