Click source to get a
compressed file containing the source code - superlu_mt.tar.gz (~480 KB).
To retrieve the source, please do the following:
gunzip superlu_mt.tar.gz
tar xvf superlu_mt.tar
Users' Guide is available, which is also in the above tar file distribution.
A brief overview of the package is given below.
SuperLU_MT (ALPHA version)
==========================
SuperLU_MT contains a set of subroutines to solve a sparse linear system
A*X=B. It uses Gaussian elimination with partial pivoting (GEPP).
The columns of A may be preordered before factorization; the
preordering for sparsity is completely separate from the factorization.
SuperLU_MT is a parallel extension to the serial SuperLU library.
SuperLU_MT is implemented in ANSI C, with multithreading extension,
for example, using POSIX threads. Currently, only the LU factorization
routine, which is the most time-consuming part of the solution process,
is parallelized on machines with a shared address space. The other
routines, such as column preordering and the forward and back substitutions
are performed sequentially. This "alpha" release contains only
double-precision real data type.
The distribution contains the following directory structure:
SuperLU_MT/README instructions on installation
SuperLU_MT/CBLAS/ needed BLAS routines in C, not necessarily fast
SuperLU_MT/EXAMPLE/ example programs
SuperLU_MT/INSTALL/ test machine dependent parameters; the Users' Guide.
SuperLU_MT/SRC/ C source code, to be compiled into superlu_mt.a library
SuperLU_MT/TESTING/ driver routines to test correctness
SuperLU_MT/Makefile top level Makefile that does installation and testing
SuperLU_MT/make.inc compiler, compile flags, library definitions and C
preprocessor definitions, included in all Makefiles.
(You may need to edit it to be suitable for your system
before compiling the whole package.)
Before installing the package, please examine the three things dependent
on your system setup:
1. Edit the make.inc include file.
This make include file is referenced inside each of the Makefiles
in the various subdirectories. As a result, there is no need to
edit the Makefiles in the subdirectories. All information that is
machine specific has been defined in this include file.
The following machine-specific SuperLU_MT/make.inc files are
provided in the top-level directory:
make.inc Platforms
-------- ---------
make.pthread Any machine that supports POSIX threads
make.sun Sun Ultra Enterprise servers
make.alpha DEC Alpha Servers
make.sgi SGI Power Challenge
make.origin SGI/Cray Origin2000
make.cray Cray C90/J90
When you have selected the machine to which you wish to install SuperLU_MT,
copy the appropriate sample include file (if one is present) into
make.inc. For example, if you wish to run SuperLU_MT on a Sun SMP system,
you can do
cp make.sun make.inc
For the systems other than listed above, some porting effort is needed
for parallel factorization routines. Please refer to the Users' Guide
for detailed instructions on porting.
2. The BLAS library.
The parallel routines in SuperLU_MT uses some sequential BLAS routines
within each process (or thread). If there is BLAS library available on
your machine, you may define the following in the file make.inc:
BLASDEF = -DUSE_VENDOR_BLAS
BLASLIB =
The CBLAS/ subdirectory contains the part of the C BLAS needed by
SuperLU_MT package. However, these codes are intended for use only if
there is no faster implementation of the BLAS already available on your
machine. In this case, you should go to the top-level SuperLU_MT/
directory and do the following:
1) In make.inc, define:
BLASLIB = ../blas$(PLAT).a
2) Type:
make blaslib
to make the BLAS library from the routines in the CBLAS/ subdirectory.
3. C preprocessor definition CDEFS.
In the header file SRC/Cnames.h, we use macros to determine how
C routines should be named so that they are callable by Fortran.
(Some vendor-supplied BLAS libraries do not have C interface. So the
re-naming is needed in order for the SuperLU BLAS calls (in C) to
interface with the Fortran-style BLAS.)
The possible options for CDEFS are:
o -DAdd_: Fortran expects a C routine to have an underscore
postfixed to the name;
o -DNoChange: Fortran expects a C routine name to be identical to
that compiled by C;
o -DUpCase: Fortran expects a C routine name to be all uppercase.
A Makefile is provided in each subdirectory. The installation can be done
completely automatically by simply typing "make" at the top level.
REFERENCES
[1] A Supernodal Approach to Sparse Partial Pivoting,
James W. Demmel, Stanley C. Eisenstat, John R. Gilbert, Xiaoye S. Li
and Joseph W.H. Liu, Tech report UCB//CSD-95-883, Computer Science
Division, U.C. Berkeley, July, 1995. To appear in SIAM J. on Matrix
Anal. and Appl.
[2] An Asynchronous Parallel Supernodal Algorithm for Sparse Gaussian
Elimination, James W. Demmel, John R. Gilbert and Xiaoye S. Li,
Tech report UCB//CSD-97-943, Computer Science Division, U.C. Berkeley.
Feburary, 1997. Submitted to SIAM J. on Matrix Anal. and Appl.
[3] Sparse Gaussian Elimination on High Performance Computers,
Xiaoye S. Li, Tech report UCB//CSD-96-919, Computer Science Division,
U.C. Berkeley, September, 1996, Ph.D dissertation.
James Demmel, Univ. of California Berkeley, demmel@cs.berkeley.edu
John R. Gilbert, Xerox Palo Alto Research Center, gilbert@parc.xerox.com
Xiaoye S. Li, Lawrence Berkeley National Lab, xiaoye@nersc.gov