self_blas_matrixop_dim2_2d Subroutine

                          subroutine self_blas_matrixop_dim2_2d(A,f,Af,beta,controldegree,targetdegree,nvars,nelems,handle)
                            type(c_ptr),intent(in) :: A
                            type(c_ptr),intent(in) :: f
                            type(c_ptr),intent(inout) :: Af
                            real(c_prec),intent(in) :: beta
                            integer,intent(in) :: controldegree,targetdegree,nvars,nelems
                            type(c_ptr),intent(in) :: handle
                            ! Local
                            integer(c_int) :: m
                            integer(c_int) :: n
                            real(c_prec) :: alpha
                            integer(c_int) :: lda

                            integer :: i
                            integer(c_int64_t) :: strideA
                            integer(c_int) :: incx
                            integer(c_int64_t) :: stridex
                            integer(c_int) :: incy
                            integer(c_int64_t) :: stridey
                            integer(c_int) :: batchCount
                            real(prec),pointer :: f_p(:,:,:,:)
                            real(prec),pointer :: Af_p(:,:,:,:)

                            m = controldegree+1 ! number of rows of A
                            n = targetdegree+1 ! number of columns of A
                            alpha = 1.0_c_prec
                            lda = m ! leading dimension of A
                            strideA = 0 ! stride for the batches of A (no stride)
                            incx = targetdegree+1 !
                            stridex = (controldegree+1)*(targetdegree+1)
                            incy = targetdegree+1
                            stridey = (targetdegree+1)*(targetdegree+1)
                            batchCount = nvars*nelems

                            call c_f_pointer(f,f_p,[targetdegree+1,controldegree+1,nelems,nvars])
                            call c_f_pointer(Af,Af_p,[targetdegree+1,targetdegree+1,nelems,nvars])

                            do i = 0,targetdegree
                              call hipblasCheck(hipblasgemvStridedBatched(handle, &
                                                                          HIPBLAS_OP_T, &
                                                                          m,n,alpha, &
                                                                          A,lda,strideA, &
                                                                          c_loc(f_p(1+i,1,1,1)),incx,stridex,beta, &
                                                                          c_loc(Af_p(1+i,1,1,1)),incy,stridey,batchCount))
                            enddo

                          endsubroutine self_blas_matrixop_dim2_2d