https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750725#M7243 yes, MKL does not support the full range of transforms, e.g. R2HC and HC2R<DIV></DIV><DIV>and in any event, the library should build, so it would be useful to know what the problem is ...</DIV><DIV></DIV> Fri, 19 Nov 2010 00:41:56 GMT Hans_J_ 2010-11-19T00:41:56Z https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750723#M7241 HI All,<DIV></DIV><DIV>I am running OS 10.6.5 on an early 2008 Mac Pro, with the recent Xcode 3.2.4.</DIV><DIV></DIV><DIV>I just installed the new version 12.0 compilers, C++ and Fortran. To set them up I use</DIV><DIV></DIV><DIV> source /opt/intel/composerxe/bin/compilervars.sh intel64</DIV><DIV></DIV><DIV>I then untarred fftw3.2.2.tar and ran</DIV><DIV></DIV><DIV>./configure CC=icc F77=ifort --with-our-malloc16 --enable-openmp --enable-sse2</DIV><DIV></DIV><DIV>Then I do</DIV><DIV></DIV><DIV> make check</DIV><DIV></DIV><DIV>Lots of things happen, but it then craps out, ending with</DIV><DIV></DIV><DIV><DIV id="_mcePaste">Executing "/Users/johnston/Numerical_Packages/fftw-3.2.2/tests/bench --verbose=1 --verify '//obc10x13x5x10v2' --verify '//ibc10x13x5x10v2' --verify '//ofc10x13x5x10v2' --verify '//ifc10x13x5x10v2' --verify 'obc10x13x5x10v2' --verify 'ibc10x13x5x10v2' --verify 'ofc10x13x5x10v2' --verify 'ifc10x13x5x10v2' --verify 'ok13bx9e11*8' --verify 'ik13bx9e11*8' --verify 'obr33x22v12' --verify 'ibr33x22v12' --verify 'ofr33x22v12' --verify 'ifr33x22v12' --verify '//obc33x22v12' --verify '//ibc33x22v12' --verify '//ofc33x22v12' --verify '//ifc33x22v12' --verify 'obc33x22v12' --verify 'ibc33x22v12' --verify 'ofc33x22v12' --verify 'ifc33x22v12' --verify 'ok120e01*168' --verify 'ik120e01*168' --verify 'obr8x5x2x13*8' --verify 'ibr8x5x2x13*8' --verify 'ofr8x5x2x13*8' --verify 'ifr8x5x2x13*8' --verify '//obc8x5x2x13*8' --verify '//ibc8x5x2x13*8' --verify '//ofc8x5x2x13*8' --verify '//ifc8x5x2x13*8' --verify 'obc8x5x2x13*8' --verify 'ibc8x5x2x13*8' --verify 'ofc8x5x2x13*8' --verify 'ifc8x5x2x13*8' --verify 'ok4o00v25' --verify 'ik4o00v25' --verify '//obr5x2x3x3' --verify '//ofr5x2x3x3' --verify 'obr5x2x3x3' --verify 'ibr5x2x3x3' --verify 'ofr5x2x3x3' --verify 'ifr5x2x3x3'"</DIV><DIV id="_mcePaste">//obc10x13x5x10v2 4.78188e-16 2.18127e-15 7.68733e-16</DIV><DIV id="_mcePaste">//ibc10x13x5x10v2 4.33957e-16 2.18127e-15 8.49837e-16</DIV><DIV id="_mcePaste">//ofc10x13x5x10v2 4.00962e-16 2.04907e-15 8.27465e-16</DIV><DIV id="_mcePaste">//ifc10x13x5x10v2 3.72694e-16 2.18127e-15 8.52547e-16</DIV><DIV id="_mcePaste">FAILED /Users/johnston/Numerical_Packages/fftw-3.2.2/tests/bench: --verify '//obc10x13x5x10v2' --verify '//ibc10x13x5x10v2' --verify '//ofc10x13x5x10v2' --verify '//ifc10x13x5x10v2' --verify 'obc10x13x5x10v2' --verify 'ibc10x13x5x10v2' --verify 'ofc10x13x5x10v2' --verify 'ifc10x13x5x10v2' --verify 'ok13bx9e11*8' --verify 'ik13bx9e11*8' --verify 'obr33x22v12' --verify 'ibr33x22v12' --verify 'ofr33x22v12' --verify 'ifr33x22v12' --verify '//obc33x22v12' --verify '//ibc33x22v12' --verify '//ofc33x22v12' --verify '//ifc33x22v12' --verify 'obc33x22v12' --verify 'ibc33x22v12' --verify 'ofc33x22v12' --verify 'ifc33x22v12' --verify 'ok120e01*168' --verify 'ik120e01*168' --verify 'obr8x5x2x13*8' --verify 'ibr8x5x2x13*8' --verify 'ofr8x5x2x13*8' --verify 'ifr8x5x2x13*8' --verify '//obc8x5x2x13*8' --verify '//ibc8x5x2x13*8' --verify '//ofc8x5x2x13*8' --verify '//ifc8x5x2x13*8' --verify 'obc8x5x2x13*8' --verify 'ibc8x5x2x13*8' --verify 'ofc8x5x2x13*8' --verify 'ifc8x5x2x13*8' --verify 'ok4o00v25' --verify 'ik4o00v25' --verify '//obr5x2x3x3' --verify '//ofr5x2x3x3' --verify 'obr5x2x3x3' --verify 'ibr5x2x3x3' --verify 'ofr5x2x3x3' --verify 'ifr5x2x3x3'</DIV><DIV id="_mcePaste">received signal 8</DIV><DIV id="_mcePaste">make[2]: *** [check-local] Error 1</DIV><DIV id="_mcePaste">make[1]: *** [check-am] Error 2</DIV><DIV id="_mcePaste">make: *** [check-recursive] Error 1</DIV><DIV id="_mcePaste">boussinesq:fftw-3.2.2 johnston$</DIV></DIV><DIV></DIV><DIV></DIV><DIV>Has anyone been successful in building this package with the new compilers?</DIV><DIV></DIV><DIV>Thanks, Hans</DIV><DIV></DIV> Thu, 18 Nov 2010 21:01:09 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750723#M7241 Hans_J_ 2010-11-18T21:01:09Z https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750724#M7242 Is there a specific reason that you do not wish to take the easier route: use the sources and makefile provided by Intel in the <INTEL_INSTALLATION_DIR>/mkl/interfaces/*fftw* directories?</INTEL_INSTALLATION_DIR> Fri, 19 Nov 2010 00:31:00 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750724#M7242 mecej4 2010-11-19T00:31:00Z https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750725#M7243 yes, MKL does not support the full range of transforms, e.g. R2HC and HC2R<DIV></DIV><DIV>and in any event, the library should build, so it would be useful to know what the problem is ...</DIV><DIV></DIV> Fri, 19 Nov 2010 00:41:56 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750725#M7243 Hans_J_ 2010-11-19T00:41:56Z https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750726#M7244 UPDATE:<DIV></DIV><DIV></DIV><DIV>NO PROBLEM IF I setup the compilers using ia32 instead of intel64, i.e.,</DIV><DIV></DIV><DIV></DIV><DIV>source /opt/intel/composerxe/bin/compilervars.sh ia32</DIV><DIV></DIV><DIV></DIV><DIV>Any idea why?</DIV> Fri, 19 Nov 2010 20:35:42 GMT https://community.intel.com/t5/Intel-Fortran-Compiler/trouble-building-fftw3-2-2-with-new-compilers/m-p/750726#M7244 Hans_J_ 2010-11-19T20:35:42Z