svn commit: r487566 - in head/math: . hipmcl hipmcl/files
Yuri Victorovich
yuri at FreeBSD.org
Sun Dec 16 09:06:15 UTC 2018
Author: yuri
Date: Sun Dec 16 09:06:13 2018
New Revision: 487566
URL: https://svnweb.freebsd.org/changeset/ports/487566
Log:
New port: math/hipmcl: High-performance parallel algorithm for large-scale network clustering
Added:
head/math/hipmcl/
head/math/hipmcl/Makefile (contents, props changed)
head/math/hipmcl/distinfo (contents, props changed)
head/math/hipmcl/files/
head/math/hipmcl/files/patch-CMakeLists.txt (contents, props changed)
head/math/hipmcl/pkg-descr (contents, props changed)
Modified:
head/math/Makefile
Modified: head/math/Makefile
==============================================================================
--- head/math/Makefile Sun Dec 16 09:01:08 2018 (r487565)
+++ head/math/Makefile Sun Dec 16 09:06:13 2018 (r487566)
@@ -264,6 +264,7 @@
SUBDIR += grpn
SUBDIR += gsl
SUBDIR += hexcalc
+ SUBDIR += hipmcl
SUBDIR += hs-Agda
SUBDIR += hs-Agda-stdlib
SUBDIR += hs-NumInstances
Added: head/math/hipmcl/Makefile
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/math/hipmcl/Makefile Sun Dec 16 09:06:13 2018 (r487566)
@@ -0,0 +1,38 @@
+# $FreeBSD$
+
+PORTNAME= HipMCL
+DISTVERSION= g20180814
+CATEGORIES= math
+MASTER_SITES= https://bitbucket.org/${BB_ACCOUNT}/${BB_PROJECT}/get/${BB_COMMIT}.tar.gz?dummy=/
+
+MAINTAINER= yuri at FreeBSD.org
+COMMENT= High-performance parallel algorithm for large-scale network clustering
+
+LICENSE= BSD3CLAUSE
+LICENSE_FILE= ${WRKSRC}/license.txt
+
+LIB_DEPENDS= libCombBLAS.so:math/combblas \
+ libmpi.so:net/mpich \
+ libomp.so:devel/openmp
+RUN_DEPENDS= ${LOCALBASE}/lib/gcc${GCC_DEFAULT}/libgcc_s.so.1:lang/gcc${GCC_DEFAULT} # because of the link flags induced by mpich
+
+USES= cmake:outsource compiler:c++14-lang dos2unix localbase:ldflags
+DOS2UNIX_FILES= CMakeLists.txt
+
+CMAKE_ON= BUILD_SHARED_LIBS
+
+BB_ACCOUNT= azadcse
+BB_PROJECT= ${PORTNAME:tl}
+BB_COMMIT= e20476acc473
+
+WRKSRC= ${WRKDIR}/${BB_ACCOUNT}-${BB_PROJECT}-${BB_COMMIT}
+
+PLIST_FILES= bin/${PORTNAME:tl}
+
+post-extract:
+ @${RM} -r ${WRKSRC}/src/CombBLAS
+
+do-install:
+ ${INSTALL_PROGRAM} ${BUILD_WRKSRC}/bin/hipmcl ${STAGEDIR}${PREFIX}/bin
+
+.include <bsd.port.mk>
Added: head/math/hipmcl/distinfo
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/math/hipmcl/distinfo Sun Dec 16 09:06:13 2018 (r487566)
@@ -0,0 +1,3 @@
+TIMESTAMP = 1544950762
+SHA256 (HipMCL-g20180814.tar.gz) = 324fbe12f3547cd1e3da38d01ea2d6882061821a7cf6a7904e1402759c01e94c
+SIZE (HipMCL-g20180814.tar.gz) = 43337935
Added: head/math/hipmcl/files/patch-CMakeLists.txt
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/math/hipmcl/files/patch-CMakeLists.txt Sun Dec 16 09:06:13 2018 (r487566)
@@ -0,0 +1,15 @@
+--- CMakeLists.txt.orig 2018-08-14 19:42:54 UTC
++++ CMakeLists.txt
+@@ -30,10 +30,10 @@ else (NOT C14)
+ endif (NOT C14)
+
+
+-ADD_SUBDIRECTORY( src/CombBLAS )
++#ADD_SUBDIRECTORY( src/CombBLAS )
+ set(CMAKE_RUNTIME_OUTPUT_DIRECTORY ${CMAKE_BINARY_DIR}/bin)
+ add_executable(hipmcl src/HipMCL.cpp)
+-target_link_libraries(hipmcl CombBLASlib )
++target_link_libraries(hipmcl CombBLAS )
+
+
+
Added: head/math/hipmcl/pkg-descr
==============================================================================
--- /dev/null 00:00:00 1970 (empty, because file is newly added)
+++ head/math/hipmcl/pkg-descr Sun Dec 16 09:06:13 2018 (r487566)
@@ -0,0 +1,16 @@
+HipMCL is a high-performance parallel algorithm for large-scale network
+clustering. HipMCL parallelizes popular Markov Cluster (MCL) algorithm that has
+been shown to be one of the most successful and widely used algorithms for
+network clustering. It is based on random walks and was initially designed to
+detect families in protein-protein interaction networks. Despite MCL's
+efficiency and multi-threading support, scalability remains a bottleneck as it
+fails to process networks of several hundred million nodes and billion edges in
+an affordable running time. HipMCL overcomes all of these challenges by
+developing massively-parallel algorithms for all components of MCL. HipMCL can
+be 1000 times faster than the original MCL without any information loss. It can
+easily cluster a network of ~75 million nodes with ~68 billion edges in ~2.4
+hours using ~2000 nodes of Cori supercomputer at NERSC. HipMCL is developed in
+C++ language and uses standard OpenMP and MPI libraries for shared- and
+distributed-memory parallelization.
+
+WWW: https://bitbucket.org/azadcse/hipmcl/wiki/Home
More information about the svn-ports-all
mailing list