Tpetra_Details_EquilibrationInfo.hpp

Go to the documentation of this file.

// @HEADER
// ***********************************************************************
//
//          Tpetra: Templated Linear Algebra Services Package
//                 Copyright (2008) Sandia Corporation
//
// Under the terms of Contract DE-AC04-94AL85000 with Sandia Corporation,
// the U.S. Government retains certain rights in this software.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// 1. Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//
// 2. Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// 3. Neither the name of the Corporation nor the names of the
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY SANDIA CORPORATION "AS IS" AND ANY
// EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
// PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL SANDIA CORPORATION OR THE
// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF
// LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING
// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Questions? Contact Michael A. Heroux (maherou@sandia.gov)
//
// ************************************************************************
// @HEADER

#ifndef TPETRA_DETAILS_EQUILIBRATIONINFO_HPP
#define TPETRA_DETAILS_EQUILIBRATIONINFO_HPP


#include "TpetraCore_config.h"
#include "Kokkos_ArithTraits.hpp"
#include "Kokkos_Core.hpp"

namespace Tpetra {
namespace Details {

template<class ScalarType, class DeviceType>
struct EquilibrationInfo {
  using val_type = typename Kokkos::ArithTraits<ScalarType>::val_type;
  using mag_type = typename Kokkos::ArithTraits<val_type>::mag_type;
  using device_type = typename DeviceType::device_type;
  using host_device_type = typename Kokkos::View<mag_type*, device_type>::HostMirror::device_type;
  using HostMirror = EquilibrationInfo<val_type, host_device_type>;

  EquilibrationInfo () :
    foundInf (false),
    foundNan (false),
    foundZeroDiag (false),
    foundZeroRowNorm (false)
  {}

  EquilibrationInfo (const std::size_t lclNumRows,
                     const std::size_t lclNumCols,
                     const bool assumeSymmetric_) :
    rowNorms (Kokkos::View<mag_type*, device_type> ("rowNorms", lclNumRows)),
    rowDiagonalEntries (Kokkos::View<val_type*, device_type> ("rowDiagonalEntries", lclNumRows)),
    colNorms (Kokkos::View<mag_type*, device_type> ("colNorms", lclNumCols)),
    colDiagonalEntries (Kokkos::View<val_type*, device_type> ("colDiagonalEntries",
                                                              assumeSymmetric_ ?
                                                              std::size_t (0) :
                                                              lclNumCols)),
    rowScaledColNorms (Kokkos::View<mag_type*, device_type> ("rowScaledColNorms",
                                                             assumeSymmetric_ ?
                                                             std::size_t (0) :
                                                             lclNumCols)),
    assumeSymmetric (assumeSymmetric_),
    foundInf (false),
    foundNan (false),
    foundZeroDiag (false),
    foundZeroRowNorm (false)
  {}

  EquilibrationInfo (const Kokkos::View<mag_type*, device_type>& rowNorms_,
                     const Kokkos::View<val_type*, device_type>& rowDiagonalEntries_,
                     const Kokkos::View<mag_type*, device_type>& colNorms_,
                     const Kokkos::View<val_type*, device_type>& colDiagonalEntries_,
                     const Kokkos::View<mag_type*, device_type>& rowScaledColNorms_,
                     const bool assumeSymmetric_,
                     const bool foundInf_,
                     const bool foundNan_,
                     const bool foundZeroDiag_,
                     const bool foundZeroRowNorm_) :
    rowNorms (rowNorms_),
    rowDiagonalEntries (rowDiagonalEntries_),
    colNorms (colNorms_),
    colDiagonalEntries (colDiagonalEntries_),
    rowScaledColNorms (rowScaledColNorms_),
    assumeSymmetric (assumeSymmetric_),
    foundInf (foundInf_),
    foundNan (foundNan_),
    foundZeroDiag (foundZeroDiag_),
    foundZeroRowNorm (foundZeroRowNorm_)
  {}

  template<class SrcDeviceType>
  void
  assign (const EquilibrationInfo<ScalarType, SrcDeviceType>& src)
  {
    Kokkos::deep_copy (rowNorms, src.rowNorms);
    Kokkos::deep_copy (rowDiagonalEntries, src.rowDiagonalEntries);
    Kokkos::deep_copy (colNorms, src.colNorms);
    if (src.colDiagonalEntries.extent (0) == 0) {
      colDiagonalEntries =
        Kokkos::View<val_type*, device_type> ("colDiagonalEntries", 0);
    }
    else {
      Kokkos::deep_copy (colDiagonalEntries, src.colDiagonalEntries);
    }
    if (src.rowScaledColNorms.extent (0) == 0) {
      rowScaledColNorms =
        Kokkos::View<mag_type*, device_type> ("rowScaledColNorms", 0);
    }
    else {
      Kokkos::deep_copy (rowScaledColNorms, src.rowScaledColNorms);
    }

    assumeSymmetric = src.assumeSymmetric;
    foundInf = src.foundInf;
    foundNan = src.foundNan;
    foundZeroDiag = src.foundZeroDiag;
    foundZeroRowNorm = src.foundZeroRowNorm;
  }

  typename EquilibrationInfo<val_type, device_type>::HostMirror
  createMirrorView ()
  {
    auto rowNorms_h = Kokkos::create_mirror_view (rowNorms);
    auto rowDiagonalEntries_h = Kokkos::create_mirror_view (rowDiagonalEntries);
    auto colNorms_h = Kokkos::create_mirror_view (colNorms);
    auto colDiagonalEntries_h = Kokkos::create_mirror_view (colDiagonalEntries);
    auto rowScaledColNorms_h = Kokkos::create_mirror_view (rowScaledColNorms);

    return HostMirror {rowNorms_h, rowDiagonalEntries_h, colNorms_h,
        colDiagonalEntries_h, rowScaledColNorms_h, assumeSymmetric,
        foundInf, foundNan, foundZeroDiag, foundZeroRowNorm};
  }

  // We call a row a "diagonally dominant row" if the absolute value
  // of the diagonal entry is >= the sum of the absolute values of the
  // off-diagonal entries.  The row norm is the sum of those two
  // things, so this means diagAbsVal >= rowNorm - diagAbsVal.  Ditto
  // for a column.

  Kokkos::View<mag_type*, device_type> rowNorms;

  Kokkos::View<val_type*, device_type> rowDiagonalEntries;

  Kokkos::View<mag_type*, device_type> colNorms;

  Kokkos::View<val_type*, device_type> colDiagonalEntries;

  Kokkos::View<mag_type*, device_type> rowScaledColNorms;

  bool assumeSymmetric;

  bool foundInf;

  bool foundNan;

  bool foundZeroDiag;

  bool foundZeroRowNorm;
};

} // namespace Details
} // namespace Tpetra

#endif // TPETRA_DETAILS_EQUILIBRATIONINFO_HPP