1# libCEED: the CEED API Library 2 3[](https://travis-ci.org/CEED/libCEED) 4[](https://codecov.io/gh/CEED/libCEED/) 5[](https://opensource.org/licenses/BSD-2-Clause) 6[](https://codedocs.xyz/CEED/libCEED/) 7 8## Code for Efficient Extensible Discretization 9 10This repository contains an initial low-level API library for the efficient 11high-order discretization methods developed by the ECP co-design [Center for 12Efficient Exascale Discretizations (CEED)](http://ceed.exascaleproject.org). 13While our focus is on high-order finite elements, the approach is mostly 14algebraic and thus applicable to other discretizations in factored form, as 15explained in the API documentation portion of the [Doxygen documentation](https://codedocs.xyz/CEED/libCEED/md_doc_libCEEDapi.html). 16 17One of the challenges with high-order methods is that a global sparse matrix is 18no longer a good representation of a high-order linear operator, both with 19respect to the FLOPs needed for its evaluation, as well as the memory transfer 20needed for a matvec. Thus, high-order methods require a new "format" that still 21represents a linear (or more generally non-linear) operator, but not through a 22sparse matrix. 23 24The goal of libCEED is to propose such a format, as well as supporting 25implementations and data structures, that enable efficient operator evaluation 26on a variety of computational device types (CPUs, GPUs, etc.). This new operator 27description is based on algebraically [factored form](https://codedocs.xyz/CEED/libCEED/md_doc_libCEEDapi.html), 28which is easy to incorporate in a wide variety of applications, without significant 29refactoring of their own discretization infrastructure. 30 31The repository is part of the [CEED software suite][ceed-soft], a collection of 32software benchmarks, miniapps, libraries and APIs for efficient exascale 33discretizations based on high-order finite element and spectral element methods. 34See http://github.com/ceed for more information and source code availability. 35 36The CEED research is supported by the [Exascale Computing Project][ecp] 37(17-SC-20-SC), a collaborative effort of two U.S. Department of Energy 38organizations (Office of Science and the National Nuclear Security 39Administration) responsible for the planning and preparation of a [capable 40exascale ecosystem](https://exascaleproject.org/what-is-exascale), including 41software, applications, hardware, advanced system engineering and early testbed 42platforms, in support of the nation’s exascale computing imperative. 43 44For more details on the CEED API see http://ceed.exascaleproject.org/ceed-code/. 45 46## Building 47 48The CEED library, `libceed`, is a C99 library with no external dependencies. It 49can be built using 50 51 make 52 53or, with optimization flags 54 55 make OPT='-O3 -march=skylake-avx512 -ffp-contract=fast' 56 57These optimization flags are used by all languages (C, C++, Fortran) and this 58makefile variable can also be set for testing and examples (below). 59 60The library attempts to automatically detect support for the AVX 61instruction set using gcc-style compiler options for the host. 62Support may need to be manually specified via 63 64 make AVX=1 65 66or 67 68 make AVX=0 69 70if your compiler does not support gcc-style options, if you are cross 71compiling, etc. 72 73## Testing 74 75The test suite produces [TAP](https://testanything.org) output and is run by: 76 77 make test 78 79or, using the `prove` tool distributed with Perl (recommended) 80 81 make prove 82 83## Backends 84 85There are multiple supported backends, which can be selected at runtime in the examples: 86 87| CEED resource | Backend | 88| :----------------------- | :------------------------------------------------ | 89| `/cpu/self/ref/serial` | Serial reference implementation | 90| `/cpu/self/ref/blocked` | Blocked refrence implementation | 91| `/cpu/self/memcheck` | Memcheck backend, undefined value checks | 92| `/cpu/self/opt/serial` | Serial optimized C implementation | 93| `/cpu/self/opt/blocked` | Blocked optimized C implementation | 94| `/cpu/self/avx/serial` | Serial AVX implementation | 95| `/cpu/self/avx/blocked` | Blocked AVX implementation | 96| `/cpu/self/xsmm/serial` | Serial LIBXSMM implementation | 97| `/cpu/self/xsmm/blocked` | Blocked LIBXSMM implementation | 98| `/cpu/occa` | Serial OCCA kernels | 99| `/gpu/occa` | CUDA OCCA kernels | 100| `/omp/occa` | OpenMP OCCA kernels | 101| `/ocl/occa` | OpenCL OCCA kernels | 102| `/gpu/cuda/ref` | Reference pure CUDA kernels | 103| `/gpu/cuda/reg` | Pure CUDA kernels using one thread per element | 104| `/gpu/cuda/shared` | Optimized pure CUDA kernels using shared memory | 105| `/gpu/cuda/gen` | Optimized pure CUDA kernels using code generation | 106| `/gpu/magma` | CUDA MAGMA kernels | 107 108The `/cpu/self/*/serial` backends process one element at a time and are intended for meshes 109with a smaller number of high order elements. The `/cpu/self/*/blocked` backends process 110blocked batches of eight interlaced elements and are intended for meshes with higher numbers 111of elements. 112 113The `/cpu/self/ref/*` backends are written in pure C and provide basic functionality. 114 115The `/cpu/self/opt/*` backends are written in pure C and use partial e-vectors to improve performance. 116 117The `/cpu/self/avx/*` backends rely upon AVX instructions to provide vectorized CPU performance. 118 119The `/cpu/self/xsmm/*` backends rely upon the [LIBXSMM](http://github.com/hfp/libxsmm) package 120to provide vectorized CPU performance. If linking MKL and LIBXSMM is desired but 121the Makefile is not detecting `MKLROOT`, linking libCEED against MKL can be 122forced by setting the environment variable `MKL=1`. 123 124The `/cpu/self/memcheck/*` backends rely upon the [Valgrind](http://valgrind.org/) Memcheck tool 125to help verify that user QFunctions have no undefined values. To use, run your code with 126Valgrind and the Memcheck backends, e.g. `valgrind ./build/ex1 -ceed /cpu/self/ref/memcheck`. A 127'development' or 'debugging' version of Valgrind with headers is required to use this backend. 128This backend can be run in serial or blocked mode and defaults to running in the serial mode 129if `/cpu/self/memcheck` is selected at runtime. 130 131The `/*/occa` backends rely upon the [OCCA](http://github.com/libocca/occa) package to provide 132cross platform performance. 133 134The `/gpu/cuda/*` backends provide GPU performance strictly using CUDA. 135 136The `/gpu/magma` backend relies upon the [MAGMA](https://bitbucket.org/icl/magma) package. 137 138## Examples 139 140libCEED comes with several examples of its usage, ranging from standalone C 141codes in the `/examples/ceed` directory to examples based on external packages, 142such as MFEM, PETSc, and Nek5000. Nek5000 v18.0 or greater is required. 143 144To build the examples, set the `MFEM_DIR`, `PETSC_DIR` and `NEK5K_DIR` variables 145and run: 146 147```console 148# libCEED examples on CPU and GPU 149cd examples/ceed 150make 151./ex1-volume -ceed /cpu/self 152./ex1-volume -ceed /gpu/occa 153./ex2-surface -ceed /cpu/self 154./ex2-surface -ceed /gpu/occa 155cd ../.. 156 157# MFEM+libCEED examples on CPU and GPU 158cd examples/mfem 159make 160./bp1 -ceed /cpu/self -no-vis 161./bp3 -ceed /gpu/occa -no-vis 162cd ../.. 163 164# Nek5000+libCEED examples on CPU and GPU 165cd examples/nek 166make 167./nek-examples.sh -e bp1 -ceed /cpu/self -b 3 168./nek-examples.sh -e bp3 -ceed /gpu/occa -b 3 169cd ../.. 170 171# PETSc+libCEED examples on CPU and GPU 172cd examples/petsc 173make 174./bps -problem bp1 -ceed /cpu/self 175./bps -problem bp2 -ceed /gpu/occa 176./bps -problem bp3 -ceed /cpu/self 177./bps -problem bp4 -ceed /gpu/occa 178./bps -problem bp5 -ceed /cpu/self 179./bps -problem bp6 -ceed /gpu/occa 180cd ../.. 181 182cd examples/petsc 183./area -problem cube -ceed /cpu/self -petscspace_degree 3 184./area -problem cube -ceed /gpu/occa -petscspace_degree 3 185./area -problem sphere -ceed /cpu/self -petscspace_degree 3 -dm_refine 2 186./area -problem sphere -ceed /gpu/occa -petscspace_degree 3 -dm_refine 2 187cd ../.. 188 189cd examples/navier-stokes 190make 191./navierstokes -ceed /cpu/self 192./navierstokes -ceed /gpu/occa 193cd ../.. 194``` 195 196The above code assumes a GPU-capable machine with the OCCA backend 197enabled. Depending on the available backends, other Ceed resource specifiers can 198be provided with the `-ceed` option. 199 200## Benchmarks 201 202A sequence of benchmarks for all enabled backends can be run using 203 204```console 205make benchmarks 206``` 207 208The results from the benchmarks are stored inside the `benchmarks/` directory 209and they can be viewed using the commands (requires python with matplotlib): 210 211```console 212cd benchmarks 213python postprocess-plot.py petsc-bps-bp1-*-output.txt 214python postprocess-plot.py petsc-bps-bp3-*-output.txt 215``` 216 217Using the `benchmarks` target runs a comprehensive set of benchmarks which may 218take some time to run. Subsets of the benchmarks can be run using the scripts in the `benchmarks` folder. 219 220For more details about the benchmarks, see 221[`benchmarks/README.md`](benchmarks/README.md) 222 223 224## Install 225 226To install libCEED, run 227 228 make install prefix=/usr/local 229 230or (e.g., if creating packages), 231 232 make install prefix=/usr DESTDIR=/packaging/path 233 234Note that along with the library, libCEED installs kernel sources, e.g. OCCA 235kernels are installed in `$prefix/lib/okl`. This allows the OCCA backend to 236build specialized kernels at run-time. In a normal setting, the kernel sources 237will be found automatically (relative to the library file `libceed.so`). 238However, if that fails (e.g. if `libceed.so` is moved), one can copy (cache) the 239kernel sources inside the user OCCA directory, `~/.occa` using 240 241 $(OCCA_DIR)/bin/occa cache ceed $(CEED_DIR)/lib/okl/*.okl 242 243This will allow OCCA to find the sources regardless of the location of the CEED 244library. One may occasionally need to clear the OCCA cache, which can be accomplished 245by removing the `~/.occa` directory or by calling `$(OCCA_DIR)/bin/occa clear -a`. 246 247To install libCEED for Python, run 248 249 python setup.py build install 250 251with the desired setuptools options, such as `--user`. 252 253Alternatively, if libCEED is installed in the directory specified by the 254environment variable `CEED_DIR`, then run 255 256 pip install . 257 258### pkg-config 259 260In addition to library and header, libCEED provides a [pkg-config][pkg-config1] 261file that can be used to easily compile and link. [For example][pkg-config2], if 262`$prefix` is a standard location or you set the environment variable 263`PKG_CONFIG_PATH`, 264 265 cc `pkg-config --cflags --libs ceed` -o myapp myapp.c 266 267will build `myapp` with libCEED. This can be used with the source or 268installed directories. Most build systems have support for pkg-config. 269 270## Contact 271 272You can reach the libCEED team by emailing [ceed-users@llnl.gov](mailto:ceed-users@llnl.gov) 273or by leaving a comment in the [issue tracker](https://github.com/CEED/libCEED/issues). 274 275## Copyright 276 277The following copyright applies to each file in the CEED software suite, unless 278otherwise stated in the file: 279 280> Copyright (c) 2017, Lawrence Livermore National Security, LLC. Produced at the 281> Lawrence Livermore National Laboratory. LLNL-CODE-734707. All Rights reserved. 282 283See files LICENSE and NOTICE for details. 284 285[ceed-soft]: http://ceed.exascaleproject.org/software/ 286[ecp]: https://exascaleproject.org/exascale-computing-project 287[pkg-config1]: https://en.wikipedia.org/wiki/Pkg-config 288[pkg-config2]: https://people.freedesktop.org/~dbn/pkg-config-guide.html#faq 289
){kind=link}
){kind=link}
){kind=link}
){kind=link}