Intel Software Network articles feed

Compiler installation path changed for Intel® Cluster Toolkit Compiler Edition 3.2.1

Mon, 22 Jun 2009 06:58:02 -0700

Intel® Cluster Toolkit Compiler Edition 3.2.1 or later installs the 11.x compiler components in different directories compared to earlier products. The compiler components in the toolkit now use the same layout as the standalone 11.x compiler products.

The previous version of the toolkit, 3.2.020, installed the 10.1 or 11.0 compiler components as follows:
./intel/fc/ <<== Fortran compiler
./intel/cc/ <<== C++ compiler

The current version of the toolkit, 3.2.1.015, now installs the 11.x compiler components with the same layout as the standalone compiler products, as follows:
./intel/Compiler/11.x/ <<== BOTH Fortran and C++ compilers

Future versions of the toolkit will continue to install the compilers with the same directory structure as the standalone compiler products.

'mpiexec -idb' fails to start the Intel GUI debugger correctly

Wed, 17 Jun 2009 11:42:09 -0700

Reference Number : DPD200136769

Version : Intel MPI 3.x, Intel C++ Linux compiler 11.x, Intel Fortran Linux compiler 11.x

Operating System : Linux

Problem Description : This is only an issue on Linux systems using Intel® 64 architecture-based processors. On such systems, the IDB debugger included with the Intel C++ and Fortran 11.x compilers runs in GUI mode by default. If you launch an MPI job under the debugger's control, for example, with 'mpiexec -idb -n number_of_processes mpi_application_name', the IDB GUI and a console will come up, but neither can effectively be used to debug the MPI application.

A possible workaround is to rename idb, and then make idb a symbolic link the idbc, the command line IDB.
In the IDB_HOME directory (by default, the location is /opt/intel/Compiler/11.x//bin/intel64), please do:
mv idb idbgui
ln -s idbc idb

As always, be sure to set environment variable IDB_HOME to the location of the IDB binaries, and set MPIEXEC_DEBUG=1 before launching a debugging session.

Resolution Status : This issue may be resolved in a future version of Intel IDB for Linux

[DISCLAIMER: The information on this web site is intended for hardware system manufacturers and software developers. Intel does not warrant the accuracy, completeness or utility of any information on this site. Intel may make changes to the information or the site at any time without notice. Intel makes no commitment to update the information at this site. ALL INFORMATION PROVIDED ON THIS WEBSITE IS PROVIDED "as is" without any express, implied, or statutory warranty of any kind including but not limited to warranties of merchantability, non-infringement of intellectual property, or fitness for any particular purpose. Independent companies manufacture the third-party products that are mentioned on this site. Intel is not responsible for the quality or performance of third-party products and makes no representation or warranty regarding such products. The third-party supplier remains solely responsible for the design, manufacture, sale and functionality of its products. Intel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. *Other names and brands may be claimed as the property of others.]

Issue with RPM installation on Itanium

Thu, 11 Jun 2009 12:38:13 -0700

Problem :
The 4.2.2 version of RPM on Red Hat Enterprise Linux 3.0 for the Itanium architecture has a broken relocation feature. This will be a serious problem for users trying to install on clusters where there are shared devices.

Secondly, on Itanium architecture for some OSes (such as Red Hat Enterprise Linux 4.0 Update 7 and Red Hat Enterprise Linux 5.0 Update 2), the installation of the Intel® Math Kernel Library non-native components "…/lib/32" and "…/lib/em64t" is affected when those components are found to be missing after a distributed install is done into local directories on each node of the cluster.

Environment :
Red Hat Enterprise Linux 3.0
Red Hat Enterprise Linux 4.0 Update 7
Red Hat Enterprise Linux 5.0 Update 2

Root Cause :
Broken Relocation Feature in RPM 4.2.2

Resolution :
A recommended solution is for the user to upgrade to the latest release of RPM. Visit http://www.redhat.com to find a recent release of RPM that resolves this problem on Itanium.

In regards to the missing MKL components, two known workarounds for this issue are:

Install the Intel® Cluster Tools at the default location (/opt/intel)
Use the non-rpm mode for installation (i.e., use the installer command-line option --nonrpm)

NOTE: The above problem does not seem to appear on SLES 10 and SLES 11 for Itanium.

Master node issue

Fri, 23 Jan 2009 12:12:46 -0800

Issue

I have a cluster, which has for example 3 nodes, where the mpd.hosts file contains:

$ cat mpd.hosts
node2
node3

The master node, which is node1 is not listed in the mpd.hosts file. The shell commands:

$ mpdboot -r ssh -n 2 -f ~/mpd.hosts
$ mpiexec -n 4 /bin/hostname

which are issued from node1 reveal the following:

node1
node1
node2
node2

Solution

By design, the master node (i.e., the node where the mpiexec command is launched from) needs to have a daemon running. If you want to run your application on node2 and node3, you can use the -host options with the mpiexec command as follows:

$ mpdboot -r ssh -n 3 -f ~/mpd.hosts
$ mpiexec -n 2 -host node2 /bin/hostname : -n 2 -host node3 /bin/hostname

Or simply to use the -nolocal option for mpiexec as follows:

$ mpdboot -r ssh -n 3 -f ~/mpd.hosts
$ mpiexec -nolocal -n 4 /bin/hostname

This will run the hostname command on nodes node2 and node3 only, even though there are three mpd daemons total.

Error: could not determine icc directory for driver: mpiifort

Fri, 19 Sep 2008 00:00:00 -0700

When I run mpiifort test.f90 I get the following error:

ERROR: could not determine icc directory for driver: mpiifort

We are not using the Intel® C Compiler on our cluster system, nor does this error affect any of the other MPI compilation drivers. The Intel® Fortran compiler works fine on its own. How can I fix this issue with Intel MPI Fortran compilation?

The mpiifort compilation script calls a shell script named compchk.sh. A short term fix is to compile the application test.f90 or test.f with -nocompchk as follows:

mpiifort -nocompchk test.f90
and
mpiifort -nocompchk test.f

Note: This issue appears for older versions of the Intel® MPI Library (1.x to 2.x). It is strongly encouraged that you upgrade to the latest version of the Intel MPI Library instead. This check has been disabled in later releases.

Operating System:

SUSE* Linux Enterprise Server 9.0, SUSE* Linux Enterprise Server 10, Red Hat* Enterprise Linux 4.0, Red Hat* Enterprise Linux 5.0, SGI ProPack 5

Error when compiling C++ aps using Intel® MPI Library compilation driver mpiicpc

Fri, 19 Sep 2008 00:00:00 -0700

Issue

When compiling certain C++ applications using the Intel® MPI Library compilation driver mpiicpc, I encounter an error message that looks something like the following:

…
/opt/intel/mpi/3.1/include/mpicxx.h(45): catastrophic error: #error directive: "SEEK_SET is #defined but must not be for the C++ binding of MPI"
#error "SEEK_SET is #defined but must not be for the C++ binding of MPI"
^

Compilation aborted for my_mpi_program.cpp (code 4)
make: *** [my_mpi_program.o] Error 4

Solution

For the application my_mpi_program.cpp, you have two options:

Try organizing the include files such that mpi.h comes before stdio.h
Use the -DMPICH_IGNORE_CXX_SEEK and -DMPICH_SKIP_MPICXX options at compile time
- Refer to http://www-unix.mcs.anl.gov/mpi/mpich/faq.htm#cxxseek† for more information.

^† This link will take you off of the Intel Web site. Intel does not control the content of the destination Web Site.