On Thu, Mar 19, 2009 at 11:24 AM, Ravi Kannan <span dir="ltr"><<a href="mailto:rxk@cfdrc.com">rxk@cfdrc.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Hi all,<br>
I am having some problems in obtaining speedups for FEM problems; I<br>
am not sure whether this is a problem related to the way I use MPI or due to<br>
the slow interconnection.<br>
<br>
So I was wondering if there are any sample benchmark FEM problems,<br>
which are guaranteed to give good speedups.</blockquote><div><br>Different architectures just do not work that way. That is why it is hard. The idea<br>of a "good" code does not even make sense. You should run with -log_summary<br>
and then compare to the model you have for your performance.<br><br> Matt<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
Thanks<br>
<font color="#888888">Ravi<br>
</font><div><div></div><div class="h5"><br>
-----Original Message-----<br>
From: <a href="mailto:petsc-users-bounces@mcs.anl.gov">petsc-users-bounces@mcs.anl.gov</a><br>
[mailto:<a href="mailto:petsc-users-bounces@mcs.anl.gov">petsc-users-bounces@mcs.anl.gov</a>]On Behalf Of Toby D. Young<br>
Sent: Thursday, March 19, 2009 8:31 AM<br>
To: PETSc users list<br>
Subject: possible error error message on "make tests"<br>
<br>
<br>
<br>
<br>
Hello,<br>
<br>
I configure and compile successfully petsc-3.0.0-p3 with:<br>
./config/configure.py --with-blas-lib=libblas.so<br>
--with-lapack-lib=liblapack.so --with-dynamic=1 --with-shared=1<br>
--with-mpi=0<br>
(linux-gnu-c-debug x86_64). I get this worrying "possible error" message<br>
running the tests.<br>
<br>
Running test examples to verify correct installation<br>
Possible error running C/C++ src/snes/examples/tutorials/ex19 with 1<br>
MPI process See<br>
<a href="http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html" target="_blank">http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html</a><br>
././ex19:<br>
symbol lookup error: ././ex19: undefined symbol: petsc_tmp_flops<br>
Possible error running Graphics examples<br>
src/snes/examples/tutorials/ex19 1 MPI process See<br>
<a href="http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html" target="_blank">http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html</a><br>
././ex19:<br>
symbol lookup error: ././ex19: undefined symbol: petsc_tmp_flops Error<br>
running Fortran example src/snes/examples/tutorials/ex5f with 1 MPI<br>
process See<br>
<a href="http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html" target="_blank">http://www.mcs.anl.gov/petsc/petsc-as/documentation/troubleshooting.html</a><br>
././ex5f:<br>
symbol lookup error: ././ex5f: undefined symbol: dmgetlocalvector_<br>
Completed test examples<br>
<br>
I have looked in the documentation and in the troubleshooting guide and<br>
found nothing. What does this "possible error" mean? ie; can I ignore it<br>
safely? I am not competent enough to understand why this happens.<br>
<br>
</div></div><div><div></div><div class="h5">Thanks.<br>
<br>
Best,<br>
Toby<br>
<br>
--<br>
<br>
Toby D. Young<br>
Adiunkt (Assistant Professor)<br>
Philosopher-Physicist<br>
Department of Computational Science<br>
Institute of Fundamental Technological Research<br>
Polish Academy of Sciences<br>
Room 206, ul. Swietokrzyska 21<br>
00-049 Warszawa, Polska<br>
<br>
+48 22 826 12 81 ext. 184<br>
<a href="http://rav.ippt.gov.pl/%7Etyoung" target="_blank">http://rav.ippt.gov.pl/~tyoung</a><br>
<br>
<br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
-- Norbert Wiener<br>