Awesome
Introduction
FleCSI is a compile-time configurable framework designed to support multi-physics application development. As such, FleCSI provides a very general set of infrastructure design patterns that can be specialized and extended to suit the needs of a broad variety of solver and data requirements. FleCSI currently supports multi-dimensional mesh topology, geometry, and adjacency information, as well as n-dimensional hashed-tree data structures, graph partitioning interfaces, and dependency closures.
FleCSI introduces a functional programming model with control, execution, and data abstractions that are consistent both with MPI and with state-of-the-art, task-based runtimes such as Legion and Charm++. The abstraction layer insulates developers from the underlying runtime, while allowing support for multiple runtime systems including conventional models like asynchronous MPI.
The intent is to provide developers with a concrete set of user-friendly programming tools that can be used now, while allowing flexibility in choosing runtime implementations and optimizations that can be applied to future architectures and runtimes.
FleCSI's control and execution models provide formal nomenclature for describing poorly understood concepts such as kernels and tasks. FleCSI's data model provides a low-buy-in approach that makes it an attractive option for many application projects, as developers are not locked into particular layouts or data structure representations.
Developers
If you are doing development of FleCSI, please take some time to read the developer README.
Requirements
The primary requirement for building FleCSI is that you have a C++17-capable compiler.
Tools
You'll need the following tools in order to build FleCSI:
- Boost >= 1.59
- CMake >= 3.12
- GCC >= 7.3.0
Install tools in the customary manner for your machine, e.g. by using apt-get on a Ubuntu system, or dnf for Fedora.
Documentation
For documentation, you'll need these as well:
- Doxygen >= 1.8
- cinch-utils >= 1.0
- Pandoc >= 1.19
Darwin
If you wish to build FleCSI on LANL's Darwin cluster, see the Darwin Cluster section later in this document.
Installing the FleCSI Third-Party Libraries
Before installing FleCSI, you should install the FleCSI third-party libraries. We'll assume that you wish to install the FleCSI third-party libraries in your home directory.
Spack
Here we talk about how to install those dependencies through Spack.
Setup your environment and load your modules such as
$ module purge
$ module load gcc/8.3.0
$ module load python/3.5.1
First, you need to download Spack if you don't already have one.
$ git clone https://github.com/spack/spack.git
Then do
$ source spack/share/spack/setup-env.sh
$ spack compiler find
==> Added 2 new compilers to /home/<user>/.spack/linux/compilers.yaml
gcc@8.3.0 gcc@4.8.5
==> Compilers are defined in the following files:
/home/<user>/.spack/linux/compilers.yaml
$ spack compiler list
==> Available compilers
-- gcc centos7-x86_64 -------------------------------------------
gcc@8.3.0 gcc@4.8.5
to get Spack into your environment and see what compilers you have that Spack can find automatically.
Get FleCSI on your desired branch following the Download section under Installing FleCSI.
Next, add the folder that contains custom flecsi spackage to Spack (you should be in flecsi folder if you followed the steps in the Download section)
$ spack repo add spack-repo
==> Added repo with namespace 'lanl_ristra_flecsi'.
$ spack repo list
==> 2 package repositories.
laristra_flecsi /home/<user>/flecsi/spack-repo
builtin /home/<user>/spack/var/spack/repos/builtin
Now, assuming you have the compiler you want recognized by Spack and added the folder, you could just do the install for a legion backend using mpich like this
$ spack install -v --only dependencies flecsi%gcc@8.3.0 +hdf5 backend=legion ^mpich@3.2.1%gcc@8.3.0
to get all the dependencies and all their dependencies installed from scratch.
[NOTE: For internal developers and users, you might need to download tar files for spack.]
After Spack finishes the installation, you can load them into your environment by doing
$ spack build-env --dump flecsi-deps.sh "flecsi%gcc@8.3.0 +hdf5 backend=legion ^mpich@3.2.1%gcc@8.3.0"
$ source flecsi-deps.sh
But if you want to save time or there is some packages that spack
has trouble installing, you could let Spack know what
packages or modules you already have on the system by adding
packages.yaml
to your ~/.spack
, which could look something like this
packages:
perl:
paths:
perl@5.16.3: /usr
numactl:
paths:
numactl@system: /usr
python:
modules:
python@2.7.3: python/2.7.3
python@3.5.1: python/3.5.1
mpich:
modules:
mpich@3.2.1%gcc@7.3.0: mpich/3.2.1-gcc_7.3.0
openmpi:
modules:
openmpi@3.1.3%gcc@7.3.0: openmpi/3.1.3-gcc_7.3.0
cmake:
modules:
cmake@3.12.4: cmake/3.12.4
Then the installation from Spack will take less time.
Either way, you can proceed to Build section under Installing FleCSI.
Build from Source
Begin by downloading the FleCSI third-party libraries:
$ cd
$ git clone --recursive https://github.com/laristra/flecsi-third-party.git
Next, enter flecsi-third-party and make a build directory:
$ cd flecsi-third-party
$ mkdir build
$ cd build
Then, for example, you can do the following for a debug-mode build:
$ cmake .. \
-DCMAKE_BUILD_TYPE=Debug \
-DCMAKE_INSTALL_PREFIX=$HOME/flecsi-third-party-debug/
Alternatively, you can run ccmake in place of cmake, and use ccmake's interface to set the options.
Finally:
$ make
$ make install
builds and installs the FleCSI third-party libraries in the prefix that you specified.
Installing FleCSI
Now that the FleCSI third-party libraries are installed, you can download and build FleCSI itself. As with the third-party libraries, we'll assume that you wish to install FleCSI in your home directory.
Download
First, download FleCSI from GitHub:
$ cd
$ git clone --recursive https://github.com/laristra/flecsi.git
By default, you'll be on the master branch. Let's say you wish to work in the branchname branch instead. Enter flecsi, switch to the relevant branch, and be sure that you have the latest updates:
$ cd flecsi
$ git checkout branchname # if you wish to work in this branch
$ git pull
$ git submodule update --recursive
Build
You can now build FleCSI. For example, a Debug build using Legion can be done like this:
$ mkdir build
$ cd build
$ cmake .. \
-DCMAKE_BUILD_TYPE=Debug \
-DENABLE_UNIT_TESTS=ON \
-DFLECSI_RUNTIME_MODEL=legion
$ make
where you should add cmake's -DCMAKE_PREFIX_PATH that points to your FleCSI third-party library installation if you built from source.
Again, you can run ccmake in place of cmake.
Building the Documentation
You can build the FleCSI User and Developer Guides, as well as the Doxygen interface documentation by specifying additional arguments to the CMake configuration line:
$ cmake .. \
-DCMAKE_BUILD_TYPE=Debug \
-DENABLE_UNIT_TESTS=ON \
-DFLECSI_RUNTIME_MODEL=legion \
-DENABLE_SPHINX=ON \
-DENABLE_DOXYGEN=ON
$ make
This will build the User and Developer Guides in the doc subdirectory.
In order to build the Doxygen interface documentation, you will need to execute:
$ make doxygen
$ make sphinx
which will build the interface documentation (also under the doc subdirectory).
Test
From within the build directory, and after running make as described above, you can run
$ make test
to run FleCSI's unit tests.
Workflow
FleCSI uses the GitHub Flow workflow pattern.
When you check-out FleCSI, you'll be on the master branch. This is the FleCSI development branch, which is protected in order to ensure that it is always deployable.
New work should always be done on a separate feature branch. When you have finished making updates to your branch, you should submit a pull request. Your changes will be tested for compliance, and reviewed by the maintainers of the project. If your changes are accepted, they will be merged into master.
Darwin Cluster
On Darwin, you can simplify some of the build requirements by using the ngc/devel-gnu environment module:
$ module load ngc # ngc/devel-gnu is the default
which will load up-to-date compiler and documentation tools. It will also load a compatible Boost & MPICH libraries.
Note: Boost & MPICH modules should match the gcc version and gcc/8.1.0 is automatically loaded.
Release
This software has been approved for open source release and has been assigned LA-CC-16-022.
Copyright
Copyright (c) 2016, Triad National Security, LLC. All rights reserved.
This program was produced under U.S. Government contract 89233218CNA000001 for Los Alamos National Laboratory (LANL), which is operated by Triad National Security, LLC for the U.S. Department of Energy/National Nuclear Security Administration.
All rights in the program are reserved by Triad National Security, LLC, and the U.S. Department of Energy/National Nuclear Security Administration. The Government is granted for itself and others acting on its behalf a nonexclusive, paid-up, irrevocable worldwide license in this material to reproduce, prepare derivative works, distribute copies to the public, perform publicly and display publicly, and to permit others to do so.
Additionally, redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
- Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
- Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
- Neither the name of Los Alamos National Security, LLC, Los Alamos National Laboratory, LANL, the U.S. Government, nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY Triad National Security, LLC. AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL Triad National Security, LLC. OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
<!-- vim: set tabstop=2 shiftwidth=2 expandtab fo=cqt tw=72 : -->