It's possible that PyCUDA didn't get entirely rebuilt. In the source directory, do
rm -Rf build and then repeat the build. A workaround for the time being is to simply use the shipped version of boost. HTH, Andreas Robert <[email protected]> writes: > Oh I posted the wrong error messages, this is the right one: > > Traceback (most recent call last): > File "numpytest.py", line 3, in <module> > import pycuda.autoinit > File "/usr/lib/python2.7/site-packages/pycuda/autoinit.py", line 9, > in <module> > context = make_default_context() > File "/usr/lib/python2.7/site-packages/pycuda/tools.py", line 199, in > make_default_context > return ctx_maker(dev) > File "/usr/lib/python2.7/site-packages/pycuda/tools.py", line 162, in > ctx_maker > return dev.make_context() > TypeError: No to_python (by-value) converter found for C++ type: > boost::shared_ptr<pycuda::context> > ------------------------------------------------------------------- > PyCUDA ERROR: The context stack was not empty upon module cleanup. > ------------------------------------------------------------------- > A context was still active when the context stack was being > cleaned up. At this point in our execution, CUDA may already > have been deinitialized, so there is no way we can finish > cleanly. The program will be aborted now. > Use Context.pop() to avoid this problem. > ------------------------------------------------------------------- > Abgebrochen (Speicherabzug geschrieben) > > > On 02.01.2016 19:15, Robert wrote: >> Hi, >> after upgrading to the newest version I encountered a problem. >> >> With python2-pycuda 2015.1.3-6 and boost-libs 1.60.0-1 the following >> error gets thrown: >> >> Traceback (most recent call last): >> File "test.py", line 1, in <module> >> import pycuda.autoinit >> File "/usr/lib/python2.7/site-packages/pycuda/autoinit.py", line 2, >> in <module> >> import pycuda.driver as cuda >> File "/usr/lib/python2.7/site-packages/pycuda/driver.py", line 5, in >> <module> >> from pycuda._driver import * # noqa >> ImportError: libboost_python.so.1.59.0: cannot open shared object >> file: No such file or directory >> [soaked@Archibald Documents]$ python2 test.py >> [soaked@Archibald Documents]$ python2 test.py >> Traceback (most recent call last): >> File "test.py", line 1, in <module> >> import pycuda.autoinit >> File "/usr/lib/python2.7/site-packages/pycuda/autoinit.py", line 9, >> in <module> >> context = make_default_context() >> File "/usr/lib/python2.7/site-packages/pycuda/tools.py", line 199, >> in make_default_context >> return ctx_maker(dev) >> File "/usr/lib/python2.7/site-packages/pycuda/tools.py", line 162, >> in ctx_maker >> return dev.make_context() >> TypeError: No to_python (by-value) converter found for C++ type: >> boost::shared_ptr<pycuda::context> >> ------------------------------------------------------------------- >> PyCUDA ERROR: The context stack was not empty upon module cleanup. >> ------------------------------------------------------------------- >> A context was still active when the context stack was being >> cleaned up. At this point in our execution, CUDA may already >> have been deinitialized, so there is no way we can finish >> cleanly. The program will be aborted now. >> Use Context.pop() to avoid this problem. >> ------------------------------------------------------------------- >> Abgebrochen (Speicherabzug geschrieben) >> >> >> the source code only contains: >> >> import pycuda.autoinit >> >> >> After downgrading to a previous version of pycuda and downgrading >> boost-libs to 1.59 everything works fine again. >> >> I'm on Arch Linux >> >> Good Evening >> Robert Zimmermann >> >> >> _______________________________________________ >> PyCUDA mailing list >> [email protected] >> http://lists.tiker.net/listinfo/pycuda > > > _______________________________________________ > PyCUDA mailing list > [email protected] > http://lists.tiker.net/listinfo/pycuda _______________________________________________ PyCUDA mailing list [email protected] http://lists.tiker.net/listinfo/pycuda
