[PyCUDA] pycuda installation problem
Paul Rigor (uci)
paul.rigor at uci.edu
Fri Jul 10 16:20:02 PDT 2009
Hi Gökhan,
Do a completely clean install. Uninstall all dependencies using yum. And
make sure that LD_LIBRARY_PATH contains the paths to the libs you know of.
I've installed pycuda on a fedora machine without much hassle a while back.
I'll look into my correspondence with other developers and directly forward
you the steps I took.
Goodluck,
Paul
On Fri, Jul 10, 2009 at 10:44 AM, Gökhan SEVER <gokhansever at gmail.com>wrote:
> On Fri, Jul 10, 2009 at 10:12 AM, Gökhan SEVER <gokhansever at gmail.com>wrote:
>
>>
>>
>> On Fri, Jul 10, 2009 at 9:43 AM, Andreas Klöckner <
>> lists at informa.tiker.net> wrote:
>>
>>> On Freitag 10 Juli 2009, Gökhan SEVER wrote:
>>> > On Thu, Jul 9, 2009 at 10:04 PM, Andreas Klöckner
>>> >
>>> > <lists at informa.tiker.net>wrote:
>>> > > On Donnerstag 09 Juli 2009, Gökhan SEVER wrote:
>>> > > > I should have bitten my tongue before saying all working fine.
>>> > > >
>>> > > > examples complain this time :(
>>> > > >
>>> > > > [gsever at ccn examples]$ python demo.py
>>> > > > RuntimeError: FATAL: module compiled as big endian, but detected
>>> > >
>>> > > different
>>> > >
>>> > > > endianness at runtime
>>> > > > RuntimeError: FATAL: module compiled as big endian, but detected
>>> > >
>>> > > different
>>> > >
>>> > > > endianness at runtime
>>> > > > Traceback (most recent call last):
>>> > > > File "demo.py", line 3, in <module>
>>> > > > import pycuda.driver as cuda
>>> > > > File
>>> > >
>>> > >
>>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/py
>>> > >cu
>>> > >
>>> > > >da/driver.py", line 1, in <module>
>>> > > > from _driver import *
>>> > > > ImportError: numpy.core.multiarray failed to import
>>> > >
>>> > > Something fishy is going on here. i686 is a little endian
>>> architecture,
>>> > > but somehow (I assume) numpy is complaining that it got compiled as
>>> big
>>> > > endian. Would you mind trying with a release numpy?
>>> > >
>>> > > Andreas
>>> > >
>>> > > _______________________________________________
>>> > > PyCUDA mailing list
>>> > > PyCUDA at tiker.net
>>> > > http://tiker.net/mailman/listinfo/pycuda_tiker.net
>>> >
>>> > Good morning,
>>> >
>>> > Rolled back to numpy.i386 0:1.2.0-3.fc10 via yum.
>>> >
>>> > However now I have even a prettier error:
>>> >
>>> > [gsever at ccn examples]$ python demo.py
>>> > Segmentation fault
>>> >
>>> > [gsever at ccn examples]$ python hello_gpu.py
>>> > Segmentation fault
>>>
>>> Have you recompiled PyCUDA against the new numpy? If yes, send the output
>>> of
>>>
>>> gdb --args python hello_gpu.py
>>> ...stuff happening...
>>> r<Enter>
>>> ...more stuff happening...
>>> Crashed with SIGSEGV.
>>> bt<Enter>
>>>
>>> Andreas
>>>
>>
>> Ehh, a long list of error strings...
>>
>> I am installing these following packages for a more descriptive listing. I
>> will update you once again.
>> debuginfo-install atlas-3.6.0-15.fc10.i386 libgcc-4.3.2-7.i386
>> libgfortran-4.3.2-7.i386 libstdc++-4.3.2-7.i386 numpy-1.2.0-3.fc10.i386
>>
>>
>> [gsever at ccn examples]$ gdb --args python hello_gpu.py
>> GNU gdb Fedora (6.8-29.fc10)
>> Copyright (C) 2008 Free Software Foundation, Inc.
>> License GPLv3+: GNU GPL version 3 or later <
>> http://gnu.org/licenses/gpl.html>
>> This is free software: you are free to change and redistribute it.
>> There is NO WARRANTY, to the extent permitted by law. Type "show copying"
>> and "show warranty" for details.
>> This GDB was configured as "i386-redhat-linux-gnu"...
>> (gdb) r
>> Starting program: /usr/bin/python hello_gpu.py
>> [Thread debugging using libthread_db enabled]
>> [New Thread 0xb7fe46c0 (LWP 13818)]
>>
>> Program received signal SIGSEGV, Segmentation fault.
>> 0x00000000 in ?? ()
>> Missing separate debuginfos, use: debuginfo-install
>> atlas-3.6.0-15.fc10.i386 libgcc-4.3.2-7.i386 libgfortran-4.3.2-7.i386
>> libstdc++-4.3.2-7.i386 numpy-1.2.0-3.fc10.i386
>> (gdb) bt
>> #0 0x00000000 in ?? ()
>> #1 0x007e3ff2 in global constructors keyed to mempool.cpp ()
>> from
>> /usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so
>> #2 0x007f0e5d in __do_global_ctors_aux ()
>> from
>> /usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so
>> #3 0x007a2960 in _init ()
>> from
>> /usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so
>> #4 0x004b61ed in call_init () at dl-init.c:70
>> #5 _dl_init (main_map=0x80a4910, argc=2, argv=0xbffff404, env=0xbffff410)
>> at dl-init.c:134
>> #6 0x004ba8d2 in dl_open_worker (a=0xbfff918c) at dl-open.c:516
>> #7 0x004b5db6 in _dl_catch_error (objname=0xbfff91b4,
>> errstring=0xbfff91b0,
>> mallocedp=0xbfff91bb, operate=0x4ba510 <dl_open_worker>,
>> args=0xbfff918c)
>> at dl-error.c:178
>> #8 0x004b9f5e in _dl_open (
>> file=0xbfff96a7
>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so",
>> mode=-2147483646, caller_dlopen=0x4776cb3,
>> nsid=-2, argc=2, argv=0xbffff404, env=0xbffff410) at dl-open.c:596
>> #9 0x00670c2c in dlopen_doit (a=0xbfff9320) at dlopen.c:67
>> #10 0x004b5db6 in _dl_catch_error (objname=0x80a8f6c, errstring=0x80a8f70,
>>
>> ---Type <return> to continue, or q <return> to quit---
>> mallocedp=0x80a8f68, operate=0x670b90 <dlopen_doit>, args=0xbfff9320)
>> at dl-error.c:178
>> #11 0x0067105c in _dlerror_run (operate=0x670b90 <dlopen_doit>,
>> args=<value optimized out>) at dlerror.c:164
>> #12 0x00670b61 in __dlopen (
>> file=0xbfff96a7
>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so",
>> mode=2) at dlopen.c:88
>> #13 0x04776cb3 in _PyImport_GetDynLoadFunc (
>> fqname=0xbfffa737 "pycuda._driver", shortname=0xbfffa73e "_driver",
>> pathname=0xbfff96a7
>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so",
>> fp=0x80a8df8) at Python/dynload_shlib.c:130
>> #14 0x04766eb1 in _PyImport_LoadDynamicModule (
>> name=0xbfffa737 "pycuda._driver",
>> pathname=0xbfff96a7
>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/_driver.so",
>> fp=0x80a8df8) at Python/importdl.c:42
>> #15 0x047647db in load_module (name=0xbfffa737 "pycuda._driver",
>> fp=0x19a640,
>> buf=0x80d8548 "\002", type=3, loader=0x0) at Python/import.c:1758
>> #16 0x04765249 in import_submodule (mod=0xb7fa444c,
>> subname=0xbfffa73e "_driver", fullname=0xbfffa737 "pycuda._driver")
>> at Python/import.c:2400
>> #17 0x047654fb in load_next (mod=0xb7fa444c, altmod=0x47d7480,
>> p_name=0xbfffa728, buf=0xbfffa737 "pycuda._driver",
>> p_buflen=0xbfffa730)
>> at Python/import.c:2220
>> ---Type <return> to continue, or q <return> to quit---
>> #18 0x04765b10 in import_module_level (name=0x0, globals=0xfffffffe,
>> locals=0xb7f8513c, fromlist=0xb7f8110c, level=-1) at
>> Python/import.c:2001
>> #19 0x04765e5b in PyImport_ImportModuleLevel (name=0xb7f8a294 "_driver",
>> globals=0xb7f8513c, locals=0xb7f8513c, fromlist=0xb7f8110c, level=-1)
>> at Python/import.c:2072
>> #20 0x04749054 in builtin___import__ (self=0x0, args=0xb7f895f4, kwds=0x0)
>> at Python/bltinmodule.c:47
>> #21 0x046ffaf9 in PyCFunction_Call (func=0xb7face6c, arg=0xb7f895f4,
>> kw=0x0)
>> at Objects/methodobject.c:77
>> #22 0x046cb397 in PyObject_Call (func=0x80d8548, arg=0xb7f895f4, kw=0x0)
>> at Objects/abstract.c:1861
>> #23 0x047496fc in PyEval_CallObjectWithKeywords (func=0xb7face6c,
>> arg=0xb7f895f4, kw=0x0) at Python/ceval.c:3442
>> #24 0x0474e289 in PyEval_EvalFrameEx (f=0x805bd9c, throwflag=0)
>> at Python/ceval.c:2067
>> #25 0x04751965 in PyEval_EvalCodeEx (co=0xb7f87800, globals=0xb7f8513c,
>> locals=0xb7f8513c, args=0x0, argcount=0, kws=0x0, kwcount=0, defs=0x0,
>>
>> defcount=0, closure=0x0) at Python/ceval.c:2836
>> #26 0x04751bc3 in PyEval_EvalCode (co=0xb7f87800, globals=0xb7f8513c,
>> locals=0xb7f8513c) at Python/ceval.c:494
>> #27 0x0476380d in PyImport_ExecCodeModuleEx (name=0xbfffdd07
>> "pycuda.driver",
>> co=0xb7f87800,
>> pathname=0xbfffbc17
>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5----Type <return> to
>> continue, or q <return> to quit---
>> linux-i686.egg/pycuda/driver.pyc") at Python/import.c:675
>> #28 0x04763abc in load_source_module (name=0xbfffdd07 "pycuda.driver",
>> pathname=0xbfffbc17
>> "/usr/lib/python2.5/site-packages/pycuda-0.94beta-py2.5-linux-i686.egg/pycuda/driver.pyc",
>> fp=0x8085650) at Python/import.c:959
>> #29 0x04765249 in import_submodule (mod=0xb7fa444c,
>> subname=0xbfffdd0e "driver", fullname=0xbfffdd07 "pycuda.driver")
>> at Python/import.c:2400
>> #30 0x047654fb in load_next (mod=0xb7fa444c, altmod=0xb7fa444c,
>> p_name=0xbfffdcf8, buf=0xbfffdd07 "pycuda.driver",
>> p_buflen=0xbfffdd00)
>> at Python/import.c:2220
>> #31 0x04765b64 in import_module_level (name=0x0,
>> globals=<value optimized out>, locals=0xb7fbdacc, fromlist=0x47d7480,
>> level=-1) at Python/import.c:2008
>> #32 0x04765e5b in PyImport_ImportModuleLevel (name=0xb7f84264
>> "pycuda.driver",
>> globals=0xb7fbdacc, locals=0xb7fbdacc, fromlist=0x47d7480, level=-1)
>> at Python/import.c:2072
>> #33 0x04749054 in builtin___import__ (self=0x0, args=0xb7f6b464, kwds=0x0)
>> at Python/bltinmodule.c:47
>> #34 0x046ffaf9 in PyCFunction_Call (func=0xb7face6c, arg=0xb7f6b464,
>> kw=0x0)
>> at Objects/methodobject.c:77
>> #35 0x046cb397 in PyObject_Call (func=0x80d8548, arg=0xb7f6b464, kw=0x0)
>> at Objects/abstract.c:1861
>> #36 0x047496fc in PyEval_CallObjectWithKeywords (func=0xb7face6c,
>> ---Type <return> to continue, or q <return> to quit---
>> arg=0xb7f6b464, kw=0x0) at Python/ceval.c:3442
>> #37 0x0474e289 in PyEval_EvalFrameEx (f=0x80a4464, throwflag=0)
>> at Python/ceval.c:2067
>> #38 0x04751965 in PyEval_EvalCodeEx (co=0xb7f695c0, globals=0xb7fbdacc,
>> locals=0xb7fbdacc, args=0x0, argcount=0, kws=0x0, kwcount=0, defs=0x0,
>>
>> defcount=0, closure=0x0) at Python/ceval.c:2836
>> #39 0x04751bc3 in PyEval_EvalCode (co=0xb7f695c0, globals=0xb7fbdacc,
>> locals=0xb7fbdacc) at Python/ceval.c:494
>> #40 0x0476d3c2 in run_mod (mod=<value optimized out>,
>> filename=<value optimized out>, globals=0xb7fbdacc, locals=0xb7fbdacc,
>>
>> flags=0xbffff338, arena=0x80710e8) at Python/pythonrun.c:1273
>> #41 0x0476d482 in PyRun_FileExFlags (fp=0x804a008,
>> filename=0xbffff59f "hello_gpu.py", start=257, globals=0xb7fbdacc,
>> locals=0xb7fbdacc, closeit=1, flags=0xbffff338) at
>> Python/pythonrun.c:1259
>> #42 0x0476ec0c in PyRun_SimpleFileExFlags (fp=0x804a008,
>> filename=0xbffff59f "hello_gpu.py", closeit=1, flags=0xbffff338)
>> at Python/pythonrun.c:879
>> #43 0x0476f37a in PyRun_AnyFileExFlags (fp=0x804a008,
>> filename=0xbffff59f "hello_gpu.py", closeit=1, flags=0xbffff338)
>> at Python/pythonrun.c:698
>> #44 0x0477943f in Py_Main (argc=1, argv=0xbffff404) at Modules/main.c:523
>> #45 0x080485d2 in main (argc=2, argv=0x47daba0) at Modules/python.c:23
>>
>>
>> --
>> Gökhan
>>
>
>
> It is still producing the same error message after the missing debub-info
> files installed.
>
> Hope this could be something curable.
>
> --
> Gökhan
>
> _______________________________________________
> PyCUDA mailing list
> PyCUDA at tiker.net
> http://tiker.net/mailman/listinfo/pycuda_tiker.net
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tiker.net/pipermail/pycuda/attachments/20090710/75444c05/attachment-0001.htm>
More information about the PyCUDA
mailing list