[f2py] f2py and lapack DLARUV

Paul Anton Letnes paul.anton.letnes at gmail.com
Fri May 20 19:12:59 EEST 2011


Aha, I see. I would still consider having a glance at how scipy wraps lapack, though. Maybe you can re-use some of their (open source!) code.

Paul

On 20. mai 2011, at 07.58, Stephane Raynaud wrote:

> I can't since my goal is to use a series of third party specific
> fortran modules that use lapack.
> However, I think about rewritting (one day...) these modules in python
> and use scipy as you suggest.
> 
> On Fri, May 20, 2011 at 4:36 PM, Paul Anton Letnes
> <paul.anton.letnes at gmail.com> wrote:
>> Just a thought, maybe I'm missing something, but: can't you simply use scipy? It contains thin wrappers for alot of LAPACK calls in:
>> from scipy.linalg.lapack import flapack as lapack
>> Now, I think this will give you a list of routine names that are wrapped:
>> for routine in dir(lapack):
>>        print routine
>> 
>> There is probably some "bloat" by using all of scipy, but it will save you a lot of compiling. Also, you could look at how scipy wraps lapack, and learn from that.
>> 
>> YMMV
>> Paul
>> 
>> On 20. mai 2011, at 05.02, Stephane Raynaud wrote:
>> 
>>> Since the only subroutine used from lapack is DLARUV with no
>>> dependencies, I folowed your excellent idea :
>>> ----
>>> f2py  -m testmod testfile.f90 -h testfile.pyf
>>> f2py testfile.pyf dlaruv.f
>>> ---
>>> 
>>> It failed on the second phase :
>>> ---
>>> |...]
>>>        Block: DLARUV
>>> Post-processing (stage 2)...
>>> Tip: If your original code is Fortran source then you must use -m option.
>>> [...]
>>> TypeError: All blocks must be python module blocks but got 'subroutine'
>>> ---
>>> 
>>> However, it work using the same idea with :
>>> ----
>>> f2py  -m testmod testfile.f90 -h testfile.pyf # setup interface only
>>> for testfunc
>>> f2py -c -m testmod testfile.f90 testfile.pyf dlaruv.f # build
>>> ---
>>> 
>>> 
>>> Thus, this solution is sufficient for me, thanks alot !!!
>>> 
>>> However, it would be great to have a solution for the when using the
>>> lapack/blas library itself.
>>> 
>>> On Fri, May 20, 2011 at 11:35 AM, Pearu Peterson
>>> <pearu.peterson at gmail.com> wrote:
>>>> 
>>>> 
>>>> On 05/20/2011 10:39 AM, Stephane Raynaud wrote:
>>>>> Hi,
>>>>> 
>>>>> trying to build a fortran extension linked to lapack (f77), I faced a
>>>>> problem that I reduced to the following example.
>>>>> 
>>>>> ------------ testfile.f90 ------------
>>>>> subroutine testfunc
>>>>>     integer,parameter :: n=2
>>>>>     integer :: iseed(4)
>>>>>     real(kind=8) :: x(n)
>>>>>     iseed = (/0,1,3,4/)
>>>>>     call DLARUV(iseed, n, x)
>>>>> end subroutine testfunc
>>>>> --------------------------------------
>>>>> 
>>>>> Then :  f2py -c -m testmod -llapack -lblas
>>>>> -L/home/raynaud/soft/lapack_gfortran/lib testfile.f90
>>>>> 
>>>>> It rises the following error (using gfortran) :
>>>>> 
>>>>>    /usr/bin/gfortran -Wall -Wall -shared
>>>>> /tmp/tmpzBIKPf/tmp/tmpzBIKPf/src.linux-x86_64-2.6/testmodule.o
>>>>> /tmp/tmpzBIKPf/tmp/tmpzBIKPf/src.linux-x86_64-2.6/fortranobject.o
>>>>> /tmp/tmpzBIKPf/test.o -L/home/raynaud/soft/lapack_gfortran/lib
>>>>> -L/soft/cdat/5.2.0/lib -llapack -lblas -lpython2.6 -lgfortran -o
>>>>> ./test.so
>>>>>    /usr/bin/ld:
>>>>> /home/raynaud/soft/lapack_gfortran/lib/liblapack.a(dlaruv.o):
>>>>> relocation R_X86_64_32S against `.data' can not be used when making a
>>>>> shared object; recompile with -fPIC
>>>>>    /home/raynaud/soft/lapack_gfortran/lib/liblapack.a: could not read
>>>>> symbols: Bad value
>>>>> 
>>>>> 
>>>>> BLAS and LAPACK where compilated with gfortran and options '-fPIC -O3'.
>>>> 
>>>> Are you absolutely sure?
>>>> 
>>>> Can you try creating the testmod without using lapack and blas libraries
>>>> but their sources? For example:
>>>> 
>>>> f2py  -m testmod testfile.f90 -h testfile.pyf
>>>> f2py testfile.pyf /path/to/lapack/src/*.f /path/to/blas/src/*.f
>>>> 
>>>> Pearu
>>>> 
>>>>> I tried to use the DATA statement directly in a subroutine that I
>>>>> converted to a python module, and it worked.
>>>>> 
>>>>> 
>>>>> Any idea ?
>>>>> 
>>>>> Thanks.
>>>>> 
>>>> 
>>>> _______________________________________________
>>>> f2py-users mailing list
>>>> f2py-users at cens.ioc.ee
>>>> http://cens.ioc.ee/mailman/listinfo/f2py-users
>>>> 
>>> 
>>> 
>>> 
>>> --
>>> Stephane Raynaud
>>> 
>>> _______________________________________________
>>> f2py-users mailing list
>>> f2py-users at cens.ioc.ee
>>> http://cens.ioc.ee/mailman/listinfo/f2py-users
>> 
>> 
>> _______________________________________________
>> f2py-users mailing list
>> f2py-users at cens.ioc.ee
>> http://cens.ioc.ee/mailman/listinfo/f2py-users
>> 
> 
> 
> 
> -- 
> Stephane Raynaud
> 
> _______________________________________________
> f2py-users mailing list
> f2py-users at cens.ioc.ee
> http://cens.ioc.ee/mailman/listinfo/f2py-users




More information about the f2py-users mailing list