2009-05-27 07:03:18 +04:00
dnl -*- shell-script -*-
dnl
dnl Copyright (c) 2004-2007 The Trustees of Indiana University and Indiana
dnl University Research and Technology
dnl Corporation. All rights reserved.
dnl Copyright (c) 2004-2005 The University of Tennessee and The University
dnl of Tennessee Research Foundation. All rights
dnl reserved.
dnl Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
dnl University of Stuttgart. All rights reserved.
dnl Copyright (c) 2004-2005 The Regents of the University of California.
dnl All rights reserved.
2014-01-11 18:06:59 +04:00
dnl Copyright (c) 2006-2014 Cisco Systems, Inc. All rights reserved.
2009-05-27 07:03:18 +04:00
dnl Copyright (c) 2007 Sun Microsystems, Inc. All rights reserved.
dnl Copyright (c) 2009 IBM Corporation. All rights reserved.
dnl Copyright (c) 2009 Los Alamos National Security, LLC. All rights
dnl reserved.
2012-06-27 05:28:28 +04:00
dnl Copyright (c) 2009 Oak Ridge National Labs. All rights reserved.
2014-01-04 02:35:20 +04:00
dnl Copyright (c) 2013 Intel, Inc. All rights reserved.
2009-05-27 07:03:18 +04:00
dnl
dnl $COPYRIGHT$
dnl
dnl Additional copyrights may follow
dnl
dnl $HEADER$
dnl
AC_DEFUN([OMPI_CONFIGURE_OPTIONS],[
2014-05-06 02:35:23 +04:00
opal_show_subtitle "OMPI Configuration options"
2009-05-27 07:03:18 +04:00
2014-01-04 02:35:20 +04:00
#
# Disable MPI layer?
#
AC_ARG_ENABLE([mpi],
[AC_HELP_STRING([--disable-mpi],
[Disable building the MPI layer (default:enabled)])])
2009-05-27 07:03:18 +04:00
#
2012-02-21 00:30:43 +04:00
# Do we want to enable MPI interface warnings (e.g. deprecated
# functionality and others)?
2009-05-27 07:03:18 +04:00
#
2012-02-21 00:30:43 +04:00
# This was disabled by default in v1.5, but will be enabled by default
# in 1.7 and beyond.
2009-05-27 07:03:18 +04:00
#
AC_MSG_CHECKING([if want compile-time warnings inside of mpi.h])
AC_ARG_ENABLE(mpi-interface-warning,
AC_HELP_STRING([--enable-mpi-interface-warning],
2012-03-23 02:59:59 +04:00
[enable compile-time warnings when deprecated MPI functions are used (default: enabled)]))
2012-02-21 00:30:43 +04:00
if test "$enable_mpi_interface_warning" != "no"; then
2009-05-27 07:03:18 +04:00
AC_MSG_RESULT([yes])
OMPI_WANT_MPI_INTERFACE_WARNING=1
else
AC_MSG_RESULT([no])
OMPI_WANT_MPI_INTERFACE_WARNING=0
fi
AC_DEFINE_UNQUOTED([OMPI_WANT_MPI_INTERFACE_WARNING], [$OMPI_WANT_MPI_INTERFACE_WARNING],
2012-02-21 00:30:43 +04:00
[Enable warnings when using deprecated MPI functions])
2009-05-27 07:03:18 +04:00
#
# Sparse Groups
#
AC_MSG_CHECKING([if want sparse process groups])
AC_ARG_ENABLE(sparse-groups,
AC_HELP_STRING([--enable-sparse-groups],
[enable sparse process groups (default: not enabled)]))
if test "$enable_sparse_groups" = "yes"; then
AC_MSG_RESULT([yes])
GROUP_SPARSE=1
else
AC_MSG_RESULT([no])
GROUP_SPARSE=0
fi
AC_DEFINE_UNQUOTED([OMPI_GROUP_SPARSE],$GROUP_SPARSE,
[Wether we want sparse process groups])
#
# Do we want to enable peruse interface?
#
AC_MSG_CHECKING([if want peruse support])
AC_ARG_ENABLE(peruse,
AC_HELP_STRING([--enable-peruse],
2009-05-27 07:24:10 +04:00
[enable PERUSE interface (default: disabled)]))
2009-05-27 07:03:18 +04:00
if test "$enable_peruse" = "yes"; then
AC_MSG_RESULT([yes])
WANT_PERUSE=1
else
AC_MSG_RESULT([no])
WANT_PERUSE=0
fi
AC_DEFINE_UNQUOTED([OMPI_WANT_PERUSE],
[$WANT_PERUSE],
[if the peruse interface should be enabled])
AM_CONDITIONAL(WANT_PERUSE, test "$WANT_PERUSE" = "1")
#
== Highlights ==
1. New mpifort wrapper compiler: you can utilize mpif.h, use mpi, and use mpi_f08 through this one wrapper compiler
1. mpif77 and mpif90 still exist, but are sym links to mpifort and may be removed in a future release
1. The mpi module has been re-implemented and is significantly "mo' bettah"
1. The mpi_f08 module offers many, many improvements over mpif.h and the mpi module
This stuff is coming from a VERY long-lived mercurial branch (3 years!); it'll almost certainly take a few SVN commits and a bunch of testing before I get it correctly committed to the SVN trunk.
== More details ==
Craig Rasmussen and I have been working with the MPI-3 Fortran WG and Fortran J3 committees for a long, long time to make a prototype MPI-3 Fortran bindings implementation. We think we're at a stable enough state to bring this stuff back to the trunk, with the goal of including it in OMPI v1.7.
Special thanks go out to everyone who has been incredibly patient and helpful to us in this journey:
* Rolf Rabenseifner/HLRS (mastermind/genius behind the entire MPI-3 Fortran effort)
* The Fortran J3 committee
* Tobias Burnus/gfortran
* Tony !Goetz/Absoft
* Terry !Donte/Oracle
* ...and probably others whom I'm forgetting :-(
There's still opportunities for optimization in the mpi_f08 implementation, but by and large, it is as far along as it can be until Fortran compilers start implementing the new F08 dimension(..) syntax.
Note that gfortran is currently unsupported for the mpi_f08 module and the new mpi module. gfortran users will a) fall back to the same mpi module implementation that is in OMPI v1.5.x, and b) not get the new mpi_f08 module. The gfortran maintainers are actively working hard to add the necessary features to support both the new mpi_f08 module and the new mpi module implementations. This will take some time.
As mentioned above, ompi/mpi/f77 and ompi/mpi/f90 no longer exist. All the fortran bindings implementations have been collated under ompi/mpi/fortran; each implementation has its own subdirectory:
{{{
ompi/mpi/fortran/
base/ - glue code
mpif-h/ - what used to be ompi/mpi/f77
use-mpi-tkr/ - what used to be ompi/mpi/f90
use-mpi-ignore-tkr/ - new mpi module implementation
use-mpi-f08/ - new mpi_f08 module implementation
}}}
There's also a prototype 6-function-MPI implementation under use-mpi-f08-desc that emulates the new F08 dimension(..) syntax that isn't fully available in Fortran compilers yet. We did that to prove it to ourselves that it could be done once the compilers fully support it. This directory/implementation will likely eventually replace the use-mpi-f08 version.
Other things that were done:
* ompi_info grew a few new output fields to describe what level of Fortran support is included
* Existing Fortran examples in examples/ were renamed; new mpi_f08 examples were added
* The old Fortran MPI libraries were renamed:
* libmpi_f77 -> libmpi_mpifh
* libmpi_f90 -> libmpi_usempi
* The configury for Fortran was consolidated and significantly slimmed down. Note that the F77 env variable is now IGNORED for configure; you should only use FC. Example:
{{{
shell$ ./configure CC=icc CXX=icpc FC=ifort ...
}}}
All of this work was done in a Mercurial branch off the SVN trunk, and hosted at Bitbucket. This branch has got to be one of OMPI's longest-running branches. Its first commit was Tue Apr 07 23:01:46 2009 -0400 -- it's over 3 years old! :-) We think we've pulled in all relevant changes from the OMPI trunk (e.g., Fortran implementations of the new MPI-3 MPROBE stuff for mpif.h, use mpi, and use mpi_f08, and the recent Fujitsu Fortran patches).
I anticipate some instability when we bring this stuff into the trunk, simply because it touches a LOT of code in the MPI layer in the OMPI code base. We'll try our best to make it as pain-free as possible, but please bear with us when it is committed.
This commit was SVN r26283.
2012-04-18 19:57:29 +04:00
# Fortran MPI bindings
2009-05-27 07:03:18 +04:00
#
== Highlights ==
1. New mpifort wrapper compiler: you can utilize mpif.h, use mpi, and use mpi_f08 through this one wrapper compiler
1. mpif77 and mpif90 still exist, but are sym links to mpifort and may be removed in a future release
1. The mpi module has been re-implemented and is significantly "mo' bettah"
1. The mpi_f08 module offers many, many improvements over mpif.h and the mpi module
This stuff is coming from a VERY long-lived mercurial branch (3 years!); it'll almost certainly take a few SVN commits and a bunch of testing before I get it correctly committed to the SVN trunk.
== More details ==
Craig Rasmussen and I have been working with the MPI-3 Fortran WG and Fortran J3 committees for a long, long time to make a prototype MPI-3 Fortran bindings implementation. We think we're at a stable enough state to bring this stuff back to the trunk, with the goal of including it in OMPI v1.7.
Special thanks go out to everyone who has been incredibly patient and helpful to us in this journey:
* Rolf Rabenseifner/HLRS (mastermind/genius behind the entire MPI-3 Fortran effort)
* The Fortran J3 committee
* Tobias Burnus/gfortran
* Tony !Goetz/Absoft
* Terry !Donte/Oracle
* ...and probably others whom I'm forgetting :-(
There's still opportunities for optimization in the mpi_f08 implementation, but by and large, it is as far along as it can be until Fortran compilers start implementing the new F08 dimension(..) syntax.
Note that gfortran is currently unsupported for the mpi_f08 module and the new mpi module. gfortran users will a) fall back to the same mpi module implementation that is in OMPI v1.5.x, and b) not get the new mpi_f08 module. The gfortran maintainers are actively working hard to add the necessary features to support both the new mpi_f08 module and the new mpi module implementations. This will take some time.
As mentioned above, ompi/mpi/f77 and ompi/mpi/f90 no longer exist. All the fortran bindings implementations have been collated under ompi/mpi/fortran; each implementation has its own subdirectory:
{{{
ompi/mpi/fortran/
base/ - glue code
mpif-h/ - what used to be ompi/mpi/f77
use-mpi-tkr/ - what used to be ompi/mpi/f90
use-mpi-ignore-tkr/ - new mpi module implementation
use-mpi-f08/ - new mpi_f08 module implementation
}}}
There's also a prototype 6-function-MPI implementation under use-mpi-f08-desc that emulates the new F08 dimension(..) syntax that isn't fully available in Fortran compilers yet. We did that to prove it to ourselves that it could be done once the compilers fully support it. This directory/implementation will likely eventually replace the use-mpi-f08 version.
Other things that were done:
* ompi_info grew a few new output fields to describe what level of Fortran support is included
* Existing Fortran examples in examples/ were renamed; new mpi_f08 examples were added
* The old Fortran MPI libraries were renamed:
* libmpi_f77 -> libmpi_mpifh
* libmpi_f90 -> libmpi_usempi
* The configury for Fortran was consolidated and significantly slimmed down. Note that the F77 env variable is now IGNORED for configure; you should only use FC. Example:
{{{
shell$ ./configure CC=icc CXX=icpc FC=ifort ...
}}}
All of this work was done in a Mercurial branch off the SVN trunk, and hosted at Bitbucket. This branch has got to be one of OMPI's longest-running branches. Its first commit was Tue Apr 07 23:01:46 2009 -0400 -- it's over 3 years old! :-) We think we've pulled in all relevant changes from the OMPI trunk (e.g., Fortran implementations of the new MPI-3 MPROBE stuff for mpif.h, use mpi, and use mpi_f08, and the recent Fujitsu Fortran patches).
I anticipate some instability when we bring this stuff into the trunk, simply because it touches a LOT of code in the MPI layer in the OMPI code base. We'll try our best to make it as pain-free as possible, but please bear with us when it is committed.
This commit was SVN r26283.
2012-04-18 19:57:29 +04:00
AC_MSG_CHECKING([if want Fortran MPI bindings])
AC_ARG_ENABLE(mpi-fortran,
AC_HELP_STRING([--enable-mpi-fortran],
2014-01-23 05:28:55 +04:00
[specify which Fortran MPI bindings to build: all (or yes), none (or no), mpifh (build only mpif.h support), usempi (build mpif.h and the mpi module), or usempif08 (build mpifh, the mpi module, and the mpi_f08 module) (default: "all" if Fortran compiler found)]))
2014-01-23 06:27:48 +04:00
OMPI_FORTRAN_USER_REQUESTED=0
2014-01-23 05:28:55 +04:00
case "x$enable_mpi_fortran" in
2014-01-23 06:27:48 +04:00
x)
AC_MSG_RESULT([yes (all/default)])
OMPI_WANT_FORTRAN_MPIFH_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPI_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS=1
;;
2014-01-23 05:28:55 +04:00
xyes|xall)
AC_MSG_RESULT([yes (all)])
2014-01-23 06:27:48 +04:00
OMPI_FORTRAN_USER_REQUESTED=1
2014-01-23 05:28:55 +04:00
OMPI_WANT_FORTRAN_MPIFH_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPI_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS=1
;;
xno|xnone)
AC_MSG_RESULT([no (none)])
OMPI_WANT_FORTRAN_MPIFH_BINDINGS=0
OMPI_WANT_FORTRAN_USEMPI_BINDINGS=0
OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS=0
;;
xmpifh)
AC_MSG_RESULT([yes (mpif.h)])
2014-01-23 06:27:48 +04:00
OMPI_FORTRAN_USER_REQUESTED=1
2014-01-23 05:28:55 +04:00
OMPI_WANT_FORTRAN_MPIFH_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPI_BINDINGS=0
OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS=0
;;
xusempi)
AC_MSG_RESULT([yes (mpif.h, mpi module)])
2014-01-23 06:27:48 +04:00
OMPI_FORTRAN_USER_REQUESTED=1
2014-01-23 05:28:55 +04:00
OMPI_WANT_FORTRAN_MPIFH_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPI_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS=0
;;
xusempif08)
AC_MSG_RESULT([yes (mpif.h, mpi and mpi_f08 modules)])
2014-01-23 06:27:48 +04:00
OMPI_FORTRAN_USER_REQUESTED=1
2014-01-23 05:28:55 +04:00
OMPI_WANT_FORTRAN_MPIFH_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPI_BINDINGS=1
OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS=1
;;
*)
AC_MSG_RESULT([unknown: $binding])
AC_MSG_WARN([--enable-mpi-fortran only one of the following values: all, none, mpifh, usempi, or usempif08])
AC_MSG_ERROR([Cannot continue])
;;
esac
== Highlights ==
1. New mpifort wrapper compiler: you can utilize mpif.h, use mpi, and use mpi_f08 through this one wrapper compiler
1. mpif77 and mpif90 still exist, but are sym links to mpifort and may be removed in a future release
1. The mpi module has been re-implemented and is significantly "mo' bettah"
1. The mpi_f08 module offers many, many improvements over mpif.h and the mpi module
This stuff is coming from a VERY long-lived mercurial branch (3 years!); it'll almost certainly take a few SVN commits and a bunch of testing before I get it correctly committed to the SVN trunk.
== More details ==
Craig Rasmussen and I have been working with the MPI-3 Fortran WG and Fortran J3 committees for a long, long time to make a prototype MPI-3 Fortran bindings implementation. We think we're at a stable enough state to bring this stuff back to the trunk, with the goal of including it in OMPI v1.7.
Special thanks go out to everyone who has been incredibly patient and helpful to us in this journey:
* Rolf Rabenseifner/HLRS (mastermind/genius behind the entire MPI-3 Fortran effort)
* The Fortran J3 committee
* Tobias Burnus/gfortran
* Tony !Goetz/Absoft
* Terry !Donte/Oracle
* ...and probably others whom I'm forgetting :-(
There's still opportunities for optimization in the mpi_f08 implementation, but by and large, it is as far along as it can be until Fortran compilers start implementing the new F08 dimension(..) syntax.
Note that gfortran is currently unsupported for the mpi_f08 module and the new mpi module. gfortran users will a) fall back to the same mpi module implementation that is in OMPI v1.5.x, and b) not get the new mpi_f08 module. The gfortran maintainers are actively working hard to add the necessary features to support both the new mpi_f08 module and the new mpi module implementations. This will take some time.
As mentioned above, ompi/mpi/f77 and ompi/mpi/f90 no longer exist. All the fortran bindings implementations have been collated under ompi/mpi/fortran; each implementation has its own subdirectory:
{{{
ompi/mpi/fortran/
base/ - glue code
mpif-h/ - what used to be ompi/mpi/f77
use-mpi-tkr/ - what used to be ompi/mpi/f90
use-mpi-ignore-tkr/ - new mpi module implementation
use-mpi-f08/ - new mpi_f08 module implementation
}}}
There's also a prototype 6-function-MPI implementation under use-mpi-f08-desc that emulates the new F08 dimension(..) syntax that isn't fully available in Fortran compilers yet. We did that to prove it to ourselves that it could be done once the compilers fully support it. This directory/implementation will likely eventually replace the use-mpi-f08 version.
Other things that were done:
* ompi_info grew a few new output fields to describe what level of Fortran support is included
* Existing Fortran examples in examples/ were renamed; new mpi_f08 examples were added
* The old Fortran MPI libraries were renamed:
* libmpi_f77 -> libmpi_mpifh
* libmpi_f90 -> libmpi_usempi
* The configury for Fortran was consolidated and significantly slimmed down. Note that the F77 env variable is now IGNORED for configure; you should only use FC. Example:
{{{
shell$ ./configure CC=icc CXX=icpc FC=ifort ...
}}}
All of this work was done in a Mercurial branch off the SVN trunk, and hosted at Bitbucket. This branch has got to be one of OMPI's longest-running branches. Its first commit was Tue Apr 07 23:01:46 2009 -0400 -- it's over 3 years old! :-) We think we've pulled in all relevant changes from the OMPI trunk (e.g., Fortran implementations of the new MPI-3 MPROBE stuff for mpif.h, use mpi, and use mpi_f08, and the recent Fujitsu Fortran patches).
I anticipate some instability when we bring this stuff into the trunk, simply because it touches a LOT of code in the MPI layer in the OMPI code base. We'll try our best to make it as pain-free as possible, but please bear with us when it is committed.
This commit was SVN r26283.
2012-04-18 19:57:29 +04:00
AS_IF([test $OMPI_WANT_FORTRAN_MPIFH_BINDINGS -eq 1 -o \
$OMPI_WANT_FORTRAN_USEMPI_BINDINGS -eq 1 -o \
$OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS -eq 1],
[OMPI_WANT_FORTRAN_BINDINGS=1],
[OMPI_WANT_FORTRAN_BINDINGS=0])
2009-05-27 07:03:18 +04:00
#
# MPI profiling
#
AC_MSG_CHECKING([if want PMPI])
AC_ARG_ENABLE(mpi-profile,
AC_HELP_STRING([--enable-mpi-profile],
[enable MPI profiling (default: enabled)]))
if test "$enable_mpi_profile" != "no"; then
AC_MSG_RESULT([yes])
WANT_MPI_PROFILING=1
MPIF_H_PMPI_W_FUNCS=", PMPI_WTICK, PMPI_WTIME"
else
AC_MSG_RESULT([no])
WANT_MPI_PROFILING=0
MPIF_H_PMPI_W_FUNCS=
fi
AC_SUBST(MPIF_H_PMPI_W_FUNCS)
2013-08-19 18:48:23 +04:00
AM_CONDITIONAL(WANT_MPI_PROFILING, test "$WANT_MPI_PROFILING" = 1)
2009-05-27 07:03:18 +04:00
#
# C++
#
AC_MSG_CHECKING([if want C++ bindings])
AC_ARG_ENABLE(mpi-cxx,
AC_HELP_STRING([--enable-mpi-cxx],
2013-01-07 23:58:19 +04:00
[enable C++ MPI bindings (default: disabled)]))
if test "$enable_mpi_cxx" = "yes"; then
2009-05-27 07:03:18 +04:00
AC_MSG_RESULT([yes])
WANT_MPI_CXX_SUPPORT=1
else
AC_MSG_RESULT([no])
WANT_MPI_CXX_SUPPORT=0
fi
AC_MSG_CHECKING([if want MPI::SEEK_SET support])
AC_ARG_ENABLE([mpi-cxx-seek],
[AC_HELP_STRING([--enable-mpi-cxx-seek],
[enable support for MPI::SEEK_SET, MPI::SEEK_END, and MPI::SEEK_POS in C++ bindings (default: enabled)])])
if test "$enable_mpi_cxx_seek" != "no" ; then
AC_MSG_RESULT([yes])
OMPI_WANT_MPI_CXX_SEEK=1
else
AC_MSG_RESULT([no])
OMPI_WANT_MPI_CXX_SEEK=0
fi
AC_DEFINE_UNQUOTED([OMPI_WANT_MPI_CXX_SEEK], [$OMPI_WANT_MPI_CXX_SEEK],
[do we want to try to work around C++ bindings SEEK_* issue?])
#
# Do we want to disable MPI parameter checking at run-time?
#
AC_MSG_CHECKING([if want run-time MPI parameter checking])
AC_ARG_WITH(mpi-param-check,
AC_HELP_STRING([--with-mpi-param-check(=VALUE)],
[behavior of MPI function parameter checking. Valid values are: always, never, runtime. If --with-mpi-param-check is specified with no VALUE argument, it is equivalent to a VALUE of "always"; --without-mpi-param-check is equivalent to "never" (default: runtime).]))
mpi_param_check=ompi_mpi_param_check
if test "$with_mpi_param_check" = "no" -o \
"$with_mpi_param_check" = "never"; then
mpi_param_check=0
2012-04-27 05:09:37 +04:00
ompi_param_check=0
2009-05-27 07:03:18 +04:00
AC_MSG_RESULT([never])
elif test "$with_mpi_param_check" = "yes" -o \
"$with_mpi_param_check" = "always"; then
mpi_param_check=1
2012-04-27 05:09:37 +04:00
ompi_param_check=1
2009-05-27 07:03:18 +04:00
AC_MSG_RESULT([always])
elif test "$with_mpi_param_check" = "runtime" -o \
-z "$with_mpi_params_check"; then
2012-04-27 05:09:37 +04:00
ompi_param_check=1
2009-05-27 07:03:18 +04:00
AC_MSG_RESULT([runtime])
else
AC_MSG_RESULT([unknown])
AC_MSG_WARN([*** Unrecognized --with-mpi-param-check value])
AC_MSG_WARN([*** See "configure --help" output])
AC_MSG_WARN([*** Defaulting to "runtime"])
fi
AC_DEFINE_UNQUOTED(MPI_PARAM_CHECK, $mpi_param_check,
[Whether we want to check MPI parameters always, never, or decide at run-time])
2012-04-27 05:09:37 +04:00
AC_DEFINE_UNQUOTED(OMPI_PARAM_CHECK, $ompi_param_check,
[Whether we want to check MPI parameters never or possible (an integer constant)])
2009-05-27 07:03:18 +04:00
#
== Highlights ==
1. New mpifort wrapper compiler: you can utilize mpif.h, use mpi, and use mpi_f08 through this one wrapper compiler
1. mpif77 and mpif90 still exist, but are sym links to mpifort and may be removed in a future release
1. The mpi module has been re-implemented and is significantly "mo' bettah"
1. The mpi_f08 module offers many, many improvements over mpif.h and the mpi module
This stuff is coming from a VERY long-lived mercurial branch (3 years!); it'll almost certainly take a few SVN commits and a bunch of testing before I get it correctly committed to the SVN trunk.
== More details ==
Craig Rasmussen and I have been working with the MPI-3 Fortran WG and Fortran J3 committees for a long, long time to make a prototype MPI-3 Fortran bindings implementation. We think we're at a stable enough state to bring this stuff back to the trunk, with the goal of including it in OMPI v1.7.
Special thanks go out to everyone who has been incredibly patient and helpful to us in this journey:
* Rolf Rabenseifner/HLRS (mastermind/genius behind the entire MPI-3 Fortran effort)
* The Fortran J3 committee
* Tobias Burnus/gfortran
* Tony !Goetz/Absoft
* Terry !Donte/Oracle
* ...and probably others whom I'm forgetting :-(
There's still opportunities for optimization in the mpi_f08 implementation, but by and large, it is as far along as it can be until Fortran compilers start implementing the new F08 dimension(..) syntax.
Note that gfortran is currently unsupported for the mpi_f08 module and the new mpi module. gfortran users will a) fall back to the same mpi module implementation that is in OMPI v1.5.x, and b) not get the new mpi_f08 module. The gfortran maintainers are actively working hard to add the necessary features to support both the new mpi_f08 module and the new mpi module implementations. This will take some time.
As mentioned above, ompi/mpi/f77 and ompi/mpi/f90 no longer exist. All the fortran bindings implementations have been collated under ompi/mpi/fortran; each implementation has its own subdirectory:
{{{
ompi/mpi/fortran/
base/ - glue code
mpif-h/ - what used to be ompi/mpi/f77
use-mpi-tkr/ - what used to be ompi/mpi/f90
use-mpi-ignore-tkr/ - new mpi module implementation
use-mpi-f08/ - new mpi_f08 module implementation
}}}
There's also a prototype 6-function-MPI implementation under use-mpi-f08-desc that emulates the new F08 dimension(..) syntax that isn't fully available in Fortran compilers yet. We did that to prove it to ourselves that it could be done once the compilers fully support it. This directory/implementation will likely eventually replace the use-mpi-f08 version.
Other things that were done:
* ompi_info grew a few new output fields to describe what level of Fortran support is included
* Existing Fortran examples in examples/ were renamed; new mpi_f08 examples were added
* The old Fortran MPI libraries were renamed:
* libmpi_f77 -> libmpi_mpifh
* libmpi_f90 -> libmpi_usempi
* The configury for Fortran was consolidated and significantly slimmed down. Note that the F77 env variable is now IGNORED for configure; you should only use FC. Example:
{{{
shell$ ./configure CC=icc CXX=icpc FC=ifort ...
}}}
All of this work was done in a Mercurial branch off the SVN trunk, and hosted at Bitbucket. This branch has got to be one of OMPI's longest-running branches. Its first commit was Tue Apr 07 23:01:46 2009 -0400 -- it's over 3 years old! :-) We think we've pulled in all relevant changes from the OMPI trunk (e.g., Fortran implementations of the new MPI-3 MPROBE stuff for mpif.h, use mpi, and use mpi_f08, and the recent Fujitsu Fortran patches).
I anticipate some instability when we bring this stuff into the trunk, simply because it touches a LOT of code in the MPI layer in the OMPI code base. We'll try our best to make it as pain-free as possible, but please bear with us when it is committed.
This commit was SVN r26283.
2012-04-18 19:57:29 +04:00
# Do we want the prototype "use mpi_f08" implementation that uses
# Fortran descriptors?
#
AC_MSG_CHECKING([which 'use mpi_f08' implementation to use])
AC_ARG_ENABLE(mpi-f08-subarray-prototype,
AC_HELP_STRING([--enable-mpi-f08-subarray-prototype],
[Use the PROTOTYPE and SEVERLY FUNCTIONALITY-LIMITED Fortran 08 'use mpi_f08' implementation that supports subarrrays (via Fortran descriptors). This option will disable the normal 'use mpi_f08' implementation and *only* build the prototype implementation.]))
OMPI_BUILD_FORTRAN_F08_SUBARRAYS=0
AS_IF([test $OMPI_WANT_FORTRAN_USEMPIF08_BINDINGS -eq 0],
[AC_MSG_RESULT([none (use mpi_f08 disabled)])],
[AS_IF([test "$enable_mpi_f08_subarray_prototype" = "yes"],
[OMPI_BUILD_FORTRAN_F08_SUBARRAYS=1
2014-06-19 17:39:56 +04:00
AC_MSG_RESULT([extra crispy (subarray prototype)])
AC_MSG_WARN([Sorry, the subarray prototype is no longer available])
AC_MSG_WARN([Contact your favorite OMPI developer and ask for it to be re-enabled])
AC_MSG_ERROR([Cannot continue])],
== Highlights ==
1. New mpifort wrapper compiler: you can utilize mpif.h, use mpi, and use mpi_f08 through this one wrapper compiler
1. mpif77 and mpif90 still exist, but are sym links to mpifort and may be removed in a future release
1. The mpi module has been re-implemented and is significantly "mo' bettah"
1. The mpi_f08 module offers many, many improvements over mpif.h and the mpi module
This stuff is coming from a VERY long-lived mercurial branch (3 years!); it'll almost certainly take a few SVN commits and a bunch of testing before I get it correctly committed to the SVN trunk.
== More details ==
Craig Rasmussen and I have been working with the MPI-3 Fortran WG and Fortran J3 committees for a long, long time to make a prototype MPI-3 Fortran bindings implementation. We think we're at a stable enough state to bring this stuff back to the trunk, with the goal of including it in OMPI v1.7.
Special thanks go out to everyone who has been incredibly patient and helpful to us in this journey:
* Rolf Rabenseifner/HLRS (mastermind/genius behind the entire MPI-3 Fortran effort)
* The Fortran J3 committee
* Tobias Burnus/gfortran
* Tony !Goetz/Absoft
* Terry !Donte/Oracle
* ...and probably others whom I'm forgetting :-(
There's still opportunities for optimization in the mpi_f08 implementation, but by and large, it is as far along as it can be until Fortran compilers start implementing the new F08 dimension(..) syntax.
Note that gfortran is currently unsupported for the mpi_f08 module and the new mpi module. gfortran users will a) fall back to the same mpi module implementation that is in OMPI v1.5.x, and b) not get the new mpi_f08 module. The gfortran maintainers are actively working hard to add the necessary features to support both the new mpi_f08 module and the new mpi module implementations. This will take some time.
As mentioned above, ompi/mpi/f77 and ompi/mpi/f90 no longer exist. All the fortran bindings implementations have been collated under ompi/mpi/fortran; each implementation has its own subdirectory:
{{{
ompi/mpi/fortran/
base/ - glue code
mpif-h/ - what used to be ompi/mpi/f77
use-mpi-tkr/ - what used to be ompi/mpi/f90
use-mpi-ignore-tkr/ - new mpi module implementation
use-mpi-f08/ - new mpi_f08 module implementation
}}}
There's also a prototype 6-function-MPI implementation under use-mpi-f08-desc that emulates the new F08 dimension(..) syntax that isn't fully available in Fortran compilers yet. We did that to prove it to ourselves that it could be done once the compilers fully support it. This directory/implementation will likely eventually replace the use-mpi-f08 version.
Other things that were done:
* ompi_info grew a few new output fields to describe what level of Fortran support is included
* Existing Fortran examples in examples/ were renamed; new mpi_f08 examples were added
* The old Fortran MPI libraries were renamed:
* libmpi_f77 -> libmpi_mpifh
* libmpi_f90 -> libmpi_usempi
* The configury for Fortran was consolidated and significantly slimmed down. Note that the F77 env variable is now IGNORED for configure; you should only use FC. Example:
{{{
shell$ ./configure CC=icc CXX=icpc FC=ifort ...
}}}
All of this work was done in a Mercurial branch off the SVN trunk, and hosted at Bitbucket. This branch has got to be one of OMPI's longest-running branches. Its first commit was Tue Apr 07 23:01:46 2009 -0400 -- it's over 3 years old! :-) We think we've pulled in all relevant changes from the OMPI trunk (e.g., Fortran implementations of the new MPI-3 MPROBE stuff for mpif.h, use mpi, and use mpi_f08, and the recent Fujitsu Fortran patches).
I anticipate some instability when we bring this stuff into the trunk, simply because it touches a LOT of code in the MPI layer in the OMPI code base. We'll try our best to make it as pain-free as possible, but please bear with us when it is committed.
This commit was SVN r26283.
2012-04-18 19:57:29 +04:00
[AC_MSG_RESULT([regular (no subarray support)])])
])
AC_DEFINE_UNQUOTED([OMPI_BUILD_FORTRAN_F08_SUBARRAYS],
[$OMPI_BUILD_FORTRAN_F08_SUBARRAYS],
[Whether we built the 'use mpi_f08' prototype subarray-based implementation or not (i.e., whether to build the use-mpi-f08-desc prototype or the regular use-mpi-f08 implementation)])
2009-05-27 07:03:18 +04:00
])dnl
