ports/openmpi
ports/openmpi
1
1
Форкнуть 0
Код Релизы Активность
bdf9aace69
openmpi/config/opal_config_pthreads.m4

736 строки
20 KiB
Plaintext
Исходник Обычный вид История

Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
dnl
dnl Copyright (c) 2004-2005 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.
== 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
dnl Copyright (c) 2012 Cisco Systems, Inc. All rights reserved.
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
dnl $COPYRIGHT$
dnl
dnl Additional copyrights may follow
dnl
dnl $HEADER$
dnl
dnl OMPI_CONFIG_POSIX_THREADS()
dnl
dnl Configure posix threads, setting the following variables (but
dnl not calling AC_SUBST on them).
# ********************************************************************
#
# Internal macros - do not call from outside OMPI_CONFIG_POSIX_THREADS
#
# ********************************************************************
AC_DEFUN([OMPI_INTL_PTHREAD_TRY_LINK], [
# BEGIN: OMPI_INTL_PTHREAD_TRY_LINK
#
# Make sure that we can run a small application in C or C++, which
# ever is the current language. Do make sure that C or C++ is the
# current language.
#
# As long as this is not being run....
# pthread_t may be anything from an int to a struct -- init with self-tid.
#
ompi_config_pthreads.m4: fix to compile with Clang 3.4 Clang 3.4 static analysis is now smart enough that it refuses to compile our POSIX threading tests because we simply passed 0 for several arguments (the test is checking for the linker presence of symbols, so the NULL arguments didn't matter). Specifically, Clang's anti-NULL compile-time checks now refuse to compile this bogus code. Hence, these 2 configure test codes now have real variables and pointers so that clang will compile it properly. Without this fix, configure determines that there is no POSIX threading support, and the osc/sm component fails to compile (which is a different defect I'll be filing shortly). As such, we need this fix in 1.8.1. Also, change from the deprecated AC_TRY_LINK to AC_LINK_IFELSE. cmr=v1.8.1:reviewer=dgoodell This commit was SVN r31458.
2014-04-19 17:13:06 +04:00
AC_LINK_IFELSE([AC_LANG_SOURCE([[
#include <pthread.h>
int i = 3;
pthread_t me, newthread;
void cleanup_routine(void *foo);
void *thread_main(void *foo);
void cleanup_routine(void *foo) { i = 4; }
void *thread_main(void *foo) { i = 2; return (void*) &i; }
int main(int argc, char* argv[])
{
pthread_attr_t attr;
me = pthread_self();
pthread_attr_init(&attr);
pthread_cleanup_push(cleanup_routine, 0);
pthread_create(&newthread, &attr, thread_main, 0);
pthread_join(newthread, 0);
pthread_cleanup_pop(0);
return 0;
}]])],
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
[$1], [$2])
# END: OMPI_INTL_PTHREAD_TRY_LINK
])dnl
== 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_DEFUN([OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN], [
# BEGIN: OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
#
# Make sure that we can run a small application in Fortran, with
# pthreads living in a C object file
# Fortran module
cat > conftestf.f <<EOF
program fpthread
call pthreadtest
end
EOF
# C module
if test -f conftest.h; then
ompi_conftest_h="#include \"conftest.h\""
else
ompi_conftest_h=""
fi
cat > conftest.c <<EOF
#include <stdio.h>
#include <stdlib.h>
#include <pthread.h>
$ompi_conftest_h
#ifdef __cplusplus
extern "C" {
#endif
ompi_config_pthreads.m4: fix to compile with Clang 3.4 Clang 3.4 static analysis is now smart enough that it refuses to compile our POSIX threading tests because we simply passed 0 for several arguments (the test is checking for the linker presence of symbols, so the NULL arguments didn't matter). Specifically, Clang's anti-NULL compile-time checks now refuse to compile this bogus code. Hence, these 2 configure test codes now have real variables and pointers so that clang will compile it properly. Without this fix, configure determines that there is no POSIX threading support, and the osc/sm component fails to compile (which is a different defect I'll be filing shortly). As such, we need this fix in 1.8.1. Also, change from the deprecated AC_TRY_LINK to AC_LINK_IFELSE. cmr=v1.8.1:reviewer=dgoodell This commit was SVN r31458.
2014-04-19 17:13:06 +04:00
int i = 3;
pthread_t me, newthread;
void cleanup_routine(void *foo);
void *thread_main(void *foo);
void pthreadtest_f(void);
void cleanup_routine(void *foo) { i = 4; }
void *thread_main(void *foo) { i = 2; return (void*) &i; }
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
void pthreadtest_f(void)
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
{
ompi_config_pthreads.m4: fix to compile with Clang 3.4 Clang 3.4 static analysis is now smart enough that it refuses to compile our POSIX threading tests because we simply passed 0 for several arguments (the test is checking for the linker presence of symbols, so the NULL arguments didn't matter). Specifically, Clang's anti-NULL compile-time checks now refuse to compile this bogus code. Hence, these 2 configure test codes now have real variables and pointers so that clang will compile it properly. Without this fix, configure determines that there is no POSIX threading support, and the osc/sm component fails to compile (which is a different defect I'll be filing shortly). As such, we need this fix in 1.8.1. Also, change from the deprecated AC_TRY_LINK to AC_LINK_IFELSE. cmr=v1.8.1:reviewer=dgoodell This commit was SVN r31458.
2014-04-19 17:13:06 +04:00
pthread_attr_t attr;
me = pthread_self();
pthread_attr_init(&attr);
pthread_cleanup_push(cleanup_routine, 0);
pthread_create(&newthread, &attr, thread_main, 0);
pthread_join(&newthread, 0);
pthread_cleanup_pop(0);
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
}
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
void pthreadtest(void)
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
{ pthreadtest_f(); }
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
void pthreadtest_(void)
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
{ pthreadtest_f(); }
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
void pthreadtest__(void)
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
{ pthreadtest_f(); }
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
void PTHREADTEST(void)
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
{ pthreadtest_f(); }
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
#ifdef __cplusplus
}
#endif
EOF
# Try the compile
Per http://www.open-mpi.org/community/lists/devel/2010/11/8671.php, change a bunch of OMPI_<foo> names to OPAL_<foo>. This commit was SVN r24046.
2010-11-13 02:22:11 +03:00
OPAL_LOG_COMMAND(
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
[$CC $CFLAGS -I. -c conftest.c],
Per http://www.open-mpi.org/community/lists/devel/2010/11/8671.php, change a bunch of OMPI_<foo> names to OPAL_<foo>. This commit was SVN r24046.
2010-11-13 02:22:11 +03:00
OPAL_LOG_COMMAND(
== 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
[$FC $FCFLAGS conftestf.f conftest.o -o conftest $LDFLAGS $LIBS],
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
[HAPPY=1],
[HAPPY=0]),
[HAPPY=0])
if test "$HAPPY" = "1"; then
$1
else
Per http://www.open-mpi.org/community/lists/devel/2010/11/8671.php, change a bunch of OMPI_<foo> names to OPAL_<foo>. This commit was SVN r24046.
2010-11-13 02:22:11 +03:00
OPAL_LOG_MSG([here is the C program:], 1)
OPAL_LOG_FILE([conftest.c])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
if test -f conftest.h; then
Per http://www.open-mpi.org/community/lists/devel/2010/11/8671.php, change a bunch of OMPI_<foo> names to OPAL_<foo>. This commit was SVN r24046.
2010-11-13 02:22:11 +03:00
OPAL_LOG_MSG([here is contest.h:], 1)
OPAL_LOG_FILE([conftest.h])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
fi
Per http://www.open-mpi.org/community/lists/devel/2010/11/8671.php, change a bunch of OMPI_<foo> names to OPAL_<foo>. This commit was SVN r24046.
2010-11-13 02:22:11 +03:00
OPAL_LOG_MSG([here is the fortran program:], 1)
OPAL_LOG_FILE([conftestf.f])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
$2
fi
unset HAPPY ompi_conftest_h
rm -rf conftest*
== 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
# END: OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
])dnl
# ********************************************************************
#
# Try to compile thread support without any special flags
#
# ********************************************************************
AC_DEFUN([OMPI_INTL_POSIX_THREADS_PLAIN_C], [
#
# C compiler
#
if test "$ompi_pthread_c_success" = "0"; then
AC_MSG_CHECKING([if C compiler and POSIX threads work as is])
if test "$HAVE_POSIX_THREADS" = "1" ; then
run_this_test=1
else
case "${host_cpu}-${host_os}" in
*solaris*)
AC_MSG_RESULT([no - Solaris, not checked])
run_this_test=0
;;
*-aix* | *-freebsd*)
if test "`echo $CPPFLAGS | $GREP 'D_THREAD_SAFE'`" = ""; then
Correct some subtle PTRHEAD_ typos (should be PTHREAD_) in config/ompi_config_pthreads.m4. Terry pointed them out. Mostly just aix/freebsd. This commit was SVN r24324.
2011-01-29 00:05:40 +03:00
PTHREAD_CPPFLAGS="-D_THREAD_SAFE"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
CPPFLAGS="$CPPFLAGS $PTHREAD_CPPFLAGS"
fi
run_this_test=1
;;
*)
if test "`echo $CPPFLAGS | $GREP 'D_REENTRANT'`" = ""; then
PTHREAD_CPPFLAGS="-D_REENTRANT"
CPPFLAGS="$CPPFLAGS $PTHREAD_CPPFLAGS"
fi
run_this_test=1
;;
esac
fi
if test "$run_this_test" = "1" ; then
AC_LANG_PUSH(C)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_c_success=1,
ompi_pthread_c_success=0)
AC_LANG_POP(C)
if test "$ompi_pthread_c_success" = "1"; then
AC_MSG_RESULT([yes])
else
PTHREAD_CPPFLAGS=
CPPFLAGS="$orig_CPPFLAGS"
AC_MSG_RESULT([no])
fi
fi
fi
])dnl
AC_DEFUN([OMPI_INTL_POSIX_THREADS_PLAIN_CXX], [
#
# C++ compiler
#
if test "$ompi_pthread_cxx_success" = "0"; then
AC_MSG_CHECKING([if C++ compiler and POSIX threads work as is])
if test "$HAVE_POSIX_THREADS" = "1" ; then
run_this_test=1
else
case "${host_cpu}-${host_os}" in
*solaris*)
AC_MSG_RESULT([no - Solaris, not checked])
run_this_test=0
;;
*-aix* | *-freebsd*)
if test "`echo $CXXCPPFLAGS | $GREP 'D_THREAD_SAFE'`" = ""; then
Correct some subtle PTRHEAD_ typos (should be PTHREAD_) in config/ompi_config_pthreads.m4. Terry pointed them out. Mostly just aix/freebsd. This commit was SVN r24324.
2011-01-29 00:05:40 +03:00
PTHREAD_CXXCPPFLAGS="-D_THREAD_SAFE"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
CXXCPPFLAGS="$CXXCPPFLAGS $PTHREAD_CXXCPPFLAGS"
fi
run_this_test=1
;;
*)
if test "`echo $CXXCPPFLAGS | $GREP 'D_REENTRANT'`" = ""; then
PTHREAD_CXXCPPFLAGS="-D_REENTRANT"
CXXCPPFLAGS="$CXXCPPFLAGS $PTHREAD_CXXCPPFLAGS"
fi
run_this_test=1
;;
esac
fi
if test "$run_this_test" = "1" ; then
AC_LANG_PUSH(C++)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_cxx_success=1,
ompi_pthread_cxx_success=0)
AC_LANG_POP(C++)
if test "$ompi_pthread_cxx_success" = "1"; then
AC_MSG_RESULT([yes])
else
PTHREAD_CXXCPPFLAGS=
CXXCPPFLAGS="$orig_CXXCPPFLAGS"
AC_MSG_RESULT([no])
fi
fi
fi
])dnl
AC_DEFUN([OMPI_INTL_POSIX_THREADS_PLAIN_FC], [
#
# Fortran compiler
#
Fix the case where there is no Fortran compiler found. Possibly fix the case where we have a really ancient Fortran compiler that does not support ISO_C_BINDING, but need to test to be sure that the new configury works. This commit was SVN r26290.
2012-04-18 23:29:07 +04:00
if test "$ompi_pthread_fortran_success" = "0" -a "$OMPI_WANT_FORTRAN_BINDINGS" = "1" -a $ompi_fortran_happy -eq 1; then
== 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 Fortran compiler and POSIX threads work as is])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
if test "$HAVE_POSIX_THREADS" = "1" ; then
run_this_test=1
else
case "${host_cpu}-${host_os}" in
*solaris*)
AC_MSG_RESULT([no - Solaris, not checked])
run_this_test=0
;;
*)
run_this_test=1
;;
esac
fi
if test "$run_this_test" = "1" ; then
AC_LANG_PUSH(C)
== 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
OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN(ompi_pthread_fortran_success=1,
ompi_pthread_fortran_success=0)
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_LANG_POP(C)
== 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
if test "$ompi_pthread_fortran_success" = "1"; then
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_MSG_RESULT([yes])
else
AC_MSG_RESULT([no])
fi
fi
fi
])dnl
AC_DEFUN([OMPI_INTL_POSIX_THREADS_PLAIN], [
# BEGIN: OMPI_INTL_POSIX_THREADS_PLAIN
#
# Check if can compile without any special flags
# we throw -D_REENTRANT or -D_THREAD_SAFE in here, just in
# case. Some systems (OS X, for example) generally don't need
# the defines, but then will on one system header here or there
# why take chances?
#
# Only run C++ and Fortran if those compilers already configured
AC_PROVIDE_IFELSE([AC_PROG_CC],
[OMPI_INTL_POSIX_THREADS_PLAIN_C],
[ompi_pthread_c_success=1])
AC_PROVIDE_IFELSE([AC_PROG_CXX],
[OMPI_INTL_POSIX_THREADS_PLAIN_CXX],
[ompi_pthread_cxx_success=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
AC_PROVIDE_IFELSE([AC_PROG_FC],
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
[OMPI_INTL_POSIX_THREADS_PLAIN_FC],
== 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
[ompi_pthread_fortran_success=1])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
# End: OMPI_INTL_POSIX_THREADS_PLAIN
])dnl
# ********************************************************************
#
# Try to compile thread support with special compiler flags
#
# ********************************************************************
AC_DEFUN([OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS_C], [
#
# C compiler
#
if test "$ompi_pthread_c_success" = "0"; then
for pf in $pflags; do
AC_MSG_CHECKING([if C compiler and POSIX threads work with $pf])
CFLAGS="$orig_CFLAGS $pf"
AC_LANG_PUSH(C)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_c_success=1,
ompi_pthread_c_success=0)
AC_LANG_POP(C)
if test "$ompi_pthread_c_success" = "1"; then
PTHREAD_CFLAGS="$pf"
AC_MSG_RESULT([yes])
break
else
PTHREAD_CFLAGS=
CFLAGS="$orig_CFLAGS"
AC_MSG_RESULT([no])
fi
done
fi
])
AC_DEFUN([OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS_CXX], [
#
# C++ compiler
#
if test "$ompi_pthread_cxx_success" = "0"; then
for pf in $pflags; do
AC_MSG_CHECKING([if C++ compiler and POSIX threads work with $pf])
CXXFLAGS="$orig_CXXFLAGS $pf"
AC_LANG_PUSH(C++)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_cxx_success=1,
ompi_pthread_cxx_success=0)
AC_LANG_POP(C++)
if test "$ompi_pthread_cxx_success" = "1"; then
PTHREAD_CXXFLAGS="$pf"
AC_MSG_RESULT([yes])
break
else
PTHREAD_CXXFLAGS=
CXXFLAGS="$orig_CXXFLAGS"
AC_MSG_RESULT([no])
fi
done
fi
])
AC_DEFUN([OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS_FC], [
#
# Fortran compiler
#
Fix fortran issues when there's no fortran compiler. This commit was SVN r27780.
2013-01-09 21:05:49 +04:00
if test "$ompi_pthread_fortran_success" = "0" -a "$OMPI_WANT_FORTRAN_BINDINGS" = "1" -a $ompi_fortran_happy -eq 1; then
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
for pf in $pflags; do
== 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 Fortran compiler and POSIX threads work with $pf])
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
FCFLAGS="$orig_FCFLAGS $pf"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_LANG_PUSH(C)
== 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
OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN(ompi_pthread_fortran_success=1,
ompi_pthread_fortran_success=0)
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_LANG_POP(C)
== 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
if test "$ompi_pthread_fortran_success" = "1"; then
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
PTHREAD_FCFLAGS="$pf"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_MSG_RESULT([yes])
break
else
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
PTHREAD_FCFLAGS=
FCFLAGS="$orig_FCFLAGS"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_MSG_RESULT([no])
fi
done
fi
])
AC_DEFUN([OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS],[
# Begin: OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS
#
# If above didn't work, try some super-special compiler flags
# that get evaluated to the "right" things.
#
# -Kthread:
# -kthread: FreeBSD kernel threads
# -pthread: Modern GCC (most all platforms)
# -pthreads: GCC on solaris
# -mthreads:
# -mt: Solaris native compilers / HP-UX aCC
#
# Put -mt before -mthreads because HP-UX aCC will properly compile
# with -mthreads (reading as -mt), but emit a warning about unknown
# flags hreads. Stupid compilers.
case "${host_cpu}-${host_os}" in
*solaris*)
pflags="-pthread -pthreads -mt"
;;
*)
pflags="-Kthread -kthread -pthread -pthreads -mt -mthreads"
;;
esac
# Only run C++ and Fortran if those compilers already configured
AC_PROVIDE_IFELSE([AC_PROG_CC],
[OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS_C],
[ompi_pthread_c_success=1])
AC_PROVIDE_IFELSE([AC_PROG_CXX],
[OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS_CXX],
[ompi_pthread_cxx_success=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
AC_PROVIDE_IFELSE([AC_PROG_FC],
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
[OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS_FC],
== 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
[ompi_pthread_fortran_success=1])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
# End: OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS
])dnl
# ********************************************************************
#
# Try to compile thread support with extra libs
#
# ********************************************************************
AC_DEFUN([OMPI_INTL_POSIX_THREADS_LIBS_C],[
#
# C compiler
#
if test "$ompi_pthread_c_success" = "0"; then
for pl in $plibs; do
AC_MSG_CHECKING([if C compiler and POSIX threads work with $pl])
case "${host_cpu}-${host-_os}" in
*-aix* | *-freebsd*)
if test "`echo $CPPFLAGS | $GREP 'D_THREAD_SAFE'`" = ""; then
Correct some subtle PTRHEAD_ typos (should be PTHREAD_) in config/ompi_config_pthreads.m4. Terry pointed them out. Mostly just aix/freebsd. This commit was SVN r24324.
2011-01-29 00:05:40 +03:00
PTHREAD_CPPFLAGS="-D_THREAD_SAFE"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
CPPFLAGS="$CPPFLAGS $PTHREAD_CPPFLAGS"
fi
;;
*)
if test "`echo $CPPFLAGS | $GREP 'D_REENTRANT'`" = ""; then
PTHREAD_CPPFLAGS="-D_REENTRANT"
CPPFLAGS="$CPPFLAGS $PTHREAD_CPPFLAGS"
fi
;;
esac
LIBS="$orig_LIBS $pl"
AC_LANG_PUSH(C)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_c_success=1,
ompi_pthread_c_success=0)
AC_LANG_POP(C)
if test "$ompi_pthread_c_success" = "1"; then
PTHREAD_LIBS="$pl"
AC_MSG_RESULT([yes])
else
PTHREAD_CPPFLAGS=
CPPFLAGS="$orig_CPPFLAGS"
LIBS="$orig_LIBS"
AC_MSG_RESULT([no])
fi
done
fi
])dnl
AC_DEFUN([OMPI_INTL_POSIX_THREADS_LIBS_CXX],[
#
# C++ compiler
#
if test "$ompi_pthread_cxx_success" = "0"; then
if test ! "$ompi_pthread_c_success" = "0" -a ! "$PTHREAD_LIBS" = "" ; then
AC_MSG_CHECKING([if C++ compiler and POSIX threads work with $PTHREAD_LIBS])
case "${host_cpu}-${host-_os}" in
*-aix* | *-freebsd*)
if test "`echo $CXXCPPFLAGS | $GREP 'D_THREAD_SAFE'`" = ""; then
Correct some subtle PTRHEAD_ typos (should be PTHREAD_) in config/ompi_config_pthreads.m4. Terry pointed them out. Mostly just aix/freebsd. This commit was SVN r24324.
2011-01-29 00:05:40 +03:00
PTHREAD_CXXCPPFLAGS="-D_THREAD_SAFE"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
CXXCPPFLAGS="$CXXCPPFLAGS $PTHREAD_CXXCPPFLAGS"
fi
;;
*)
if test "`echo $CXXCPPFLAGS | $GREP 'D_REENTRANT'`" = ""; then
PTHREAD_CXXCPPFLAGS="-D_REENTRANT"
CXXCPPFLAGS="$CXXCPPFLAGS $PTHREAD_CXXCPPFLAGS"
fi
;;
esac
LIBS="$orig_LIBS $PTHREAD_LIBS"
AC_LANG_PUSH(C++)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_cxx_success=1,
ompi_pthread_cxx_success=0)
AC_LANG_POP(C++)
if test "$ompi_pthread_cxx_success" = "1"; then
AC_MSG_RESULT([yes])
else
CXXCPPFLAGS="$orig_CXXCPPFLAGS"
LIBS="$orig_LIBS"
AC_MSG_RESULT([no])
AC_MSG_ERROR([Can not find working threads configuration. aborting])
fi
else
for pl in $plibs; do
AC_MSG_CHECKING([if C++ compiler and POSIX threads work with $pl])
case "${host_cpu}-${host-_os}" in
*-aix* | *-freebsd*)
if test "`echo $CXXCPPFLAGS | $GREP 'D_THREAD_SAFE'`" = ""; then
Correct some subtle PTRHEAD_ typos (should be PTHREAD_) in config/ompi_config_pthreads.m4. Terry pointed them out. Mostly just aix/freebsd. This commit was SVN r24324.
2011-01-29 00:05:40 +03:00
PTHREAD_CXXCPPFLAGS="-D_THREAD_SAFE"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
CXXCPPFLAGS="$CXXCPPFLAGS $PTHREAD_CXXCPPFLAGS"
fi
;;
*)
if test "`echo $CXXCPPFLAGS | $GREP 'D_REENTRANT'`" = ""; then
PTHREAD_CXXCPPFLAGS="-D_REENTRANT"
CXXCPPFLAGS="$CXXCPPFLAGS $PTHREAD_CXXCPPFLAGS"
fi
;;
esac
LIBS="$orig_LIBS $pl"
AC_LANG_PUSH(C++)
OMPI_INTL_PTHREAD_TRY_LINK(ompi_pthread_cxx_success=1,
ompi_pthread_cxx_success=0)
AC_LANG_POP(C++)
if test "$ompi_pthread_cxx_success" = "1"; then
PTHREAD_LIBS="$pl"
AC_MSG_RESULT([yes])
else
PTHREAD_CXXCPPFLAGS=
CXXCPPFLAGS="$orig_CXXCPPFLAGS"
LIBS="$orig_LIBS"
AC_MSG_RESULT([no])
fi
done
fi
fi
])dnl
AC_DEFUN([OMPI_INTL_POSIX_THREADS_LIBS_FC],[
#
# Fortran compiler
#
Fix fortran issues when there's no fortran compiler. This commit was SVN r27780.
2013-01-09 21:05:49 +04:00
if test "$ompi_pthread_fortran_success" = "0" -a "$OMPI_WANT_FORTRAN_BINDINGS" = "1" -a $ompi_fortran_happy -eq 1; then
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
if test ! "$ompi_pthread_c_success" = "0" -a ! "$PTHREAD_LIBS" = "" ; then
== 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 Fortran compiler and POSIX threads work with $PTHREAD_LIBS])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
LIBS="$orig_LIBS $PTHREAD_LIBS"
AC_LANG_PUSH(C)
== 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
OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN(ompi_pthread_fortran_success=1,
ompi_pthread_fortran_success=0)
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_LANG_POP(C)
== 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
if test "$ompi_pthread_fortran_success" = "1"; then
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_MSG_RESULT([yes])
else
LIBS="$orig_LIBS"
AC_MSG_RESULT([no])
AC_MSG_ERROR([Can not find working threads configuration. aborting])
fi
else
for pl in $plibs; do
== 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 Fortran compiler and POSIX threads work with $pl])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
LIBS="$orig_LIBS $pl"
AC_LANG_PUSH(C)
== 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
OMPI_INTL_PTHREAD_TRY_LINK_FORTRAN(ompi_pthread_fortran_success=1,
ompi_pthread_fortran_success=0)
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
AC_LANG_POP(C)
== 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
if test "$ompi_pthread_fortran_success" = "1"; then
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
PTHREAD_LIBS="$pl"
AC_MSG_RESULT([yes])
break
else
LIBS="$orig_LIBS"
AC_MSG_RESULT([no])
fi
done
fi
fi
])dnl
AC_DEFUN([OMPI_INTL_POSIX_THREADS_LIBS],[
# Begin: OMPI_INTL_POSIX_THREADS_LIBS
#
# if we can't find a super-special compiler flags, try some libraries.
# we throw -D_REENTRANT or -D_THREAD_SAFE in here, just in case. Some
# systems (OS X, for example) generally don't need the defines, but
# then will on one system header here or there why take chances?
#
# libpthreads: AIX - must check before libpthread
# liblthread: LinuxThreads on FreeBSD
# libpthread: The usual place (like we can define usual!)
plibs="-lpthreads -llthread -lpthread"
# Only run C++ and Fortran if those compilers already configured
AC_PROVIDE_IFELSE([AC_PROG_CC],
[OMPI_INTL_POSIX_THREADS_LIBS_C],
[ompi_pthread_c_success=1])
AC_PROVIDE_IFELSE([AC_PROG_CXX],
[OMPI_INTL_POSIX_THREADS_LIBS_CXX],
[ompi_pthread_cxx_success=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
AC_PROVIDE_IFELSE([AC_PROG_FC],
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
[OMPI_INTL_POSIX_THREADS_LIBS_FC],
== 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
[ompi_pthread_fortran_success=1])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
# End: OMPI_INTL_POSIX_THREADS_LIBS]
)dnl
#********************************************************************
#
# External macro (aka, the real thing)
#
#********************************************************************
AC_DEFUN([OMPI_CONFIG_POSIX_THREADS],[
AC_REQUIRE([AC_PROG_GREP])
ompi_pthread_c_success=0
== 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
ompi_pthread_fortran_success=0
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
ompi_pthread_cxx_success=0
orig_CFLAGS="$CFLAGS"
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
orig_FCFLAGS="$FCFLAGS"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
orig_CXXFLAGS="$CXXFLAGS"
orig_CPPFLAGS="$CPPFLAGS"
orig_CXXCPPFLAGS="$CXXCPPFLAGS"
orig_LDFLAGS="$LDFLAGS"
orig_LIBS="$LIBS"
Correct some subtle PTRHEAD_ typos (should be PTHREAD_) in config/ompi_config_pthreads.m4. Terry pointed them out. Mostly just aix/freebsd. This commit was SVN r24324.
2011-01-29 00:05:40 +03:00
PTHREAD_CFLAGS=
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
PTHREAD_FCFLAGS=
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
PTHREAD_CXXFLAGS=
PTHREAD_CPPFLAGS=
PTHREAD_CXXCPPFLAGS=
PTHREAD_LDFLAGS=
PTHREAD_LIBS=
# Try with the basics, mam.
OMPI_INTL_POSIX_THREADS_PLAIN
# Try the super-special compiler flags.
OMPI_INTL_POSIX_THREADS_SPECIAL_FLAGS
# Try the normal linking methods (that's no fun)
OMPI_INTL_POSIX_THREADS_LIBS
osc/sm: check for pthread_condattr_setpshared and pthread_mutexattr_setpshared. fall back on barrier if either function doesn't exist. cmr=v1.7.5:reviewer=jsquyres This commit was SVN r30911.
2014-03-03 21:09:09 +04:00
#
# check to see if we can create shared memory mutexes and conditions
#
AC_CHECK_FUNCS([pthread_mutexattr_setpshared pthread_condattr_setpshared])
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
#
# check to see if we can set error checking mutexes
#
# LinuxThreads
AC_MSG_CHECKING([for PTHREAD_MUTEX_ERRORCHECK_NP])
AC_LINK_IFELSE(
[AC_LANG_PROGRAM(
[[#include <pthread.h>]],
[[pthread_mutexattr_settype(NULL, PTHREAD_MUTEX_ERRORCHECK_NP);]])],
[result="yes" defval=1], [result="no" defval=0])
AC_MSG_RESULT([$result])
AC_DEFINE_UNQUOTED([OMPI_HAVE_PTHREAD_MUTEX_ERRORCHECK_NP], [$defval],
[If PTHREADS implementation supports PTHREAD_MUTEX_ERRORCHECK_NP])
# Mac OS X
AC_MSG_CHECKING([for PTHREAD_MUTEX_ERRORCHECK])
AC_LINK_IFELSE(
[AC_LANG_PROGRAM(
[[#include <pthread.h>]],
[[pthread_mutexattr_settype(NULL, PTHREAD_MUTEX_ERRORCHECK);]])],
[result="yes" defval=1], [result="no" defval=0])
AC_MSG_RESULT([$result])
AC_DEFINE_UNQUOTED([OMPI_HAVE_PTHREAD_MUTEX_ERRORCHECK], [$defval],
[If PTHREADS implementation supports PTHREAD_MUTEX_ERRORCHECK])
CFLAGS="$orig_CFLAGS"
Fix some fortran flags variable names, and fortran configury C test problems (i.e., don't return a value from a void() function!). Thanks to Eugene for identifying these issues. Hopefully this will fix up the problems Oracle is having with compiling the new Fortran stuff. This commit was SVN r26310.
2012-04-23 19:28:26 +04:00
FCFLAGS="$orig_FCFLAGS"
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
CXXFLAGS="$orig_CXXFLAGS"
CPPFLAGS="$orig_CPPFLAGS"
CXXCPPFLAGS="$orig_CXXCPPFLAGS"
LDFLAGS="$orig_LDFLAGS"
LIBS="$orig_LIBS"
Fix fortran issues when there's no fortran compiler. This commit was SVN r27780.
2013-01-09 21:05:49 +04:00
if test "$OMPI_WANT_FORTRAN_BINDINGS" != "1" -o $ompi_fortran_happy -ne 1; then
== 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
ompi_pthread_fortran_success=1
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
fi
if test "$ompi_pthread_c_success" = "1" -a \
"$ompi_pthread_cxx_success" = "1" -a \
== 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
"$ompi_pthread_fortran_success" = "1"; then
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
internal_useless=1
$1
else
internal_useless=1
$2
fi
== 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
unset ompi_pthread_c_success ompi_pthread_fortran_success ompi_pthread_cxx_success
Rollback the last two commits. svn merge -r 19823:19821 https://svn.open-mpi.org/svn/ompi/trunk ./ I copied the config directory in another location, and the original .svn directory was replaced by the Open MPI one ... As a result my first commit, applied the changes on the Open MPI trunk instead of the other project. This commit was SVN r19824.
2008-10-28 20:22:29 +03:00
unset internal_useless
])dnl
Копировать постоянную ссылку