Problem with building LBPM

[haghanire]

Dear James,

I faced the following problem while installing LBPM. The configure file is :

cmake
-D CMAKE_BUILD_TYPE:STRING=Release
-D CMAKE_C_COMPILER:PATH=$MPI_DIR/bin/mpicc
-D CMAKE_CXX_COMPILER:PATH=$MPI_DIR/bin/mpicxx
-D CMAKE_C_FLAGS="-O3 -fPIC"
-D CMAKE_CXX_FLAGS="-fPIC -O3 -std=c++14"
-D CMAKE_CXX_STANDARD=14
-D MPIEXEC=$MPI_DIR/bin/mpirun
-D CUDA_FLAGS="-arch sm_35"
-D CUDA_HOST_COMPILER="/usr/bin/gcc"
-D HDF5_DIRECTORY=$LBPM_HDF5_DIR
-D HDF5_LIB=$LBPM_HDF5_DIR/lib/libhdf5.a
-D USE_SILO=1
-D SILO_LIB=$LBPM_SILO_DIR/lib/libsiloh5.a
-D SILO_DIRECTORY=$LBPM_SILO_DIR
-D USE_MPI=1
-D USE_NETCDF=0
-D NETCDF_DIRECTORY="/opt/netcdf/4.6.1"
-D USE_CUDA=0
-D USE_TIMER=0
$LBPM_SOURCE

and I get the following errors while compiling the source code:

[ 0%] Built target copy-LBPM--home-rezaha-lbpm-LBPM-include
[ 0%] Built target copy-LBPM-include
Scanning dependencies of target lbpm-wia-library
[ 0%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/Array.cpp.o
[ 1%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/Communication.cpp.o
[ 1%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/Database.cpp.o
[ 2%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/Domain.cpp.o
[ 3%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/FunctionTable.cpp.o
[ 3%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/MPI.cpp.o
/home/rezaha/lbpm/LBPM/common/MPI.cpp: In function ‘constexpr long unsigned int Utilities::offset_long()’:
/home/rezaha/lbpm/LBPM/common/MPI.cpp:156:73: warning: integer overflow in expression of type ‘long int’ results in ‘-9223372036854775808’ [-Woverflow]
156 | return ~static_cast(std::numeric_limits::min()) + 1;
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~
/home/rezaha/lbpm/LBPM/common/MPI.cpp: In function ‘constexpr long long unsigned int Utilities::offset_long_long()’:
/home/rezaha/lbpm/LBPM/common/MPI.cpp:160:59: warning: integer overflow in expression of type ‘long long int’ results in ‘-9223372036854775808’ [-Woverflow]
159 | return ~static_cast(
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
160 | std::numeric_limits::min()) +
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^
161 | 1;
| ~
[ 4%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/ReadMicroCT.cpp.o
[ 4%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/ScaLBL.cpp.o
[ 5%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/SpherePack.cpp.o
[ 5%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/UnitTest.cpp.o
[ 6%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/Units.cpp.o
[ 6%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/Utilities.cpp.o
[ 7%] Building CXX object CMakeFiles/lbpm-wia-library.dir/common/WideHalo.cpp.o
[ 7%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/ElectroChemistry.cpp.o
[ 8%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/FlowAdaptor.cpp.o
[ 9%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/FreeEnergy.cpp.o
[ 9%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/GreyPhase.cpp.o
[ 10%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/Minkowski.cpp.o
[ 10%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/SubPhase.cpp.o
[ 11%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/TwoPhase.cpp.o
[ 11%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/analysis.cpp.o
[ 12%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/dcel.cpp.o
[ 12%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/distance.cpp.o
[ 13%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/filters.cpp.o
[ 14%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/morphology.cpp.o
[ 14%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/runAnalysis.cpp.o
[ 15%] Building CXX object CMakeFiles/lbpm-wia-library.dir/analysis/uCT.cpp.o
[ 15%] Building CXX object CMakeFiles/lbpm-wia-library.dir/IO/HDF5Writer.cpp.o
[ 16%] Building CXX object CMakeFiles/lbpm-wia-library.dir/IO/HDF5_IO.cpp.o
[ 16%] Building CXX object CMakeFiles/lbpm-wia-library.dir/IO/Mesh.cpp.o
[ 17%] Building CXX object CMakeFiles/lbpm-wia-library.dir/IO/MeshDatabase.cpp.o
[ 17%] Building CXX object CMakeFiles/lbpm-wia-library.dir/IO/PIO.cpp.o
[ 18%] Building CXX object CMakeFiles/lbpm-wia-library.dir/IO/PackData.cpp.o
In file included from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:12:1: error: ‘size_t’ does not name a type
12 | size_t packsize( const TYPE &rhs );
| ^~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:8:1: note: ‘size_t’ is defined in header ‘’; did you forget to ‘#include ’?
7 | #include
+++ |+#include
8 |
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:25:1: error: ‘size_t’ does not name a type
25 | size_t packsize( const std::vector &rhs );
| ^~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:25:1: note: ‘size_t’ is defined in header ‘’; did you forget to ‘#include ’?
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:38:1: error: ‘size_t’ does not name a type
38 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs );
| ^~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:38:1: note: ‘size_t’ is defined in header ‘’; did you forget to ‘#include ’?
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:51:1: error: ‘size_t’ does not name a type
51 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs );
| ^~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:51:1: note: ‘size_t’ is defined in header ‘’; did you forget to ‘#include ’?
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:64:1: error: ‘size_t’ does not name a type
64 | size_t packsize( const std::set &rhs );
| ^~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:64:1: note: ‘size_t’ is defined in header ‘’; did you forget to ‘#include ’?
/home/rezaha/lbpm/LBPM/IO/PackData.cpp:11:8: error: template-id ‘packsize’ for ‘size_t packsize(const unsigned char&)’ does not match any template declaration
11 | size_t packsize( const unsigned char &rhs )
| ^~~~~~~~~~~~~~~~~~~~~~~
In file included from /home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:75,
from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:122:8: note: candidates are: ‘template size_t packsize(const std::set&)’
122 | size_t packsize( const std::set &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:76:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::map<TYPE1, TYPE2>&)’
76 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:54:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::pair<_T1, _T2>&)’
54 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:17:8: note: ‘template size_t packsize(const std::vector&)’
17 | size_t packsize( const std::vector &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM/IO/PackData.cpp:27:8: error: template-id ‘packsize’ for ‘size_t packsize(const char&)’ does not match any template declaration
27 | size_t packsize( const char &rhs )
| ^~~~~~~~~~~~~~
In file included from /home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:75,
from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:122:8: note: candidates are: ‘template size_t packsize(const std::set&)’
122 | size_t packsize( const std::set &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:76:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::map<TYPE1, TYPE2>&)’
76 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:54:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::pair<_T1, _T2>&)’
54 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:17:8: note: ‘template size_t packsize(const std::vector&)’
17 | size_t packsize( const std::vector &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM/IO/PackData.cpp:43:8: error: template-id ‘packsize’ for ‘size_t packsize(const int&)’ does not match any template declaration
43 | size_t packsize( const int &rhs )
| ^~~~~~~~~~~~~
In file included from /home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:75,
from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:122:8: note: candidates are: ‘template size_t packsize(const std::set&)’
122 | size_t packsize( const std::set &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:76:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::map<TYPE1, TYPE2>&)’
76 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:54:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::pair<_T1, _T2>&)’
54 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:17:8: note: ‘template size_t packsize(const std::vector&)’
17 | size_t packsize( const std::vector &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM/IO/PackData.cpp:59:8: error: template-id ‘packsize’ for ‘size_t packsize(const unsigned int&)’ does not match any template declaration
59 | size_t packsize( const unsigned int &rhs )
| ^~~~~~~~~~~~~~~~~~~~~~
In file included from /home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:75,
from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:122:8: note: candidates are: ‘template size_t packsize(const std::set&)’
122 | size_t packsize( const std::set &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:76:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::map<TYPE1, TYPE2>&)’
76 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:54:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::pair<_T1, _T2>&)’
54 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:17:8: note: ‘template size_t packsize(const std::vector&)’
17 | size_t packsize( const std::vector &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM/IO/PackData.cpp:75:8: error: template-id ‘packsize<size_t>’ for ‘size_t packsize(const size_t&)’ does not match any template declaration
75 | size_t packsize<size_t>( const size_t &rhs )
| ^~~~~~~~~~~~~~~~
In file included from /home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:75,
from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:122:8: note: candidates are: ‘template size_t packsize(const std::set&)’
122 | size_t packsize( const std::set &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:76:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::map<TYPE1, TYPE2>&)’
76 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:54:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::pair<_T1, _T2>&)’
54 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:17:8: note: ‘template size_t packsize(const std::vector&)’
17 | size_t packsize( const std::vector &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM/IO/PackData.cpp:91:8: error: template-id ‘packsizestd::string’ for ‘size_t packsize(const string&)’ does not match any template declaration
91 | size_t packsizestd::string( const std::string &rhs )
| ^~~~~~~~~~~~~~~~~~~~~
In file included from /home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.h:75,
from /home/rezaha/lbpm/LBPM/IO/PackData.cpp:1:
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:122:8: note: candidates are: ‘template size_t packsize(const std::set&)’
122 | size_t packsize( const std::set &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:76:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::map<TYPE1, TYPE2>&)’
76 | size_t packsize( const std::map<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:54:8: note: ‘template<class TYPE1, class TYPE2> size_t packsize(const std::pair<_T1, _T2>&)’
54 | size_t packsize( const std::pair<TYPE1, TYPE2> &rhs )
| ^~~~~~~~
/home/rezaha/lbpm/LBPM_BUILD/include/IO/PackData.hpp:17:8: note: ‘template size_t packsize(const std::vector&)’
17 | size_t packsize( const std::vector &rhs )
| ^~~~~~~~
make[2]: *** [CMakeFiles/lbpm-wia-library.dir/build.make:498: CMakeFiles/lbpm-wia-library.dir/IO/PackData.cpp.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:336: CMakeFiles/lbpm-wia-library.dir/all] Error 2
make: *** [Makefile:160: all] Error 2

Could you please let me know what is wrong? I use the same configure and install LBPM successfully on another pc, however, it does not work on this one.

Thank you so much.

[JamesEMcClure]

Check to make sure that the gcc compiler passed to CUDA is the same one that MPI is using. First use

mpicc --show

This should tell you how mpi is calling the compiler. Then get the full path to gcc,

which gcc

Then provide this path to CUDA_HOST_COMPILER

I can't guarantee this will fix it, but this is my initial hunch. Let me know if it doesn't succeed.

[haghanire]

Hi, I got the following:

$ mpicc -show
x86_64-conda_cos6-linux-gnu-cc -I/home/rezaha/anaconda3/include -I/home/rezaha/anaconda3/include -L/home/rezaha/anaconda3/lib -Wl,-rpath,/home/rezaha/anaconda3/lib -I/home/rezaha/anaconda3/include -L/home/rezaha/anaconda3/lib -lmpi
$ which gcc
/usr/bin/gcc

I do not know what to do about the first command, however, the CUDA_HOST_COMPILER path is already the same as gcc path. Could you elaborate on the first comment?. Thank you so much for your time.

[JamesEMcClure]

I would check gcc --version and make sure that it is compatible with the version of cuda that you are using.

I think the anaconda compiler is probably not the best choice for MPI compiler. It may work fine for c/c++ code but I'm not sure how nicely it will play with the gpu code. If you are using ubuntu on a local workstation the compiler installed with sudo apt install build-essential should work with cuda. If you not on a shared system let me know and I can share more info.

[haghanire]

A new Ubuntu was installed and I followed the steps and installed build-essential. Unfortunately, it didn't work, so I installed anaconda compiler as well and it didn't either. It a local workstation with Intel Core i7 vPRO.

[JamesEMcClure]

I suspect you will need to rebuild the dependencies, if you have not done so already, since these may contain symbols that depend on the previous version of the OS.

[JamesEMcClure]

Can you try adding the header file below to the file IO/PackData.h and see if it fixes the issue on your build?

#include <cstddef>

[haghanire]

Hi,
By downgrading Ubuntu from 21.10 to 20.04.4 LTS, the problem was solved. I think the problem that I faced was a matter of code compatibility with new Ubuntu version.
Thanks