feat(cuda): Extend tdsops_t for CUDA backend.

Author: semi-h

src/CMakeLists.txt

@@ -15,6 +15,7 @@ set(CUDASRC
   cuda/common.f90
   cuda/cuda_allocator.f90
   cuda/kernels_dist.f90
+  cuda/tdsops.f90
 )
 
 if(${CMAKE_Fortran_COMPILER_ID} STREQUAL "PGI")

src/cuda/tdsops.f90

@@ -0,0 +1,73 @@
+module m_cuda_tdsops
+   use iso_fortran_env, only: stderr => error_unit
+
+   use m_common, only: dp
+   use m_tdsops, only: tdsops_t, tdsops_init
+
+   implicit none
+
+   type, extends(tdsops_t) :: cuda_tdsops_t
+      !! CUDA extension of the Tridiagonal Solver Operators class.
+      !!
+      !! Regular tdsops_t class is initiated and the coefficient arrays are
+      !! copied into device arrays so that cuda kernels can use them.
+      real(dp), device, allocatable :: dist_fw_dev(:), dist_bw_dev(:), &
+                                       dist_sa_dev(:), dist_sc_dev(:), &
+                                       dist_af_dev(:)
+      real(dp), device, allocatable :: coeffs_dev(:), &
+                                       coeffs_s_dev(:, :), coeffs_e_dev(:, :)
+   contains
+   end type cuda_tdsops_t
+
+   interface cuda_tdsops_t
+      module procedure cuda_tdsops_init
+   end interface cuda_tdsops_t
+
+contains
+
+   function cuda_tdsops_init(n, delta, operation, scheme, n_halo, from_to, &
+                             bc_start, bc_end, sym, c_nu, nu0_nu) &
+      result(tdsops)
+      !! Constructor function for the cuda_tdsops_t class.
+      !! See tdsops_t for details.
+      implicit none
+
+      type(cuda_tdsops_t) :: tdsops !! return value of the function
+
+      integer, intent(in) :: n
+      real(dp), intent(in) :: delta
+      character(*), intent(in) :: operation, scheme
+      integer, optional, intent(in) :: n_halo
+      character(*), optional, intent(in) :: from_to, bc_start, bc_end
+      logical, optional, intent(in) :: sym
+      real(dp), optional, intent(in) :: c_nu, nu0_nu
+
+      integer :: n_stencil
+
+      tdsops%tdsops_t = tdsops_init(n, delta, operation, scheme, n_halo, &
+                                    from_to, bc_start, bc_end, sym, &
+                                    c_nu, nu0_nu)
+
+      n_stencil = 2*tdsops%n_halo + 1
+
+      allocate(tdsops%dist_fw_dev(n), tdsops%dist_bw_dev(n))
+      allocate(tdsops%dist_sa_dev(n), tdsops%dist_sc_dev(n))
+      allocate(tdsops%dist_af_dev(n))
+      allocate(tdsops%coeffs_dev(n_stencil))
+      allocate(tdsops%coeffs_s_dev(n_stencil, tdsops%n_halo))
+      allocate(tdsops%coeffs_e_dev(n_stencil, tdsops%n_halo))
+
+      tdsops%dist_fw_dev(:) = tdsops%dist_fw(:)
+      tdsops%dist_bw_dev(:) = tdsops%dist_bw(:)
+      tdsops%dist_sa_dev(:) = tdsops%dist_sa(:)
+      tdsops%dist_sc_dev(:) = tdsops%dist_sc(:)
+      tdsops%dist_af_dev(:) = tdsops%dist_af(:)
+
+      tdsops%coeffs_dev(:) = tdsops%coeffs(:)
+      tdsops%coeffs_s_dev(:, :) = tdsops%coeffs_s(:, :)
+      tdsops%coeffs_e_dev(:, :) = tdsops%coeffs_e(:, :)
+
+   end function cuda_tdsops_init
+
+end module m_cuda_tdsops
+