delete cutensor

Author: wilfonba

CMakeLists.txt

@@ -479,17 +479,6 @@ function(MFC_SETUP_TARGET)
                     "-foffload-options=-lgfortran\ -lm"
                     "-fno-exceptions")
             elseif(CMAKE_Fortran_COMPILER_ID STREQUAL "NVHPC" OR CMAKE_Fortran_COMPILER_ID STREQUAL "PGI")
-                find_package(cuTENSOR)
-                if (NOT cuTENSOR_FOUND)
-                    message(WARNING
-                        "Failed to locate the NVIDIA cuTENSOR library. MFC will be "
-                        "built without support for it, disallowing the use of "
-                        "cu_tensor=T. This can result in degraded performance.")
-                else()
-                    target_link_libraries     (${a_target} PRIVATE cuTENSOR::cuTENSOR)
-                    target_compile_definitions(${a_target} PRIVATE MFC_cuTENSOR)
-                endif()
-
                 foreach (cc ${MFC_CUDA_CC})
                     target_compile_options(${a_target}
                         PRIVATE -gpu=cc${cc}

src/simulation/m_checker.fpp

@@ -57,11 +57,6 @@ contains
 #if !defined(MFC_OpenACC) && !(defined(__PGI) || defined(_CRAYFTN))
         @:PROHIBIT(rdma_mpi, "Unsupported value of rdma_mpi for the current compiler")
 #endif
-
-#ifndef MFC_cuTENSOR
-        @:PROHIBIT(cu_tensor, "MFC was not built with the NVIDIA cuTENSOR library")
-#endif
-
     end subroutine s_check_inputs_compilers
 
     impure subroutine s_check_inputs_igr

src/simulation/m_weno.fpp

@@ -1142,69 +1142,32 @@ contains
         if (n == 0) return
 
         if (weno_dir == 2) then
-#if MFC_cuTENSOR
-            if (cu_tensor) then
-                if (p == 0) then
-                    block
-                        use CuTensorEx
-
-                        #:call GPU_HOST_DATA(use_device='[v_rs_ws_x, v_rs_ws_y]')
-                            v_rs_ws_y = reshape(v_rs_ws_x, shape=[n + 1 + 2*buff_size, m + 2*buff_size + 1, p + 1, sys_size], order=[2, 1, 3, 4])
-                        #:endcall GPU_HOST_DATA
-                    end block
-                else
-                    block
-                        use CuTensorEx
-
-                        #:call GPU_HOST_DATA(use_device='[v_rs_ws_x, v_rs_ws_y]')
-                            v_rs_ws_y = reshape(v_rs_ws_x, shape=[n + 1 + 2*buff_size, m + 2*buff_size + 1, p + 1 + 2*buff_size, sys_size], order=[2, 1, 3, 4])
-                        #:endcall GPU_HOST_DATA
-                    end block
-                end if
-            else
-#endif
-                $:GPU_PARALLEL_LOOP(collapse=4)
-                do j = 1, v_size
-                    do q = is3_weno%beg, is3_weno%end
-                        do l = is2_weno%beg, is2_weno%end
-                            do k = is1_weno%beg - weno_polyn, is1_weno%end + weno_polyn
-                                v_rs_ws_y(k, l, q, j) = v_vf(j)%sf(l, k, q)
-                            end do
+            $:GPU_PARALLEL_LOOP(collapse=4)
+            do j = 1, v_size
+                do q = is3_weno%beg, is3_weno%end
+                    do l = is2_weno%beg, is2_weno%end
+                        do k = is1_weno%beg - weno_polyn, is1_weno%end + weno_polyn
+                            v_rs_ws_y(k, l, q, j) = v_vf(j)%sf(l, k, q)
                         end do
                     end do
                 end do
-#if MFC_cuTENSOR
-            end if
-#endif
+            end do
         end if
 
         ! Reshaping/Projecting onto Characteristic Fields in z-direction
         if (p == 0) return
+
         if (weno_dir == 3) then
-#if MFC_cuTENSOR
-            if (cu_tensor) then
-                block
-                    use CuTensorEx
-
-                    #:call GPU_HOST_DATA(use_device='[v_rs_ws_x, v_rs_ws_z]')
-                        v_rs_ws_z = reshape(v_rs_ws_x, shape=[p + 1 + 2*buff_size, n + 2*buff_size + 1, m + 2*buff_size + 1, sys_size], order=[3, 2, 1, 4])
-                    #:endcall
-                end block
-            else
-#endif
-                $:GPU_PARALLEL_LOOP(collapse=4)
-                do j = 1, v_size
-                    do q = is3_weno%beg, is3_weno%end
-                        do l = is2_weno%beg, is2_weno%end
-                            do k = is1_weno%beg - weno_polyn, is1_weno%end + weno_polyn
-                                v_rs_ws_z(k, l, q, j) = v_vf(j)%sf(q, l, k)
-                            end do
+            $:GPU_PARALLEL_LOOP(collapse=4)
+            do j = 1, v_size
+                do q = is3_weno%beg, is3_weno%end
+                    do l = is2_weno%beg, is2_weno%end
+                        do k = is1_weno%beg - weno_polyn, is1_weno%end + weno_polyn
+                            v_rs_ws_z(k, l, q, j) = v_vf(j)%sf(q, l, k)
                         end do
                     end do
                 end do
-#if MFC_cuTENSOR
-            end if
-#endif
+            end do
         end if
 
     end subroutine s_initialize_weno