54b2cf747e
The commit has been tested for C/R and Cray operations, and on Odin (SLURM, rsh) and RoadRunner (TM). I tried to update all environments, but obviously could not test them. I know that Windows needs some work, and have highlighted what is know to be needed in the odls process component. This represents a lot of work by Brian, Tim P, Josh, and myself, with much advice from Jeff and others. For posterity, I have appended a copy of the email describing the work that was done: As we have repeatedly noted, the modex operation in MPI_Init is the single greatest consumer of time during startup. To-date, we have executed that operation as an ORTE stage gate that held the process until a startup message containing all required modex (and OOB contact info - see #3 below) info could be sent to it. Each process would send its data to the HNP's registry, which assembled and sent the message when all processes had reported in. In addition, ORTE had taken responsibility for monitoring process status as it progressed through a series of "stage gates". The process reported its status at each gate, and ORTE would then send a "release" message once all procs had reported in. The incoming changes revamp these procedures in three ways: 1. eliminating the ORTE stage gate system and cleanly delineating responsibility between the OMPI and ORTE layers for MPI init/finalize. The modex stage gate (STG1) has been replaced by a collective operation in the modex itself that performs an allgather on the required modex info. The allgather is implemented using the orte_grpcomm framework since the BTL's are not active at that point. At the moment, the grpcomm framework only has a "basic" component analogous to OMPI's "basic" coll framework - I would recommend that the MPI team create additional, more advanced components to improve performance of this step. The other stage gates have been replaced by orte_grpcomm barrier functions. We tried to use MPI barriers instead (since the BTL's are active at that point), but - as we discussed on the telecon - these are not currently true barriers so the job would hang when we fell through while messages were still in process. Note that the grpcomm barrier doesn't actually resolve that problem, but Brian has pointed out that we are unlikely to ever see it violated. Again, you might want to spend a little time on an advanced barrier algorithm as the one in "basic" is very simplistic. Summarizing this change: ORTE no longer tracks process state nor has direct responsibility for synchronizing jobs. This is now done via collective operations within the MPI layer, albeit using ORTE collective communication services. I -strongly- urge the MPI team to implement advanced collective algorithms to improve the performance of this critical procedure. 2. reducing the volume of data exchanged during modex. Data in the modex consisted of the process name, the name of the node where that process is located (expressed as a string), plus a string representation of all contact info. The nodename was required in order for the modex to determine if the process was local or not - in addition, some people like to have it to print pretty error messages when a connection failed. The size of this data has been reduced in three ways: (a) reducing the size of the process name itself. The process name consisted of two 32-bit fields for the jobid and vpid. This is far larger than any current system, or system likely to exist in the near future, can support. Accordingly, the default size of these fields has been reduced to 16-bits, which means you can have 32k procs in each of 32k jobs. Since the daemons must have a vpid, and we require one daemon/node, this also restricts the default configuration to 32k nodes. To support any future "mega-clusters", a configuration option --enable-jumbo-apps has been added. This option increases the jobid and vpid field sizes to 32-bits. Someday, if necessary, someone can add yet another option to increase them to 64-bits, I suppose. (b) replacing the string nodename with an integer nodeid. Since we have one daemon/node, the nodeid corresponds to the local daemon's vpid. This replaces an often lengthy string with only 2 (or at most 4) bytes, a substantial reduction. (c) when the mca param requesting that nodenames be sent to support pretty error messages, a second mca param is now used to request FQDN - otherwise, the domain name is stripped (by default) from the message to save space. If someone wants to combine those into a single param somehow (perhaps with an argument?), they are welcome to do so - I didn't want to alter what people are already using. While these may seem like small savings, they actually amount to a significant impact when aggregated across the entire modex operation. Since every proc must receive the modex data regardless of the collective used to send it, just reducing the size of the process name removes nearly 400MBytes of communication from a 32k proc job (admittedly, much of this comm may occur in parallel). So it does add up pretty quickly. 3. routing RML messages to reduce connections. The default messaging system remains point-to-point - i.e., each proc opens a socket to every proc it communicates with and sends its messages directly. A new option uses the orteds as routers - i.e., each proc only opens a single socket to its local orted. All messages are sent from the proc to the orted, which forwards the message to the orted on the node where the intended recipient proc is located - that orted then forwards the message to its local proc (the recipient). This greatly reduces the connection storm we have encountered during startup. It also has the benefit of removing the sharing of every proc's OOB contact with every other proc. The orted routing tables are populated during launch since every orted gets a map of where every proc is being placed. Each proc, therefore, only needs to know the contact info for its local daemon, which is passed in via the environment when the proc is fork/exec'd by the daemon. This alone removes ~50 bytes/process of communication that was in the current STG1 startup message - so for our 32k proc job, this saves us roughly 32k*50 = 1.6MBytes sent to 32k procs = 51GBytes of messaging. Note that you can use the new routing method by specifying -mca routed tree - if you so desire. This mode will become the default at some point in the future. There are a few minor additional changes in the commit that I'll just note in passing: * propagation of command line mca params to the orteds - fixes ticket #1073. See note there for details. * requiring of "finalize" prior to "exit" for MPI procs - fixes ticket #1144. See note there for details. * cleanup of some stale header files This commit was SVN r16364.
708 строки
24 KiB
Bash
708 строки
24 KiB
Bash
dnl -*- shell-script -*-
|
|
dnl
|
|
dnl Copyright (c) 2004-2007 The Trustees of Indiana University and Indiana
|
|
dnl University Research and Technology
|
|
dnl Corporation. All rights reserved.
|
|
dnl Copyright (c) 2004-2005 The University of Tennessee and The University
|
|
dnl of Tennessee Research Foundation. All rights
|
|
dnl reserved.
|
|
dnl Copyright (c) 2004-2005 High Performance Computing Center Stuttgart,
|
|
dnl University of Stuttgart. All rights reserved.
|
|
dnl Copyright (c) 2004-2005 The Regents of the University of California.
|
|
dnl All rights reserved.
|
|
dnl Copyright (c) 2006 Cisco Systems, Inc. All rights reserved.
|
|
dnl $COPYRIGHT$
|
|
dnl
|
|
dnl Additional copyrights may follow
|
|
dnl
|
|
dnl $HEADER$
|
|
dnl
|
|
|
|
AC_DEFUN([OMPI_CONFIGURE_OPTIONS],[
|
|
ompi_show_subtitle "Configuration options"
|
|
|
|
#
|
|
# Code coverage options
|
|
#
|
|
AC_MSG_CHECKING([whether to run code coverage])
|
|
AC_ARG_ENABLE(coverage,
|
|
AC_HELP_STRING([--enable-coverage],
|
|
[enable code coverage files to be generated]))
|
|
if test "$enable_coverage" = "yes"; then
|
|
if test "$enable_shared" = "yes"; then
|
|
AC_MSG_WARN([Code coverage can run only with static libraries. Please
|
|
run configure with --enable-static --disable-shared if
|
|
you want code coverage. Also ensure that you execute
|
|
make clean too ensure removal of all leftover shared
|
|
mpi libraries])
|
|
AC_MSG_ERROR([Cannot continue processing])
|
|
fi
|
|
AC_MSG_RESULT([yes])
|
|
WANT_COVERAGE=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_COVERAGE=0
|
|
fi
|
|
|
|
|
|
#
|
|
# Branch Probabilities options
|
|
#
|
|
AC_MSG_CHECKING([whether to compile with branch probabilities])
|
|
AC_ARG_ENABLE(branch-probabilities,
|
|
AC_HELP_STRING([--enable-branch-probabilities],
|
|
[enable profile arcs and branch probability optimization]))
|
|
if test "$enable_branch_probabilities" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_BRANCH_PROBABILITIES=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_BRANCH_PROBABILITIES=0
|
|
fi
|
|
|
|
|
|
#
|
|
# Memory debugging
|
|
#
|
|
|
|
AC_MSG_CHECKING([whether to debug memory usage])
|
|
AC_ARG_ENABLE(mem-debug,
|
|
AC_HELP_STRING([--enable-mem-debug],
|
|
[enable memory debugging (debugging only) (default: disabled)]))
|
|
if test "$enable_mem_debug" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_MEM_DEBUG=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_MEM_DEBUG=0
|
|
fi
|
|
#################### Early development override ####################
|
|
if test "$WANT_MEM_DEBUG" = "0" -a -z "$enable_mem_debug" -a -d .svn; then
|
|
WANT_MEM_DEBUG=1
|
|
echo "--> developer override: enable mem profiling by default"
|
|
fi
|
|
#################### Early development override ####################
|
|
AC_DEFINE_UNQUOTED(OMPI_ENABLE_MEM_DEBUG, $WANT_MEM_DEBUG,
|
|
[Whether we want the memory profiling or not])
|
|
|
|
#
|
|
# Memory profiling
|
|
#
|
|
|
|
AC_MSG_CHECKING([whether to profile memory usage])
|
|
AC_ARG_ENABLE(mem-profile,
|
|
AC_HELP_STRING([--enable-mem-profile],
|
|
[enable memory profiling (debugging only) (default: disabled)]))
|
|
if test "$enable_mem_profile" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_MEM_PROFILE=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_MEM_PROFILE=0
|
|
fi
|
|
#################### Early development override ####################
|
|
if test "$WANT_MEM_PROFILE" = "0" -a -z "$enable_mem_profile" -a -d .svn; then
|
|
WANT_MEM_PROFILE=1
|
|
echo "--> developer override: enable mem profiling by default"
|
|
fi
|
|
#################### Early development override ####################
|
|
AC_DEFINE_UNQUOTED(OMPI_ENABLE_MEM_PROFILE, $WANT_MEM_PROFILE,
|
|
[Whether we want the memory profiling or not])
|
|
|
|
#
|
|
# Developer picky compiler options
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want developer-level compiler pickyness])
|
|
AC_ARG_ENABLE(picky,
|
|
AC_HELP_STRING([--enable-picky],
|
|
[enable developer-level compiler pickyness when building Open MPI (default: disabled)]))
|
|
if test "$enable_picky" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_PICKY_COMPILER=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_PICKY_COMPILER=0
|
|
fi
|
|
#################### Early development override ####################
|
|
if test "$WANT_PICKY_COMPILER" = "0" -a -z "$enable_picky" -a -d .svn; then
|
|
WANT_PICKY_COMPILER=1
|
|
echo "--> developer override: enable picky compiler by default"
|
|
fi
|
|
#################### Early development override ####################
|
|
|
|
#
|
|
# Developer debugging
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want developer-level debugging code])
|
|
AC_ARG_ENABLE(debug,
|
|
AC_HELP_STRING([--enable-debug],
|
|
[enable developer-level debugging code (not for general MPI users!) (default: disabled)]))
|
|
if test "$enable_debug" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_DEBUG=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_DEBUG=0
|
|
fi
|
|
#################### Early development override ####################
|
|
if test "$WANT_DEBUG" = "0" -a -z "$enable_debug" -a -d .svn; then
|
|
WANT_DEBUG=1
|
|
echo "--> developer override: enable debugging code by default"
|
|
fi
|
|
#################### Early development override ####################
|
|
if test "$WANT_DEBUG" = "0"; then
|
|
CFLAGS="-DNDEBUG $CFLAGS"
|
|
CXXFLAGS="-DNDEBUG $CXXFLAGS"
|
|
fi
|
|
AC_DEFINE_UNQUOTED(OMPI_ENABLE_DEBUG, $WANT_DEBUG,
|
|
[Whether we want developer-level debugging code or not])
|
|
|
|
AC_ARG_ENABLE(debug-symbols,
|
|
AC_HELP_STRING([--disable-debug-symbols],
|
|
[Disable adding compiler flags to enable debugging symbols if --enable-debug is specified. For non-debugging builds, this flag has no effect.]))
|
|
|
|
#
|
|
# Sparse Groups
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want sparse process groups])
|
|
AC_ARG_ENABLE(sparse-groups,
|
|
AC_HELP_STRING([--enable-sparse-groups],
|
|
[enable sparse process groups (default: not enabled)]))
|
|
if test "$enable_sparse_groups" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
GROUP_SPARSE=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
GROUP_SPARSE=0
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OMPI_GROUP_SPARSE],$GROUP_SPARSE,
|
|
[Wether we want sparse process groups])
|
|
|
|
#
|
|
# Fortran 77
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want Fortran 77 bindings])
|
|
AC_ARG_ENABLE(mpi-f77,
|
|
AC_HELP_STRING([--enable-mpi-f77],
|
|
[enable f77 MPI bindings (default: enabled)]))
|
|
if test "$enable_mpi_f77" != "no"; then
|
|
AC_MSG_RESULT([yes])
|
|
OMPI_WANT_F77_BINDINGS=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
OMPI_WANT_F77_BINDINGS=0
|
|
fi
|
|
|
|
|
|
#
|
|
# Fortran 90
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want Fortran 90 bindings])
|
|
AC_ARG_ENABLE(mpi-f90,
|
|
AC_HELP_STRING([--enable-mpi-f90],
|
|
[enable f90 MPI bindings (default: enabled)]))
|
|
if test "$enable_mpi_f90" != "no"; then
|
|
AC_MSG_RESULT([yes])
|
|
OMPI_WANT_F90_BINDINGS=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
OMPI_WANT_F90_BINDINGS=0
|
|
fi
|
|
|
|
AC_MSG_CHECKING([desired Fortran 90 bindings "size"])
|
|
AC_ARG_WITH(mpi-f90-size,
|
|
AC_HELP_STRING([--with-mpi-f90-size=SIZE],
|
|
[specify the types of functions in the Fortran 90 MPI module, where size is one of: trivial (MPI-2 F90-specific functions only), small (trivial + all MPI functions without choice buffers), medium (small + all MPI functions with one choice buffer), large (medium + all MPI functions with 2 choice buffers, but only when both buffers are the same type)]))
|
|
|
|
if test "$OMPI_WANT_F90_BINDINGS" = "0"; then
|
|
AC_MSG_RESULT([disabled (Fortran 90 bindings disabled)])
|
|
elif test "$with_mpi_f90_size" = "no"; then
|
|
OMPI_WANT_F90_BINDINGS=0
|
|
AC_MSG_RESULT([disabling F90 MPI module (used specified)])
|
|
else
|
|
# Default value
|
|
if test "$with_mpi_f90_size" = ""; then
|
|
with_mpi_f90_size=small
|
|
fi
|
|
|
|
# Check for each of the sizes
|
|
if test "$with_mpi_f90_size" = "trivial"; then
|
|
OMPI_F90_BUILD_SIZE=trivial
|
|
elif test "$with_mpi_f90_size" = "small"; then
|
|
OMPI_F90_BUILD_SIZE=small
|
|
elif test "$with_mpi_f90_size" = "medium"; then
|
|
OMPI_F90_BUILD_SIZE=medium
|
|
elif test "$with_mpi_f90_size" = "large"; then
|
|
OMPI_F90_BUILD_SIZE=large
|
|
else
|
|
AC_MSG_RESULT([Unrecognized size: $with_mpi_f90_size])
|
|
AC_MSG_ERROR([Cannot continue])
|
|
fi
|
|
fi
|
|
|
|
AM_CONDITIONAL([OMPI_WANT_BUILD_F90_TRIVIAL],
|
|
[test "$OMPI_F90_BUILD_SIZE" = "trivial"])
|
|
AM_CONDITIONAL([OMPI_WANT_BUILD_F90_SMALL],
|
|
[test "$OMPI_F90_BUILD_SIZE" = "small"])
|
|
AM_CONDITIONAL([OMPI_WANT_BUILD_F90_MEDIUM],
|
|
[test "$OMPI_F90_BUILD_SIZE" = "medium"])
|
|
AM_CONDITIONAL([OMPI_WANT_BUILD_F90_LARGE],
|
|
[test "$OMPI_F90_BUILD_SIZE" = "large"])
|
|
|
|
AC_SUBST(OMPI_F90_BUILD_SIZE)
|
|
AC_MSG_RESULT([$OMPI_F90_BUILD_SIZE])
|
|
|
|
#
|
|
# MPI profiling
|
|
#
|
|
|
|
AC_MSG_CHECKING([whether to enable PMPI])
|
|
AC_ARG_ENABLE(mpi-profile,
|
|
AC_HELP_STRING([--enable-mpi-profile],
|
|
[enable MPI profiling (default: enabled)]))
|
|
if test "$enable_mpi_profile" != "no"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_MPI_PROFILING=1
|
|
MPIF_H_PMPI_W_FUNCS=", PMPI_WTICK, PMPI_WTIME"
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_MPI_PROFILING=0
|
|
MPIF_H_PMPI_W_FUNCS=
|
|
fi
|
|
AC_SUBST(MPIF_H_PMPI_W_FUNCS)
|
|
|
|
|
|
#
|
|
# C++
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want C++ bindings])
|
|
AC_ARG_ENABLE(mpi-cxx,
|
|
AC_HELP_STRING([--enable-mpi-cxx],
|
|
[enable C++ MPI bindings (default: enabled)]))
|
|
if test "$enable_mpi_cxx" != "no"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_MPI_CXX_SUPPORT=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_MPI_CXX_SUPPORT=0
|
|
fi
|
|
|
|
AC_MSG_CHECKING([if want MPI::SEEK_SET support])
|
|
AC_ARG_ENABLE([mpi-cxx-seek],
|
|
[AC_HELP_STRING([--enable-mpi-cxx-seek],
|
|
[enable support for MPI::SEEK_SET, MPI::SEEK_END, and MPI::SEEK_POS in C++ bindings (default: enabled)])])
|
|
if test "$enable_mpi_cxx_seek" != "no" ; then
|
|
AC_MSG_RESULT([yes])
|
|
OMPI_WANT_MPI_CXX_SEEK=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
OMPI_WANT_MPI_CXX_SEEK=0
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OMPI_WANT_MPI_CXX_SEEK], [$OMPI_WANT_MPI_CXX_SEEK],
|
|
[do we want to try to work around C++ bindings SEEK_* issue?])
|
|
|
|
#
|
|
# Do we want to disable weak symbols for some reason?
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want to enable weak symbol support])
|
|
AC_ARG_ENABLE(weak-symbols,
|
|
AC_HELP_STRING([--enable-weak-symbols],
|
|
[use weak symbols, if available (default: enabled)]))
|
|
if test "$enable_weak_symbols" != "no"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_WEAK_SYMBOLS=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_WEAK_SYMBOLS=0
|
|
fi
|
|
|
|
#
|
|
# Do we want to disable MPI parameter checking at run-time?
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want run-time MPI parameter checking])
|
|
AC_ARG_WITH(mpi-param-check,
|
|
AC_HELP_STRING([--with-mpi-param-check(=VALUE)],
|
|
[behavior of MPI function parameter checking. Valid values are: always, never, runtime. If --with-mpi-param-check is specified with no VALUE argument, it is equivalent to a VALUE of "always"; --without-mpi-param-check is equivalent to "never" (default: runtime).]))
|
|
mpi_param_check=ompi_mpi_param_check
|
|
if test "$with_mpi_param_check" = "no" -o \
|
|
"$with_mpi_param_check" = "never"; then
|
|
mpi_param_check=0
|
|
AC_MSG_RESULT([never])
|
|
elif test "$with_mpi_param_check" = "yes" -o \
|
|
"$with_mpi_param_check" = "always"; then
|
|
mpi_param_check=1
|
|
AC_MSG_RESULT([always])
|
|
elif test "$with_mpi_param_check" = "runtime" -o \
|
|
-z "$with_mpi_params_check"; then
|
|
AC_MSG_RESULT([runtime])
|
|
else
|
|
AC_MSG_RESULT([unknown])
|
|
AC_MSG_WARN([*** Unrecognized --with-mpi-param-check value])
|
|
AC_MSG_WARN([*** See "configure --help" output])
|
|
AC_MSG_WARN([*** Defaulting to "runtime"])
|
|
fi
|
|
AC_DEFINE_UNQUOTED(MPI_PARAM_CHECK, $mpi_param_check,
|
|
[Whether we want to check MPI parameters always, never, or decide at run-time])
|
|
|
|
|
|
#
|
|
# Do we want to install all of OMPI's header files?
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want to install OMPI header files])
|
|
AC_ARG_WITH(devel-headers,
|
|
AC_HELP_STRING([--with-devel-headers],
|
|
[normal MPI users/applications do not need this (mpi.h and mpif.h are ALWAYS installed). Developer headers are only necessary for MCA module authors (default: disabled).]))
|
|
if test "$with_devel_headers" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_INSTALL_HEADERS=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_INSTALL_HEADERS=0
|
|
fi
|
|
AM_CONDITIONAL(WANT_INSTALL_HEADERS, test "$WANT_INSTALL_HEADERS" = 1)
|
|
|
|
#
|
|
# Do we want the pretty-print stack trace feature?
|
|
#
|
|
AC_MSG_CHECKING([if want pretty-print stacktrace])
|
|
AC_ARG_ENABLE([pretty-print-stacktrace],
|
|
[AC_HELP_STRING([--enable-pretty-print-stacktrace],
|
|
[Pretty print stacktrace on process signal])])
|
|
if test "$enable_pretty_print_stacktrace" = "no" ; then
|
|
AC_MSG_RESULT([no])
|
|
WANT_PRETTY_PRINT_STACKTRACE=0
|
|
else
|
|
AC_MSG_RESULT([yes])
|
|
WANT_PRETTY_PRINT_STACKTRACE=1
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OMPI_WANT_PRETTY_PRINT_STACKTRACE],
|
|
[$WANT_PRETTY_PRINT_STACKTRACE],
|
|
[if want pretty-print stack trace feature])
|
|
|
|
#
|
|
# Do we want deprecated executable names ?
|
|
#
|
|
AC_MSG_CHECKING(if want deprecated executable names)
|
|
AC_ARG_ENABLE(deprecated-executable-names,
|
|
AC_HELP_STRING([--enable-deprecated-executable-names], [make sym links to deprecated OMPI executables (e.g., hcc, hcp, hf77, wipe) (default: disabled)]))
|
|
if test "$enable_deprecated_executable_names" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_DEN=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_DEN=0
|
|
fi
|
|
|
|
AM_CONDITIONAL(WANT_DEPRECATED_EXECUTABLE_NAMES, test "$WANT_DEN" = "1")
|
|
|
|
#
|
|
# Do we want to enable peruse interface?
|
|
#
|
|
|
|
AC_MSG_CHECKING([if peruse support is required])
|
|
AC_ARG_ENABLE(peruse,
|
|
AC_HELP_STRING([--enable-peruse],
|
|
[Support PERUSE interface (default: disabled)]))
|
|
if test "$enable_peruse" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
WANT_PERUSE=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
WANT_PERUSE=0
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OMPI_WANT_PERUSE],
|
|
[$WANT_PERUSE],
|
|
[if the peruse interface should be enabled])
|
|
AM_CONDITIONAL(WANT_PERUSE, test "$WANT_PERUSE" = "1")
|
|
|
|
#
|
|
# What is the max array rank that we want to support in the f90 bindings?
|
|
#
|
|
|
|
OMPI_FORTRAN_MAX_ARRAY_RANK=4
|
|
AC_MSG_CHECKING([max supported array dimension in F90 MPI bindings])
|
|
AC_ARG_WITH(f90-max-array-dim,
|
|
AC_HELP_STRING([--with-f90-max-array-dim=<DIM>],
|
|
[The maximum array dimension supported in the F90 MPI bindings (default: $OMPI_FORTRAN_MAX_ARRAY_RANK).]))
|
|
if test ! -z "$with_f90_max_array_dim" -a "$with_f90_max_array_dim" != "no"; then
|
|
# Ensure it's a number; hopefully a integer...
|
|
expr $with_f90_max_array_dim + 1 > /dev/null 2> /dev/null
|
|
if test "$?" = "0"; then
|
|
OMPI_FORTRAN_MAX_ARRAY_RANK="$with_f90_max_array_dim"
|
|
fi
|
|
fi
|
|
AC_MSG_RESULT([$OMPI_FORTRAN_MAX_ARRAY_RANK])
|
|
AC_SUBST(OMPI_FORTRAN_MAX_ARRAY_RANK)
|
|
|
|
# do we want PTY support?
|
|
AC_MSG_CHECKING([if pty support should be enabled])
|
|
AC_ARG_ENABLE(pty-support,
|
|
AC_HELP_STRING([--enable-pty-support],
|
|
[Enable/disable PTY support for STDIO forwarding. default: enabled]))
|
|
if test "$enable_pty_support" = "no" ; then
|
|
AC_MSG_RESULT([no])
|
|
OMPI_ENABLE_PTY_SUPPORT=0
|
|
else
|
|
AC_MSG_RESULT([yes])
|
|
OMPI_ENABLE_PTY_SUPPORT=1
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OMPI_ENABLE_PTY_SUPPORT], [$OMPI_ENABLE_PTY_SUPPORT],
|
|
[Whether user wants PTY support or not])
|
|
|
|
#
|
|
# Do we want to allow DLOPEN?
|
|
#
|
|
AC_MSG_CHECKING([if user wants dlopen support])
|
|
AC_ARG_ENABLE([dlopen],
|
|
[AC_HELP_STRING([--enable-dlopen],
|
|
[Whether build should attempt to use dlopen (or
|
|
similar) to dynamically load components.
|
|
Disabling dlopen implies --disable-mca-dso.
|
|
(default: enabled)])])
|
|
if test "$enable_dlopen" = "no" ; then
|
|
enable_mca_dso="no"
|
|
enable_mca_static="yes"
|
|
OMPI_ENABLE_DLOPEN_SUPPORT=0
|
|
AC_MSG_RESULT([no])
|
|
else
|
|
OMPI_ENABLE_DLOPEN_SUPPORT=1
|
|
AC_MSG_RESULT([yes])
|
|
fi
|
|
|
|
#
|
|
# Heterogeneous support
|
|
#
|
|
|
|
AC_MSG_CHECKING([if heterogeneous support should be enabled])
|
|
AC_ARG_ENABLE([heterogeneous],
|
|
[AC_HELP_STRING([--enable-heterogeneous],
|
|
[Enable features required for heterogeneous
|
|
platform support (default: enabled)])])
|
|
if test "$enable_heterogeneous" = "no" ; then
|
|
AC_MSG_RESULT([no])
|
|
ompi_want_heterogeneous=0
|
|
else
|
|
AC_MSG_RESULT([yes])
|
|
ompi_want_heterogeneous=1
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OMPI_ENABLE_HETEROGENEOUS_SUPPORT],
|
|
[$ompi_want_heterogeneous],
|
|
[Enable features required for heterogeneous support])
|
|
|
|
#
|
|
# Internal trace file logging (debugging)
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want trace file debugging])
|
|
AC_ARG_ENABLE([trace],
|
|
[AC_HELP_STRING([--enable-trace],
|
|
[Enable internal tracing of OMPI/ORTE/OPAL calls -- used only for developer debugging, not tracing of MPI applications (default: disabled)])])
|
|
if test "$enable_trace" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
opal_want_trace=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
opal_want_trace=0
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OPAL_ENABLE_TRACE], [$opal_want_trace],
|
|
[Enable run-time tracing of internal functions])
|
|
|
|
|
|
#
|
|
# Jumbo application support
|
|
#
|
|
|
|
AC_MSG_CHECKING([if want jumbo app support])
|
|
AC_ARG_ENABLE([jumbo-apps],
|
|
[AC_HELP_STRING([--enable-jumbo-apps],
|
|
[Enable support for applications in excess of 32K processes and/or 32K jobs, or running on clusters in excess of 32k nodes (default: disabled)])])
|
|
if test "$enable_jumbo_apps" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
orte_want_jumbo_apps=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
orte_want_jumbo_apps=0
|
|
fi
|
|
AC_DEFINE_UNQUOTED([ORTE_ENABLE_JUMBO_APPS], [$orte_want_jumbo_apps],
|
|
[Enable support for applications in excess of 32K processes and/or 32K jobs, or running on clusters in excess of 32k nodes])
|
|
|
|
|
|
#
|
|
# Cross-compile data
|
|
#
|
|
AC_ARG_WITH([cross],
|
|
[AC_HELP_STRING([--with-cross=FILE],
|
|
[Specify configure values that can not be determined in a cross-compilation environment. See the Open MPI FAQ.])])
|
|
if test "$with_cross" = "yes" ; then
|
|
AC_MSG_ERROR([--with-cross argument must include FILE option])
|
|
elif test "$with_cross" = "no" ; then
|
|
AC_MSG_ERROR([--without-cross is not a valid argument])
|
|
elif test "$with_cross" != "" ; then
|
|
if test ! -r $with_cross ; then
|
|
AC_MSG_ERROR([could not find cross-compile data file $with_cross])
|
|
fi
|
|
|
|
# eval into environment
|
|
OMPI_LOG_MSG([Loading cross-compile file $with_cross, with contents below])
|
|
OMPI_LOG_FILE([$with_cross])
|
|
. "$with_cross"
|
|
fi
|
|
|
|
#
|
|
# --with-ft=TYPE
|
|
# TYPE:
|
|
# - LAM (synonym for 'cr' currently)
|
|
# - cr
|
|
# /* General FT sections */
|
|
# #if OPAL_ENABLE_FT == 0 /* FT Disabled globaly */
|
|
# #if OPAL_ENABLE_FT == 1 /* FT Enabled globaly */
|
|
# /* CR Specific sections */
|
|
# #if OPAL_ENABLE_FT_CR == 0 /* FT Ckpt/Restart Disabled */
|
|
# #if OPAL_ENABLE_FT_CR == 1 /* FT Ckpt/Restart Enabled */
|
|
#
|
|
AC_MSG_CHECKING([if want fault tolerance])
|
|
AC_ARG_WITH(ft,
|
|
[AC_HELP_STRING([--with-ft=TYPE],
|
|
[Specify the type of fault tolerance to enable. Options: LAM (LAM/MPI-like), cr (Checkpoint/Restart) (default: disabled)])],
|
|
[ompi_want_ft=1],
|
|
[ompi_want_ft=0])
|
|
if test "$with_ft" = "no" -o "$ompi_want_ft" = "0"; then
|
|
ompi_want_ft=0
|
|
ompi_want_ft_cr=0
|
|
AC_MSG_RESULT([Disabled fault tolerance])
|
|
else
|
|
ompi_want_ft=1
|
|
ompi_want_ft_cr=0
|
|
ompi_want_ft_type=none
|
|
|
|
# Default value
|
|
if test "$with_ft" = "" -o "$with_ft" = "yes"; then
|
|
ompi_want_ft_type=cr
|
|
ompi_want_ft_cr=1
|
|
elif test "$with_ft" = "LAM"; then
|
|
ompi_want_ft_type=lam
|
|
ompi_want_ft_cr=1
|
|
elif test "$with_ft" = "lam"; then
|
|
ompi_want_ft_type=lam
|
|
ompi_want_ft_cr=1
|
|
elif test "$with_ft" = "CR"; then
|
|
ompi_want_ft_type=cr
|
|
ompi_want_ft_cr=1
|
|
elif test "$with_ft" = "cr"; then
|
|
ompi_want_ft_type=cr
|
|
ompi_want_ft_cr=1
|
|
else
|
|
AC_MSG_RESULT([Unrecognized FT TYPE: $with_ft])
|
|
AC_MSG_ERROR([Cannot continue])
|
|
fi
|
|
AC_MSG_RESULT([Enabled $with_ft ($ompi_want_ft_type)])
|
|
AC_MSG_WARN([**************************************************])
|
|
AC_MSG_WARN([*** Fault Tolerance Integration into Open MPI is *])
|
|
AC_MSG_WARN([*** a research quality implementation, and care *])
|
|
AC_MSG_WARN([*** should be used when choosing to enable it. *])
|
|
AC_MSG_WARN([**************************************************])
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OPAL_ENABLE_FT], [$ompi_want_ft],
|
|
[Enable fault tolerance general components and logic])
|
|
AC_DEFINE_UNQUOTED([OPAL_ENABLE_FT_CR], [$ompi_want_ft_cr],
|
|
[Enable fault tolerance checkpoint/restart components and logic])
|
|
AM_CONDITIONAL(WANT_FT, test "$ompi_want_ft" = "1")
|
|
|
|
#
|
|
# Fault Tolerance Components and Logic
|
|
#
|
|
# --enable-ft-thread
|
|
# #if OPAL_ENABLE_FT_THREAD == 0 /* Disabled */
|
|
# #if OPAL_ENABLE_FT_THREAD == 1 /* Enabled */
|
|
#
|
|
AC_MSG_CHECKING([if want fault tolerance thread])
|
|
AC_ARG_ENABLE([ft_thread],
|
|
[AC_HELP_STRING([--enable-ft-thread],
|
|
[Enable fault tolerance thread running inside all processes. Requires progress threads (default: disabled)])])
|
|
if test "$ompi_want_ft" = "0"; then
|
|
ompi_want_ft_thread=0
|
|
AC_MSG_RESULT([Disabled (fault tolerance disabled --without-ft-style)])
|
|
elif test "$enable_ft_thread" = "yes"; then
|
|
# This check may not fire since progress threads are checked after this section :/
|
|
if test "$OMPI_ENABLE_PROGRESS_THREADS" = "0"; then
|
|
AC_MSG_RESULT([Must enable progress threads to use this option])
|
|
AC_MSG_ERROR([Cannot continue])
|
|
else
|
|
AC_MSG_RESULT([yes])
|
|
ompi_want_ft_thread=1
|
|
AC_MSG_WARN([**************************************************])
|
|
AC_MSG_WARN([*** Fault Tolerance with a thread in Open MPI *])
|
|
AC_MSG_WARN([*** is an experimental, research quality option. *])
|
|
AC_MSG_WARN([*** It requires progress threads to be used, and *])
|
|
AC_MSG_WARN([*** care should be used when enabling these *])
|
|
AC_MSG_WARN([*** options. *])
|
|
AC_MSG_WARN([**************************************************])
|
|
fi
|
|
else
|
|
ompi_want_ft_thread=0
|
|
AC_MSG_RESULT([Disabled])
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OPAL_ENABLE_FT_THREAD], [$ompi_want_ft_thread],
|
|
[Enable fault tolerance thread in Open PAL])
|
|
AM_CONDITIONAL(WANT_FT_THREAD, test "$ompi_want_ft_thread" = "1")
|
|
|
|
|
|
#
|
|
# Do we want to install binaries?
|
|
#
|
|
AC_ARG_ENABLE([binaries],
|
|
[AC_HELP_STRING([--enable-binaries],
|
|
[Build and install binaries required for Open MPI, such as the wrapper compilers. Useful for multi-lib installations. (default: enabled)])])
|
|
AM_CONDITIONAL([OMPI_INSTALL_BINARIES], [test "$enable_binaries" != "no"])
|
|
|
|
#
|
|
# Do we want to disable IPv6 support?
|
|
#
|
|
AC_MSG_CHECKING([if want IPv6 support])
|
|
AC_ARG_ENABLE([ipv6],
|
|
[AC_HELP_STRING([--disable-ipv6],
|
|
[Disable IPv6 support (default: enabled, but only if the underlying system supports it)])])
|
|
if test "$enable_ipv6" = "no"; then
|
|
AC_MSG_RESULT([no])
|
|
opal_want_ipv6=0
|
|
else
|
|
AC_MSG_RESULT([yes (if underlying system supports it)])
|
|
opal_want_ipv6=1
|
|
fi
|
|
AC_DEFINE_UNQUOTED([OPAL_ENABLE_IPV6], [$opal_want_ipv6],
|
|
[Enable IPv6 support, but only if the underlying system supports it])
|
|
|
|
#
|
|
# Do we want orterun's --prefix behavior to be enabled by default?
|
|
#
|
|
AC_MSG_CHECKING([if want orterun "--prefix" behavior to be enabled by default])
|
|
AC_ARG_ENABLE([orterun-prefix-by-default],
|
|
[AC_HELP_STRING([--enable-orterun-prefix-by-default],
|
|
[Make "orterun ..." behave exactly the same as "orterun --prefix \$prefix" (where \$prefix is the value given to --prefix in configure)])])
|
|
AC_ARG_ENABLE([mpirun-prefix-by-default],
|
|
[AC_HELP_STRING([--enable-mpirun-prefix-by-default],
|
|
[Synonym for --enable-orterun-prefix-by-default])])
|
|
if test "$enable_orterun_prefix_by_default" = ""; then
|
|
enable_orterun_prefix_by_default=$enable_mpirun_prefix_by_default
|
|
fi
|
|
if test "$enable_orterun_prefix_by_default" = "yes"; then
|
|
AC_MSG_RESULT([yes])
|
|
orte_want_orterun_prefix_by_default=1
|
|
else
|
|
AC_MSG_RESULT([no])
|
|
orte_want_orterun_prefix_by_default=0
|
|
fi
|
|
AC_DEFINE_UNQUOTED([ORTE_WANT_ORTERUN_PREFIX_BY_DEFAULT],
|
|
[$orte_want_orterun_prefix_by_default],
|
|
[Whether we want orterun to effect "--prefix $prefix" by default])
|
|
])
|