The idea of IMB is to provide a concise set of elementary MPI
benchmark kernels. With one executable, all of the supported
benchmarks, or a subset specified by the command line, can be run.
The rules, such as time measurement (including a repetitive call
of the kernels for better clock synchronization), message lengths,
selection of communicators to run a particular benchmark (inside
the group of all started processes) are program parameters.
===> The following configuration options are available for IMB-2021.11_2:
CHECK_RESULTS=off: Check results (not valid as real benchmark data)
DOCS=on: Build and/or install documentation
===> Use 'make config' to modify these settings
benchmarks/imb: Fix P_Write_Shared segfault
With CHECK_RESULTS enabled, IMB-IO-openmpi P_Write_Shared crashes on
the first non-zero message size because IMB_chk_distr() uses reccnt,
while IMB_init_pointers() leaves it NULL for MPIIO builds.
Allocate the arrays needed for MPIIO builds.
PR: 296182
Reviewed by: thierry (mentor)
Approved by: thierry (mentor)
benchmarks/imb: not for 32 bit architectures
The code bases uses %lu or PRIu64 to print size_t values in various
places. Elsewhere it assumes size_t is 64 bits, such as when
evaluating ((size_t)1) << IMB_P2P_MAX_MSG_LOG where
IMB_P2P_MAX_MSG_LOG is greater than 31.
Approved by: portmgr (build fix blanket)
MFH: 2026Q2
Bump PORTREVISION for ports depending on the canonical version of GCC
as defined in Mk/bsd.default-versions.mk which has moved from GCC 8.3
to GCC 9.1 under most circumstances now after revision 507371.
This includes ports
- with USE_GCC=yes or USE_GCC=any,
- with USES=fortran,
- using Mk/bsd.octave.mk which in turn features USES=fortran, and
- with USES=compiler specifying openmp, nestedfct, c11, c++0x, c++11-lang,
c++11-lib, c++14-lang, c++17-lang, or gcc-c++11-lib
plus, everything INDEX-11 shows with a dependency on lang/gcc9 now.
PR: 238330
Bump PORTREVISION for ports depending on the canonical version of GCC
defined via Mk/bsd.default-versions.mk which has moved from GCC 7.4 t
GCC 8.2 under most circumstances.
This includes ports
- with USE_GCC=yes or USE_GCC=any,
- with USES=fortran,
- using Mk/bsd.octave.mk which in turn features USES=fortran, and
- with USES=compiler specifying openmp, nestedfct, c11, c++0x, c++11-lang,
c++11-lib, c++14-lang, c++17-lang, or gcc-c++11-lib
plus, as a double check, everything INDEX-11 showed depending on lang/gcc7.
PR: 231590
Change 4 ports from net/mpich2 to net/mpich
mpich2 is the obsolete version 2.x, mpich is at the current version 3.x
Ports:
biology/phyml benchmarks/mdtest benchmarks/imb math/parmetis
Convert ports to use the options helpers in categories [abc]*, and minor fixes.
Sponsored by: Absolight
Differential Revision: https://reviews.freebsd.org/D3412?
MASTER_SITES cleanup.
- Replace ${MASTER_SITE_FOO} with FOO.
- Merge MASTER_SITE_SUBDIR into MASTER_SITES when possible. (This means 99.9%
of the time.)
- Remove occurrences of MASTER_SITE_LOCAL when no subdirectory was present and
no hint of what it should be was present.
- Fix some logic.
- And generally, make things more simple and easy to understand.
While there, add magic values to the FESTIVAL, GENTOO, GIMP, GNUPG, QT and
SAMBA macros.
Also, replace some EXTRACT_SUFX occurences with USES=tar:*.
Checked by: make fetch-urlall-list
With hat: portmgr
Sponsored by: Absolight
Update the default version of GCC in the Ports Collection from GCC 4.7.4
to GCC 4.8.3.
Part II, Bump PORTREVISIONs.
PR: 192025
Tested by: antoine (-exp runs)
Approved by: portmgr (implicit)
Update the default version of GCC used in the Ports Collection from
GCC 4.6.4 to GCC 4.7.3. This entails updating the lang/gcc port as
well as changing the default in Mk/bsd.default-versions.mk.
Part II, Bump PORTREVISIONs.
PR: 182136
Supported by: Christoph Moench-Tegeder <cmt@burggraben.net> (fixing many ports)
Tested by: bdrewery (two -exp runs)
Allow the port to go a bit further in the build process.
Instead of complaining that `gmake' does not exist, the port fails a bit
later and at least starts building.
The previous failure was caused because we used USES=gmake but then re-set
BUILD_DEPENDS and MAKE_ENV a few lines below in the Makefile.
Update to libmpc version 1.0.1 which brings the following fixes:
- Switched to automake 1.11.6, see CVE-2012-3386.
- #14669: Fixed extraction of CC from gmp.h.
- Fixed case of intermediate zero real or imaginary part in mpc_fma,
found by hydra with GMP_CHECK_RANDOMIZE=1346362345.
This is on top of the following changes from version 1.0
- Licence change towards LGPLv3+ for the code and GFDLv1.3+ (with no
invariant sections) for the documentation.
- 100% of all lines are covered by tests
- Renamed functions
. mpc_mul_2exp to mpc_mul_2ui
. mpc_div_2exp to mpc_div_2ui
(Only the first 15 lines of the commit message are shown above )
Bump portrevision due to upgrade of devel/gettext.
The affected ports are the ones with gettext as a run-dependency
according to ports/INDEX-7 (5007 of them) and the ones with USE_GETTEXT
in Makefile (29 of them).
PR: ports/124340
Submitted by: edwin@
Approved by: portmgr (pav)
The idea of IMB is to provide a concise set of elementary MPI
benchmark kernels. With one executable, all of the supported
benchmarks, or a subset specified by the command line, can be run.
The rules, such as time measurement (including a repetitive call
of the kernels for better clock synchronization), message lengths,
selection of communicators to run a particular benchmark (inside
the group of all started processes) are program parameters.
WWW:
http://www.intel.com/cd/software/products/asmo-na/eng/cluster/mpi/219848.htm
PR: ports/105665
Submitted by: trasz <trasz at pin.if.uz.zgora.pl>
