[scalapack] SIGSEGV when running the example1

Open discussion regarding features, bugs, issues, vendors, etc.

[scalapack] SIGSEGV when running the example1

Postby jorgeu » Wed Jul 12, 2006 6:44 pm

Hi, i just installed a little beowulf cluster using ubuntu server 6.06 LTS amd64. I installed precompiled debian packages for lam-mpi-7.1, blacs, atlas-3 and scalapack. Also installed the test programs for scalapack and they run without any notable problem.

I downloaded the classic scalapack example: example1.f, and compiled with:
Code: Select all
$ mpif77 -o example example1.f /usr/lib/libscalapack.a /usr/lib/libblacsF77init.a /usr/lib/libblacs.a /usr/lib/libblacsF77init.a /usr/lib/atlas/libblas.a

some of that .a files are symbolic links to the specific archive. That compilation finish without any problem.

When i try to run it with:
Code: Select all
$ mpirun -np 6 example

Then i get the following error message:
Code: Select all
One of the processes started by mpirun has exited with a nonzero exit
code.  This typically indicates that the process finished in error.
If your process did not finish in error, be sure to include a "return
0" or "exit(0)" in your C code before exiting the application.

PID 18303 failed on node n0 ( due to signal 11.

As i can remember it is a segmentation fault .

I added some debuging output to see where it stops and the line responsable here is:
Code: Select all

near line 47.

I have seen a similar problem when trying to use scalapack from R (RScaLAPACK), it always outputs an error message indicating that it can't spawn child proceses when trying to create some slaves

Running other mpi programs work fine. I really apreciate if someone can help on that problem.
Posts: 1
Joined: Wed Jul 12, 2006 6:25 pm
Location: Maracaibo, Venezuela

Return to User Discussion

Who is online

Users browsing this forum: Google [Bot] and 3 guests