I'm currently waiting for the admins to Install OpenMPI1.3, but in the
meantime I recompiled and linked against Mesa (rather than the nvidia
drivers).<br><br>Tada, all works. So there's definately some problem with NVidia Drivers + OpenMPI1.2.x <br clear="all">--<br>Randall Hand<br>Visualization Scientist<br><a href="http://www.yeraze.com">http://www.yeraze.com</a><br>
<a href="http://www.vizworld.com">http://www.vizworld.com</a><br>
<br><br><div class="gmail_quote">On Tue, Aug 11, 2009 at 8:52 AM, Berk Geveci <span dir="ltr"><<a href="mailto:berk.geveci@kitware.com">berk.geveci@kitware.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
My guess is that there is a symbol conflict between something in<br>
ParaView and OpenMPI 1.2.6. If switching to 1.3 does not fix the<br>
problem, we'll have to resort to debugging this some other way.<br>
<font color="#888888"><br>
-berk<br>
</font><div><div></div><div class="h5"><br>
On Fri, Aug 7, 2009 at 8:59 AM, Rick Angelini<<a href="mailto:angel@arl.army.mil">angel@arl.army.mil</a>> wrote:<br>
> Randall - try compiling against Openmpi-1.3 and see if you get different<br>
> results. We saw issues with compiling Paraview 3.x against Openmpi-1.2.6<br>
> that went away when we switch to Openmpi-1.3.<br>
><br>
><br>
><br>
> j s wrote:<br>
>><br>
>> Looking online, that is a nasty FPE, since it is a result of an integer<br>
>> operation, not a floating point operation.<br>
>><br>
>> This is occuring at address NULL, which should not be possible.<br>
>> #8 0x0000000000000000 in ?? ()<br>
>><br>
>> It looks like a function at address 0 is being called. Do you have more<br>
>> of the trace, or is that the top? If that is the top, can you see if this<br>
>> function is being invoked from a different thread?<br>
>><br>
>> Juan<br>
>><br>
>><br>
>> On Thu, Aug 6, 2009 at 3:47 PM, Randall Hand <<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>>> wrote:<br>
>><br>
>> Ok, more information.<br>
>><br>
>> If I compile ParaView without Python and without MPI, it works<br>
>> beautifully.<br>
>><br>
>> Enabling MPI, however, causes the problem.. a GDB trace:<br>
>><br>
>><br>
>> (gdb) run<br>
>> Starting program: /viz/work/rhand/pvezviz/install-MPI/bin/paraview<br>
>><br>
>> Program received signal SIGFPE, Arithmetic exception.<br>
>> 0x00000030fea6911c in _int_free () from /lib64/tls/libc.so.6<br>
>> (gdb) bt<br>
>> #0 0x00000030fea6911c in _int_free () from /lib64/tls/libc.so.6<br>
>> #1 0x00000030fea69596 in free () from /lib64/tls/libc.so.6<br>
>> #2 0x00000030fe800a2b in ?? ()<br>
>> #3 0x0000007fbfffc8b0 in ?? ()<br>
>> #4 0x0000002a9758a11e in ?? ()<br>
>> #5 0x000000000000fd00 in ?? ()<br>
>> #6 0x000000000128502b in ?? ()<br>
>> #7 0x0000002aa267372a in ?? ()<br>
>> #8 0x0000000000000000 in ?? ()<br>
>> (gdb)<br>
>><br>
>> The MPI libraries I'm using:<br>
>><br>
>> MPI_EXTRA_LIBRARY<br>
>> /usr/local/ofed/mpi/gcc/openmpi-1.2.8/lib64/libmpi.so;/usr/local/ofed/mpi/gcc/openmpi-1.2.8/lib64/libopen-rte.so;/usr/local/ofed/mpi/gcc/openmpi-1.2.8/lib64/libopen-pal.so;/usr/lib64/libdl.so;/usr/lib64/libnsl.so;/usr/lib64/libutil.so;/usr/lib64/libm.so;/usr/lib64/libdl.so<br>
>> MPI_INCLUDE_PATH<br>
>> /usr/local/ofed/mpi/gcc/openmpi-1.2.8/include<br>
>> MPI_LIBRARY<br>
>> /usr/local/ofed/mpi/gcc/openmpi-1.2.8/lib64/libmpi_cxx.so<br>
>><br>
>> Anyone see anything obviously wrong?<br>
>><br>
>><br>
>> --<br>
>> Randall Hand<br>
>> Visualization Scientist<br>
>> <a href="http://www.yeraze.com" target="_blank">http://www.yeraze.com</a><br>
>> <a href="http://www.vizworld.com" target="_blank">http://www.vizworld.com</a><br>
>><br>
>><br>
>><br>
>> On Thu, Jul 30, 2009 at 10:49 AM, Rick Angelini<br>
>> <<a href="mailto:angel@arl.army.mil">angel@arl.army.mil</a> <mailto:<a href="mailto:angel@arl.army.mil">angel@arl.army.mil</a>>> wrote:<br>
>><br>
>> Randall - we saw some execution-time failures of 3.6.1 that<br>
>> was related to an older version of the nVidia driver. We<br>
>> didn't see floating point exceptions, though - we saw socket<br>
>> errors related to the client-server connection. Moving to a<br>
>> newer version of the nVidia driver on the client side fixed<br>
>> the problem. It doesn't make sense, but that's what<br>
>> happened! 8-)<br>
>><br>
>><br>
>><br>
>> j s wrote:<br>
>><br>
>> I had meant to send this to the list, instead of just Randall.<br>
>><br>
>> ---------- Forwarded message ----------<br>
>> From: *j s* <<a href="mailto:j.s4403@gmail.com">j.s4403@gmail.com</a> <mailto:<a href="mailto:j.s4403@gmail.com">j.s4403@gmail.com</a>><br>
>> <mailto:<a href="mailto:j.s4403@gmail.com">j.s4403@gmail.com</a> <mailto:<a href="mailto:j.s4403@gmail.com">j.s4403@gmail.com</a>>>><br>
>> Date: Thu, Jul 30, 2009 at 9:03 AM<br>
>> Subject: Re: [Paraview] Baffled by "floating point<br>
>> exceptions" in ParaView 3.6.1<br>
>> To: Randall Hand <<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>>>><br>
>><br>
>><br>
>> If you are running it in gdb, you might be able to do:<br>
>><br>
>> (gdb) handle SIGFPE nopass<br>
>> Signal Stop Print Pass to program Description<br>
>> SIGFPE Yes Yes No Arithmetic<br>
>> exception<br>
>><br>
>> or<br>
>><br>
>> (gdb) handle SIGFPE nostop nopass<br>
>> Signal Stop Print Pass to program Description<br>
>> SIGFPE No Yes No Arithmetic<br>
>> exception<br>
>><br>
>><br>
>><br>
>> On Thu, Jul 30, 2009 at 8:52 AM, Randall Hand<br>
>> <<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>>>> wrote:<br>
>><br>
>> I was unable to compile 3.4 against these libraries,<br>
>> compilation<br>
>> would fail with errors in the Python libraries about<br>
>> "-fPIC".<br>
>><br>
>> I'll try again without MPI & Python, and get back with you.<br>
>><br>
>> --<br>
>> Randall Hand<br>
>> Visualization Scientist<br>
>> <a href="http://www.yeraze.com" target="_blank">http://www.yeraze.com</a> <<a href="http://www.yeraze.com/" target="_blank">http://www.yeraze.com/</a>><br>
>> <a href="http://www.vizworld.com" target="_blank">http://www.vizworld.com</a> <<a href="http://www.vizworld.com/" target="_blank">http://www.vizworld.com/</a>><br>
>><br>
>><br>
>><br>
>> On Thu, Jul 30, 2009 at 8:47 AM, Berk Geveci<br>
>> <<a href="mailto:berk.geveci@kitware.com">berk.geveci@kitware.com</a><br>
>> <mailto:<a href="mailto:berk.geveci@kitware.com">berk.geveci@kitware.com</a>><br>
>> <mailto:<a href="mailto:berk.geveci@kitware.com">berk.geveci@kitware.com</a><br>
>> <mailto:<a href="mailto:berk.geveci@kitware.com">berk.geveci@kitware.com</a>>>> wrote:<br>
>><br>
>> Can you try compiling without Python and MPI? Also,<br>
>> can you<br>
>> try with<br>
>> Mesa? I am guessing some system library is causing this<br>
>> problem but<br>
>> without a stack trace, it will be trial-and-error<br>
>> to find out<br>
>> which.<br>
>> Also, do older versions work with the same libraries?<br>
>><br>
>> -berk<br>
>><br>
>> On Thu, Jul 30, 2009 at 9:43 AM, Randall<br>
>> Hand<<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a><br>
>> <mailto:<a href="mailto:randall.hand@gmail.com">randall.hand@gmail.com</a>>>><br>
>><br>
>> wrote:<br>
>> > I've compiled ParaView 3.6.1 three times so far,<br>
>> and every<br>
>> time it compiles<br>
>> > successfully but when I attempt to run it, it<br>
>> dies with<br>
>> "Floating point<br>
>> > exception".<br>
>> ><br>
>> > I'm compiling it on a Linux system (RHEL4), and<br>
>> linking<br>
>> against MPI (OpenMPE<br>
>> > 1.2.8) and python (2.5.1). I'm using cmake 2.6.4<br>
>> . Is<br>
>> there some known<br>
>> > problem here?<br>
>> ><br>
>> > I compiled Paraview in "Debug" mode, but a<br>
>> backtrace doesn't<br>
>> show anything<br>
>> > but question marks.<br>
>> > --<br>
>> > Randall Hand<br>
>> > Visualization Scientist<br>
>> > <a href="http://www.yeraze.com" target="_blank">http://www.yeraze.com</a> <<a href="http://www.yeraze.com/" target="_blank">http://www.yeraze.com/</a>><br>
>> > <a href="http://www.vizworld.com" target="_blank">http://www.vizworld.com</a> <<a href="http://www.vizworld.com/" target="_blank">http://www.vizworld.com/</a>><br>
>> ><br>
>> > _______________________________________________<br>
>> > Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a><br>
>> <<a href="http://www.kitware.com" target="_blank">http://www.kitware.com</a>> <<a href="http://www.kitware.com/" target="_blank">http://www.kitware.com/</a>><br>
>><br>
>> ><br>
>> > Visit other Kitware open-source projects at<br>
>> > <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
>> ><br>
>> > Please keep messages on-topic and check the<br>
>> ParaView Wiki at:<br>
>> > <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
>> ><br>
>> > Follow this link to subscribe/unsubscribe:<br>
>> > <a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
>> ><br>
>> ><br>
>><br>
>><br>
>><br>
>> _______________________________________________<br>
>> Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a> <<a href="http://www.kitware.com" target="_blank">http://www.kitware.com</a>><br>
>> <<a href="http://www.kitware.com/" target="_blank">http://www.kitware.com/</a>><br>
>><br>
>><br>
>> Visit other Kitware open-source projects at<br>
>> <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
>><br>
>> Please keep messages on-topic and check the ParaView<br>
>> Wiki at:<br>
>> <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
>><br>
>> Follow this link to subscribe/unsubscribe:<br>
>> <a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
>><br>
>><br>
>><br>
>><br>
>> ------------------------------------------------------------------------<br>
>><br>
>> _______________________________________________<br>
>> Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a> <<a href="http://www.kitware.com" target="_blank">http://www.kitware.com</a>><br>
>><br>
>> Visit other Kitware open-source projects at<br>
>> <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
>><br>
>> Please keep messages on-topic and check the ParaView Wiki<br>
>> at: <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
>><br>
>> Follow this link to subscribe/unsubscribe:<br>
>> <a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
>><br>
>> _______________________________________________<br>
>> Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a> <<a href="http://www.kitware.com" target="_blank">http://www.kitware.com</a>><br>
>><br>
>> Visit other Kitware open-source projects at<br>
>> <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
>><br>
>> Please keep messages on-topic and check the ParaView Wiki at:<br>
>> <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
>><br>
>> Follow this link to subscribe/unsubscribe:<br>
>> <a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
>><br>
>><br>
>><br>
> _______________________________________________<br>
> Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a><br>
><br>
</div></div><div><div></div><div class="h5">> Visit other Kitware open-source projects at<br>
> <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
><br>
> Please keep messages on-topic and check the ParaView Wiki at:<br>
> <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
><br>
> Follow this link to subscribe/unsubscribe:<br>
> <a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
><br>
_______________________________________________<br>
Powered by <a href="http://www.kitware.com" target="_blank">www.kitware.com</a><br>
<br>
Visit other Kitware open-source projects at <a href="http://www.kitware.com/opensource/opensource.html" target="_blank">http://www.kitware.com/opensource/opensource.html</a><br>
<br>
Please keep messages on-topic and check the ParaView Wiki at: <a href="http://paraview.org/Wiki/ParaView" target="_blank">http://paraview.org/Wiki/ParaView</a><br>
<br>
Follow this link to subscribe/unsubscribe:<br>
<a href="http://www.paraview.org/mailman/listinfo/paraview" target="_blank">http://www.paraview.org/mailman/listinfo/paraview</a><br>
</div></div></blockquote></div><br>