alpscore / alpscore Goto Github PK

subj

Maybe cloudbees/jenkins can call up windows to build?

It might be good to write a README, containing info of the design of alpscore.

Are you sure you want to get rid of boost::array and boost::multiarray in alps core?

• boost::array is quite usefull, since it has a compiletime length (and the speed of a stack allocated array) and a container interface
• how are you going to store multidimensional measurements? Do you create your own datastructure?

Make deb / ppa / rpm / macports / homebrew.

We have close to 324 tests but jenkins seems to run only a few of these. What is going on?

make NGS compile using MPI, using OpenMP, and using both

sic, so that unknown pragma disappears.

CMake Error at mc/tutorials/02-mcbase/CMakeLists.txt:11 (add_executable):
Cannot find source file:

pscan.cpp

...could you please check it in?

Thanks!

py_test --junitxml=path should give xmls. Write and parse a demo test for python.

It used to use the hdf5-1.8.9 library but it seems to be not compatible with executables which were compiled with hdf5-1.8.13

if [[ -z "$HDF5_ROOT" ]]
then
echo
else
export DYLD_LIBRARY_PATH=$BOOST_ROOT/lib:$HDF5_ROOT/lib:$DYLD_LIBRARY_PATH
fi

-> why? And why is it checked when hdf5_root is given?

Make a jenkins test that compiles and runs with the clang address sanitizer to detect memory problems (factor of two).

Cmake produces the following error for doxygen:

CMake Warning (dev) at common/cmake/UseDoxygen.cmake:138 (get_target_property):
Policy CMP0045 is not set: Error on non-existent target in
get_target_property. Run "cmake --help-policy CMP0045" for policy details.
Use the cmake_policy command to set the policy and suppress this warning.

get_target_property() called with non-existent target "doc".

change scripts so that google test sources are only fetched once.

I experience random crashes when I try to reduce RealVectorObservables. Here's a code to test:
https://gist.github.com/aeantipov/f0e0a2af5a47c6b24171

Sometimes everything goes smoothly, sometimes I get the error. Here's the shell output (sorry, it's long)

(sci) [18:40] antipov@aantipov ~/code/alpscoretest/build $ mpirun --np 3 ./reduce_test02
e1: [1,..5..,1] #29855 +/-[0.0217015,..5..,0.0217015] Tau: [31.6243,..5..,31.6243] Bins: [[1,..5..,1],..232..,[0,..5..,0]]#128
e2: [1,..5..,1] #29855 +/-[0.0217015,..5..,0.0217015] Tau: [31.6243,..5..,31.6243] Bins: [[1,..5..,1],..232..,[0,..5..,0]]#128

(sci) [18:40] antipov@aantipov ~/code/alpscoretest/build $ mpirun --np 3 ./reduce_test02
e1: [1,..5..,1] #31224 +/-[0.0214054,..5..,0.0214054] Tau: [31.6809,..5..,31.6809] Bins: [[1,..5..,1],..242..,[0,..5..,0]]#128
e2: [1,..5..,1] #31224 +/-[0.0214054,..5..,0.0214054] Tau: [31.6809,..5..,31.6809] Bins: [[1,..5..,1],..242..,[0,..5..,0]]#128

(sci) [18:40] antipov@aantipov ~/code/alpscoretest/build $ mpirun --np 3 ./reduce_test02
libc++abi.dylib: terminating with uncaught exception of type std::logic_error: No alps::mpi::reduce available for this type NSt3__16vectorINS0_IdNS_9allocatorIdEEEENS1_IS3_EEEE
In /Users/antipov/_local/include/alps/accumulator/mpi.hpp on 89 in reduce_impl
    1   reduce_test02                       0x000000010089911c void alps::mpi::detail::reduce_impl >, std::__1::allocator > > >, std::__1::plus >(boost::mpi::communicator const&, std::__1::vector >, std::__1::allocator > > > const&, std::__1::plus, int, boost::integral_constant, boost::integral_constant) + 1260
    2   reduce_test02                       0x000000010089852b alps::accumulator::impl::Accumulator >, alps::accumulator::binning_analysis_tag, alps::accumulator::detail::simple_observable_type > > >::collective_merge(boost::mpi::communicator const&, int) const + 603
    3   reduce_test02                       0x0000000100897cf0 alps::accumulator::impl::Accumulator >, alps::accumulator::max_num_binning_tag, alps::accumulator::impl::Accumulator >, alps::accumulator::binning_analysis_tag, alps::accumulator::detail::simple_observable_type > > > >::collective_merge(boost::mpi::communicator const&, int) const + 32
    4   reduce_test02                       0x0000000100896e8c alps::accumulator::impl::Accumulator >, alps::accumulator::max_num_binning_tag, alps::accumulator::impl::Accumulator >, alps::accumulator::binning_analysis_tag, alps::accumulator::detail::simple_observable_type > > > >::collective_merge(boost::mpi::communicator const&, int) + 364
    5   reduce_test02                       0x000000010086163b alps::mcmpiadapter::collect_results(std::__1::vector, std::__1::allocator >, std::__1::allocator, std::__1::allocator > > > const&) const + 683
    6   reduce_test02                       0x00000001008612fd alps::mcmpiadapter::collect_results() const + 45
    7   reduce_test02                       0x000000010085f7a1 main + 513
    8   libdyld.dylib                       0x00007fff9139e5fd start + 1

[phys-098-161:38639] *** Process received signal ***
[phys-098-161:38639] Signal: Abort trap: 6 (6)
[phys-098-161:38639] Signal code:  (0)
[phys-098-161:38639] [ 0] libc++abi.dylib: terminating with uncaught exception of type std::logic_error: No alps::mpi::reduce available for this type NSt3__16vectorINS0_IdNS_9allocatorIdEEEENS1_IS3_EEEE
In /Users/antipov/_local/include/alps/accumulator/mpi.hpp on 89 in reduce_impl
    1   reduce_test02                       0x000000010412911c void alps::mpi::detail::reduce_impl >, std::__1::allocator > > >, std::__1::plus >(boost::mpi::communicator const&, std::__1::vector >, std::__1::allocator > > > const&, std::__1::plus, int, boost::integral_constant, boost::integral_constant) + 1260
    2   reduce_test02                       0x000000010412852b alps::accumulator::impl::Accumulator >, alps::accumulator::binning_analysis_tag, alps::accumulator::detail::simple_observable_type > > >::collective_merge(boost::mpi::communicator const&, int) const + 603
    3   reduce_test02                       0x0000000104127cf0 alps::accumulator::impl::Accumulator >, alps::accumulator::max_num_binning_tag, alps::accumulator::impl::Accumulator >, alps::accumulator::binning_analysis_tag, alps::accumulator::detail::simple_observable_type > > > >::collective_merge(boost::mpi::communicator const&, int) const + 32
    4   reduce_test02                       0x0000000104126e8c alps::accumulator::impl::Accumulator >, alps::accumulator::max_num_binning_tag, alps::accumulator::impl::Accumulator >, alps::accumulator::binning_analysis_tag, alps::accumulator::detail::simple_observable_type > > > >::collective_merge(boost::mpi::communicator const&, int) + 364
    5   reduce_test02                       0x00000001040f163b alps::mcmpiadapter::collect_results(std::__1::vector, std::__1::allocator >, std::__1::allocator, std::__1::allocator > > > const&) const + 683
    6   reduce_test02                       0x00000001040f12fd alps::mcmpiadapter::collect_results() const + 45
    7   reduce_test02                       0x00000001040ef7a1 main + 513
    8   libdyld.dylib                       0x00007fff9139e5fd start + 1

[phys-098-161:38638] *** Process received signal ***
[phys-098-161:38638] Signal: Abort trap: 6 (6)
[phys-098-161:38638] Signal code:  (0)
[phys-098-161:38638] [ 0] 0   libsystem_platform.dylib            0x00007fff89fb35aa _sigtramp + 26
[phys-098-161:38638] [ 1] 0   ???                                 0x0000000000000000 0x0 + 0
[phys-098-161:38638] [ 2] 0   libsystem_c.dylib                   0x00007fff901dfb1a abort + 125
[phys-098-161:38638] [ 3] 0   libc++abi.dylib                     0x00007fff94b17f31 __cxa_bad_cast + 0
[phys-098-161:38638] [ 4] 0   libsystem_platform.dylib            0x00007fff89fb35aa _sigtramp + 26
[phys-098-161:38639] [ 1] 0   ???                                 0x0000000000000000 0x0 + 0
[phys-098-161:38639] [ 2] 0   libsystem_c.dylib                   0x00007fff901dfb1a abort + 125
[phys-098-161:38639] [ 3] 0   libc++abi.dylib                     0x00007fff94b17f31 __cxa_bad_cast + 0
[phys-098-161:38639] [ 4] 0   libc++abi.dylib                     0x00007fff94b3d93a _ZL25default_terminate_handlerv + 240
[phys-098-161:38639] [ 5] 0   libobjc.A.dylib                     0x00007fff8838a322 _ZL15_objc_terminatev + 124
[phys-098-161:38639] [ 6] 0   libc++abi.dylib                     0x00007fff94b3b1d1 _ZSt11__terminatePFvvE + 8
[phys-098-161:38639] [ 7] 0   libc++abi.dylib                     0x00007fff94b3d93a _ZL25default_terminate_handlerv + 240
[phys-098-161:38638] [ 5] 0   libobjc.A.dylib                     0x00007fff8838a322 _ZL15_objc_terminatev + 124
[phys-098-161:38638] [ 6] 0   libc++abi.dylib                     0x00007fff94b3b1d1 _ZSt11__terminatePFvvE + 8
[phys-098-161:38638] [ 7] 0   libc++abi.dylib                     0x00007fff94b3ac5b _ZN10__cxxabiv1L22exception_cleanup_funcE19_Unwind_Reason_CodeP17_Unwind_Exception + 0
[phys-098-161:38638] [ 8] 0   libc++abi.dylib                     0x00007fff94b3ac5b _ZN10__cxxabiv1L22exception_cleanup_funcE19_Unwind_Reason_CodeP17_Unwind_Exception + 0
[phys-098-161:38639] [ 8] 0   reduce_test02                       0x000000010089932e _ZN4alps3mpi6detail11reduce_implINSt3__16vectorINS4_IdNS3_9allocatorIdEEEENS5_IS7_EEEENS3_4plusIdEEEEvRKN5boost3mpi12communicatorERKT_T0_iNSC_17integral_constantIbLb0EEESM_ + 1790
[phys-098-161:38639] [ 9] 0   reduce_test02                       0x000000010089852b _ZNK4alps11accumulator4impl11AccumulatorINSt3__16vectorIdNS3_9allocatorIdEEEENS0_20binning_analysis_tagENS0_6detail22simple_observable_typeIS7_EEE16collective_mergeERKN5boost3mpi12communicatorEi + 603
[phys-098-161:38639] [10] 0   reduce_test02                       0x0000000100897cf0 _ZNK4alps11accumulator4impl11AccumulatorINSt3__16vectorIdNS3_9allocatorIdEEEENS0_19max_num_binning_tagENS2_IS7_NS0_20binning_analysis_tagENS0_6detail22simple_observable_typeIS7_EEEEE16collective_mergeERKN5boost3mpi12communicatorEi + 32
[phys-098-161:38639] [11] 0   reduce_test02                       0x0000000100896e8c _ZN4alps11accumulator4impl11AccumulatorINSt3__16vectorIdNS3_9allocatorIdEEEENS0_19max_num_binning_tagENS2_IS7_NS0_20binning_analysis_tagENS0_6detail22simple_observable_typeIS7_EEEEE16collective_mergeERKN5boost3mpi12communicatorEi + 364
[phys-098-161:38639] [12] 0   reduce_test02                       0x000000010412932e _ZN4alps3mpi6detail11reduce_implINSt3__16vectorINS4_IdNS3_9allocatorIdEEEENS5_IS7_EEEENS3_4plusIdEEEEvRKN5boost3mpi12communicatorERKT_T0_iNSC_17integral_constantIbLb0EEESM_ + 1790
[phys-098-161:38638] [ 9] 0   reduce_test02                       0x000000010412852b _ZNK4alps11accumulator4impl11AccumulatorINSt3__16vectorIdNS3_9allocatorIdEEEENS0_20binning_analysis_tagENS0_6detail22simple_observable_typeIS7_EEE16collective_mergeERKN5boost3mpi12communicatorEi + 603
[phys-098-161:38638] [10] 0   reduce_test02                       0x0000000104127cf0 _ZNK4alps11accumulator4impl11AccumulatorINSt3__16vectorIdNS3_9allocatorIdEEEENS0_19max_num_binning_tagENS2_IS7_NS0_20binning_analysis_tagENS0_6detail22simple_observable_typeIS7_EEEEE16collective_mergeERKN5boost3mpi12communicatorEi + 32
[phys-098-161:38638] [11] 0   reduce_test02                       0x0000000104126e8c _ZN4alps11accumulator4impl11AccumulatorINSt3__16vectorIdNS3_9allocatorIdEEEENS0_19max_num_binning_tagENS2_IS7_NS0_20binning_analysis_tagENS0_6detail22simple_observable_typeIS7_EEEEE16collective_mergeERKN5boost3mpi12communicatorEi + 364
[phys-098-161:38638] [12] 0   reduce_test02                       0x00000001040f163b _ZNK4alps12mcmpiadapterI4sim1NS_14check_scheduleEE15collect_resultsERKNSt3__16vectorINS4_12basic_stringIcNS4_11char_traitsIcEENS4_9allocatorIcEEEENS9_ISB_EEEE + 683
[phys-098-161:38638] [13] 0   reduce_test02                       0x00000001040f12fd _ZNK4alps12mcmpiadapterI4sim1NS_14check_scheduleEE15collect_resultsEv + 45
[phys-098-161:38638] [14] 0   reduce_test02                       0x000000010086163b _ZNK4alps12mcmpiadapterI4sim1NS_14check_scheduleEE15collect_resultsERKNSt3__16vectorINS4_12basic_stringIcNS4_11char_traitsIcEENS4_9allocatorIcEEEENS9_ISB_EEEE + 683
[phys-098-161:38639] [13] 0   reduce_test02                       0x00000001008612fd _ZNK4alps12mcmpiadapterI4sim1NS_14check_scheduleEE15collect_resultsEv + 45
[phys-098-161:38639] [14] 0   reduce_test02                       0x000000010085f7a1 main + 513
[phys-098-161:38639] [15] 0   libdyld.dylib                       0x00007fff9139e5fd start + 1
[phys-098-161:38639] *** End of error message ***
0   reduce_test02                       0x00000001040ef7a1 main + 513
[phys-098-161:38638] [15] 0   libdyld.dylib                       0x00007fff9139e5fd start + 1
[phys-098-161:38638] *** End of error message ***
Received signal 15
--------------------------------------------------------------------------
mpirun noticed that process rank 2 with PID 38639 on node phys-098-161 exited on signal 6 (Abort trap: 6).
--------------------------------------------------------------------------

In legacy alps, boost can be compiled from source using custom cmake scripts from alps. This is not portable, but has one major advantage - it guarantees the same compiler flags.

An example of a problem with system boost - it is compiled with c++11, but alps-core is not -> clash in boost::filesystem.

One way to allow for source compilation of boost is to provide a module, that can be used as an optional dependence.

This is another intel warning... I don't understand the syntax here. @gamperl : can you help me out? This, too, showed up as part of Guy's #27 ...

builddir/ALPSCore/params/include/alps/params/paramiterator.hpp(37): warning #597: "alps::detail::paramiterator::operator alps::detail::paramiterator() const [with params_type=const alps::params, value_type=const alps::params::iterator_value_type]" will not be called for implicit or explicit conversions
operator paramiterator() const {
^
detected during instantiation of class "alps::detail::paramiterator [with params_type=const alps::params, value_type=const alps::params::iterator_value_type]" at line 22 of "builddir/ALPSCore/params/test/ordering.cpp"

On {Ubuntu 12.04, CMake 2.8.7, gcc 4.6.3} CMake fails with:

CMake Error at common/cmake/ALPSCompilerVersionCheck.cmake:6 
Insufficient gcc version

The variable CMAKE_CXX_COMPILER_VERSION appears to be empty.

Please support this, or if you don't warn with a friendly message.

Different compilers and compiler flags (such as c++11).
Figure out how do it.

get rid of the ifdefs:

if (NOT DEFINED ALPS_EXTENSIVE_HDF5_TESTS)

...and replace them with advanced options (ON/OFF)

make output:

Scanning dependencies of target alps-utility
Scanning dependencies of target gtest
[ 2%] [ 4%] [ 11%] [ 11%] [ 11%] [ 13%] Building CXX object utility/src/CMakeFiles/alps-utility.dir/encode.cpp.o
Building CXX object gtest/CMakeFiles/gtest.dir/src/gtest-all.cc.o
[ 15%] Building CXX object utility/src/CMakeFiles/alps-utility.dir/os.cpp.o
Building CXX object utility/src/CMakeFiles/alps-utility.dir/vmusage.cpp.o
Building CXX object utility/src/CMakeFiles/alps-utility.dir/temporary_filename.cpp.o
Building CXX object utility/src/CMakeFiles/alps-utility.dir/signal.cpp.o
Building CXX object utility/src/CMakeFiles/alps-utility.dir/stacktrace.cpp.o
[ 18%] [ 20%] Building CXX object utility/src/CMakeFiles/alps-utility.dir/stop_callback.cpp.o
Building CXX object utility/src/CMakeFiles/alps-utility.dir/parseargs.cpp.o
Linking CXX static library libalps-utility.a
[ 20%] Built target alps-utility
Scanning dependencies of target alps-hdf5
[ 22%] Building CXX object hdf5/src/CMakeFiles/alps-hdf5.dir/archive.cpp.o
Linking CXX static library libgtest.a
[ 22%] Built target gtest
Scanning dependencies of target gtest_main
[ 25%] Building CXX object gtest/CMakeFiles/gtest_main.dir/src/gtest_main.cc.o
Linking CXX static library libgtest_main.a
[ 25%] Built target gtest_main
Scanning dependencies of target temporary_filename
Scanning dependencies of target bitops
Scanning dependencies of target vmusage
[ 29%] [ 29%] [ 31%] Building CXX object utility/test/CMakeFiles/bitops.dir/bitops.cpp.o
Building CXX object utility/test/CMakeFiles/temporary_filename.dir/temporary_filename.cpp.o
Building CXX object utility/test/CMakeFiles/vmusage.dir/vmusage.cpp.o
Linking CXX executable temporary_filename
Linking CXX executable bitops
[ 31%] Built target temporary_filename
[ 31%] Built target bitops
Linking CXX executable vmusage
[ 31%] Built target vmusage
Linking CXX static library libalps-hdf5.a
[ 31%] Built target alps-hdf5
Scanning dependencies of target hdf5_complex
Scanning dependencies of target hdf5_family
Scanning dependencies of target hdf5_omp
Scanning dependencies of target hdf5_exceptions
Scanning dependencies of target hdf5_real_complex_vec
Scanning dependencies of target hdf5_memory
Scanning dependencies of target hdf5_bool
Scanning dependencies of target hdf5_pair
Scanning dependencies of target hdf5_multiarchive
Scanning dependencies of target hdf5_replace
Scanning dependencies of target hdf5_copy
Scanning dependencies of target hdf5_misc
[ 34%] [ 36%] [ 40%] [ 40%] [ 43%] [ 45%] Building CXX object hdf5/test/CMakeFiles/hdf5_memory.dir/hdf5_memory.cpp.o
[ 47%] [ 50%] [ 52%] Building CXX object hdf5/test/CMakeFiles/hdf5_bool.dir/hdf5_bool.cpp.o
[ 54%] Building CXX object hdf5/test/CMakeFiles/hdf5_exceptions.dir/hdf5_exceptions.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_complex.dir/hdf5_complex.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_family.dir/hdf5_family.cpp.o
[ 59%] [ 59%] Building CXX object hdf5/test/CMakeFiles/hdf5_omp.dir/hdf5_omp.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_real_complex_vec.dir/hdf5_real_complex_vec.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_multiarchive.dir/hdf5_multiarchive.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_copy.dir/hdf5_copy.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_replace.dir/hdf5_replace.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_pair.dir/hdf5_pair.cpp.o
Building CXX object hdf5/test/CMakeFiles/hdf5_misc.dir/hdf5_misc.cpp.o
builddir/ALPSCore/hdf5/test/hdf5_omp.cpp(17): warning #161: unrecognized #pragma
#pragma omp parallel for
^

Linking CXX executable hdf5_bool
[ 59%] Built target hdf5_bool
Scanning dependencies of target hdf5_vecveccplx
[ 61%] Building CXX object hdf5/test/CMakeFiles/hdf5_vecveccplx.dir/hdf5_vecveccplx.cpp.o
Linking CXX executable hdf5_omp
[ 61%] Built target hdf5_omp
Scanning dependencies of target hdf5_vecvecdbl
[ 63%] Linking CXX executable hdf5_complex
Building CXX object hdf5/test/CMakeFiles/hdf5_vecvecdbl.dir/hdf5_vecvecdbl.cpp.o
Linking CXX executable hdf5_memory
Linking CXX executable hdf5_misc
[ 63%] Built target hdf5_complex
Scanning dependencies of target alps-accumulator
[ 65%] Building CXX object accumulator/src/CMakeFiles/alps-accumulator.dir/accumulator.cpp.o
Linking CXX executable hdf5_replace
Linking CXX executable hdf5_pair
Linking CXX executable hdf5_real_complex_vec
[ 65%] Built target hdf5_misc
Scanning dependencies of target alps-params
[ 68%] Building CXX object params/src/CMakeFiles/alps-params.dir/params.cpp.o
[ 68%] Built target hdf5_memory
[ 70%] Building CXX object params/src/CMakeFiles/alps-params.dir/paramproxy.cpp.o
[ 70%] Built target hdf5_pair
[ 72%] [ 72%] Built target hdf5_replace
Building CXX object params/src/CMakeFiles/alps-params.dir/paramvalue.cpp.o
[ 72%] Built target hdf5_real_complex_vec
Linking CXX executable hdf5_multiarchive
Linking CXX executable hdf5_family
[ 72%] Built target hdf5_multiarchive
Linking CXX executable hdf5_exceptions
[ 72%] Built target hdf5_family
[ 72%] Built target hdf5_exceptions
Linking CXX executable hdf5_vecveccplx
[ 72%] Built target hdf5_vecveccplx
Linking CXX executable hdf5_vecvecdbl
[ 72%] Built target hdf5_vecvecdbl
builddir/ALPSCore/params/include/alps/params/paramiterator.hpp(37): warning #597: "alps::detail::paramiterator<params_type, value_type>::operator alps::detail::paramiterator<const alps::params, const alps::params::iterator_value_type>() const [with params_type=const alps::params, value_type=const alps::params::iterator_value_type]" will not be called for implicit or explicit conversions
operator paramiterator<const params_type, const value_type>() const {
^
detected during instantiation of class "alps::detail::paramiterator<params_type, value_type> [with params_type=const alps::params, value_type=const alps::params::iterator_value_type]" at line 85 of "builddir/ALPSCore/params/src/params.cpp"

Linking CXX static library libalps-accumulator.a
[ 72%] Built target alps-accumulator
Scanning dependencies of target count
Scanning dependencies of target mean
[ 75%] [ 77%] Building CXX object accumulator/test/CMakeFiles/count.dir/count.cpp.o
Building CXX object accumulator/test/CMakeFiles/mean.dir/mean.cpp.o
Linking CXX executable hdf5_copy
[ 77%] Built target hdf5_copy
Linking CXX static library libalps-params.a
[ 77%] Built target alps-params
Scanning dependencies of target stream
Scanning dependencies of target ordering
Scanning dependencies of target param
Scanning dependencies of target assign
Scanning dependencies of target default
Scanning dependencies of target not_found
Scanning dependencies of target alps-mc
[ 79%] [ 81%] [ 84%] [ 86%] Building CXX object params/test/CMakeFiles/not_found.dir/not_found.cpp.o
Building CXX object params/test/CMakeFiles/default.dir/default.cpp.o
Building CXX object params/test/CMakeFiles/ordering.dir/ordering.cpp.o
[ 88%] Building CXX object params/test/CMakeFiles/stream.dir/stream.cpp.o
[ 90%] Building CXX object params/test/CMakeFiles/assign.dir/assign.cpp.o
[ 93%] [ 95%] Building CXX object params/test/CMakeFiles/param.dir/param.cpp.o
[ 97%] Building CXX object mc/src/CMakeFiles/alps-mc.dir/mcoptions.cpp.o
Building CXX object mc/src/CMakeFiles/alps-mc.dir/mcbase.cpp.o
Building CXX object mc/src/CMakeFiles/alps-mc.dir/api.cpp.o
builddir/ALPSCore/params/include/alps/params/paramiterator.hpp(37): warning #597: "alps::detail::paramiterator<params_type, value_type>::operator alps::detail::paramiterator<const alps::params, const alps::params::iterator_value_type>() const [with params_type=const alps::params, value_type=const alps::params::iterator_value_type]" will not be called for implicit or explicit conversions
operator paramiterator<const params_type, const value_type>() const {
^
detected during instantiation of class "alps::detail::paramiterator<params_type, value_type> [with params_type=const alps::params, value_type=const alps::params::iterator_value_type]" at line 22 of "builddir/ALPSCore/params/test/ordering.cpp"

Linking CXX executable ordering
Linking CXX executable default
Linking CXX executable not_found
Linking CXX executable stream
[ 97%] Built target ordering
[ 97%] Built target not_found
[ 97%] Built target stream
[ 97%] Built target default
Linking CXX executable assign
[ 97%] Built target assign
Linking CXX executable param
[ 97%] Built target param
Linking CXX executable mean
Linking CXX executable count
[ 97%] [ 97%] Built target count
Built target mean
Linking CXX static library libalps-mc.a
[ 97%] Built target alps-mc
Scanning dependencies of target sum_single
[100%] Building CXX object mc/test/CMakeFiles/sum_single.dir/sum_single.cpp.o
Linking CXX executable sum_single
[100%] Built target sum_single

Should the build be run from cmake?
how should the copyright notice be called?
Should it be called from within cmake?
What happens to the master build.jenkins.sh script?

How about moving tutorials to a separate repo, say alpscore-tutorials? The pros are:

• Useful for demonstrations at schools and hands-on sessions.
• Gives a pattern for developers to start writing their application.
• Smaller repo is easier to maintain. Tutorials have a slower development cycle, in comparison with alpscore libs.
• More modular, 1 repo per 1 goal.

Cons are as well present:

• Need to compile alpscore first. Tutorials won't get compiled automatically.
• Separate maintenance : automated builds, etc.

MPICXX or similar should be used, rather than overwriting CXX as compiler.

It would be nice in future to have an interface for alps::params to populate it with the values from command line.

This is a typical (harmless) warning caused by our use of ublas in clang:

In file included from /Users/egull/ALPSCore/ALPSCore/install/include/alps/accumulator/accumulator.hpp:18:
In file included from /Users/egull/ALPSCore/ALPSCore/install/include/alps/accumulator/wrappers.hpp:16:
In file included from /Users/egull/ALPSCore/ALPSCore/install/include/alps/accumulator/feature/max_num_binning.hpp:24:
In file included from /Users/egull/ALPSCore/ALPSCore/install/include/alps/numeric/outer_product.hpp:13:
In file included from /Users/egull/ALPSCore/ALPSCore/install/include/alps/type_traits/covariance_type.hpp:15:
In file included from /opt/local/include/boost/numeric/ublas/matrix.hpp:16:
In file included from /opt/local/include/boost/numeric/ublas/vector.hpp:19:
In file included from /opt/local/include/boost/numeric/ublas/storage.hpp:26:
/opt/local/include/boost/numeric/ublas/traits.hpp:45:27: warning: unused function 'boost_numeric_ublas_abs' [-Wunused-function]
template<> unsigned int boost_numeric_ublas_abs (const unsigned int& t) {

how to get rid of it?

Looking at this:

http://schneide.wordpress.com/2014/01/27/integrating-googletest-in-cmake-projects-and-jenkins/

it seems to be possible to have the output of gtest use the JUnit format, which then can be understood by Jenkins.

Is there a possibility to do it?

Please see the following:
https://gist.github.com/guycohen/bc2804963763088dd8ed

Some builds fail with 'tar: unable to open archive' which indicates some filesystem bug.

as discussed. With acknowledgments.txt and check scripts

Did this:
4 git pull
11 mkdir utility_build
12 cd utility_build/
13 cmake ../utility

and got this:
...
-- gtest is in /Users/egull/ALPSCore/ALPSCore/utility_build/gtest
CMake Error at CMakeLists.txt:39 (add_subdirectory):
add_subdirectory not given a binary directory but the given source
directory "/Users/egull/ALPSCore/ALPSCore/utility_build/gtest" is not a
subdirectory of "/Users/egull/ALPSCore/ALPSCore/utility". When specifying
an out-of-tree source a binary directory must be explicitly specified.

-- Configuring incomplete, errors occurred!
See also "/Users/egull/ALPSCore/ALPSCore/utility_build/CMakeFiles/CMakeOutput.log".

Doxygen is set up, but commenting the files is to be done.

on edison, NERSC

The method alps::mcmpiadapter::collect_results on SignedRealObservable and SignedRealVectorObservable raises an error

MPI_Reduce: MPI_ERR_ARG: invalid argument of some other kind

This error disappears when RealObservable and RealVectorObservable are used.

Subj: compilation time of the library on a laptop with 5 threads without typed hdf5 tests costs 42s.
With tests: 4m 23s.

So, it might be useful to mark some specific tests as advanced and have them turned on for automated builds and debugging.

{GCC 4.6.3, Boost 1.48} from Ubuntu 12.04:

Building CXX object accumulator/src/CMakeFiles/alps-accumulator.dir/accumulator.cpp.o
......
.../accumulator/include/alps/accumulator/wrappers.hpp:57:41: error: ‘value_type’ is not a type

(and several more error messages like this)

...to be consistent with boost.

We have #pragma omp lurking around at exactly one place in our code:
../hdf5/test/hdf5_omp.cpp
nowhere else do we test or use openmp. The omp causes warnings under intel (see #27)
Given that we only have a single reference to openmp in the code, and even that one is only in a test: should we just remove all mention of openmp and have user codes worry about it? Any opinions? @aeantipov ? @gamperl?

move covariance out of ublas and somewhere else as external function or similar