[Numpy-discussion] ANN: Numpy 1.6.0 beta 1

Derek Homeier derek@astro.physik.uni-goettingen...
Thu Mar 24 18:44:09 CDT 2011


On 24.03.2011, at 9:11AM, Pearu Peterson wrote:
>  
> Intel-64bit:
> ERROR: test_assumed_shape.TestAssumedShapeSumExample.test_all
> ----------------------------------------------------------------------
> Traceback (most recent call last):
>  File "/sw/lib/python3.2/site-packages/nose/case.py", line 372, in setUp
>    try_run(self.inst, ('setup', 'setUp'))
>  File "/sw/lib/python3.2/site-packages/nose/util.py", line 478, in try_run
>    return func()
>  File "/sw/lib/python3.2/site-packages/numpy/f2py/tests/util.py", line 352, in setUp
>    module_name=self.module_name)
>  File "/sw/lib/python3.2/site-packages/numpy/f2py/tests/util.py", line 73, in wrapper
>    memo[key] = func(*a, **kw)
>  File "/sw/lib/python3.2/site-packages/numpy/f2py/tests/util.py", line 134, in build_module
>    % (cmd[4:], asstr(out)))
> RuntimeError: Running f2py failed: ['-m', '_test_ext_module_5403', '/var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpfiy1jn/foo_free.f90', '/var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpfiy1jn/foo_use.f90', '/var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpfiy1jn/precision.f90']
>    Reading .f2py_f2cmap ...
>    Mapping "real(kind=rk)" to "double"
> 
> Hmm, this should not happen as real(kind=rk) should be mapped to "float".
> It seems that you have .f2py_f2cmap file lying around. Could you remove it and try again.

Yes, it's in the tarball and was installed together with the f2py tests!
tar tzvf /sw/src/numpy-1.6.0b1.tar.gz  | grep f2py_f2                                                      
-rw-r--r-- zouzoujing/staff      29 2011-03-15 06:22 numpy-1.6.0b1/numpy/f2py/tests/src/assumed_shape/.f2py_f2cmap
cat /sw/lib/python3.2/site-packages/numpy/f2py/tests/src/assumed_shape/.f2py_f2cmap                                                                  
dict(real=dict(rk="double"))

But still after removing all installed instances of the file and even removing it from the 
tarball prior to rebuilding numpy, the test apparently keeps re-creating it in its own tmpdir:

building extension "_test_ext_module_5403" sources
f2py options: []
f2py:> /var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpS4KGum/src.macosx-10.6-x86_64-2.7/_test_ext_module_5403module.c
creating /var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpS4KGum
creating /var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpS4KGum/src.macosx-10.6-x86_64-2.7
getctype: "real(kind=rk)" is mapped to C "float" (to override define dict(real = dict(rk="<C typespec>")) in /private/var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpPo744G/.f2py_f2cmap file).
...
gfortran:f77: /var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpS4KGum/src.macosx-10.6-x86_64-2.7/_test_ext_module_5403-f2pywrappers.f
/var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpS4KGum/src.macosx-10.6-x86_64-2.7/_test_ext_module_5403-f2pywrappers.f:26.21:

          real :: res                                                   
                     1
Error: Symbol 'res' at (1) already has basic type of REAL
/var/folders/DC/DC7g9UNr2RWkb++8ZSn1J+++0Dk/-Tmp-/tmpS4KGum/src.macosx-10.6-x86_64-2.7/_test_ext_module_5403-f2pywrappers.f:26.21:

          real :: res                                                   
                     1
Error: Symbol 'res' at (1) already has basic type of REAL
...

f2py is the one installed with this numpy version, gfortran is 
COLLECT_GCC=/sw/bin/gfortran
COLLECT_LTO_WRAPPER=/sw/lib/gcc4.5/libexec/gcc/x86_64-apple-darwin10.6.0/4.5.2/lto-wrapper
Ziel: x86_64-apple-darwin10.6.0
Konfiguriert mit: ../gcc-4.5.2/configure --prefix=/sw --prefix=/sw/lib/gcc4.5 --mandir=/sw/share/man --infodir=/sw/lib/gcc4.5/info --enable-languages=c,c++,fortran,objc,obj-c++,java --with-gmp=/sw --with-libiconv-prefix=/sw --with-ppl=/sw --with-cloog=/sw --with-mpc=/sw --with-system-zlib --x-includes=/usr/X11R6/include --x-libraries=/usr/X11R6/lib --program-suffix=-fsf-4.5 --enable-lto
Thread-Modell: posix
gcc-Version 4.5.2 (GCC) 

Any more ideas?

Thanks,
							Derek


More information about the NumPy-Discussion mailing list