/usr/include/trilinos/Thyra_SpmdVectorSpaceDefaultBase_decl.hpp is in libtrilinos-thyra-dev 12.10.1-3.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 | // @HEADER
// ***********************************************************************
//
// Thyra: Interfaces and Support for Abstract Numerical Algorithms
// Copyright (2004) Sandia Corporation
//
// Under terms of Contract DE-AC04-94AL85000, there is a non-exclusive
// license for use of this work by or on behalf of the U.S. Government.
//
// 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 Roscoe A. Bartlett (bartlettra@ornl.gov)
//
// ***********************************************************************
// @HEADER
#ifndef THYRA_SPMD_VECTOR_SPACE_BASE_DECL_HPP
#define THYRA_SPMD_VECTOR_SPACE_BASE_DECL_HPP
#include "Thyra_SpmdVectorSpaceBase.hpp"
#include "Thyra_ScalarProdVectorSpaceBase_decl.hpp"
namespace Thyra {
/** \brief Base <tt>%VectorSpaceBase</tt> class for all SPMD vector spaces
* with contiguous local-to-global indexing.
*
* See <tt>SpmdVectorSpaceBase</tt> for details on what this class represents
* in an abstract way.
*
* <b>Notes to subclass developers:</b>
*
* The pure virtual methods <tt>mpiComm()</tt>, <tt>localSubDim()</tt> and
* <tt>createMember()</tt> are the only functions that must be overridden.
*
* If <tt>this</tt> this is in an uninitialized state then
* <tt>localSubDim()</tt> should return <tt>0</tt>.
*
* It should never be necessary to override the virtual functions
* <tt>mapCode()</tt> and <tt>isCompatible()</tt> as these functions
* have very good and very general implementations.
*
* If optimized implementations of multi-vectors can be supported,
* then the <tt>createMembers()</tt> method should also be overridden.
*
* This class defines a very general default implementation for
* <tt>smallVecSpcFcty()</tt> that returns a
* <tt>DefaultSpmdVectorSpaceFactory</tt> object which in turn can be used to
* create <tt>DefaultSpmdVectorSpace</tt> objects. These are the vector space
* objects that are used by the domain space of multi-vectors created by
* <tt>createMembers()</tt>. The class <tt>DefaultSpmdVectorSpace</tt>
* creates <tt>DefaultSpmdVector</tt> and <tt>DefaultSpmdMultiVector</tt>
* objects. This implementation of <tt>smallVecSpcFcty()</tt> is very general
* should be very appropriate for many different concrete implementations.
*
* <b>Note:</b> It is very important that subclasses call the
* <tt>updateState()</tt> function whenever the state of <tt>*this</tt>
* changes in a way that might affect the behavior of any of the public member
* functions. For example, if a different value of <tt>localSubDim()</tt>
* will be returned the next time it is called by a client, then
* <tt>%updateState()</tt> needs to be called by the subclass. External
* clients should never need to worry about this function and that is why
* <tt>%updateState()</tt> is declared protected.
*
* \ingroup Thyra_Op_Vec_adapters_Spmd_support_grp
*/
template<class Scalar>
class SpmdVectorSpaceDefaultBase
: virtual public SpmdVectorSpaceBase<Scalar>
, virtual public ScalarProdVectorSpaceBase<Scalar>
{
public:
/** \brief . */
SpmdVectorSpaceDefaultBase();
/** @name Overridden from SpmdVectorSpaceBase */
//@{
/** \brief .
*
* This method has a default implementation which just assigns this offset
* based on counting up <tt>localSubDim()</tt> on each process and then
* setting <tt>localOffset()</tt> by the rank of the process. For
* example, if there are 5 elements in process 0 and 4 elements in process
* rank, then <tt>localOffset</tt> on each of these processes will be set
* as: <tt>localOffset=0</tt> on process 0, <tt>localOffset=5</tt> on
* process 1, <tt>localOffset=9</tt> on process 2 and so on.
*/
Ordinal localOffset() const;
/** \brief .
*
* This method takes the data <tt>getComm()</tt>, <tt>numProc</tt> (where
* <tt>numProc=this->getComm()->size()</tt>), <tt>localOffset()</tt> or
* <tt>localSubDim()</tt> on each process and then uses it to compute a
* value for <tt>mapCode</tt> (using a single global reduction if
* <tt>numProc > 1</tt>) which is returned from this function.
*
* The value returned from this default implementation of this
* method must not be changed or this approach breaks down. The
* only reason for overriding this method is for the subclass to
* be alerted of <em>when</em> this method is called but not
* <em>what</em> is returned from this method. If a subclass
* developer does not understand what this means then <b>don't</b>
* override this method!
*
* The default implementation will always return <tt>return >
* 0</tt> (unless <tt>this</tt> is uninitialized) so that if this
* method is overridden to return <tt>return <= </tt> then this is
* a flag that the underlying vector map does not satisfy the
* assumptions of this vector space interface and vectors that are
* in <tt>*this</tt> vector space can not collaborate with other
* Spmd-based vector implementations.
*/
Ordinal mapCode() const;
/** \brief Returns true if vector space is locally replicated space. */
bool isLocallyReplicated() const;
//@}
/** @name Overridden form Teuchos::Describable */
//@{
/** \brief . */
std::string description() const;
//@}
/** @name Overridden from VectorSpaceBase */
//@{
/** \brief Returns the sum of the local number of elements on every process.
*/
Ordinal dim() const;
/** \brief Returns a <tt>DefaultSpmdVectorSpaceFactory</tt> object that has been given <tt>getComm()</tt>.
*/
Teuchos::RCP< const VectorSpaceFactoryBase<Scalar> > smallVecSpcFcty() const;
/** \brief Checks the general compatibility of parallel (or serial on one
* process) Spmd-based vector spaces.
*
* @return Returns true if <tt>*this</tt> and <tt>vecSpace</tt>
* are both serial in-core vectors or if <tt>vecSpc</tt> is of
* type <tt>SpmdVectorSpaceDefaultBase<Scalar></tt> and both <tt>*this</tt>
* and <tt>vecSpc</tt> have the same Spmd communicators and the same
* mapping of elements to processes.
*
* Postconditions:<ul>
* <li> <tt>return = ( this->hasInCoreView() &&
* vecSpc->hasInCoreView() ) || ( (mpiVecSpc = dynamic_cast<const
* SpmdVectorSpaceDefaultBase<Scalar>*>(&vecSpc)) && this->getComm() ==
* mpiVecSpc->getComm() && this->mapCode() ==
* mpiVecSpc->getComm())</tt>.
*
* </ul>
*
* If the mapping of vector elements to processes is not as
* described
* <A HREF="classThyra_1_1SpmdVectorSpaceBase.html#SpmdVectorSpaceBase_Vector_layout">above</A>
* then this method should be overridden in a way that is specific
* to the vector implementation.
*/
bool isCompatible(const VectorSpaceBase<Scalar>& vecSpc) const;
//@}
protected:
/** \brief This function must be called whenever the state of
* <tt>this</tt> changes and some internal state must be updated.
*
* @param globalDim [in] If <tt>globalDim > 0</tt> then this determines the
* global dimension of the vector space. If
* <tt>globalDim==this->localSubDim()</tt> and <tt>this->localSubDim()</tt>
* is the same on every process and is > 0 then this is a locally replicated
* vector space. If <tt>globalDim < 0</tt> then the global dimension is
* computed using a global reduction. If
* <tt>this->getComm()->size()==1</tt> then this argument is ignored.
*
* @param isLocallyReplicated [in] If <tt>isLocallyReplicated==true</tt>,
* then tt>globalDim</tt> must be equal to <tt>this->localSubDim()</tt> on
* every process but no global reduction will be done to verify this. If
* <tt>false</tt>, an global reduction is used to automatically determine if
* <tt>globalDim == this->localSubDim()</tt> on every process and therefore
* construct a locally replicated vector space.
*
* Note that calling this function may involve one or more global reductions
* being called if this is parallel vector space so it should only be called
* when needed by subclasses and RCP should be use to share a constructed VS
* with lots of clients.
*
* Usually, this operation only needs to be called once for every
* *new* parallel vector space constructed and very few parallel
* vector spaces will be created per application usually.
*/
virtual void updateState(const Ordinal globalDim,
const bool isLocallyReplicated = false);
private:
// //////////////////////////////////////
// Private data members
Ordinal mapCode_; // < 0 is a flag that everything needs initialized
Ordinal defaultLocalOffset_;
Ordinal defaultGlobalDim_;
Ordinal localSubDim_;
bool isLocallyReplicated_;
Teuchos::RCP< const VectorSpaceFactoryBase<Scalar> > smallVecSpcFcty_;
}; // end class SpmdVectorSpaceDefaultBase
} // end namespace Thyra
#endif // THYRA_SPMD_VECTOR_SPACE_BASE_DECL_HPP
|