elpa1_tridiag_template.F90

#if 0
!    This file is part of ELPA.
!
!    The ELPA library was originally created by the ELPA consortium,
!    consisting of the following organizations:
!
!    - Max Planck Computing and Data Facility (MPCDF), formerly known as
!      Rechenzentrum Garching der Max-Planck-Gesellschaft (RZG),
!    - Bergische Universität Wuppertal, Lehrstuhl für angewandte
!      Informatik,
!    - Technische Universität München, Lehrstuhl für Informatik mit
!      Schwerpunkt Wissenschaftliches Rechnen ,
!    - Fritz-Haber-Institut, Berlin, Abt. Theorie,
!    - Max-Plack-Institut für Mathematik in den Naturwissenschaften,
!      Leipzig, Abt. Komplexe Strukutren in Biologie und Kognition,
!      and
!    - IBM Deutschland GmbH
!
!    This particular source code file contains additions, changes and
!    enhancements authored by Intel Corporation which is not part of
!    the ELPA consortium.
!
!    More information can be found here:
!    http://elpa.mpcdf.mpg.de/
!
!    ELPA is free software: you can redistribute it and/or modify
!    it under the terms of the version 3 of the license of the
!    GNU Lesser General Public License as published by the Free
!    Software Foundation.
!
!    ELPA is distributed in the hope that it will be useful,
!    but WITHOUT ANY WARRANTY; without even the implied warranty of
!    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
!    GNU Lesser General Public License for more details.
!
!    You should have received a copy of the GNU Lesser General Public License
!    along with ELPA.  If not, see <http://www.gnu.org/licenses/>
!
!    ELPA reflects a substantial effort on the part of the original
!    ELPA consortium, and we ask you to respect the spirit of the
!    license that we chose: i.e., please contribute any changes you
!    may have back to the original ELPA library distribution, and keep
!    any derivatives of ELPA under the same license that we chose for
!    the original distribution, the GNU Lesser General Public License.
!
!
! ELPA1 -- Faster replacements for ScaLAPACK symmetric eigenvalue routines
!
! Copyright of the original code rests with the authors inside the ELPA
! consortium. The copyright of any additional modifications shall rest
! with their original authors, but shall adhere to the licensing terms
! distributed along with the original code in the file "COPYING".
#endif

#include "../general/sanity.F90"

#undef SAVE_MATR
#ifdef DOUBLE_PRECISION_REAL
#define SAVE_MATR(name, iteration) \
call prmat(na,useGpu,a_mat,a_dev,lda,matrixCols,nblk,my_prow,my_pcol,np_rows,np_cols,name,iteration)
#else
#define SAVE_MATR(name, iteration)
#endif

!> \brief Reduces a distributed symmetric matrix to tridiagonal form (like Scalapack Routine PDSYTRD)
!>
!  Parameters
!
!> \param obj	      object of elpa_type
!> \param na          Order of matrix
!>
!> \param a_mat(lda,matrixCols)    Distributed matrix which should be reduced.
!>              Distribution is like in Scalapack.
!>              Opposed to PDSYTRD, a(:,:) must be set completely (upper and lower half)
!>              a(:,:) is overwritten on exit with the Householder vectors
!>
!> \param lda         Leading dimension of a
!>
!> \param nblk        blocksize of cyclic distribution, must be the same in both directions!
!>
!> \param matrixCols  local columns of matrix
!>
!> \param mpi_comm_rows        MPI-Communicator for rows
!> \param mpi_comm_cols        MPI-Communicator for columns
!>
!> \param d_vec(na)       Diagonal elements (returned), identical on all processors
!>
!> \param e_vec(na)       Off-Diagonal elements (returned), identical on all processors
!>
!> \param tau(na)     Factors for the Householder vectors (returned), needed for back transformation
!>
!> \param useGPU      If true,  GPU version of the subroutine will be used
!> \param wantDebug   if true more debug information
!>
    subroutine tridiag_&
    &MATH_DATATYPE&
    &_&
    &PRECISION &
    (obj, na, a_mat, lda, nblk, matrixCols, mpi_comm_rows, mpi_comm_cols, d_vec, e_vec, tau, useGPU, wantDebug, max_threads)
      use cuda_functions
      use iso_c_binding
      use precision
      use elpa_abstract_impl
      use matrix_plot
      use elpa_omp
      use elpa_blas_interfaces

      implicit none
#include "../general/precision_kinds.F90"
      class(elpa_abstract_impl_t), intent(inout)    :: obj
      integer(kind=ik), intent(in)                  :: na, lda, nblk, matrixCols, mpi_comm_rows, mpi_comm_cols
      logical, intent(in)                           :: useGPU, wantDebug
      integer(kind=c_int)                           :: skewsymmetric
      logical                                       :: isSkewsymmetric

      MATH_DATATYPE(kind=rck), intent(out)          :: tau(na)
#ifdef USE_ASSUMED_SIZE
      MATH_DATATYPE(kind=rck), intent(inout)        :: a_mat(lda,*)
#else
      MATH_DATATYPE(kind=rck), intent(inout)        :: a_mat(lda,matrixCols)
#endif
      real(kind=rk), intent(out)                    :: d_vec(na)
      real(kind=rk), intent(out)                    :: e_vec(na)
      integer(kind=ik), parameter                   :: max_stored_uv = 32
      logical,          parameter                   :: mat_vec_as_one_block = .true.

      ! id in processor row and column and total numbers of processor rows and columns
      integer(kind=ik)                              :: my_prow, my_pcol, np_rows, np_cols
      integer(kind=MPI_KIND)                        :: my_prowMPI, my_pcolMPI, np_rowsMPI, np_colsMPI
      integer(kind=MPI_KIND)                        :: mpierr
      integer(kind=ik)                              :: totalblocks, max_loc_block_rows, max_loc_block_cols, max_local_rows, &
                                                       max_local_cols
      ! updated after each istep (in the main cycle) to contain number of
      ! local columns and rows of the remaining part of the matrix
      !integer(kind=ik)                             :: l_cols, l_rows
      integer(kind=ik)                              :: l_cols, l_rows
      integer(kind=ik)                              :: n_stored_vecs

      integer(kind=C_intptr_T)                      :: a_dev, v_row_dev, v_col_dev, u_row_dev, u_col_dev, vu_stored_rows_dev, &
                                                       uv_stored_cols_dev
      logical                                       :: successCUDA

      integer(kind=ik)                              :: istep, i, j, l_col_beg, l_col_end, l_row_beg, l_row_end
      integer(kind=ik)                              :: tile_size, l_rows_per_tile, l_cols_per_tile
      integer(kind=c_intptr_t)                      :: a_offset

      integer(kind=ik), intent(in)                  :: max_threads
#ifdef WITH_OPENMP
      integer(kind=ik)                              :: my_thread, n_threads, n_iter
#endif

      real(kind=rk)                                 :: vnorm2
      MATH_DATATYPE(kind=rck)                       :: vav, x, aux(2*max_stored_uv), aux1(2), aux2(2), vrl, xf
#if COMPLEXCASE == 1
      complex(kind=rck)                             :: aux3(1)
#endif

      integer(kind=c_intptr_t)                      :: num
      MATH_DATATYPE(kind=rck), allocatable          :: tmp(:)
      MATH_DATATYPE(kind=rck), pointer              :: v_row(:), & ! used to store calculated Householder Vector
                                                       v_col(:)   ! the same Vector, but transposed 
      MATH_DATATYPE(kind=rck), pointer              :: u_col(:), u_row(:)

      ! the following two matrices store pairs of vectors v and u calculated in each step
      ! at most max_stored_uv Vector pairs are stored, than the matrix A_i is explicitli updated
      ! u and v are stored both in row and Vector forms
      ! pattern: v1,u1,v2,u2,v3,u3,....
      ! todo: It is little bit confusing, I think, that variables _row actually store columns and vice versa
      MATH_DATATYPE(kind=rck), pointer             :: vu_stored_rows(:,:)
      ! pattern: u1,v1,u2,v2,u3,v3,....
      MATH_DATATYPE(kind=rck), allocatable         :: uv_stored_cols(:,:)

#ifdef WITH_OPENMP
      MATH_DATATYPE(kind=rck), allocatable         :: ur_p(:,:), uc_p(:,:)
#endif

      type(c_ptr)                                   :: v_row_host, v_col_host
      type(c_ptr)                                   :: u_row_host, u_col_host
      type(c_ptr)                                   :: vu_stored_rows_host, uv_stored_cols_host
      real(kind=rk), allocatable                    :: tmp_real(:)
      integer(kind=ik)                              :: min_tile_size, error
      integer(kind=ik)                              :: istat
      character(200)                                :: errorMessage
      character(20)                                 :: gpuString
      integer(kind=c_intptr_t), parameter           :: size_of_datatype = size_of_&
                                                                          &PRECISION&
                                                                          &_&
                                                                          &MATH_DATATYPE
      call obj%get("is_skewsymmetric",skewsymmetric,istat)
      if (istat .ne. ELPA_OK) then
           print *,"Problem getting option for skewsymmetric settings. Aborting..."
           stop
      endif
      isSkewsymmetric = (skewsymmetric == 1)

      if(useGPU) then
        gpuString = "_gpu"
      else
        gpuString = ""
      endif

      call obj%timer%start("tridiag_&
      &MATH_DATATYPE&
      &" // &
      PRECISION_SUFFIX // &
      gpuString )

      if (wantDebug) call obj%timer%start("mpi_communication")
      call mpi_comm_rank(int(mpi_comm_rows,kind=MPI_KIND), my_prowMPI, mpierr)
      call mpi_comm_size(int(mpi_comm_rows,kind=MPI_KIND), np_rowsMPI, mpierr)
      call mpi_comm_rank(int(mpi_comm_cols,kind=MPI_KIND), my_pcolMPI, mpierr)
      call mpi_comm_size(int(mpi_comm_cols,kind=MPI_KIND), np_colsMPI, mpierr)

      my_prow = int(my_prowMPI, kind=c_int)
      np_rows = int(np_rowsMPI, kind=c_int)
      my_pcol = int(my_pcolMPI, kind=c_int)
      np_cols = int(np_colsMPI, kind=c_int)
      if (wantDebug) call obj%timer%stop("mpi_communication")

      ! Matrix is split into tiles; work is done only for tiles on the diagonal or above
      ! seems that tile is a square submatrix, consisting by several blocks
      ! it is a smallest possible square submatrix, where blocks being distributed among
      ! processors are "aligned" in both rows and columns
      !  -----------------
      ! | 1 4 | 1 4 | 1 4 | ...
      ! | 2 5 | 2 5 | 2 5 | ...
      ! | 3 6 | 3 6 | 3 6 | ...
      !  ----------------- ...
      ! | 1 4 | 1 4 | 1 4 | ...
      ! | 2 5 | 2 5 | 2 5 | ...
      ! | 3 6 | 3 6 | 3 6 | ...
      !  ----------------- .
      !   : :   : :   : :    .
      !   : :   : :   : :      .
      !
      ! this is a tile, where each number represents block, assigned to a processor with the shown number
      ! size of this small block is nblk
      ! Image is for situation with 6 processors, 3 processor rows and 2 columns
      ! tile_size is thus nblk * 6
      !
      tile_size = nblk*least_common_multiple(np_rows,np_cols) ! minimum global tile size

      ! make tile_size a smallest possible multiple of previously defined tile size, such that it is
      ! larger or equal to min_tile_size
      ! min_tile_size has been originally hardcoded as 128 * max(np_rows, np_cols), so it is now the implicit value
      ! it can, however, be set by the user
      call obj%get("min_tile_size", min_tile_size ,error)
      if (error .ne. ELPA_OK) then
        print *,"Problem setting option for min_tile_size. Aborting..."
        stop
      endif
      if(min_tile_size == 0) then
        ! not set by the user, use the default value
        min_tile_size = 128*max(np_rows, np_cols)
      endif
      tile_size = ((min_tile_size-1)/tile_size+1)*tile_size

      l_rows_per_tile = tile_size/np_rows ! local rows of a tile
      l_cols_per_tile = tile_size/np_cols ! local cols of a tile

      totalblocks = (na-1)/nblk + 1
      max_loc_block_rows = (totalblocks-1)/np_rows + 1
      max_loc_block_cols = (totalblocks-1)/np_cols + 1

      ! localy owned submatrix has size at most max_local_rows x max_local_cols at each processor
      max_local_rows = max_loc_block_rows*nblk
      max_local_cols = max_loc_block_cols*nblk

      ! allocate memmory for vectors
      ! todo: It is little bit confusing, I think, that variables _row actually store columns and vice versa
      ! todo: if something has length max_local_rows, it is actually a column, no?
      ! todo: probably one should read it as v_row = Vector v distributed among rows
      !
      allocate(tmp(MAX(max_local_rows,max_local_cols)), stat=istat, errmsg=errorMessage)
      call check_alloc("tridiag_&
      &MATH_DATATYPE ", "tmp", istat, errorMessage)

      ! allocate v_row 1 element longer to allow store and broadcast tau together with it
        allocate(uv_stored_cols(max_local_cols,2*max_stored_uv), stat=istat, errmsg=errorMessage)
        call check_alloc("tridiag_&
        &MATH_DATATYPE ", "uv_stored_cols", istat, errorMessage)

      if (useGPU) then
        num = (max_local_rows+1) * size_of_datatype
        successCUDA = cuda_malloc_host(v_row_host,num)
        check_alloc_cuda("tridiag: v_row_host", successCUDA)

        call c_f_pointer(v_row_host,v_row,(/num/))

        num = (max_local_cols) * size_of_datatype
        successCUDA = cuda_malloc_host(v_col_host,num)
        check_alloc_cuda("tridiag: v_col_host", successCUDA)

        call c_f_pointer(v_col_host,v_col,(/num/))

        num = (max_local_cols) * size_of_datatype
        successCUDA = cuda_malloc_host(u_col_host,num)
        check_alloc_cuda("tridiag: u_col_host", successCUDA)

        call c_f_pointer(u_col_host,u_col,(/num/))

        num = (max_local_rows) * size_of_datatype
        successCUDA = cuda_malloc_host(u_row_host,num)
        check_alloc_cuda("tridiag: u_row_host", successCUDA)

        call c_f_pointer(u_row_host,u_row,(/num/))

        num = (max_local_rows * 2*max_stored_uv) * size_of_datatype
        successCUDA = cuda_malloc_host(vu_stored_rows_host,num)
        check_alloc_cuda("tridiag: vu_stored_rows_host", successCUDA)

        call c_f_pointer(vu_stored_rows_host,vu_stored_rows,(/max_local_rows,2*max_stored_uv/))

        num = (max_local_cols * 2*max_stored_uv) * size_of_datatype
        !successCUDA = cuda_malloc_host(uv_stored_cols_host,num)
        !check_alloc_cuda("tridiag: uv_stored_cols_host", successCUDA)

        !call c_f_pointer(uv_stored_cols_host,uv_stored_cols,(/max_local_cols,2*max_stored_uv/))

        successCUDA = cuda_host_register(int(loc(uv_stored_cols),kind=c_intptr_t),num,&
                      cudaHostRegisterDefault)
        check_host_register_cuda("tridiag: uv_stored_cols", successCUDA)

#if defined(DOUBLE_PRECISION_REAL) || defined(DOUBLE_PRECISION_COMPLEX)
        num = na * 8
#else
        num = na * 4
#endif
        successCUDA = cuda_host_register(int(loc(e_vec),kind=c_intptr_t),num,&
                      cudaHostRegisterDefault)
        check_host_register_cuda("tridiag: e_vec", successCUDA)

#if defined(DOUBLE_PRECISION_REAL) || defined(DOUBLE_PRECISION_COMPLEX)
        num = na * 8
#else
        num = na * 4
#endif
        successCUDA = cuda_host_register(int(loc(d_vec),kind=c_intptr_t),num,&
                      cudaHostRegisterDefault)
        check_host_register_cuda("tridiag: d_vec", successCUDA)

      else
        allocate(v_row(max_local_rows+1), stat=istat, errmsg=errorMessage)
        call check_alloc("tridiag_&
        &MATH_DATATYPE ", "v_row", istat, errorMessage)

        allocate(v_col(max_local_cols), stat=istat, errmsg=errorMessage)
        call check_alloc("tridiag_&
         &MATH_DATATYPE ", "v_col", istat, errorMessage)

        allocate(u_col(max_local_cols), stat=istat, errmsg=errorMessage)
        call check_alloc("tridiag_&
        &MATH_DATATYPE ", "u_col", istat, errorMessage)

        allocate(u_row(max_local_rows), stat=istat, errmsg=errorMessage)
        call check_alloc("tridiag_&
        &MATH_DATATYPE ", "u_row", istat, errorMessage)

        allocate(vu_stored_rows(max_local_rows,2*max_stored_uv), stat=istat, errmsg=errorMessage)
        call check_alloc("tridiag_&
        &MATH_DATATYPE ", "vu_stored_rows", istat, errorMessage)
        
      endif

#ifdef WITH_OPENMP
      allocate(ur_p(max_local_rows,0:max_threads-1), stat=istat, errmsg=errorMessage)
      call check_alloc("tridiag_&
      &MATH_DATATYPE ", "ur_p", istat, errorMessage)

      allocate(uc_p(max_local_cols,0:max_threads-1), stat=istat, errmsg=errorMessage)
      call check_alloc("tridiag_&
      &MATH_DATATYPE ", "uc_p", istat, errorMessage)
#endif /* WITH_OPENMP */

      tmp = 0
      v_row = 0
      u_row = 0
      v_col = 0
      u_col = 0

      if (useGPU) then
         successCUDA = cuda_malloc(v_row_dev, max_local_rows * size_of_datatype)
         check_alloc_cuda("tridiag: v_row_dev", successCUDA)

         successCUDA = cuda_malloc(u_row_dev, max_local_rows * size_of_datatype)

         check_alloc_cuda("tridiag: u_row_dev", successCUDA)

         successCUDA = cuda_malloc(v_col_dev, max_local_cols * size_of_datatype)
         check_alloc_cuda("tridiag: v_col_dev", successCUDA)

         successCUDA = cuda_malloc(u_col_dev, max_local_cols * size_of_datatype)
         check_alloc_cuda("tridiag: u_col_dev", successCUDA)

         successCUDA = cuda_malloc(vu_stored_rows_dev, max_local_rows * 2 * max_stored_uv * size_of_datatype)
         check_alloc_cuda("tridiag: vu_stored_rows_dev", successCUDA)

         successCUDA = cuda_malloc(uv_stored_cols_dev, max_local_cols * 2 * max_stored_uv * size_of_datatype)
         check_alloc_cuda("tridiag: vu_stored_rows_dev", successCUDA)
      endif !useGPU


      d_vec(:) = 0
      e_vec(:) = 0
      tau(:) = 0

      n_stored_vecs = 0

      l_rows = local_index(na, my_prow, np_rows, nblk, -1) ! Local rows of a_mat
      l_cols = local_index(na, my_pcol, np_cols, nblk, -1) ! Local cols of a_mat

      if (my_prow == prow(na, nblk, np_rows) .and. my_pcol == pcol(na, nblk, np_cols)) &
#if COMPLEXCASE == 1
        d_vec(na) = real(a_mat(l_rows,l_cols), kind=rk)
#endif
#if REALCASE == 1
        d_vec(na) = a_mat(l_rows,l_cols)
#endif

      if (useGPU) then
        ! allocate memmory for matrix A on the device and than copy the matrix

        num = lda * matrixCols * size_of_datatype

        successCUDA = cuda_malloc(a_dev, num)
        check_alloc_cuda("tridiag: a_dev", successCUDA)

        successCUDA = cuda_host_register(int(loc(a_mat),kind=c_intptr_t),num,&
                      cudaHostRegisterDefault)
        check_host_register_cuda("tridiag: a_mat", successCUDA)

        successCUDA = cuda_memcpy(a_dev, int(loc(a_mat(1,1)),kind=c_intptr_t), &
                                  num, cudaMemcpyHostToDevice)
        check_memcpy_cuda("tridiag: a_dev", successCUDA)
      endif

      ! main cycle of tridiagonalization
      ! in each step, 1 Householder Vector is calculated
      do istep = na, 3 ,-1

        ! Calculate number of local rows and columns of the still remaining matrix
        ! on the local processor
        l_rows = local_index(istep-1, my_prow, np_rows, nblk, -1)
        l_cols = local_index(istep-1, my_pcol, np_cols, nblk, -1)

        ! Calculate Vector for Householder transformation on all procs
        ! owning column istep

        if (my_pcol == pcol(istep, nblk, np_cols)) then

          ! Get Vector to be transformed; distribute last element and norm of
          ! remaining elements to all procs in current column

          ! copy l_cols + 1 column of A to v_row
          if (useGPU) then
            a_offset = l_cols * lda * size_of_datatype
            ! we use v_row on the host at the moment! successCUDA = cuda_memcpy(v_row_dev, a_dev + a_offset, 
            ! (l_rows)*size_of_PRECISION_real, cudaMemcpyDeviceToDevice)

            successCUDA = cuda_memcpy(int(loc(v_row),kind=c_intptr_t), &
                                      a_dev + a_offset, (l_rows)* size_of_datatype, cudaMemcpyDeviceToHost)
            check_memcpy_cuda("tridiag a_dev 1", successCUDA)
          else
            v_row(1:l_rows) = a_mat(1:l_rows,l_cols+1)
          endif

          if (n_stored_vecs > 0 .and. l_rows > 0) then
            if (wantDebug) call obj%timer%start("blas")
#if COMPLEXCASE == 1
            aux(1:2*n_stored_vecs) = conjg(uv_stored_cols(l_cols+1,1:2*n_stored_vecs))
#endif
            call PRECISION_GEMV('N',   &
                                  int(l_rows,kind=BLAS_KIND), int(2*n_stored_vecs,kind=BLAS_KIND), &
                                  ONE, vu_stored_rows, int(ubound(vu_stored_rows,dim=1),kind=BLAS_KIND), &
#if REALCASE == 1
                                  uv_stored_cols(l_cols+1,1), &
                                          int(ubound(uv_stored_cols,dim=1),kind=BLAS_KIND), &
#endif
#if COMPLEXCASE == 1
                                   aux, 1_BLAS_KIND,  &

#endif
                                  ONE, v_row, 1_BLAS_KIND)
            if (wantDebug) call obj%timer%stop("blas")

          endif

          if (my_prow == prow(istep-1, nblk, np_rows)) then
            aux1(1) = dot_product(v_row(1:l_rows-1),v_row(1:l_rows-1))
            aux1(2) = v_row(l_rows)
          else
            aux1(1) = dot_product(v_row(1:l_rows),v_row(1:l_rows))
            aux1(2) = 0.
          endif

#ifdef WITH_MPI
          if (wantDebug) call obj%timer%start("mpi_communication")
          call mpi_allreduce(aux1, aux2, 2_MPI_KIND, MPI_MATH_DATATYPE_PRECISION, &
                               MPI_SUM, int(mpi_comm_rows,kind=MPI_KIND), mpierr)
          if (wantDebug) call obj%timer%stop("mpi_communication")
#else /* WITH_MPI */
          aux2 = aux1
#endif /* WITH_MPI */

#if REALCASE == 1
          vnorm2 = aux2(1)
#endif
#if COMPLEXCASE == 1
          vnorm2 = real(aux2(1),kind=rk)
#endif
          vrl    = aux2(2)

          ! Householder transformation
#if REALCASE == 1
          call hh_transform_real_&
#endif
#if COMPLEXCASE == 1
          call hh_transform_complex_&
#endif
               &PRECISION &
                             (obj, vrl, vnorm2, xf, tau(istep), wantDebug)
          ! Scale v_row and store Householder Vector for back transformation

          v_row(1:l_rows) = v_row(1:l_rows) * xf
          if (my_prow == prow(istep-1, nblk, np_rows)) then
            v_row(l_rows) = 1.

            ! vrl is newly computed off-diagonal element of the final tridiagonal matrix
#if REALCASE == 1
            e_vec(istep-1) = vrl
#endif
#if COMPLEXCASE == 1
            e_vec(istep-1) = real(vrl,kind=rk)
#endif
          endif

          ! store Householder Vector for back transformation
          a_mat(1:l_rows,l_cols+1) = v_row(1:l_rows)

          ! add tau after the end of actuall v_row, to be broadcasted with it
          v_row(l_rows+1) = tau(istep)
         endif !(my_pcol == pcol(istep, nblk, np_cols))

!          SAVE_MATR("HH vec stored", na - istep + 1)

#ifdef WITH_MPI
         if (wantDebug) call obj%timer%start("mpi_communication")
         ! Broadcast the Householder Vector (and tau) along columns
         call MPI_Bcast(v_row, int(l_rows+1,kind=MPI_KIND), MPI_MATH_DATATYPE_PRECISION,    &
                        int(pcol(istep, nblk, np_cols),kind=MPI_KIND), int(mpi_comm_cols,kind=MPI_KIND), mpierr)
         if (wantDebug) call obj%timer%stop("mpi_communication")
#endif /* WITH_MPI */

         !recover tau, which has been broadcasted together with v_row
         tau(istep) =  v_row(l_rows+1)

         ! Transpose Householder Vector v_row -> v_col
         call elpa_transpose_vectors_&
             &MATH_DATATYPE&
             &_&
             &PRECISION &
                   (obj, v_row, ubound(v_row,dim=1), mpi_comm_rows, v_col, ubound(v_col,dim=1), mpi_comm_cols, &
                    1, istep-1, 1, nblk, max_threads)

         ! Calculate u = (A + VU**T + UV**T)*v

         ! For cache efficiency, we use only the upper half of the matrix tiles for this,
         ! thus the result is partly in u_col(:) and partly in u_row(:)

         u_col(1:l_cols) = 0
         u_row(1:l_rows) = 0
         if (l_rows > 0 .and. l_cols> 0 ) then
          if (useGPU) then
            successCUDA = cuda_memset(u_col_dev, 0, l_cols * size_of_datatype)
            check_memcpy_cuda("tridiag: u_col_dev", successCUDA)

            successCUDA = cuda_memset(u_row_dev, 0, l_rows * size_of_datatype)
            check_memcpy_cuda("tridiag: u_row_dev", successCUDA)

            successCUDA = cuda_memcpy(v_col_dev, int(loc(v_col(1)),kind=c_intptr_t), &
                          l_cols * size_of_datatype, cudaMemcpyHostToDevice)

            check_memcpy_cuda("tridiag: v_col_dev", successCUDA)

            successCUDA = cuda_memcpy(v_row_dev, int(loc(v_row(1)),kind=c_intptr_t), &
                                      l_rows * size_of_datatype, cudaMemcpyHostToDevice)
            check_memcpy_cuda("tridiag: v_row_dev", successCUDA)
          endif ! useGU

#ifdef WITH_OPENMP
          call obj%timer%start("OpenMP parallel")
!$OMP PARALLEL PRIVATE(my_thread,n_threads,n_iter,i,l_col_beg,l_col_end,j,l_row_beg,l_row_end)

          my_thread = omp_get_thread_num()
          
          n_threads = omp_get_num_threads()

          n_iter = 0

          ! first calculate A*v part of (A + VU**T + UV**T)*v
          uc_p(1:l_cols,my_thread) = 0.
          ur_p(1:l_rows,my_thread) = 0.
#endif /* WITH_OPENMP */
          do i= 0, (istep-2)/tile_size
            l_col_beg = i*l_cols_per_tile+1
            l_col_end = min(l_cols,(i+1)*l_cols_per_tile)
            if (l_col_end < l_col_beg) cycle
            do j = 0, i
              l_row_beg = j*l_rows_per_tile+1
              l_row_end = min(l_rows,(j+1)*l_rows_per_tile)
              if (l_row_end < l_row_beg) cycle
#ifdef WITH_OPENMP
              if (mod(n_iter,n_threads) == my_thread) then
                if (wantDebug) call obj%timer%start("blas")
                call PRECISION_GEMV(BLAS_TRANS_OR_CONJ, &
                                    int(l_row_end-l_row_beg+1,kind=BLAS_KIND), int(l_col_end-l_col_beg+1,kind=BLAS_KIND), &
                                    ONE, a_mat(l_row_beg,l_col_beg), int(lda,kind=BLAS_KIND),         &
                                    v_row(l_row_beg:max_local_rows+1), 1_BLAS_KIND, ONE, uc_p(l_col_beg,my_thread), 1_BLAS_KIND)

                if (i/=j) then
                  if (isSkewsymmetric) then
                    call PRECISION_GEMV('N', int(l_row_end-l_row_beg+1,kind=BLAS_KIND), &
                                        int(l_col_end-l_col_beg+1,kind=BLAS_KIND), &
                                        -ONE, a_mat(l_row_beg,l_col_beg), int(lda,kind=BLAS_KIND), &
                                        v_col(l_col_beg:max_local_cols), 1_BLAS_KIND,  &
                                        ONE, ur_p(l_row_beg,my_thread), 1_BLAS_KIND)

                  else
                    call PRECISION_GEMV('N', int(l_row_end-l_row_beg+1,kind=BLAS_KIND), &
                                        int(l_col_end-l_col_beg+1,kind=BLAS_KIND), &
                                        ONE, a_mat(l_row_beg,l_col_beg), int(lda,kind=BLAS_KIND), &
                                        v_col(l_col_beg:max_local_cols), 1_BLAS_KIND,  &
                                        ONE, ur_p(l_row_beg,my_thread), 1_BLAS_KIND)
                  endif
                endif
                if (wantDebug) call obj%timer%stop("blas")
              endif
              n_iter = n_iter+1
#else /* WITH_OPENMP */

              ! multiplication by blocks is efficient only for CPU
              ! for GPU we introduced 2 other ways, either by stripes (more simmilar to the original
              ! CPU implementation) or by one large matrix Vector multiply
              if (.not. useGPU) then
                if (wantDebug) call obj%timer%start("blas")
                call PRECISION_GEMV(BLAS_TRANS_OR_CONJ,  &
                            int(l_row_end-l_row_beg+1,kind=BLAS_KIND), int(l_col_end-l_col_beg+1,kind=BLAS_KIND), &
                            ONE, a_mat(l_row_beg, l_col_beg), int(lda,kind=BLAS_KIND),         &
                            v_row(l_row_beg:max_local_rows+1), 1_BLAS_KIND,                           &
                            ONE, u_col(l_col_beg:max_local_cols), 1_BLAS_KIND)

                if (i/=j) then
                  if (isSkewsymmetric) then
                    call PRECISION_GEMV('N',int(l_row_end-l_row_beg+1,kind=BLAS_KIND), int(l_col_end-l_col_beg+1,kind=BLAS_KIND),  &
                                        -ONE, a_mat(l_row_beg,l_col_beg), int(lda,kind=BLAS_KIND),               &
                                        v_col(l_col_beg:max_local_cols), 1_BLAS_KIND, ONE, u_row(l_row_beg:max_local_rows), &
                                        1_BLAS_KIND)

                  else
                    call PRECISION_GEMV('N',int(l_row_end-l_row_beg+1,kind=BLAS_KIND), int(l_col_end-l_col_beg+1,kind=BLAS_KIND),  &
                                        ONE, a_mat(l_row_beg,l_col_beg), int(lda,kind=BLAS_KIND),               &
                                        v_col(l_col_beg:max_local_cols), 1_BLAS_KIND, ONE, u_row(l_row_beg:max_local_rows), &
                                        1_BLAS_KIND)
                  endif
                endif
                if (wantDebug) call obj%timer%stop("blas")
              endif ! not useGPU

#endif /* WITH_OPENMP */
            enddo  ! j=0,i
          enddo  ! i=0,(istep-2)/tile_size

          if (useGPU) then
            if(mat_vec_as_one_block) then
              ! Unlike for CPU, we (for each MPI thread) do just one large mat-vec multiplication
              ! this requires altering of the algorithm when later explicitly updating the matrix
              ! after max_stored_uv is reached : we need to update all tiles, not only those above diagonal
              if (wantDebug) call obj%timer%start("cublas")
              call cublas_PRECISION_GEMV(BLAS_TRANS_OR_CONJ, l_rows,l_cols,  &
                                        ONE, a_dev, lda,                   &
                                        v_row_dev , 1,                          &
                                        ONE, u_col_dev, 1)

       ! todo: try with non transposed!!!
!                 if(i/=j) then
!                   call cublas_PRECISION_GEMV('N', l_row_end-l_row_beg+1,l_col_end-l_col_beg+1,  &
!                                             ONE, a_dev + a_offset, lda,                        &
!                                             v_col_dev + (l_col_beg - 1) *                      &
!                                             size_of_datatype, 1,                          &
!                                             ONE, u_row_dev + (l_row_beg - 1) *                 &
!                                             size_of_datatype, 1)
!                 endif
              if (wantDebug) call obj%timer%stop("cublas")

            else
              !perform multiplication by stripes - it is faster than by blocks, since we call cublas with
              !larger matrices. In general, however, this algorithm is very simmilar to the one with CPU
              do i=0,(istep-2)/tile_size
                  l_col_beg = i*l_cols_per_tile+1
                  l_col_end = min(l_cols,(i+1)*l_cols_per_tile)
                  if(l_col_end<l_col_beg) cycle

                  l_row_beg = 1
                  l_row_end = min(l_rows,(i+1)*l_rows_per_tile)

                  a_offset = ((l_row_beg-1) + (l_col_beg - 1) * lda) * &
                          size_of_datatype

                  call cublas_PRECISION_GEMV(BLAS_TRANS_OR_CONJ, &
                              l_row_end-l_row_beg+1, l_col_end-l_col_beg+1, &
                              ONE, a_dev + a_offset, lda,  &
                              v_row_dev + (l_row_beg - 1) * size_of_datatype, 1,  &
                              ONE, u_col_dev + (l_col_beg - 1) * size_of_datatype, 1)
              enddo

              do i=0,(istep-2)/tile_size
                  l_col_beg = i*l_cols_per_tile+1
                  l_col_end = min(l_cols,(i+1)*l_cols_per_tile)
                  if(l_col_end<l_col_beg) cycle

                  l_row_beg = 1
                  l_row_end = min(l_rows,i*l_rows_per_tile)

                  a_offset = ((l_row_beg-1) + (l_col_beg - 1) * lda) * &
                          size_of_datatype
                  if (isSkewsymmetric) then
                     call cublas_PRECISION_GEMV('N', l_row_end-l_row_beg+1, l_col_end-l_col_beg+1, &
                                 -ONE, a_dev + a_offset, lda, &
                                 v_col_dev + (l_col_beg - 1) * size_of_datatype,1, &
                                 ONE, u_row_dev + (l_row_beg - 1) * size_of_datatype, 1)
                  else
                     call cublas_PRECISION_GEMV('N', l_row_end-l_row_beg+1, l_col_end-l_col_beg+1, &
                                 ONE, a_dev + a_offset, lda, &
                                 v_col_dev + (l_col_beg - 1) * size_of_datatype,1, &
                                 ONE, u_row_dev + (l_row_beg - 1) * size_of_datatype, 1)
                  endif
              enddo
            end if !multiplication as one block / per stripes

            successCUDA = cuda_memcpy(int(loc(u_col(1)),kind=c_intptr_t), &
                          u_col_dev, l_cols * size_of_datatype, cudaMemcpyDeviceToHost)
            check_memcpy_cuda("tridiag: u_col_dev 1", successCUDA)

            successCUDA = cuda_memcpy(int(loc(u_row(1)),kind=c_intptr_t), &
                          u_row_dev, l_rows * size_of_datatype, cudaMemcpyDeviceToHost)
            check_memcpy_cuda("tridiag: u_row_dev 1", successCUDA)
          endif

!              call PRECISION_SYMV('U', l_cols,  &
!                         1.d0, a_mat, ubound(a_mat,1),  &
!                         v_row, 1,  &
!                         0.d0, u_col, 1)

!            endif ! useGPU

#ifdef WITH_OPENMP
!$OMP END PARALLEL
          call obj%timer%stop("OpenMP parallel")

          do i=0,max_threads-1
            u_col(1:l_cols) = u_col(1:l_cols) + uc_p(1:l_cols,i)
            u_row(1:l_rows) = u_row(1:l_rows) + ur_p(1:l_rows,i)
          enddo
#endif /* WITH_OPENMP */

          ! second calculate (VU**T + UV**T)*v part of (A + VU**T + UV**T)*v
          if (n_stored_vecs > 0) then
            if (wantDebug) call obj%timer%start("blas")
#if REALCASE == 1
            call PRECISION_GEMV('T',     &
#endif
#if COMPLEXCASE == 1
            call PRECISION_GEMV('C',     &
#endif
                                int(l_rows,kind=BLAS_KIND), int(2*n_stored_vecs,kind=BLAS_KIND),   &
                                ONE, vu_stored_rows, int(ubound(vu_stored_rows,dim=1),kind=BLAS_KIND),   &
                                v_row,  1_BLAS_KIND, ZERO, aux, 1_BLAS_KIND)

            call PRECISION_GEMV('N', int(l_cols,kind=BLAS_KIND), int(2*n_stored_vecs,kind=BLAS_KIND),   &
                                ONE, uv_stored_cols, int(ubound(uv_stored_cols,dim=1),kind=BLAS_KIND),   &
                                aux, 1_BLAS_KIND, ONE, u_col,  1_BLAS_KIND)
            if (wantDebug) call obj%timer%stop("blas")
          endif

        endif  ! (l_rows>0 .and. l_cols>0)

        ! Sum up all u_row(:) parts along rows and add them to the u_col(:) parts
        ! on the processors containing the diagonal
        ! This is only necessary if u_row has been calculated, i.e. if the
        ! global tile size is smaller than the global remaining matrix

        if (tile_size < istep-1) then

          call elpa_reduce_add_vectors_&
          &MATH_DATATYPE&
          &_&
          &PRECISION &
          (obj, u_row, ubound(u_row,dim=1), mpi_comm_rows, u_col, ubound(u_col,dim=1), &
           mpi_comm_cols, istep-1, 1, nblk, max_threads)

        endif

        ! Sum up all the u_col(:) parts, transpose u_col -> u_row

        if (l_cols>0) then
          tmp(1:l_cols) = u_col(1:l_cols)
#ifdef WITH_MPI
          if (wantDebug) call obj%timer%start("mpi_communication")
          call mpi_allreduce(tmp, u_col, int(l_cols,kind=MPI_KIND), MPI_MATH_DATATYPE_PRECISION,    &
                             MPI_SUM, int(mpi_comm_rows,kind=MPI_KIND), mpierr)
          if (wantDebug) call obj%timer%stop("mpi_communication")
#else /* WITH_MPI */
          u_col = tmp
#endif /* WITH_MPI */
        endif
        if (isSkewsymmetric) then
           call elpa_transpose_vectors_ss_&
           &MATH_DATATYPE&
           &_&
           &PRECISION &
           (obj, u_col, ubound(u_col,dim=1), mpi_comm_cols, u_row, ubound(u_row,dim=1), &
            mpi_comm_rows, 1, istep-1, 1, nblk, max_threads)
        else
           call elpa_transpose_vectors_&
           &MATH_DATATYPE&
           &_&
           &PRECISION &
           (obj, u_col, ubound(u_col,dim=1), mpi_comm_cols, u_row, ubound(u_row,dim=1), &
            mpi_comm_rows, 1, istep-1, 1, nblk, max_threads)
        endif

        ! calculate u**T * v (same as v**T * (A + VU**T + UV**T) * v )
        x = 0
        if (l_cols>0)  &
           x = dot_product(v_col(1:l_cols),u_col(1:l_cols))

#ifdef WITH_MPI
        if (wantDebug) call obj%timer%start("mpi_communication")