PRACE NPT

      Program mod2g

! **********************************************************************

! *** This program is part of the EuroBen Benchmark                  ***

! *** Copyright: EuroBen Group p/o                                   ***

! ***            Utrecht University, Computational Physics Dept.     ***

! ***            P.O. Box 80195                                      ***

! ***            3508 TD Utrecht                                     ***

! ***            The Netherlands                                     ***

! ***                                                                ***

! *** Author of this program: Aad J. van der Steen                   ***

! *** Contributed:            Autumn 1998                            ***

! *** Improved communicaton:  Spring 2005                            ***

! **********************************************************************

!  Version 2.1

!

! --- Purpose of this module

!     ----------------------

!     This program performs 2-D Haar wavelet transforms as used in data

!     compression. Data are generated internally and, subsequently, a

!     wavelet analysis and sythesis are performed. The time and Mflop/s

!     for an analysis/synthesis pair are determined.

! ----------------------------------------------------------------------

      Use                      numerics

      Use                      dist_module ! Contains # of proc.s &

                                           ! proc. numbers.

      Implicit                 None

      Include                 'mpif.h'

      Real(l_), Allocatable :: a1(:,:), a2(:,:), w1(:,:), w2(:,:)

      Real(l_)              :: speed, time, timeg

      Real(l_)              :: ctime, ctimeg, pctime

      Real(l_)              :: start_time, end_time

      Integer, Allocatable  :: actsiz(:,:), base(:,:)

      Integer               :: hsize, i, ierr, n1, n2, nflops, nrep,

     &                         vsize

      Logical               :: ok, okg

! ----------------------------------------------------------------------

      Call csetup

      start_time = MPI_Wtime()

      If ( me == 0 ) Call state( 'mod2g   ' )

      Open( 1, File = 'mod2g.in' )

      If ( me == 0 ) Print 1000, nodes

   10 Read( 1, *, End = 30 ) n1, n2, nrep

      Allocate( actsiz(0:nodes-1,2), base(0:nodes-1,2) )

      Call sizoff( n1, n2, actsiz, base )

      vsize  = Maxval( actsiz(:,1) )

      hsize  = Maxval( actsiz(:,2) )

      Allocate( a1(n1,hsize), a2(n2,vsize), w1(n1,hsize), w2(n2,vsize) )

      nflops = 5*(n2*(n1 - 1) + n1*(n2 - 1))

      Call gendat( a1, n1, actsiz(me,2), actsiz, base )

      time = MPI_Wtime()

      ctime = 0.0_l_

      Do i = 1, nrep

         Call anhr2d( a1, a2, w1, w2, n1, n2, vsize, hsize, ctime )

         Call synhr2d( a1, a2, w1, w2, n1, n2, vsize, hsize, ctime )

      End Do

      time = MPI_Wtime() - time

      Call MPI_Reduce( time, timeg, 1, MPI_Real8, MPI_Max, 0,

     &                 MPI_Comm_World, ierr )

      Call MPI_Reduce( ctime, ctimeg, 1, MPI_Real8, MPI_Max, 0,

     &                 MPI_Comm_World, ierr )

      pctime = 100.0_l_*(ctimeg/timeg)

      timeg  = timeg/Real( nrep, l_ )

      ctimeg = ctimeg/Real( nrep, l_ )

      speed = ( 1.0E-6_l_*nflops )/Max( timeg, 1.0E-9_l_ )

      Call gendat( w1, n1, actsiz(me,2), actsiz, base )    

      ok = .TRUE.

      Call check( w1, a1, hsize, n1, actsiz, ok )

      Call MPI_Reduce( ok, okg, 1, MPI_Logical, MPI_Land, 0,

     &                 MPI_Comm_World, ierr )

      If ( me == 0 ) Print 1010, n1, n2, timeg, speed, ctimeg, pctime,

     &                           okg    

      Deallocate( a1, a2, w1, w2 , actsiz, base )

      Go To 10

   30 If ( me == 0 ) Print 1020

      end_time =  MPI_Wtime() - start_time

      If (me == 0) Then

         Write(6,22) 'Walltime: ', end_time, " s"

 22      Format(A,F9.3,A)

      End If

      Call MPI_Finalize( ierr )

! ----------------------------------------------------------------------

 1000 Format( //, 'Haar wavelet transform test: No. of procs. = ', i3/

     &        '-------------------------------------------------------',

     &        '------------------------'/

     &        '    Order   | Exec. Time  |    Speed    | Comm. time  |',

     &        '  Frac.  | Correctness |'/

     &        '  n1 |  n2  |     (s)     |  (Mflop/s)  |     (s)     |',

     &        '   (%)   |      OK     |'/

     &        '-------------------------------------------------------',

     &        '------------------------' )

 1010 Format( i4,' | ',  i4,' |', g12.5,' |', g13.5, '|', g13.5,

     &        '| ', f7.3, ' |    ', l3, '      |' )

 1020 Format( '-------------------------------------------------------',

     &        '------------------------' )

! ----------------------------------------------------------------------

      End Program mod2g