<?xml version="1.0" encoding="utf-8" ?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <meta name="generator" content="Docutils 0.6: http://docutils.sourceforge.net/" /> <title>Parallel BGL MPI BSP Process Group</title> <link rel="stylesheet" href="../../../../rst.css" type="text/css" /> </head> <body> <div class="document" id="logo-mpi-bsp-process-group"> <h1 class="title"><a class="reference external" href="http://www.osl.iu.edu/research/pbgl"><img align="middle" alt="Parallel BGL" class="align-middle" src="pbgl-logo.png" /></a> MPI BSP Process Group</h1> <!-- Copyright (C) 2004-2009 The Trustees of Indiana University. Use, modification and distribution is subject to the Boost Software License, Version 1.0. (See accompanying file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) --> <div class="contents topic" id="contents"> <p class="topic-title first">Contents</p> <ul class="simple"> <li><a class="reference internal" href="#introduction" id="id1">Introduction</a></li> <li><a class="reference internal" href="#where-defined" id="id2">Where Defined</a></li> <li><a class="reference internal" href="#reference" id="id3">Reference</a></li> </ul> </div> <div class="section" id="introduction"> <h1><a class="toc-backref" href="#id1">Introduction</a></h1> <p>The MPI <tt class="docutils literal"><span class="pre">mpi_process_group</span></tt> is an implementation of the <a class="reference external" href="process_group.html">process group</a> interface using the Message Passing Interface (MPI). It is the primary process group used in the Parallel BGL at this time.</p> </div> <div class="section" id="where-defined"> <h1><a class="toc-backref" href="#id2">Where Defined</a></h1> <p>Header <tt class="docutils literal"><span class="pre"><boost/graph/distributed/mpi_process_group.hpp></span></tt></p> </div> <div class="section" id="reference"> <h1><a class="toc-backref" href="#id3">Reference</a></h1> <pre class="literal-block"> namespace boost { namespace graph { namespace distributed { class mpi_process_group { public: typedef boost::mpi::communicator communicator_type; // Process group constructors mpi_process_group(communicator_type comm = communicator_type()); mpi_process_group(std::size_t num_headers, std::size_t buffer_size, communicator_type comm = communicator_type()); mpi_process_group(); mpi_process_group(const mpi_process_group&, boost::parallel::attach_distributed_object); // Triggers template<typename Type, typename Handler> void trigger(int tag, const Handler& handler); template<typename Type, typename Handler> void trigger_with_reply(int tag, const Handler& handler); trigger_receive_context trigger_context() const; // Helper operations void poll(); mpi_process_group base() const; }; // Process query int process_id(const mpi_process_group&); int num_processes(const mpi_process_group&); // Message transmission template<typename T> void send(const mpi_process_group& pg, int dest, int tag, const T& value); template<typename T> void receive(const mpi_process_group& pg, int source, int tag, T& value); optional<std::pair<int, int> > probe(const mpi_process_group& pg); // Synchronization void synchronize(const mpi_process_group& pg); // Out-of-band communication template<typename T> void send_oob(const mpi_process_group& pg, int dest, int tag, const T& value); template<typename T, typename U> void send_oob_with_reply(const mpi_process_group& pg, int dest, int tag, const T& send_value, U& receive_value); template<typename T> void receive_oob(const mpi_process_group& pg, int source, int tag, T& value); } } } </pre> <p>Since the <tt class="docutils literal"><span class="pre">mpi_process_group</span></tt> is an implementation of the <a class="reference external" href="process_group.html">process group</a> interface, we omit the description of most of the functions in the prototype. Two constructors need special mentioning:</p> <pre class="literal-block"> mpi_process_group(communicator_type comm = communicator_type()); </pre> <p>The constructor can take an optional MPI communicator. As default a communicator constructed from MPI_COMM_WORLD is used.</p> <pre class="literal-block"> mpi_process_group(std::size_t num_headers, std::size_t buffer_size, communicator_type comm = communicator_type()); </pre> <p>For performance fine tuning the maximum number of headers in a message batch (num_headers) and the maximum combined size of batched messages (buffer_size) can be specified. The maximum message size of a batch is 16*num_headers+buffer_size. Sensible default values have been found by optimizing a typical application on a cluster with Ethernet network, and are num_header=64k and buffer_size=1MB, for a total maximum batches message size of 2MB.</p> <hr class="docutils" /> <p>Copyright (C) 2007 Douglas Gregor</p> <p>Copyright (C) 2007 Matthias Troyer</p> </div> </div> <div class="footer"> <hr class="footer" /> Generated on: 2009-05-31 00:22 UTC. Generated by <a class="reference external" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference external" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source. </div> </body> </html>