mpich: now 3.0.4 breaks on freebsd. [workarround by using 3.0.3 on it]
https://trac.mpich.org/projects/mpich/ticket/1820

Merge branch 'jed/sf-mpich1'

The last release of MPICH1 had some MPI-2 functions without essential
parts. They strictly adhere to MPIX_* naming now so that this problem
does not occur.

* jed/sf-mp

Merge branch 'jed/sf-mpich1'

The last release of MPICH1 had some MPI-2 functions without essential
parts. They strictly adhere to MPIX_* naming now so that this problem
does not occur.

* jed/sf-mpich1:
SF: define MPIU_REPLACE to work around missing MPI_REPLACE
SF: define and work around missing MPI_COMBINER_DUP [reconfigure-needed]

show more ...

BuildSystem: remove old files; unused since 2010 (7526b813c934d)

blaslapack: check for atlas in default locations [and not /usr/local/lib]

blaslapack: fix checkLapack(routines) - fixes macports atlas detection.

when [optional] 'routines' is modified inplace - subsequent call to
checkLapack() uses this modified values resulting in tests

blaslapack: fix checkLapack(routines) - fixes macports atlas detection.

when [optional] 'routines' is modified inplace - subsequent call to
checkLapack() uses this modified values resulting in tests for
ddgetrs, dddgetrs etc.. breaking the tests. Fix this by explicitly
making a copy of 'routines' - before modifying inplace.

Similarly fix checkBlas()

show more ...

mpich: 3.0.4 now works on opensloaris - so download_mpich_sol is not needed
http://trac.mpich.org/projects/mpich/ticket/1772

mpich: update to mpich-3.0.4.tar.gz

SF: define and work around missing MPI_COMBINER_DUP [reconfigure-needed]

A standards-compliant MPI cannot provide MPI_Type_get_envelope without
also defining MPI_COMBINER_DUP, because the latter is

SF: define and work around missing MPI_COMBINER_DUP [reconfigure-needed]

A standards-compliant MPI cannot provide MPI_Type_get_envelope without
also defining MPI_COMBINER_DUP, because the latter is necessary to
interpret the results of the former. However, a version of MPICH1 does
exactly this so we need a configure test and a workaround in the code.

show more ...

Merge branch 'balay/openmpi-1.6.4'

Merge branch 'balay/hdf5-1.8.10-patch1'

Merge branch 'balay/fix-downlad-fblas-on-win-ifort'

fblaslapack: set NOOPT correctly for download-fblas on win32/ifort #159403

hdf5: update to hdf5-1.8.10-patch1.tar.gz

configure: --download-openmpi=URL should also skip looking for mpicc - similar to --download-openmpi=1

openmpi: update tarball to openmpi-1.6.4.tar.gz

configure: remove debug print statement

Git download: recover gracefully from failed Git clones

Merge branch 'jed/cxxmangle-implies-cxx'

* jed/cxxmangle-implies-cxx:
libraries.check: mangleCxx implies compilation as C++

Merge branch 'jed/mpich-3.0.3'

* jed/mpich-3.0.3:
mpich-3.0.3

Merge branch 'jed/pkg-download-git'

* jed/pkg-download-git:
SuperLU_DIST: point gitcommit at v3.2
Git download: ML, SuperLU_DIST, MUMPS, ctetgen
package.py: add support for cloning from Git

mpich-3.0.3

libraries.check: mangleCxx implies compilation as C++

When a package sets functionsCxx, it implies that a C++ compiler must be
used. Formerly, the tests always compiled using the default clanguage
(

libraries.check: mangleCxx implies compilation as C++

When a package sets functionsCxx, it implies that a C++ compiler must be
used. Formerly, the tests always compiled using the default clanguage
(usually C). That can never work, so use a C++ compiler when necessary.

show more ...

gnupackage: fixup for moab [and also netcdf]
- say hdf5 is not configured - then pass in --disable-hdf5 to moab configure
- moab uses FC/FCFLAGS [not F77 etc.].
- add --disable-fortran for --with-fc=

gnupackage: fixup for moab [and also netcdf]
- say hdf5 is not configured - then pass in --disable-hdf5 to moab configure
- moab uses FC/FCFLAGS [not F77 etc.].
- add --disable-fortran for --with-fc=0
- use --enable-shared/--disable-shared correctly
[and remove mpich specific sharedlibrary configure options]

show more ...

moab: test for a proper c++ symbol [not the unused/fortran symbol]

netcdf5: setup the optional hdf5 options correctly