I attached a site.cfg file for numpy 1.3 compiled with MKL on some Linux 64 bit architecture, in case it might help.<br>I always had trouble getting programs (other than numpy though) to link and execute properly with MKL.<br>
You might also try to play with LD_PRELOAD.<br>Good luck,<br><br>-=- Olivier<br><br><div class="gmail_quote">2011/11/23 Xavier Barthelemy <span dir="ltr"><<a href="mailto:xabart@gmail.com">xabart@gmail.com</a>></span><br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Dear everyone<br><br>I have been a bit better<br><br>patching the files I finally get something<br><br>in .basrc<br><br>
source /opt/intel/composer_xe_2011_sp1.7.256/bin/compilervars.sh intel64<br>source /opt/intel/composer_xe_2011_sp1.7.256/mkl/bin/mklvars.sh intel64 lp64<br>
export OMP_NUM_THREADS=4<br>export KMP_AFFINITY=verbose,compact<br><br>env variables:<br>export LD_LIBRARY_PATH=/opt/intel/composer_xe_2011_sp1.7.256/mkl/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64/:/opt/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/intel64/lp64:/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/:$LD_LIBRARY_PATH<br>

<br>site.cfg<br><br>[mkl]<br>library_dirs = /opt/intel/composer_xe_2011_sp1.7.256/mkl/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64/:/opt/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/intel64/lp64/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64/<br>

include_dirs = /opt/intel/composer_xe_2011_sp1.7.256/mkl/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64/:/opt/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/intel64/lp64/:/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/<br>

lapack_libs =  <br>mkl_libs = mkl_def, mkl_intel_lp64, mkl_core , mkl_intel_thread, mkl_mc<br><br>distutils/intelcompiler.py<br><br>cc_exe = 'icc -m64 -O2 -fomit-frame-pointer -openmp -lpthread -fPIC -xHost -fp-model strict -L/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64 -I/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/intel64/lp64 -lm'<br>

 ccxx_exe = 'icpc -m64 -O2 -fomit-frame-pointer -openmp -lpthread -fPIC -xHost -fp-model strict -L/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64 -I/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/intel64/lp64 -lm'<br>

<br><br>distutils/fcompiler/intel.py<br><br>return ['-O3 -openmp -fpe0 -L/opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64 -lpthread -xHost -I/opt/intel/composer_xe_2011_sp1.7.256/mkl/include/intel64/lp64 -I/opt/intel/composer_xe_2011_sp1.7.256/mkl/include -lm']<br>

<br>in numpy.test("full")<br><br>i get test_double (test_linalg.TestCond2) ... python: symbol lookup error: /opt/intel/composer_xe_2011_sp1.7.256/mkl/lib/intel64/libmkl_sequential.so: undefined symbol: mkl_serv_lock<br>

<br> <br><br>I am trying to look for mkl_serv_lock, but impossible to find.<br><br>some ideas, someone?<span class="HOEnZb"><font color="#888888"><br><br>Xavier<br clear="all"><br>
</font></span><br>_______________________________________________<br>
NumPy-Discussion mailing list<br>
<a href="mailto:NumPy-Discussion@scipy.org">NumPy-Discussion@scipy.org</a><br>
<a href="http://mail.scipy.org/mailman/listinfo/numpy-discussion" target="_blank">http://mail.scipy.org/mailman/listinfo/numpy-discussion</a><br>
<br></blockquote></div><br>