MPI_STUBS
Dummy MPI Library

MPI_STUBS is a FORTRAN90 library which provides "stub" versions of some MPI routines.

MPI_STUBS is intended to include stubs for the most commonly called MPI routines. Most of the stub routines don't do anything. In a few cases, where it makes sense, they do some simple action or return a value that is appropriate for the serial processing case.

MPI_STUBS can be used as a convenience, when a real MPI implementation is not available, and the user simply wants to test-compile a code. It may also be useful in those occasions when a code has been so carefully written that it will still execute correctly on a single processor.

MPI_STUBS is based on a similar package supplied as part of the LAMMPS program, which allow that program to be compiled, linked and run on a single processor machine, although it is normally intended for parallel execution.

Licensing:

The computer code and data files described and made available on this web page are distributed under the GNU LGPL license.

Related Data and Programs:

HEAT_MPI is a FORTRAN90 program which solves the 1D Time Dependent Heat Equation using MPI.

LAMMPS is a FORTRAN77 program which uses MPI for a molecular dynamics simulation. A serial version can be created by linking the program with the MPI_STUBS library.

MPI is an library of message passing routines which enables parallel processing on a variety of machine architectures, and with a varying number of processors.

MPI_INTRODUCTION is a one page introduction to MPI.

MPI_MORE_INFO contains a list of references, web sites, examples and tutorials on MPI.

MPI_STUBS is available in a C version and a C++ version and a FORTRAN77 version and a FORTRAN90 version.

PETSC is a scientific programming library for parallel programming, which requires MPI in order to run.

PLTMG_SINGLE is a FORTRAN77 which implements the finite element method, and can be compiled and run with the MPI library.

QUAD_MPI is a FORTRAN90 program which approximates an integral using a quadrature rule, and carries out the computation in parallel using MPI.

RANDOM_MPI, a FORTRAN90 program which demonstrates one way to generate the same sequence of random numbers for both sequential execution and parallel execution under MPI.

SATISFIABILITY_MPI is a FORTRAN90 program which demonstrates, for a particular circuit, an exhaustive search for solutions of the circuit satisfiability problem, using MPI to carry out the calculation in parallel.

Reference:

1. William Gropp, Steven Huss-Lederman, Andrew Lumsdaine, Ewing Lusk, Bill Nitzberg, William Saphir, Marc Snir,
   MPI: The Complete Reference,
   Volume II: The MPI-2 Extensions,
   Second Edition,
   MIT Press, 1998,
   ISBN13: 978-0-262-57123-4,
   LC: QA76.642.M65.

Source Code:

• mpi_stubs.f90, the source code.
• mpi_stubs_f90.h, the "include" file.
• mpi_stubs.csh, commands to compile the source code.

Examples and Tests:

BUFFON_LAPLACE is an "embarrassingly parallel" Monte Carlo simulation of the Buffon-Laplace needle dropping process.

• buffon_laplace.f90, a sample calling program.
• buffon_laplace.csh, commands to compile and run the sample program.
• buffon_laplace_output.txt, the output file.

HELLO is a simple program that says "Hello, world!".

• hello.f90, a sample calling program.
• hello.csh, commands to compile and run the sample program.
• hello_output.txt, the output file.

QUADRATURE is a program that estimates an integral using the random sampling.

• quadrature.f90, a sample calling program.
• quadrature.csh, commands to compile and run the sample program.
• quadrature_output.txt, the output file.

List of Routines:

• MPI_ABORT shuts down the processes in a given communicator.
• MPI_ALLGATHER gathers data from all the processes in a communicator.
• MPI_ALLGATHERV gathers data from all the processes in a communicator.
• MPI_ALLREDUCE carries out a reduction operation.
• MPI_BARRIER forces processes within a communicator to wait together.
• MPI_BCAST broadcasts data from one process to all others.
• MPI_BSEND sends data from one process to another, using buffering.
• MPI_CART_CREATE creates a communicator for a Cartesian topology.
• MPI_CART_GET returns the "Cartesian coordinates" of the calling process.
• MPI_CART_SHIFT finds the destination and source for Cartesian shifts.
• MPI_COMM_DUP duplicates a communicator.
• MPI_COMM_FREE "frees" a communicator.
• MPI_COMM_RANK reports the rank of the calling process.
• MPI_COMM_SIZE reports the number of processes in a communicator.
• MPI_COMM_SPLIT splits up a communicator based on a key.
• MPI_COPY_DOUBLE copies a double precision vector.
• MPI_COPY_INTEGER copies an integer vector.
• MPI_COPY_REAL copies a real vector.
• MPI_FINALIZE shuts down the MPI library.
• MPI_GET_COUNT reports the actual number of items transmitted.
• MPI_INIT initializes the MPI library.
• MPI_IRECV receives data from another process.
• MPI_ISEND sends data from one process to another using nonblocking transmission.
• MPI_RECV receives data from another process within a communicator.
• MPI_REDUCE carries out a reduction operation.
• MPI_REDUCE_DOUBLE_PRECISION carries out a reduction operation on double precision values.
• MPI_REDUCE_INTEGER carries out a reduction operation on integers.
• MPI_REDUCE_REAL carries out a reduction operation on reals.
• MPI_REDUCE_SCATTER collects a message of the same length from each process.
• MPI_RSEND "ready sends" data from one process to another.
• MPI_SEND sends data from one process to another.
• MPI_WAIT waits for an I/O request to complete.
• MPI_WAITALL waits until all I/O requests have completed.
• MPI_WAITANY waits until one I/O requests has completed.
• MPI_WTICK returns the number of seconds per clock tick.
• MPI_WTIME returns the elapsed wall clock time.
• TIMESTRING writes the current YMDHMS date into a string.

You can go up one level to the FORTRAN90 source codes.