Make sure you have the latest version of the Nvidia drivers installed. In my case with ubuntu i had to get the upstream drivers. ____________ Bryan Absher
On Sat, Jan 8, 2011 at 12:00 PM, <[email protected]> wrote: > Send PyCUDA mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.tiker.net/listinfo/pycuda > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of PyCUDA digest..." > > > Today's Topics: > > 1. Re: Problems compling (Patrick Herron) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 7 Jan 2011 20:27:51 +0000 (UTC) > From: Patrick Herron <[email protected]> > To: [email protected] > Subject: Re: [PyCUDA] Problems compling > Message-ID: <[email protected]> > Content-Type: text/plain; charset=us-ascii > > Andreas Kloeckner <li...@...> writes: > >> >> Hi Bryan, >> >> On Fri, 1 Oct 2010 21:10:27 -0700, Bryan Absher >> <bryan.abs...@...> wrote: >> > I have heard great things about the pycuda, if only it would run. Here >> > is the output to the test_driver.py >> > >> > $ python ./test_driver.py >> > Traceback (most recent call last): > [...] >> >> My suspicion would be that /usr/lib/nvidia-current/libcuda.so.1 is not >> the current (3.2) version of libcuda.so. (But it seems you're compiling >> against 3.2 headers, which gives you unavailable symbol errors. You can >> easily check for a 3.2 libcuda yourself: >> >> I get: >> nm -D /usr/lib/libcuda.so | grep cuMemAllocPitch_v2 >> 000000000015c0e0 T cuMemAllocPitch_v2 >> >> Try >> >> nm -D /usr/lib/nvidia-current/libcuda.so.1 | grep cuMemAllocPitch_v2 >> >> If you get empty output, you have an old version. >> >> HTH, >> Andreas >> >> > > > I too am very eager to use PyCUDA. After some effort I was finally able to > install on my RHEL5.5 system using > [r...@netreg-536608 doc]# CC=gcc44 make install > This solved a host of build errors. Your advice on this and other forums has > been very helpful. > Unfortunately when I run the test code I get a very similar error to the one > described above. Further, when I use the above nm command I get empty output, > which you suggest means I don't have 3.2 drivers. Yet I (only) have 3.2 > drivers > installed. > > Error: > > [r...@netreg-536608 test]# python test_driver.py > Traceback (most recent call last): > File "test_driver.py", line 4, in ? > from pycuda.tools import mark_cuda_test > File > "/usr/lib64/python2.4/site-packages/pycuda-0.94.2-py2.4-linux-x86_64.egg/ > pycuda/tools.py", > line 30, in ? > import pycuda.driver as cuda > File > "/usr/lib64/python2.4/site-packages/pycuda-0.94.2-py2.4-linux-x86_64.egg/ > pycuda/driver.py", > line 1, in ? > from pycuda._driver import * > ImportError: > /usr/lib64/python2.4/site-packages/pycuda-0.94.2-py2.4-linux-x86_64.egg/ > pycuda/_driver.so: > undefined symbol: _ZNKSt5ctypeIcE13_M_widen_initEv > > > [r...@netreg-536608 test]# ldd > /usr/lib64/python2.4/site-packages/pycuda-0.94.2-py2.4-linux-x86_64.egg/ > pycuda/_driver.so > > libcuda.so.1 => /usr/lib64/libcuda.so.1 (0x00002b095e94d000) > libstdc++.so.6 => /usr/lib64/libstdc++.so.6 (0x00002b095f3b4000) > libm.so.6 => /lib64/libm.so.6 (0x00002b095f6b4000) > libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00002b095f938000) > libpthread.so.0 => /lib64/libpthread.so.0 (0x00002b095fb46000) > libc.so.6 => /lib64/libc.so.6 (0x00002b095fd61000) > libz.so.1 => /usr/lib64/libz.so.1 (0x00002b09600b9000) > libdl.so.2 => /lib64/libdl.so.2 (0x00002b09602cd000) > /lib64/ld-linux-x86-64.so.2 (0x0000003d81600000) > > [r...@netreg-536608 doc]# nm -D /usr/lib64/libcuda.so | grep > _ZNKSt5ctypeIcE13_M_widen_initEv > > > > [r...@netreg-536608 doc]# > /root/NVIDIA_GPU_Computing_SDK/C/bin/linux/release/deviceQuery > Device 0: "Tesla C2070" > CUDA Driver Version: 3.20 > CUDA Runtime Version: 3.20 > [...] > > > Thanks in advance! > > > > > ------------------------------ > > _______________________________________________ > PyCUDA mailing list > [email protected] > http://lists.tiker.net/listinfo/pycuda > > > End of PyCUDA Digest, Vol 31, Issue 3 > ************************************* > _______________________________________________ PyCUDA mailing list [email protected] http://lists.tiker.net/listinfo/pycuda
