[Numpy-discussion] libmkl_lapack error in numpy

akshar bhosale akshar.bhosale at gmail.com
Sun Oct 23 03:35:07 EDT 2011


Hi,

libmkl_lapack.so is added in site.cfg and now the matrix function is not
giving an error, but numpy.test hangs.


On Sun, Oct 23, 2011 at 11:24 AM, akshar bhosale
<akshar.bhosale at gmail.com>wrote:

> i am getting following error.
>    python -c 'import numpy;numpy.matrix([[1, 5, 10], [1.0, 3j, 4]],
>    numpy.complex128).T.I.H'
>    MKL FATAL ERROR: Cannot load libmkl_lapack.so
>
>    have installed numpy 1.6.0 with python 2.6.
>    i have intel cluster toolkit installed on my system. (11/069 version and
> mlk=10.3). i have machine having intel xeon processor and rhel 5.2 x86_64
>    platform. i am trying with intel compilers.
>    if i do
>    python -c 'import numpy;numpy.matrix([[1, 5, 10], [1.0, 3j, 4]],
>    numpy.complex128).T.I.H'
>    python: symbol lookup error:
>    /opt/intel/Compiler/11.0/069/
> mkl/lib/em64/libmkl_lapack.so: undefined
>    symbol: mkl_lapack_zgeqrf
>    my site.cfg is :
>    ####################
>    [mkl]
>    mkl_libs =3D mkl_def, mkl_intel_lp64, mkl_intel_thread, mkl_core, mkl_mc
>
> - Ignored:
>    lapack_libs =3D mkl_lapack95_lp64
>
>    library_dirs =3D /opt/intel/Compiler/11.0/069/
>    mkl/lib/em64t:/opt/intel/Compiler/11.0/069/lib/intel64/
>    include_dirs =3D
>
>  /opt/intel/Compiler/11.0/069/mkl/include:/opt/intel/Compiler/11.0/069/inclu=
>    de/
>    ####################
>    and intelcompiler.py is :
>    ############################
>    from distutils.unixccompiler import UnixCCompiler
>    from numpy.distutils.exec_command import find_executable
>    import sys
>
>    class IntelCCompiler(UnixCCompiler):
>        """ A modified Intel compiler compatible with an gcc built
> Python."""
>        compiler_type =3D 'intel'
>        cc_exe =3D 'icc'
>        cc_args =3D 'fPIC'
>
>        def __init__ (self, verbose=3D0, dry_run=3D0, force=3D0):
>            sys.exit(0)
>            UnixCCompiler.__init__ (self, verbose,dry_run, force)
>            self.cc_exe =3D 'icc -fPIC '
>            compiler =3D self.cc_exe
>            self.set_executables(compiler=3Dcompiler,
>                                 compiler_so=3Dcompiler,
>                                 compiler_cxx=3Dcompiler,
>                                 linker_exe=3Dcompiler,
>                                 linker_so=3Dcompiler + ' -shared -lstdc++')
>
>    class IntelItaniumCCompiler(IntelCCompiler):
>        compiler_type =3D 'intele'
>
>        # On Itanium, the Intel Compiler used to be called ecc, let's search
> fo=
>    r
>        # it (now it's also icc, so ecc is last in the search).
>        for cc_exe in map(find_executable,['icc','ecc']):
>            if cc_exe:
>                break
>
>    class IntelEM64TCCompiler(UnixCCompiler):
>        """ A modified Intel x86_64 compiler compatible with a 64bit gcc
> built
>    Python.
>        """
>        compiler_type =3D 'intelem'
>        cc_exe =3D 'icc -m64 -fPIC'
>        cc_args =3D "-fPIC -openmp"
>        def __init__ (self, verbose=3D0, dry_run=3D0, force=3D0):
>            UnixCCompiler.__init__ (self, verbose,dry_run, force)
>            self.cc_exe =3D 'icc -m64 -fPIC -openmp '
>            compiler =3D self.cc_exe
>            self.set_executables(compiler=3Dcompiler,
>                                 compiler_so=3Dcompiler,
>                                 compiler_cxx=3Dcompiler,
>                                 linker_exe=3Dcompiler,
>                                 linker_so=3Dcompiler + ' -shared -lstdc++')
>    ##########################
>    LD_LIBRARY_PATH is :
>    #########################
>
>  /opt/intel/Compiler/11.0/069/mkl/lib/em64t:/opt/scalasca-1.3.3/lib:/opt/PBS=
>
>  /lib:/opt/intel/mpi/lib64:/opt/maui/lib:/opt/jdk1.6.0_23/lib:/opt/intel/Com=
>
>  piler/11.0/069/ipp/em64t/sharedlib:/opt/intel/Compiler/11.0/069/mkl/lib/em6=
>
>  4t:/opt/intel/Compiler/11.0/069/tbb/em64t/cc4.1.0_libc2.4_kernel2.6.16.21/l=
>
>  ib:/opt/intel/Compiler/11.0/069/lib/intel64:/opt/intel/Compiler/11.0/069/ip=
>
>  p/em64t/sharedlib:/opt/intel/Compiler/11.0/069/mkl/lib/em64t:/opt/intel/Com=
>
>  piler/11.0/069/tbb/em64t/cc4.1.0_libc2.4_kernel2.6.16.21/lib:/opt/intel/Com=
>    piler/11.0/069/lib/intel64:/usr/local/lib
>    #########################
>
> - Done.
>
>
>
> ---------- Forwarded message ----------
> From: akshar bhosale <akshar.bhosale at gmail.com>
> To: numpy-discussion-request at scipy.org
> Date: Sun, 23 Oct 2011 11:19:27 +0530
> Subject: libmkl_lapack error
> Hi,
> i am getting following error.
> python -c 'import numpy;numpy.matrix([[1, 5, 10], [1.0, 3j, 4]],
> numpy.complex128).T.I.H'
> MKL FATAL ERROR: Cannot load libmkl_lapack.so
>
> have installed numpy 1.6.0 with python 2.6.
>
> i have intel cluster toolkit installed on my system. (11/069 version and
> mkl 10.3). i have machine having intel xeon processor and rhel 5.2 x86_64
> platform. i am trying with intel compilers.
> if i do
>
> python -c 'import numpy;numpy.matrix([[1, 5, 10], [1.0, 3j, 4]],
> numpy.complex128).T.I.H'
> python: symbol lookup error:
> /opt/intel/Compiler/11.0/069/mkl/lib/em64/libmkl_lapack.so: undefined
> symbol: mkl_lapack_zgeqrf
>
> my site.cfg is :
> ####################
> [mkl]
>
> mkl_libs = mkl_def, mkl_intel_lp64, mkl_intel_thread, mkl_core, mkl_mc
> lapack_libs = mkl_lapack95_lp64
>
> library_dirs = /opt/intel/Compiler/11.0/069/
> mkl/lib/em64t:/opt/intel/Compiler/11.0/069/lib/intel64/
> include_dirs =
> /opt/intel/Compiler/11.0/069/mkl/include:/opt/intel/Compiler/11.0/069/include/
> ####################
> and intelcompiler.py is :
> ############################
> from distutils.unixccompiler import UnixCCompiler
> from numpy.distutils.exec_command import find_executable
> import sys
>
> class IntelCCompiler(UnixCCompiler):
>     """ A modified Intel compiler compatible with an gcc built Python."""
>     compiler_type = 'intel'
>     cc_exe = 'icc'
>     cc_args = 'fPIC'
>
>     def __init__ (self, verbose=0, dry_run=0, force=0):
>         sys.exit(0)
>         UnixCCompiler.__init__ (self, verbose,dry_run, force)
>         self.cc_exe = 'icc -fPIC '
>         compiler = self.cc_exe
>         self.set_executables(compiler=compiler,
>                              compiler_so=compiler,
>                              compiler_cxx=compiler,
>                              linker_exe=compiler,
>                              linker_so=compiler + ' -shared -lstdc++')
>
> class IntelItaniumCCompiler(IntelCCompiler):
>     compiler_type = 'intele'
>
>     # On Itanium, the Intel Compiler used to be called ecc, let's search
> for
>     # it (now it's also icc, so ecc is last in the search).
>     for cc_exe in map(find_executable,['icc','ecc']):
>         if cc_exe:
>             break
>
> class IntelEM64TCCompiler(UnixCCompiler):
>     """ A modified Intel x86_64 compiler compatible with a 64bit gcc built
> Python.
>     """
>     compiler_type = 'intelem'
>     cc_exe = 'icc -m64 -fPIC'
>     cc_args = "-fPIC -openmp"
>     def __init__ (self, verbose=0, dry_run=0, force=0):
>         UnixCCompiler.__init__ (self, verbose,dry_run, force)
>         self.cc_exe = 'icc -m64 -fPIC -openmp '
>         compiler = self.cc_exe
>         self.set_executables(compiler=compiler,
>                              compiler_so=compiler,
>                              compiler_cxx=compiler,
>                              linker_exe=compiler,
>                              linker_so=compiler + ' -shared -lstdc++')
> ##########################
> LD_LIBRARY_PATH is :
> #########################
>
> /opt/intel/Compiler/11.0/069/mkl/lib/em64t:/opt/scalasca-1.3.3/lib:/opt/PBS/lib:/opt/intel/mpi/lib64:/opt/maui/lib:/opt/jdk1.6.0_23/lib:/opt/intel/Compiler/11.0/069/ipp/em64t/sharedlib:/opt/intel/Compiler/11.0/069/mkl/lib/em64t:/opt/intel/Compiler/11.0/069/tbb/em64t/cc4.1.0_libc2.4_kernel2.6.16.21/lib:/opt/intel/Compiler/11.0/069/lib/intel64:/opt/intel/Compiler/11.0/069/ipp/em64t/sharedlib:/opt/intel/Compiler/11.0/069/mkl/lib/em64t:/opt/intel/Compiler/11.0/069/tbb/em64t/cc4.1.0_libc2.4_kernel2.6.16.21/lib:/opt/intel/Compiler/11.0/069/lib/intel64:/usr/local/lib
> #########################
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20111023/b1a99892/attachment.html>


More information about the NumPy-Discussion mailing list