Problems with Intel compilers for Linux and MPICH-GM

Jeff Layton <[email protected]> Mon, 03 Feb 2003 07:42:53 -0500
Newsgroups gmane.network.myrinet.general
Organization Lockheed-Martin Aeronautics Company
Message-ID <[email protected]>
Good morning,

   I'm running my MPI code on a Linux cluster with Myrinet using
the Intel compilers for Linux (Fortran and C) and the PGI compilers
for Linux. I'm using both v. 7.0 of the Intel compilers. The PGI
compilers are v. 4.0.2.
   The cluster nodes are dual Xeon/2.4 GHz CPUs. MPICH-GM
1.2.4..8a was built and installed but the vendor (both for v 7.0 and
the PGI compilers).
   My code runs correctly using the PGI compilers and MPICH-GM.
However, when I run using the Intel compilers, my code hits a logic
check for the input data and exits. However, before it exits, it prints
the rank of each process. Here's a snippet of the output:


 Before call to rg_one: myrank -1073743036
 Error: partition with max. cells not found
 STOPPING

I get the same rank for all of the processes (the error message
is from my code). When I use the PGI compilers, the code runs
normally.
   Is this a bug in my code or in the compilers or in this version of
MPICH-GM? (BTW, the code runs fine using the Intel compiler
and MPI/Pro-GM).

Thanks!

Jeff



--

Jeff Layton
Senior Engineer
Lockheed-Martin Aeronautical Company - Marietta
Aerodynamics & CFD

"Is it possible to overclock a cattle prod?" - Irv Mullins

This email may contain confidential information. If you have received this
email in error, please delete it immediately, and inform me of the mistake by
return email. Any form of reproduction, or further dissemination of this
email is strictly prohibited. Also, please note that opinions expressed in
this email are those of the author, and are not necessarily those of the
Lockheed-Martin Corporation.