This code has been tested under Ubuntu 20.04. Other Linux Distributions will probably work. However, we do not recommend using Windows or MacOS. If you do not have Linux installed on your computer, please use a Virtual Machine or similar.
- Under Ubuntu you can simply run
apt install libopenmpi-devto install OpenMPI.
- Please install PETSc on your system.
Recommended: Use PETSc from Ubuntu 20.04 package repositories
We recommend to use apt install libpetsc3.12-dev-common to install PETSc 3.12 under Ubuntu 20.04. See https://packages.ubuntu.com/focal/libpetsc3.12-dev-common
Alternative: Building PETSc from source
You may also use other PETSc Versions. But this might break. This code should work with PETSc 3.7.4 or newer. If not, feel free to open an issue.
If you are building PETSc from source, the specific PETSc version might be incompatible with your system MPI. To fix this issue, please make sure to use the following options, when configuring PETSc:
python2 ./configure --download-mpich --download-f2cblaslapack
This will tell PETSc to take care of downloading a compatible MPI version. Additionally, please set the following environment variables before running CMake:
export PETSC_DIR=<Folder Containing PETSc>/petsc-x.y.z/
export PETSC_ARCH=arch-linux2-c-debug
export MPI_HOME=${PETSC_DIR}/${PETSC_ARCH}
When running cmake .., please check the output to ensure that the correct PETSc version and MPI is used. You should see something like this:
/hpc_turbulence_code/build$ cmake ..
...
-- Detecting additional PETSc prefixes
-- Detected <Folder Containing PETSc>/petsc-x.y.z/arch-linux2-c-debug
-- Skipped the following invalid prefixes: <Folder Containing PETSc>/petsc-x.y.z/
-- pkg-config will search the following paths:
CMAKE_PREFIX_PATH
<Folder Containing PETSc>/petsc-x.y.z/arch-linux2-c-debug
-- Checking for module 'PETSc'
-- Found PETSc, version x.y.z
-- Found MPI_CXX: <Folder Containing PETSc>/petsc-x.y.z/arch-linux2-c-debug/lib/libmpicxx.so (found version "a.b")
-- Found MPI: TRUE (found version "a.b")
-- Found PETSc: TRUE (found version "x.y.z")
-- Configuring done
-- Generating done
-- Build files have been written to: /hpc_turbulence_code/build
If you are building PETSc from source, make sure to use the correct mpirun, when running the program! Use $MPI_HOME/bin/mpirun, not just mpirun.
To run the project, you have to install the Eigen library. Execute the following commands in the root directory:
- Create a library directory:
mkdir lib - Switch to this directory:
cd lib - Clone Eigen:
git clone [email protected]:libigl/eigen.git eigen - Switch to eigen directory:
cd eigen - Create build directory:
mkdir build - Switch to this directory:
cd build - Set eigen install dir in cmake (cmake-gui can come in handy):
sudo apt-get install cmake-guicmake-gui- Set CMAKE_INSTALL_PREFIX to
<project_root_dir_here>/lib/eigen
- Then make the Eigen project by running the following commands in
<project_root_dir_here>/lib/eigen/build:make && make install <project_root_dir_here>/lib/eigen/folder should contain bothshareandinclude. If not, it means that the Eigen is installed on the system directly. In that case, you have to copyshareandincludeto<project_root_dir_here>/lib/eigen/or set the EIGEN_PARENT_DIR in the project's CMake to make it point to the folder containingshareandinclude.
Run apt install cmake to install CMake on your system.
You can also use the Dockerfile included in the repository. This will create an image where everything is set up and you can just begin with compiling the code.
Of course, you can also just look at the Dockerfile as a recipe on how to set up your environment.
As build system configurator we use CMake. To compile the code execute the following commands in this directory:
- Create a build directory:
mkdir build - Switch to this directory:
cd build - (Optional): Choose the compiler being used (if you want to use a specific MPI compiler/version):
export CXX=mpic++ - Run CMake:
cmake ..(this configures a Release build, which is default. For a Debug build runcmake .. -DCMAKE_BUILD_TYPE=Debug) - Run Make:
make(ormake -jfor compiling with multiple cores) - (Optional): Run
make testto validate your build (this will execute some unit tests and run some simulations)
In the code skeleton, some basic unit tests have been implemented (make test). Feel free to add your own test cases inside the Tests folder.
- Run the code in serial via
./build/ns path/to/your/configuration- Example:
./build/ns ExampleCases/Cavity2D.xml(for a Debug build, the executable is./build/nsd)
- Example:
- Run the code in parallel via
mpirun -np n_proc ./build/ns path/to/your/configuration- Example:
mpirun -np 4 ./build/ns ExampleCases/Cavity2DParallel.xml(in the skeleton version of the code this is expected to not work!)
- Example:
You can add new source files by just creating them somewhere within the Source folder. CMake automatically detects these files and adds them to the build.
Probably you are talkling about the include guard. More information can be found here.
Note: If you forget to rename the include guard, when you copy-paste an existing class, the new class will not be compiled properly, since the class is not included properly. It is guarded by the include guard, even though it should not.
Example: If you used RHSStencil.hpp as a template your original include guard will be
#ifndef __RHS_STENCIL_HPP_
#define __RHS_STENCIL_HPP_
for DPStencil.hpp, you should change it to
#ifndef __DP_STENCIL_HPP__
#define __DP_STENCIL_HPP__
Make sure to use make clean before you use make. Sometimes there are build artifacts from previous build processes that spoil your current compilation process. make clean takes care of deleting everything that should not be there and allows the compiler to start from scratch.
Sometimes it is also helpful to delete the build folder and create a new one, following the steps from the Compilation section above.
Instead of using make, run make VERBOSE=1. You can also run make -n to invoke a dry run where you see what the Makefile would do in case of compilation.
Instead of using make test, run ctest --verbose.