HPC Challenge 0.1.5

pts/hpcc-0.1.5 - 04 November 2013 - Allow HOSTFILE to be passed to MPI, allow overriding OMP_NUM_THREADS, and remove disabling of certain BLAS path checks. Almost ready for HPC Challenge benchmarking...

downloads.xml

<?xml version="1.0"?>
<!--Phoronix Test Suite v4.8.3 (Sokndal)-->
<PhoronixTestSuite>
  <Downloads>
    <Package>
      <URL>http://icl.cs.utk.edu/projectsfiles/hpcc/download/hpcc-1.4.3.tar.gz</URL>
      <MD5>dd4c2b6b3ec766c469c59d8310778f49</MD5>
      <SHA256>57cd4fd00065ebb63f5b298e54510be84aecba4b3c4788a0fb8b9fa79c752c79</SHA256>
      <FileSize>654049</FileSize>
    </Package>
  </Downloads>
</PhoronixTestSuite>

install.sh

#!/bin/sh

tar -zxvf hpcc-1.4.3.tar.gz
cd hpcc-1.4.3

# Find MPI To Use
if [ ! "X$MPI_PATH" = "X" ] && [ -d $MPI_PATH ] && [ -d $MPI_INCLUDE ] && [ -x $MPI_CC ] && [ -e $MPI_LIBS ]
then
	# PRE-SET MPI
	echo "Using pre-set environment variables."
elif [ -d /usr/lib/openmpi/include ]
then
	# OpenMPI
	MPI_PATH=/usr/lib/openmpi
	MPI_INCLUDE=/usr/lib/openmpi/include
	MPI_LIBS=/usr/lib/openmpi/lib/libmpi.so
	MPI_CC=/usr/bin/mpicc.openmpi
elif [ -d /usr/lib/mpich/include ]
then
	# MPICH
	MPI_PATH=/usr/lib/mpich
	MPI_INCLUDE=/usr/lib/mpich/include
	MPI_LIBS=/usr/lib/libmpich.so.1.0
	MPI_CC=/usr/lib/mpich/bin/mpicc
elif [ -d /usr/include/mpich2 ]
then
	# MPICH2
	MPI_PATH=/usr/include/mpich2
	MPI_INCLUDE=/usr/include/mpich2
	MPI_LIBS=/usr/lib/mpich2/lib/libmpich.so
	MPI_CC=/usr/bin/mpicc.mpich2
fi

# Find Linear Algebra Package To Use
if [ ! "X$LA_PATH" = "X" ] && [ -d $LA_PATH ] && [ -d $LA_INCLUDE ] && [ -e $LA_LIBS ]
then
	# PRE-SET MPI
	echo "Using pre-set environment variables."
elif [ -d /usr/lib/libblas ]
then
	# libblas
	LA_PATH=/usr/lib
	LA_INCLUDE=/usr/include
	LA_LIBS="-lblas"
elif [ -d /usr/lib/openblas-base ]
then
	# OpenBLAS
	LA_PATH=/usr/lib/openblas-base
	LA_INCLUDE=/usr/include
	LA_LIBS=/usr/lib/openblas-base/libopenblas.so.0
elif [ -d /usr/lib/atlas-base ]
then
	# ATLAS
	LA_PATH=/usr/lib/atlas-base
	LA_INCLUDE=/usr/include
	LA_LIBS="-llapack -lf77blas -lcblas -latlas"
fi

# Make.pts generation
echo "
SHELL        = /bin/sh
CD           = cd
CP           = cp
LN_S         = ln -s
MKDIR        = mkdir
RM           = /bin/rm -f
TOUCH        = touch
ARCH         = \$(arch)
TOPdir       = ../../..
INCdir       = \$(TOPdir)/include
BINdir       = \$(TOPdir)/bin/\$(ARCH)
LIBdir       = \$(TOPdir)/lib/\$(ARCH)
HPLlib       = \$(LIBdir)/libhpl.a 

# MPI

MPdir        = $MPI_PATH
MPinc        = -I$MPI_INCLUDE
MPlib        = $MPI_LIBS

# BLAS or VSIPL

LAdir        = $LA_PATH
LAinc        = -I$LA_INCLUDE
LAlib        = $LA_LIBS

# F77 / C interface 

F2CDEFS      =

# HPL includes / libraries / specifics

HPL_INCLUDES = -I\$(INCdir) -I\$(INCdir)/\$(ARCH) \$(LAinc) \$(MPinc)
HPL_LIBS     = \$(HPLlib) \$(LAlib) \$(MPlib) -lm
HPL_OPTS     = -DHPL_CALL_CBLAS
HPL_DEFS     = \$(F2CDEFS) \$(HPL_OPTS) \$(HPL_INCLUDES)
CC           = $MPI_CC
CCNOOPT      = \$(HPL_DEFS)
CCFLAGS      = \$(HPL_DEFS) -fomit-frame-pointer -O3 -march=native -funroll-loops
LINKER       = $MPI_CC
LINKFLAGS    = \$(CCFLAGS)
ARCHIVER     = ar
ARFLAGS      = r
RANLIB       = echo
" > hpl/Make.pts

# HPL.dat generation
N=$(echo "scale=0;sqrt(${NUM_CPU_CORES}*${SYS_MEMORY}*128)" |bc -l)
PQ=0
P=$(echo "scale=0;sqrt($NUM_CPU_CORES)" |bc -l)
Q=$P
PQ=$(($P*$Q))

while [ $PQ -ne $NUM_CPU_CORES ]; do
    Q=$(($NUM_CPU_CORES/$P))
    PQ=$(($P*$Q))
    if [ $PQ -ne $NUM_CPU_CORES ] && [ $P -gt 1 ]; then P=$(($P-1)); fi
done

echo "HPLinpack benchmark input file
Innovative Computing Laboratory, University of Tennessee
HPL.out      output file name (if any)
6            device out (6=stdout,7=stderr,file)
1            # of problems sizes (N)
$N
1            # of NBs
100          NBs
0            PMAP process mapping (0=Row-,1=Column-major)
1            # of process grids (P x Q)
$P           Ps
$Q           Qs
16.0         threshold
1            # of panel fact
2            PFACTs (0=left, 1=Crout, 2=Right)
1            # of recursive stopping criterium
4            NBMINs (>= 1)
1            # of panels in recursion
2            NDIVs
1            # of recursive panel fact.
2            RFACTs (0=left, 1=Crout, 2=Right)
1            # of broadcast
1            BCASTs (0=1rg,1=1rM,2=2rg,3=2rM,4=Lng,5=LnM)
1            # of lookahead depth
0            DEPTHs (>=0)
1            SWAP (0=bin-exch,1=long,2=mix)
64           swapping threshold
0            L1 in (0=transposed,1=no-transposed) form
0            U  in (0=transposed,1=no-transposed) form
1            Equilibration (0=no,1=yes)
8            memory alignment in double (> 0)
" > HPL.dat

cd hpl/
make arch=pts
cd ..
make arch=pts

echo $? > ~/install-exit-status

cd ~
echo "#!/bin/sh
cd hpcc-1.4.3

if [ \"X\$OMP_NUM_THREADS\" = \"X\" ]
then
	OMP_NUM_THREADS=\$NUM_CPU_CORES
fi

if [ ! \"X\$HOSTFILE\" = \"X\" ] && [ -f \$HOSTFILE ]
then
	\$HOSTFILE=\"--hostfile \$HOSTFILE\"
elif [ -f /etc/hostfile ]
then
	\$HOSTFILE=\"--hostfile /etc/hostfile\"
else
	\$HOSTFILE=\"\"
fi

mpirun -np \$OMP_NUM_THREADS \$HOSTFILE hpcc
echo \$? > ~/test-exit-status
cat hpccoutf.txt > \$LOG_FILE" > hpcc
chmod +x hpcc

results-definition.xml

<?xml version="1.0"?>
<!--Phoronix Test Suite v4.8.3 (Sokndal)-->
<PhoronixTestSuite>
  <ResultsParser>
    <OutputTemplate> TODO     =                  #_RESULT_#</OutputTemplate>
    <LineHint>TODO</LineHint>
  </ResultsParser>
</PhoronixTestSuite>

test-definition.xml

<?xml version="1.0"?>
<!--Phoronix Test Suite v4.8.3 (Sokndal)-->
<PhoronixTestSuite>
  <TestInformation>
    <Title>HPC Challenge</Title>
    <AppVersion>1.4.3</AppVersion>
    <Description>HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAm, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Most of this HPCC super-computing testing has been done on Ubuntu and Red Hat Enterprise Linux installations.</Description>
    <ResultScale>TODO</ResultScale>
    <Proportion>HIB</Proportion>
    <TimesToRun>3</TimesToRun>
  </TestInformation>
  <TestProfile>
    <Version>0.1.5</Version>
    <SupportedPlatforms>Linux</SupportedPlatforms>
    <SoftwareType>Benchmark</SoftwareType>
    <TestType>Processor</TestType>
    <License>Free</License>
    <Status>Unverified</Status>
    <ExternalDependencies>build-utilities, fortran-compiler, openmpi-development, blas-development, atlas-development</ExternalDependencies>
    <EnvironmentSize>9</EnvironmentSize>
    <ProjectURL>http://icl.cs.utk.edu/hpcc/index.html</ProjectURL>
    <InternalTags>SMP, MPI</InternalTags>
    <Maintainer>Michael Larabel</Maintainer>
  </TestProfile>
  <TestSettings>
    <Option>
      <DisplayName>Test / Class</DisplayName>
      <Identifier>run-test</Identifier>
      <Menu>
        <Entry>
          <Name>TODO</Name>
          <Value>TODO</Value>
        </Entry>
      </Menu>
    </Option>
  </TestSettings>
</PhoronixTestSuite>