![[Menu Bar]](/images/header.gif)
| Newsletter Index | Quick-Tip Index | Search Newsletters |
[By: Oralee Nudson]
The MPI compilers on midnight include the "-mpitrace" flag. This can be a quick and easy option for debugging MPI applications. Adding this flag to program compilation directs each MPI task to print an announcement to stdout every time an MPI call is made. Reviewing this "trace" of MPI calls may be helpful in determining the cause of a program deadlock, segmentation fault, or unexplainable errors in program output.
Here's an example compile statement:
mg56 % mpicc mpi_example.c -mpitrace -o mpi_example
The following output gives an example of this "verbose" processor output:
mt350 % mpirun -np 2 ./mpi_example Starting MPI_Init... [0] Ending MPI_Init [1] Ending MPI_Init [1] Starting MPI_Comm_size... [1] Ending MPI_Comm_size [1] Starting MPI_Comm_rank... [1] Ending MPI_Comm_rank [1] Starting MPI_Alltoall.. [0] Starting MPI_Comm_size... [0] Ending MPI_Comm_size [0] Starting MPI_Comm_rank... [0] Ending MPI_Comm_rank [0] Starting MPI_Alltoall... [1] Ending MPI_Alltoall [0] Ending MPI_Alltoall [1] Starting MPI_Finalize... [0] Starting MPI_Finalize... [0] Ending MPI_Finalize [1] Ending MPI_Finalize Cleaning up all processes ... done.
The "-mpitrace" option is available for all MPI compilers on midnight (i.e. C, C++ and Fortran).
The PathScale compilers are aware of several environment variables that are used at compile time. These environment variables allow you to pass flags to a compile, and include:
PSC_CFLAGS: used by pathcc during compilation and linkingPSC_CXXFLAGS: used by pathCC during compilation and linkingPSC_FFLAGS: used by pathf90 and pathf95 during compilation and
linking PSC_GENFLAGS: used by pathcc, pathCC, path90, and pathf95 during
compilation and linkingHere's an example which will tell the compiler to generate C intermediate code:
mg56 % export PSC_CFLAGS="-CLIST:=ON:show=ON:emit_pfetch=ON" mg56 % mpicc hello.c -o hello.exe /import/local/pkg/pathscale/pathscale-3.0/lib/3.0/be translates ...
Since these environment variables are interpreted by the compiler they can be convenient when a configure script isn't doing what you expect, or when you don't want to change a Makefile to try different options.
If you have a code which uses more than one programming language, you can set the PSC_GENFLAGS variable to ensure each compiler gets the same options (e.g. include and library directories).
There will be two talks by members of the ARSC Technology Panel next week.
Presenter: Becky Verastegui, CIO Oak Ridge National Laboratory Title: Building a World Class IT Infrastructure to enhance R&D at Oak Ridge National Laboratory Time: Friday, April 11th, 2008 -- 1:00 PM Location: West Ridge Research Building Room 010 Abstract: Over the last 12 months IT at ORNL has rapidly deployed many new technologies to enhance the work of the staff at ORNL. These technologies include: portals, blogs, Instant Messaging, Voice Over IP, desktop video conferencing, 3GB inboxes, executive dashboards, expertise locators, visual search engines, mapping software, and a lot of chaos. The presentation will give an overview of the IT modernization initiative at ORNL. It will also include live demonstrations of how the integrated information technology is being used to transform the R&D at ORNL.
Presenter: Keith Bergeron, USAF Academy Instructor and ARSC User Time: Friday, April 11th 2008 -- 1:45 PM Location: West Ridge Research Building Room 010 Title: Modeling and Simulation Research at the U.S. Air Force Academy Abstract: The USAF Academy's (USAFA) Modeling and Simulation Research Center (M&SRC) provides the computational foundation in expertise, equipment, and personnel to facilitate M&S and High Performance Computing (HPC) research and education. Researchers bring M&S tools into the classroom and laboratories for students to learn the capabilities, and limitations, of modern computational tools. Working with the DoD HPC Modernization Program, the M&SRC enriches cadet experience and improves support of the Air Force Research Laboratory and other USAF organizations in completing their missions. Descriptions of how M&S is performed at the Academy are included, as well as details about the computational resources being used. In addition, descriptions of the methods used to bring these tools to Department of Aeronautics classes are described in detail.
A:[[ Here's one proposed a long time ago by friend and former
[[ co-editor, Guy Robinson.
[[
[[ There are some interesting collective nouns in the English
[[ language.
[[ E.g., a "pride" of lions, a "knot" of toads.
[[
[[ The question is, what collectives would you propose for terms in
[[ the supercomputing vernacular? For instance, we want collectives
[[ for terms such as cores, nodes, infinite loops, bugs, compiler
[[ options, defunct HPC vendors, ARSC consultants, irrelevent
[[ benchmarks, etc.,
[[ etc. Where there is a known collective, maybe you can improve
[[ on it. This should be interesting... maybe even fun.
[[
#
# Here's a response from Derek Bastille
#
A cluster of cores
A stack of nodes
A tangle of infinite loops
A gaggle of consultants
A swarm of bugs
A (hungry) pack of HPC vendors
A morass of compiler options
A brace of compilers
A thicket of libraries
An assemblage of engineers
A camp of users
A bestiary of benchmarks
#
# Thanks to Rick Roberts for this response.
#
suggestion what it actually refers to
==================== ===========================================
Clowder of cores clowder is a set of cats
Nodule of nodes
Murder of Crays set of HPC vendors; murder is a set of crows
YiMP set of defunct HPC vendors
Scalin-marks a set of irrelevent benchmarks
Do-Mobius infinite loop
Blasphemous Bill cost of new system
Hypo-thermiants ARSC consultants
ice-worms ARSC consultants (sorry, I couldn't help myself!)
Guy de Part fondly remembered ARSC consultants.
#
# Here's are several suggestions from Lee Higbie.
#
A blyght of programmers.
Maybe we also have flops of supercomputers?
And, of course, it's the blyght of programmers using the flop of
supercomputers that HPCMP has.
Q: I just generated a bunch of image files that are sequentially named:
image1.png, image2.png, ..., image100.png. Unfortunately when I
do an "ls" on the directory, the files aren't sorted the way I
would like.
mg56 % ls -1
image100.png
image10.png
image11.png
image12.png
image13.png
image14.png
...
...
How can rename the files with padded zeros so the files show up
in numerical order? For example, I would like the file image1.png
to become image001.png and image33.jpg to become image033.png, etc.
Show me how to do this shell gurus!
[[ Answers, Questions, and Tips Graciously Accepted ]]
Contact:
Thomas J. Baring ARSC Web Specialist ph: 907-450-8619 Donald Bahls ARSC User Consultant ph: 907-450-8674 Arctic Region Supercomputing Center University of Alaska Fairbanks PO Box 756020 Fairbanks AK 99775-6020
Send comments and questions to the current editors using this Contact Form.Email Subscriptions:
| Newsletter Index | Quick-Tip Index | Search Newsletters |
Arctic Region Supercomputing Center
PO Box 756020, Fairbanks, AK 99775 |
voice: 907-474-6935 |
email: 
home | search | about | support | news | science | resources