[Insight-users] Build error on OSX: wrap_VXLNumericsPythonPython.o
Gaëtan Lehmann
gaetan.lehmann at jouy.inra.fr
Sat May 16 03:15:48 EDT 2009
Le 14 mai 09 à 22:30, Darren Weber a écrit :
>
>
> 2009/5/14 Gaëtan Lehmann <gaetan.lehmann at jouy.inra.fr>
>
> Le 12 mai 09 à 20:26, Darren Weber a écrit :
>
>
>
>
> 2009/5/12 Gaëtan Lehmann <gaetan.lehmann at jouy.inra.fr>
>
> Le 12 mai 09 à 19:46, Darren Weber a écrit :
>
>
> Hi,
>
> What's the fix for this issue with dynamic lib builds with python
> wrapping in ITK (3.12)? The linker appears to be building a dynamic
> lib (or shared lib, *.so), but perhaps the linker option "-
> dynamiclib" is missing somewhere in the cmake config? Please tell
> me if there is simple patch to fix this. Could it be an issue with
> using "-Wno-long-double" for the vnl numerics wrapping?
>
> This build uses CableSwig build options (ie, cmake ... -D
> ITK_CSWIG_PYTHON:BOOL=ON ...), rather than USE_WRAP_ITK. What is
> the most preferred wrapping method for ITK @3.12? I noticed mention
> of a new release, 3.14, in the wings and nearly available. Does the
> new release (3.14) resolve build problems for wrappers (or is the
> focus of that release elsewhere)?
>
>
> Hi,
>
> I can't recall any problem while building WrapITK on Mac OS X.
> Have you tried with WrapITK?
>
> Regards,
>
> Gaëtan
>
>
> I posted a build error the other day when USE_WRAP_ITK was ON. The
> title of that post was "ITK wrap build error on
> itkLaplacianRecursiveGaussianImageFilterPython.o" The error
> message was obtuse. I've had no feedback on that post.
>
>
> ok, I've missed that one. The message is not much informative for me
> as well.
> It seems that you are using stuff from macport. Can you try to build
> cable swig and itk with the compiler provided with xcode and the
> default compiler options? That's what I'm doing here without trouble.
>
> You can disable most of the WRAP_* options in cmake for the tests,
> and keep only one dimension in WRAP_ITK_DIMS, to make things a bit
> faster.
>
>
> As a MacPorts developer, working on a Portfile for ITK called
> InsightToolkit, I need to and want to work with MacPorts.
ok, sounds reasonable :-)
> I'm currently updating this port to ITK @3.12.0. MacPorts
> provides a way to document and automate build configurations,
> including dependencies. It helps me, while simultaneously making my
> work available to anyone using MacPorts. AFAIK, the MacPorts
> compiler is provided by an xcode installation (I can double check
> the defaults or change to a specific compiler at will). Most of the
> ITK build is fine (as expected), but I've run into troubles with
> wrapping ITK. It looks like it's best to stick with python wrapping
> (leave out java and tcl, or whatever else, for now). As a python
> programmer myself, that preference is understandable, although I
> thought tcl would be fine too, but maybe not. I have too much to
> learn about wrapping - ignorance is bliss, until you need something
> to work when it doesn't work.
>
> BTW, there is a VTK 5.4.0 port in vtk-devel in macports. In that
> port, just about everything is worked out, including wrapping, MPI,
> Boost, databases, etc. I'm also working on a cableswig port, which
> builds and installs fine, but there's an issue with finding the SWIG
> Lib directory for cswig.
VTK and ITK wrapping process are very different. Having one working
perfectly gives not much information on how the other will work.
>
>
> So, this port development is working with the wrapITK in the ITK
> 3.12.0 release. It also pulls in the CableSwig-ITK-3.12.0 release
> and puts it into <ITK-srcPath>/Utilities/CableSwig and assumes that
> this cableswig is built by default (i.e., no cmake variables are
> explicitly set to switch on the build for it). The cmake vars for
> the wrapITK are specified:
>
> -DUSE_WRAP_ITK:BOOL=ON \
> -DITK_USE_REVIEW:BOOL=ON \
> -DGUILE_EXECUTABLE:FILEPATH=${prefix}/bin/guile \
> -DGUILE_LIBRARY:FILEPATH=${prefix}/lib/libguile.dylib \
> -DGUILE_INCLUDE_PATH:PATH=${prefix}/include/guile \
> -DSWIG_BUILD_EXAMPLES:BOOL=ON
>
> The following are set by default (maybe a cmake module does this?):
>
> #WRAP_BinaryMorphology ON
> #WRAP_Calculators ON
> #WRAP_Compose ON
> #WRAP_DeformableTransforms ON
> #WRAP_Denoising ON
> #WRAP_DistanceMap ON
> #WRAP_EdgesAndContours ON
> #WRAP_FFT ON
> #WRAP_FFTW OFF
> #WRAP_Filtering ON
> #WRAP_IO ON
> #WRAP_ITK_DIMS 2;3
> #WRAP_ITK_INSTALL_PREFIX /lib/InsightToolkit/WrapITK
> #WRAP_ITK_JAVA ON
> #WRAP_ITK_PYTHON ON
> #WRAP_ITK_TCL ON
> #WRAP_IntensityFilters ON
> #WRAP_Interpolators ON
> #WRAP_Iterators OFF
> #WRAP_LevelSet ON
> #WRAP_Morphology ON
> #WRAP_PixelMath ON
> #WRAP_Registration ON
> #WRAP_Resize ON
> #WRAP_Review ON
> #WRAP_SegmentationAndThreshold ON
> #WRAP_SegmentationValidation ON
> #WRAP_SimpleFilters ON
> #WRAP_UnaryPixelMath ON
> #WRAP_complex_double OFF
> #WRAP_complex_float ON
> #WRAP_covariant_vector_double OFF
> #WRAP_covariant_vector_float ON
> #WRAP_double OFF
> #WRAP_float ON
> #WRAP_rgb_unsigned_char OFF
> #WRAP_rgb_unsigned_short ON
> #WRAP_rgba_unsigned_char OFF
> #WRAP_rgba_unsigned_short ON
> #WRAP_signed_char OFF
> #WRAP_signed_long OFF
> #WRAP_signed_short OFF
> #WRAP_unsigned_char OFF
> #WRAP_unsigned_long OFF
> #WRAP_unsigned_short ON
> #WRAP_vector_double OFF
> #WRAP_vector_float ON
>
That's the default options for wrapitk.
It looks ok.
Gaëtan
--
Gaëtan Lehmann
Biologie du Développement et de la Reproduction
INRA de Jouy-en-Josas (France)
tel: +33 1 34 65 29 66 fax: 01 34 65 29 09
http://voxel.jouy.inra.fr http://www.mandriva.org
http://www.itk.org http://www.clavier-dvorak.org
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 203 bytes
Desc: Ceci est une signature ?lectronique PGP
URL: <http://www.itk.org/pipermail/insight-users/attachments/20090516/a1635a59/attachment.pgp>
More information about the Insight-users
mailing list