Open MPI logo

Open MPI User's Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Open MPI User's mailing list

Subject: Re: [OMPI users] segmentation fault in openmpi-1.9a1r27285
From: Ralph Castain (rhc_at_[hidden])
Date: 2012-09-14 09:54:59


If you look at the ompi_info output, you'll see this:

  FT Checkpoint support: no (checkpoint thread: no)
  C/R Enabled Debugging: no

FT is correctly disabled. As I said before, I believe your prior test was using the previously-built version where you had explicitly enabled FT. This sometimes happens because people forget to clear out the prefix location before doing another configure/build. This leaves old libraries laying around, which get picked up when we execute. Remember, when we install, we don't remove anything - so if we don't build something, we can still pickup the library from a prior build and attempt to use it.

On Sep 14, 2012, at 5:51 AM, Siegmar Gross <Siegmar.Gross_at_[hidden]> wrote:

> Hi,
>
> in openmpi-1.9a1r27285 I got a segmentation fault while somehow
> "FT Checkpoint support" was enabled. When I compile openmpi-1.7a1r27338
> with the same configuration options "FT Checkpoint" is disabled so that
> in my opinion it will be automatically enabled in openmpi-1.9a1r27285.
>
> ../openmpi-1.7a1r27338/configure --prefix=/usr/local/openmpi-1.7_64_gcc \
> --libdir=/usr/local/openmpi-1.7_64_gcc/lib64 \
> LDFLAGS="-m64 -L/usr/local/gcc-4.6.2/lib/sparcv9" \
> CC="gcc" CXX="g++" F77="gfortran" FC="gfortran" \
> CFLAGS="-m64" CXXFLAGS="-m64" FFLAGS="-m64" FCFLAGS="-m64" \
> CPP="cpp" CXXCPP="cpp" \
> CPPFLAGS="" CXXCPPFLAGS="" \
> C_INCL_PATH="" C_INCLUDE_PATH="" CPLUS_INCLUDE_PATH="" \
> OBJC_INCLUDE_PATH="" MPIHOME="" \
> --with-wrapper-cflags=-m64 \
> --without-udapl --without-openib \
> --enable-mpi-f90 --with-mpi-f90-size=small \
> --enable-heterogeneous --enable-cxx-exceptions \
> --enable-orterun-prefix-by-default \
> --with-threads=posix --enable-mpi-thread-multiple \
> --enable-opal-multi-threads \
> --with-hwloc=internal \
> |& tee log.configure.$SYSTEM_ENV.$MACHINE_ENV.64_gcc
>
>
> tyr small_prog 122 ompi_info
> Package: Open MPI fd1026_at_[hidden]
> Distribution
> Open MPI: 1.7a1r27338
> Open MPI repo revision: r27338
> Open MPI release date: Sep 13, 2012 (nightly snapshot tarball)
> Open RTE: 1.7a1r27338
> Open RTE repo revision: r27338
> Open RTE release date: Sep 13, 2012 (nightly snapshot tarball)
> OPAL: 1.7a1r27338
> OPAL repo revision: r27338
> OPAL release date: Sep 13, 2012 (nightly snapshot tarball)
> MPI API: 2.1
> Ident string: 1.7a1r27338
> Prefix: /usr/local/openmpi-1.7_64_cc
> Configured architecture: sparc-sun-solaris2.10
> Configure host: tyr.informatik.hs-fulda.de
> Configured by: fd1026
> Configured on: Fri Sep 14 10:48:04 CEST 2012
> Configure host: tyr.informatik.hs-fulda.de
> Built by: fd1026
> Built on: Fri Sep 14 12:04:50 CEST 2012
> Built host: tyr.informatik.hs-fulda.de
> C bindings: yes
> C++ bindings: yes
> Fort mpif.h: yes (all)
> Fort use mpi: yes (full: ignore TKR)
> Fort use mpi size: deprecated-ompi-info-value
> Fort use mpi_f08: yes
> Fort mpi_f08 compliance: The mpi_f08 module is available, but due to
> limitations in the f95 compiler, does not support
> the following: array subsections, ABSTRACT
> INTERFACE function pointers, Fortran '08-specified
> ASYNCHRONOUS behavior, PROCEDUREs, direct passthru
> (where possible) to underlying Open MPI's C
> functionality
> Fort mpi_f08 subarrays: no
> Java bindings: no
> C compiler: cc
> C compiler absolute: /opt/solstudio12.3/bin/cc
> C compiler family name: SUN
> C compiler version: 0x5120
> C++ compiler: CC
> C++ compiler absolute: /opt/solstudio12.3/bin/CC
> Fort compiler: f95
> Fort compiler abs: /opt/solstudio12.3/bin/f95
> Fort ignore TKR: yes (!$PRAGMA IGNORE_TKR)
> Fort 08 assumed shape: no
> Fort optional args: yes
> Fort BIND(C): yes
> Fort PRIVATE: yes
> Fort ABSTRACT: no
> Fort ASYNCHRONOUS: no
> Fort PROCEDURE: no
> Fort f08 using wrappers: yes
> C profiling: yes
> C++ profiling: yes
> Fort mpif.h profiling: yes
> Fort use mpi profiling: yes
> Fort use mpi_f08 prof: yes
> C++ exceptions: yes
> Thread support: posix (MPI_THREAD_MULTIPLE: yes, progress: no)
> Sparse Groups: no
> Internal debug support: no
> MPI interface warnings: yes
> MPI parameter check: runtime
> Memory profiling support: no
> Memory debugging support: no
> libltdl support: yes
> Heterogeneous support: yes
> mpirun default --prefix: yes
> MPI I/O support: yes
> MPI_WTIME support: native
> Symbol vis. support: yes
> Host topology support: yes
> MPI extensions:
> FT Checkpoint support: no (checkpoint thread: no)
> C/R Enabled Debugging: no
> VampirTrace support: yes
> MPI_MAX_PROCESSOR_NAME: 256
> MPI_MAX_ERROR_STRING: 256
> MPI_MAX_OBJECT_NAME: 64
> MPI_MAX_INFO_KEY: 36
> MPI_MAX_INFO_VAL: 256
> MPI_MAX_PORT_NAME: 1024
> MPI_MAX_DATAREP_STRING: 128
> MCA backtrace: printstack (MCA v2.0, API v2.0, Component v1.7)
> ...
>
>
>>> Afraid I have no earthly idea what is going on with your
>>> installation - like I said, that code cannot build with the
>>> configure line you've shown me, so something must be getting confused.
>>>
>>> Afraid I'll have to punt :-(
> ...
>> I played with some other commands as well.
>>
>> tyr bin 140 orte-info
>> Package: Open MPI fd1026_at_[hidden]
>> Distribution
>> Open RTE: 1.9a1r27285
>> Open RTE repo revision: r27285
>> Open RTE release date: Sep 10, 2012 (nightly snapshot tarball)
>> OPAL: 1.9a1r27285
> ...
>> FT Checkpoint support: yes (checkpoint thread: yes)
> ...
>
>> Do you see where I can have enabled "FT Checkpoint support" in my
>> configure command?
>>
>> ../openmpi-1.9a1r27285/configure --prefix=/usr/local/openmpi-1.9_64_cc \
>> --libdir=/usr/local/openmpi-1.9_64_cc/lib64 \
>> LDFLAGS="-m64" \
>> CC="cc" CXX="CC" F77="f77" FC="f95" \
>> CFLAGS="-m64" CXXFLAGS="-m64 -library=stlport4" FFLAGS="-m64" \
>> FCFLAGS="-m64" \
>> CPP="cpp" CXXCPP="cpp" \
>> CPPFLAGS="" CXXCPPFLAGS="" \
>> C_INCL_PATH="" C_INCLUDE_PATH="" CPLUS_INCLUDE_PATH="" \
>> OBJC_INCLUDE_PATH="" MPIHOME="" \
>> --without-udapl --without-openib \
>> --enable-mpi-f90 --with-mpi-f90-size=small \
>> --enable-heterogeneous --enable-cxx-exceptions \
>> --enable-orterun-prefix-by-default \
>> --with-threads=posix --enable-mpi-thread-multiple \
>> --enable-opal-multi-threads \
>> --with-hwloc=internal \
>> |& tee log.configure.$SYSTEM_ENV.$MACHINE_ENV.64_cc
>
> Does anybody know why "FT Checkpoint" is enabled in openmpi-1.9a1r27285?
> Thank you very much for any help in advance.
>
>
> Kind regards
>
> Siegmar
>