diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index 7534033e923c90154cbc872c71e9ffd0057398a1..b84d22038b3159fd5aedb951fb48c35b99babe94 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -1,4 +1,4 @@
-image: irmb/virtualfluids-python-deps
+image: irmb/virtualfluids-python-deps:latest
stages:
- build
@@ -11,7 +11,6 @@ stages:
###############################################################################
## Builds ##
###############################################################################
-
.gnu_build_template:
stage: build
@@ -42,8 +41,8 @@ stages:
- mkdir -p $CI_PROJECT_DIR/$BUILD_FOLDER
- cd $CI_PROJECT_DIR/$BUILD_FOLDER
- cmake ..
- -DBUILD_VF_CPU=ON
- -DBUILD_VF_GPU=ON
+ --preset=all_make
+ -DCMAKE_CUDA_ARCHITECTURES=60
- cmake . -LAH
- make -j4
@@ -55,6 +54,7 @@ gcc_9:
extends: .gnu_build_template
###############################################################################
+
clang_10:
extends: .gnu_build_template
@@ -83,11 +83,8 @@ gcc_9_rebuild:
- cd $CI_PROJECT_DIR/build
- rm -r -f ./*
- cmake ..
- -DBUILD_VF_CPU=ON
- -DBUILD_VF_GPU=ON
- -DCMAKE_CXX_COMPILER_LAUNCHER=ccache
- -DCMAKE_CUDA_COMPILER_LAUNCHER=ccache
- -DCMAKE_C_COMPILER_LAUNCHER=ccache
+ --preset=all_make_ccache
+ -DCMAKE_CUDA_ARCHITECTURES=60
- make -j4 2>&1 | tee gcc_warnings.txt
- ccache -s
@@ -100,6 +97,37 @@ gcc_9_rebuild:
paths:
- $CI_PROJECT_DIR/cache
+
+###############################################################################
+gcc_9_cpu_warning_like_errors:
+ stage: build
+
+ image: irmb/virtualfluids-deps-ubuntu20.04
+
+ tags:
+ - gpu
+ - linux
+
+ before_script:
+ - export CCACHE_BASEDIR=$CI_PROJECT_DIR
+ - export CCACHE_DIR=$CI_PROJECT_DIR/cache
+ - ccache -s
+
+ script:
+ - mkdir -p $CI_PROJECT_DIR/build
+ - cd $CI_PROJECT_DIR/build
+ - rm -r -f ./*
+ - cmake ..
+ --preset=cpu_make_ccache
+ -DBUILD_WARNINGS_AS_ERRORS=ON
+ - make -j4
+ - ccache -s
+
+ cache:
+ key: "$CI_JOB_NAME-$CI_COMMIT_REF_SLUG"
+ paths:
+ - $CI_PROJECT_DIR/cache
+
###############################################################################
msvc_16:
stage: build
@@ -109,12 +137,8 @@ msvc_16:
- gpu
variables:
- CMAKE_GENERATOR: "Visual Studio 16 2019"
BUILD_CONFIGURATION: "Release"
BUILD_FOLDER: "build"
- BUILD_VF_CPU: "ON"
- BUILD_VF_GPU: "ON"
- BUILD_VF_UNIT_TESTS: "ON"
# add cmake and MSBuild.exe to the path.
# This Needs to be adapted when moved to a new build machine.
@@ -129,7 +153,7 @@ msvc_16:
- cd $CI_PROJECT_DIR
- md -force $env:BUILD_FOLDER
- cd $env:BUILD_FOLDER
- - cmake .. -DBUILD_VF_CPU=$env:BUILD_VF_CPU -DBUILD_VF_GPU=$env:BUILD_VF_GPU -DBUILD_VF_UNIT_TESTS=$env:BUILD_VF_UNIT_TESTS -G "$env:CMAKE_GENERATOR" ..
+ - cmake .. --preset=all_msvc
- MSBuild.exe VirtualFluids.sln /property:Configuration=$env:BUILD_CONFIGURATION /verbosity:minimal /maxcpucount:4
cache:
@@ -144,6 +168,11 @@ msvc_16:
###############################################################################
build_singularity_image:
stage: build
+
+ tags:
+ - priviliged
+ - linux
+
rules:
- if: $CI_COMMIT_TAG
@@ -152,13 +181,24 @@ build_singularity_image:
- Containers/VirtualFluidsOpenMPI.sif
script:
- - singularity build Containers/VirtualFluidsOpenMPI.sif Containers/VirtualFluidsOpenMPI.def
+ - singularity build --fakeroot Containers/VirtualFluidsOpenMPI.sif Containers/VirtualFluidsOpenMPI.def
- ls -sh Containers/VirtualFluidsOpenMPI.sif
###############################################################################
## Tests ##
###############################################################################
+gcc_9_unit_tests:
+ stage: test
+
+ needs: ["gcc_9"]
+ before_script:
+ - cd $CI_PROJECT_DIR/build
+
+ script:
+ - ./bin/basicsTests
+
+###############################################################################
msvc_16_unit_tests:
stage: test
@@ -203,7 +243,7 @@ clang_build_analyzer_clang_10:
- cmake ..
-DBUILD_VF_CPU=ON
-DBUILD_VF_GPU=ON
- -DUSE_OPENMP=OFF
+ -DCMAKE_CUDA_ARCHITECTURES=60
-DCMAKE_CXX_FLAGS=-ftime-trace
- ClangBuildAnalyzer --start .
- make
@@ -238,7 +278,7 @@ include_what_you_use_clang_10:
- cmake ..
-DBUILD_VF_CPU=ON
-DBUILD_VF_GPU=ON
- -DUSE_OPENMP=OFF
+ -DCMAKE_CUDA_ARCHITECTURES=60
-DBUILD_VF_INCLUDE_WHAT_YOU_USE=ON
- make
@@ -246,6 +286,9 @@ include_what_you_use_clang_10:
cppcheck:
stage: analyze
+ only:
+ - open_source@irmb/VirtualFluids_dev
+
image: irmb/virtualfluids-deps-ubuntu20.04
needs: []
@@ -272,6 +315,9 @@ cppcheck:
lizard:
stage: analyze
+ only:
+ - open_source@irmb/VirtualFluids_dev
+
image: irmb/virtualfluids-python-deps-ubuntu20.04
needs: []
@@ -285,7 +331,7 @@ lizard:
script:
- cd $CI_PROJECT_DIR
- - lizard -l cpp src/ > lizard.txt --ignore_warnings 191
+ - lizard -l cpp src/ > lizard.txt --warnings_only --ignore_warnings 400
artifacts:
paths:
@@ -296,6 +342,9 @@ lizard:
gcov_gcc_9:
stage: analyze
+ only:
+ - open_source@irmb/VirtualFluids_dev
+
image: irmb/virtualfluids-python-deps-ubuntu20.04
needs: []
@@ -310,10 +359,9 @@ gcov_gcc_9:
- mkdir -p $CI_PROJECT_DIR/build
- cd $CI_PROJECT_DIR/build
- cmake ..
- -DBUILD_VF_CPU=ON
- -DBUILD_VF_GPU=ON
+ --preset=all_make
+ -DCMAKE_CUDA_ARCHITECTURES=60
-DBUILD_VF_COVERAGE=ON
- -DBUILD_VF_UNIT_TESTS=ON
- make -j4
- ./bin/basicsTests
- cd ..
@@ -338,6 +386,9 @@ gcov_gcc_9:
clang-tidy:
stage: analyze
+ only:
+ - open_source@irmb/VirtualFluids_dev
+
image: irmb/virtualfluids-python-deps-ubuntu20.04
needs: []
@@ -369,6 +420,8 @@ clang-tidy:
# the reports in this file needs to match the artifacts.
sonar-scanner:
stage: deploy
+ tags:
+ - linux
only:
- open_source@irmb/VirtualFluids_dev
@@ -412,4 +465,4 @@ create_release:
--ref "$CI_COMMIT_SHA" \
--job-token "$CI_JOB_TOKEN" \
--assets-link="{'name':'VirtualFluidsSingularityImage_OpenMPI','url':'','type':'other','filepath':'Containers/VirtualFluidsOpenMPI.sif'}"
- - build/bin/basicsTests
\ No newline at end of file
+ - build/bin/basicsTests
diff --git a/3rdParty/MuParser/CMakeLists.txt b/3rdParty/MuParser/CMakeLists.txt
index 49aa3ac25552f6050de4eb5ae08928fddb8a6cf4..1634a2f15a71aeec53bd0ffb5f14c22aec7893aa 100644
--- a/3rdParty/MuParser/CMakeLists.txt
+++ b/3rdParty/MuParser/CMakeLists.txt
@@ -14,15 +14,9 @@ set(MUPARSER_VERSION ${MUPARSER_VERSION_MAJOR}.${MUPARSER_VERSION_MINOR}.${MUPAR
# Build options
option(ENABLE_SAMPLES "Build the samples" OFF)
-option(ENABLE_OPENMP "Enable OpenMP for multithreading" ON)
+#option(ENABLE_OPENMP "Enable OpenMP for multithreading" ON)
#option(BUILD_SHARED_LIBS "Build shared/static libs" ON)
-if(ENABLE_OPENMP)
- find_package(OpenMP REQUIRED)
- set(CMAKE_CXX_FLAGS "${OpenMP_CXX_FLAGS} ${CMAKE_CXX_FLAGS}")
- set(CMAKE_SHARED_LIBRARY_CXX_FLAGS "${OpenMP_CXX_FLAGS} ${CMAKE_SHARED_LIBRARY_CXX_FLAGS}")
-endif()
-
# Credit: https://stackoverflow.com/questions/2368811/how-to-set-warning-level-in-cmake/3818084
if(MSVC)
@@ -50,7 +44,9 @@ add_library(muparser
src/muParserTokenReader.cpp
)
-target_link_libraries(muparser PUBLIC OpenMP::OpenMP_CXX)
+if(BUILD_USE_OPENMP)
+ target_link_libraries(muparser PUBLIC OpenMP::OpenMP_CXX)
+endif()
target_include_directories(muparser PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}/include)
# this compiles the "DLL" interface (C API)
@@ -66,7 +62,7 @@ if (CMAKE_BUILD_TYPE STREQUAL Debug)
target_compile_definitions(muparser PRIVATE _DEBUG)
endif ()
-if(ENABLE_OPENMP)
+if(BUILD_USE_OPENMP)
target_compile_definitions(muparser PRIVATE MUP_USE_OPENMP)
endif()
set_target_properties(muparser PROPERTIES
diff --git a/3rdParty/MuParser/include/muParser.h b/3rdParty/MuParser/include/muParser.h
index ab1e21e5b2a7506593de90d4845b6c2e5e39904a..3a2bedc39180c678a63cfa81b893a339b85324f7 100644
--- a/3rdParty/MuParser/include/muParser.h
+++ b/3rdParty/MuParser/include/muParser.h
@@ -29,7 +29,9 @@
#ifndef MU_PARSER_H
#define MU_PARSER_H
+#ifdef __clang__
#pragma clang system_header
+#endif
//--- Standard includes ------------------------------------------------------------------------
#include <vector>
diff --git a/3rdParty/MuParser/include/muParserBase.h b/3rdParty/MuParser/include/muParserBase.h
index e42aca8b121c95b3d61e4ef2414f2f0bea76cab1..5ccadd99d1d33b6e69a123480a31b82079234945 100644
--- a/3rdParty/MuParser/include/muParserBase.h
+++ b/3rdParty/MuParser/include/muParserBase.h
@@ -29,7 +29,9 @@
#ifndef MU_PARSER_BASE_H
#define MU_PARSER_BASE_H
+#ifdef __clang__
#pragma clang system_header
+#endif
//--- Standard includes ------------------------------------------------------------------------
#include <cmath>
diff --git a/3rdParty/MuParser/include/muParserDLL.h b/3rdParty/MuParser/include/muParserDLL.h
index 18051c4d7835b266a463bd36f974d66d5729b13d..14c65b48a463c6bfb52fb92cca2daf62988df937 100644
--- a/3rdParty/MuParser/include/muParserDLL.h
+++ b/3rdParty/MuParser/include/muParserDLL.h
@@ -29,6 +29,10 @@
#ifndef MU_PARSER_DLL_H
#define MU_PARSER_DLL_H
+#ifdef __clang__
+#pragma clang system_header
+#endif
+
#include "muParserFixes.h"
#ifdef __cplusplus
diff --git a/3rdParty/cuda_samples/README b/3rdParty/cuda_samples/README
new file mode 100644
index 0000000000000000000000000000000000000000..5db13e7bda8365d792e3e68840d02c442348596e
--- /dev/null
+++ b/3rdParty/cuda_samples/README
@@ -0,0 +1,2 @@
+# 3rd party cuda
+The files in this folder are added from here https://github.com/NVIDIA/cuda-samples/blob/v11.2/Common/.
\ No newline at end of file
diff --git a/3rdParty/cuda_samples/exception.h b/3rdParty/cuda_samples/exception.h
new file mode 100644
index 0000000000000000000000000000000000000000..84e348b59fb6892439e8057b03093bac48cadcd4
--- /dev/null
+++ b/3rdParty/cuda_samples/exception.h
@@ -0,0 +1,151 @@
+/* Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
+ *
+ * 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 NVIDIA CORPORATION 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 THE COPYRIGHT HOLDERS ``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 THE COPYRIGHT OWNER 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.
+ */
+
+/* CUda UTility Library */
+#ifndef COMMON_EXCEPTION_H_
+#define COMMON_EXCEPTION_H_
+
+// includes, system
+#include <stdlib.h>
+#include <exception>
+#include <iostream>
+#include <stdexcept>
+#include <string>
+
+//! Exception wrapper.
+//! @param Std_Exception Exception out of namespace std for easy typing.
+template <class Std_Exception>
+class Exception : public Std_Exception {
+public:
+ //! @brief Static construction interface
+ //! @return Alwayss throws ( Located_Exception<Exception>)
+ //! @param file file in which the Exception occurs
+ //! @param line line in which the Exception occurs
+ //! @param detailed details on the code fragment causing the Exception
+ static void throw_it(const char *file, const int line,
+ const char *detailed = "-");
+
+ //! Static construction interface
+ //! @return Alwayss throws ( Located_Exception<Exception>)
+ //! @param file file in which the Exception occurs
+ //! @param line line in which the Exception occurs
+ //! @param detailed details on the code fragment causing the Exception
+ static void throw_it(const char *file, const int line,
+ const std::string &detailed);
+
+ //! Destructor
+ virtual ~Exception() throw();
+
+private:
+ //! Constructor, default (private)
+ Exception();
+
+ //! Constructor, standard
+ //! @param str string returned by what()
+ explicit Exception(const std::string &str);
+};
+
+////////////////////////////////////////////////////////////////////////////////
+//! Exception handler function for arbitrary exceptions
+//! @param ex exception to handle
+////////////////////////////////////////////////////////////////////////////////
+template <class Exception_Typ>
+inline void handleException(const Exception_Typ &ex) {
+ std::cerr << ex.what() << std::endl;
+
+ exit(EXIT_FAILURE);
+}
+
+//! Convenience macros
+
+//! Exception caused by dynamic program behavior, e.g. file does not exist
+#define RUNTIME_EXCEPTION(msg) \
+ Exception<std::runtime_error>::throw_it(__FILE__, __LINE__, msg)
+
+//! Logic exception in program, e.g. an assert failed
+#define LOGIC_EXCEPTION(msg) \
+ Exception<std::logic_error>::throw_it(__FILE__, __LINE__, msg)
+
+//! Out of range exception
+#define RANGE_EXCEPTION(msg) \
+ Exception<std::range_error>::throw_it(__FILE__, __LINE__, msg)
+
+////////////////////////////////////////////////////////////////////////////////
+//! Implementation
+
+// includes, system
+#include <sstream>
+
+////////////////////////////////////////////////////////////////////////////////
+//! Static construction interface.
+//! @param Exception causing code fragment (file and line) and detailed infos.
+////////////////////////////////////////////////////////////////////////////////
+/*static*/ template <class Std_Exception>
+void Exception<Std_Exception>::throw_it(const char *file, const int line,
+ const char *detailed) {
+ std::stringstream s;
+
+ // Quiet heavy-weight but exceptions are not for
+ // performance / release versions
+ s << "Exception in file '" << file << "' in line " << line << "\n"
+ << "Detailed description: " << detailed << "\n";
+
+ throw Exception(s.str());
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Static construction interface.
+//! @param Exception causing code fragment (file and line) and detailed infos.
+////////////////////////////////////////////////////////////////////////////////
+/*static*/ template <class Std_Exception>
+void Exception<Std_Exception>::throw_it(const char *file, const int line,
+ const std::string &msg) {
+ throw_it(file, line, msg.c_str());
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Constructor, default (private).
+////////////////////////////////////////////////////////////////////////////////
+template <class Std_Exception>
+Exception<Std_Exception>::Exception() : Std_Exception("Unknown Exception.\n") {}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Constructor, standard (private).
+//! String returned by what().
+////////////////////////////////////////////////////////////////////////////////
+template <class Std_Exception>
+Exception<Std_Exception>::Exception(const std::string &s) : Std_Exception(s) {}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Destructor
+////////////////////////////////////////////////////////////////////////////////
+template <class Std_Exception>
+Exception<Std_Exception>::~Exception() throw() {}
+
+// functions, exported
+
+#endif // COMMON_EXCEPTION_H_
\ No newline at end of file
diff --git a/3rdParty/cuda_samples/helper_cuda.h b/3rdParty/cuda_samples/helper_cuda.h
new file mode 100644
index 0000000000000000000000000000000000000000..81d3f9e76983a32daff4c2649ab0880e29f9881a
--- /dev/null
+++ b/3rdParty/cuda_samples/helper_cuda.h
@@ -0,0 +1,967 @@
+/* Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
+ *
+ * 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 NVIDIA CORPORATION 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 THE COPYRIGHT HOLDERS ``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 THE COPYRIGHT OWNER 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.
+ */
+
+////////////////////////////////////////////////////////////////////////////////
+// These are CUDA Helper functions for initialization and error checking
+
+#ifndef COMMON_HELPER_CUDA_H_
+#define COMMON_HELPER_CUDA_H_
+
+#pragma once
+
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include <helper_string.h>
+
+#ifndef EXIT_WAIVED
+#define EXIT_WAIVED 2
+#endif
+
+// Note, it is required that your SDK sample to include the proper header
+// files, please refer the CUDA examples for examples of the needed CUDA
+// headers, which may change depending on which CUDA functions are used.
+
+// CUDA Runtime error messages
+#ifdef __DRIVER_TYPES_H__
+static const char *_cudaGetErrorEnum(cudaError_t error) {
+ return cudaGetErrorName(error);
+}
+#endif
+
+#ifdef CUDA_DRIVER_API
+// CUDA Driver API errors
+static const char *_cudaGetErrorEnum(CUresult error) {
+ static char unknown[] = "<unknown>";
+ const char *ret = NULL;
+ cuGetErrorName(error, &ret);
+ return ret ? ret : unknown;
+}
+#endif
+
+#ifdef CUBLAS_API_H_
+// cuBLAS API errors
+static const char *_cudaGetErrorEnum(cublasStatus_t error) {
+ switch (error) {
+ case CUBLAS_STATUS_SUCCESS:
+ return "CUBLAS_STATUS_SUCCESS";
+
+ case CUBLAS_STATUS_NOT_INITIALIZED:
+ return "CUBLAS_STATUS_NOT_INITIALIZED";
+
+ case CUBLAS_STATUS_ALLOC_FAILED:
+ return "CUBLAS_STATUS_ALLOC_FAILED";
+
+ case CUBLAS_STATUS_INVALID_VALUE:
+ return "CUBLAS_STATUS_INVALID_VALUE";
+
+ case CUBLAS_STATUS_ARCH_MISMATCH:
+ return "CUBLAS_STATUS_ARCH_MISMATCH";
+
+ case CUBLAS_STATUS_MAPPING_ERROR:
+ return "CUBLAS_STATUS_MAPPING_ERROR";
+
+ case CUBLAS_STATUS_EXECUTION_FAILED:
+ return "CUBLAS_STATUS_EXECUTION_FAILED";
+
+ case CUBLAS_STATUS_INTERNAL_ERROR:
+ return "CUBLAS_STATUS_INTERNAL_ERROR";
+
+ case CUBLAS_STATUS_NOT_SUPPORTED:
+ return "CUBLAS_STATUS_NOT_SUPPORTED";
+
+ case CUBLAS_STATUS_LICENSE_ERROR:
+ return "CUBLAS_STATUS_LICENSE_ERROR";
+ }
+
+ return "<unknown>";
+}
+#endif
+
+#ifdef _CUFFT_H_
+// cuFFT API errors
+static const char *_cudaGetErrorEnum(cufftResult error) {
+ switch (error) {
+ case CUFFT_SUCCESS:
+ return "CUFFT_SUCCESS";
+
+ case CUFFT_INVALID_PLAN:
+ return "CUFFT_INVALID_PLAN";
+
+ case CUFFT_ALLOC_FAILED:
+ return "CUFFT_ALLOC_FAILED";
+
+ case CUFFT_INVALID_TYPE:
+ return "CUFFT_INVALID_TYPE";
+
+ case CUFFT_INVALID_VALUE:
+ return "CUFFT_INVALID_VALUE";
+
+ case CUFFT_INTERNAL_ERROR:
+ return "CUFFT_INTERNAL_ERROR";
+
+ case CUFFT_EXEC_FAILED:
+ return "CUFFT_EXEC_FAILED";
+
+ case CUFFT_SETUP_FAILED:
+ return "CUFFT_SETUP_FAILED";
+
+ case CUFFT_INVALID_SIZE:
+ return "CUFFT_INVALID_SIZE";
+
+ case CUFFT_UNALIGNED_DATA:
+ return "CUFFT_UNALIGNED_DATA";
+
+ case CUFFT_INCOMPLETE_PARAMETER_LIST:
+ return "CUFFT_INCOMPLETE_PARAMETER_LIST";
+
+ case CUFFT_INVALID_DEVICE:
+ return "CUFFT_INVALID_DEVICE";
+
+ case CUFFT_PARSE_ERROR:
+ return "CUFFT_PARSE_ERROR";
+
+ case CUFFT_NO_WORKSPACE:
+ return "CUFFT_NO_WORKSPACE";
+
+ case CUFFT_NOT_IMPLEMENTED:
+ return "CUFFT_NOT_IMPLEMENTED";
+
+ case CUFFT_LICENSE_ERROR:
+ return "CUFFT_LICENSE_ERROR";
+
+ case CUFFT_NOT_SUPPORTED:
+ return "CUFFT_NOT_SUPPORTED";
+ }
+
+ return "<unknown>";
+}
+#endif
+
+#ifdef CUSPARSEAPI
+// cuSPARSE API errors
+static const char *_cudaGetErrorEnum(cusparseStatus_t error) {
+ switch (error) {
+ case CUSPARSE_STATUS_SUCCESS:
+ return "CUSPARSE_STATUS_SUCCESS";
+
+ case CUSPARSE_STATUS_NOT_INITIALIZED:
+ return "CUSPARSE_STATUS_NOT_INITIALIZED";
+
+ case CUSPARSE_STATUS_ALLOC_FAILED:
+ return "CUSPARSE_STATUS_ALLOC_FAILED";
+
+ case CUSPARSE_STATUS_INVALID_VALUE:
+ return "CUSPARSE_STATUS_INVALID_VALUE";
+
+ case CUSPARSE_STATUS_ARCH_MISMATCH:
+ return "CUSPARSE_STATUS_ARCH_MISMATCH";
+
+ case CUSPARSE_STATUS_MAPPING_ERROR:
+ return "CUSPARSE_STATUS_MAPPING_ERROR";
+
+ case CUSPARSE_STATUS_EXECUTION_FAILED:
+ return "CUSPARSE_STATUS_EXECUTION_FAILED";
+
+ case CUSPARSE_STATUS_INTERNAL_ERROR:
+ return "CUSPARSE_STATUS_INTERNAL_ERROR";
+
+ case CUSPARSE_STATUS_MATRIX_TYPE_NOT_SUPPORTED:
+ return "CUSPARSE_STATUS_MATRIX_TYPE_NOT_SUPPORTED";
+ }
+
+ return "<unknown>";
+}
+#endif
+
+#ifdef CUSOLVER_COMMON_H_
+// cuSOLVER API errors
+static const char *_cudaGetErrorEnum(cusolverStatus_t error) {
+ switch (error) {
+ case CUSOLVER_STATUS_SUCCESS:
+ return "CUSOLVER_STATUS_SUCCESS";
+ case CUSOLVER_STATUS_NOT_INITIALIZED:
+ return "CUSOLVER_STATUS_NOT_INITIALIZED";
+ case CUSOLVER_STATUS_ALLOC_FAILED:
+ return "CUSOLVER_STATUS_ALLOC_FAILED";
+ case CUSOLVER_STATUS_INVALID_VALUE:
+ return "CUSOLVER_STATUS_INVALID_VALUE";
+ case CUSOLVER_STATUS_ARCH_MISMATCH:
+ return "CUSOLVER_STATUS_ARCH_MISMATCH";
+ case CUSOLVER_STATUS_MAPPING_ERROR:
+ return "CUSOLVER_STATUS_MAPPING_ERROR";
+ case CUSOLVER_STATUS_EXECUTION_FAILED:
+ return "CUSOLVER_STATUS_EXECUTION_FAILED";
+ case CUSOLVER_STATUS_INTERNAL_ERROR:
+ return "CUSOLVER_STATUS_INTERNAL_ERROR";
+ case CUSOLVER_STATUS_MATRIX_TYPE_NOT_SUPPORTED:
+ return "CUSOLVER_STATUS_MATRIX_TYPE_NOT_SUPPORTED";
+ case CUSOLVER_STATUS_NOT_SUPPORTED:
+ return "CUSOLVER_STATUS_NOT_SUPPORTED ";
+ case CUSOLVER_STATUS_ZERO_PIVOT:
+ return "CUSOLVER_STATUS_ZERO_PIVOT";
+ case CUSOLVER_STATUS_INVALID_LICENSE:
+ return "CUSOLVER_STATUS_INVALID_LICENSE";
+ }
+
+ return "<unknown>";
+}
+#endif
+
+#ifdef CURAND_H_
+// cuRAND API errors
+static const char *_cudaGetErrorEnum(curandStatus_t error) {
+ switch (error) {
+ case CURAND_STATUS_SUCCESS:
+ return "CURAND_STATUS_SUCCESS";
+
+ case CURAND_STATUS_VERSION_MISMATCH:
+ return "CURAND_STATUS_VERSION_MISMATCH";
+
+ case CURAND_STATUS_NOT_INITIALIZED:
+ return "CURAND_STATUS_NOT_INITIALIZED";
+
+ case CURAND_STATUS_ALLOCATION_FAILED:
+ return "CURAND_STATUS_ALLOCATION_FAILED";
+
+ case CURAND_STATUS_TYPE_ERROR:
+ return "CURAND_STATUS_TYPE_ERROR";
+
+ case CURAND_STATUS_OUT_OF_RANGE:
+ return "CURAND_STATUS_OUT_OF_RANGE";
+
+ case CURAND_STATUS_LENGTH_NOT_MULTIPLE:
+ return "CURAND_STATUS_LENGTH_NOT_MULTIPLE";
+
+ case CURAND_STATUS_DOUBLE_PRECISION_REQUIRED:
+ return "CURAND_STATUS_DOUBLE_PRECISION_REQUIRED";
+
+ case CURAND_STATUS_LAUNCH_FAILURE:
+ return "CURAND_STATUS_LAUNCH_FAILURE";
+
+ case CURAND_STATUS_PREEXISTING_FAILURE:
+ return "CURAND_STATUS_PREEXISTING_FAILURE";
+
+ case CURAND_STATUS_INITIALIZATION_FAILED:
+ return "CURAND_STATUS_INITIALIZATION_FAILED";
+
+ case CURAND_STATUS_ARCH_MISMATCH:
+ return "CURAND_STATUS_ARCH_MISMATCH";
+
+ case CURAND_STATUS_INTERNAL_ERROR:
+ return "CURAND_STATUS_INTERNAL_ERROR";
+ }
+
+ return "<unknown>";
+}
+#endif
+
+#ifdef NVJPEGAPI
+// nvJPEG API errors
+static const char *_cudaGetErrorEnum(nvjpegStatus_t error) {
+ switch (error) {
+ case NVJPEG_STATUS_SUCCESS:
+ return "NVJPEG_STATUS_SUCCESS";
+
+ case NVJPEG_STATUS_NOT_INITIALIZED:
+ return "NVJPEG_STATUS_NOT_INITIALIZED";
+
+ case NVJPEG_STATUS_INVALID_PARAMETER:
+ return "NVJPEG_STATUS_INVALID_PARAMETER";
+
+ case NVJPEG_STATUS_BAD_JPEG:
+ return "NVJPEG_STATUS_BAD_JPEG";
+
+ case NVJPEG_STATUS_JPEG_NOT_SUPPORTED:
+ return "NVJPEG_STATUS_JPEG_NOT_SUPPORTED";
+
+ case NVJPEG_STATUS_ALLOCATOR_FAILURE:
+ return "NVJPEG_STATUS_ALLOCATOR_FAILURE";
+
+ case NVJPEG_STATUS_EXECUTION_FAILED:
+ return "NVJPEG_STATUS_EXECUTION_FAILED";
+
+ case NVJPEG_STATUS_ARCH_MISMATCH:
+ return "NVJPEG_STATUS_ARCH_MISMATCH";
+
+ case NVJPEG_STATUS_INTERNAL_ERROR:
+ return "NVJPEG_STATUS_INTERNAL_ERROR";
+ }
+
+ return "<unknown>";
+}
+#endif
+
+#ifdef NV_NPPIDEFS_H
+// NPP API errors
+static const char *_cudaGetErrorEnum(NppStatus error) {
+ switch (error) {
+ case NPP_NOT_SUPPORTED_MODE_ERROR:
+ return "NPP_NOT_SUPPORTED_MODE_ERROR";
+
+ case NPP_ROUND_MODE_NOT_SUPPORTED_ERROR:
+ return "NPP_ROUND_MODE_NOT_SUPPORTED_ERROR";
+
+ case NPP_RESIZE_NO_OPERATION_ERROR:
+ return "NPP_RESIZE_NO_OPERATION_ERROR";
+
+ case NPP_NOT_SUFFICIENT_COMPUTE_CAPABILITY:
+ return "NPP_NOT_SUFFICIENT_COMPUTE_CAPABILITY";
+
+#if ((NPP_VERSION_MAJOR << 12) + (NPP_VERSION_MINOR << 4)) <= 0x5000
+
+ case NPP_BAD_ARG_ERROR:
+ return "NPP_BAD_ARGUMENT_ERROR";
+
+ case NPP_COEFF_ERROR:
+ return "NPP_COEFFICIENT_ERROR";
+
+ case NPP_RECT_ERROR:
+ return "NPP_RECTANGLE_ERROR";
+
+ case NPP_QUAD_ERROR:
+ return "NPP_QUADRANGLE_ERROR";
+
+ case NPP_MEM_ALLOC_ERR:
+ return "NPP_MEMORY_ALLOCATION_ERROR";
+
+ case NPP_HISTO_NUMBER_OF_LEVELS_ERROR:
+ return "NPP_HISTOGRAM_NUMBER_OF_LEVELS_ERROR";
+
+ case NPP_INVALID_INPUT:
+ return "NPP_INVALID_INPUT";
+
+ case NPP_POINTER_ERROR:
+ return "NPP_POINTER_ERROR";
+
+ case NPP_WARNING:
+ return "NPP_WARNING";
+
+ case NPP_ODD_ROI_WARNING:
+ return "NPP_ODD_ROI_WARNING";
+#else
+
+ // These are for CUDA 5.5 or higher
+ case NPP_BAD_ARGUMENT_ERROR:
+ return "NPP_BAD_ARGUMENT_ERROR";
+
+ case NPP_COEFFICIENT_ERROR:
+ return "NPP_COEFFICIENT_ERROR";
+
+ case NPP_RECTANGLE_ERROR:
+ return "NPP_RECTANGLE_ERROR";
+
+ case NPP_QUADRANGLE_ERROR:
+ return "NPP_QUADRANGLE_ERROR";
+
+ case NPP_MEMORY_ALLOCATION_ERR:
+ return "NPP_MEMORY_ALLOCATION_ERROR";
+
+ case NPP_HISTOGRAM_NUMBER_OF_LEVELS_ERROR:
+ return "NPP_HISTOGRAM_NUMBER_OF_LEVELS_ERROR";
+
+ case NPP_INVALID_HOST_POINTER_ERROR:
+ return "NPP_INVALID_HOST_POINTER_ERROR";
+
+ case NPP_INVALID_DEVICE_POINTER_ERROR:
+ return "NPP_INVALID_DEVICE_POINTER_ERROR";
+#endif
+
+ case NPP_LUT_NUMBER_OF_LEVELS_ERROR:
+ return "NPP_LUT_NUMBER_OF_LEVELS_ERROR";
+
+ case NPP_TEXTURE_BIND_ERROR:
+ return "NPP_TEXTURE_BIND_ERROR";
+
+ case NPP_WRONG_INTERSECTION_ROI_ERROR:
+ return "NPP_WRONG_INTERSECTION_ROI_ERROR";
+
+ case NPP_NOT_EVEN_STEP_ERROR:
+ return "NPP_NOT_EVEN_STEP_ERROR";
+
+ case NPP_INTERPOLATION_ERROR:
+ return "NPP_INTERPOLATION_ERROR";
+
+ case NPP_RESIZE_FACTOR_ERROR:
+ return "NPP_RESIZE_FACTOR_ERROR";
+
+ case NPP_HAAR_CLASSIFIER_PIXEL_MATCH_ERROR:
+ return "NPP_HAAR_CLASSIFIER_PIXEL_MATCH_ERROR";
+
+#if ((NPP_VERSION_MAJOR << 12) + (NPP_VERSION_MINOR << 4)) <= 0x5000
+
+ case NPP_MEMFREE_ERR:
+ return "NPP_MEMFREE_ERR";
+
+ case NPP_MEMSET_ERR:
+ return "NPP_MEMSET_ERR";
+
+ case NPP_MEMCPY_ERR:
+ return "NPP_MEMCPY_ERROR";
+
+ case NPP_MIRROR_FLIP_ERR:
+ return "NPP_MIRROR_FLIP_ERR";
+#else
+
+ case NPP_MEMFREE_ERROR:
+ return "NPP_MEMFREE_ERROR";
+
+ case NPP_MEMSET_ERROR:
+ return "NPP_MEMSET_ERROR";
+
+ case NPP_MEMCPY_ERROR:
+ return "NPP_MEMCPY_ERROR";
+
+ case NPP_MIRROR_FLIP_ERROR:
+ return "NPP_MIRROR_FLIP_ERROR";
+#endif
+
+ case NPP_ALIGNMENT_ERROR:
+ return "NPP_ALIGNMENT_ERROR";
+
+ case NPP_STEP_ERROR:
+ return "NPP_STEP_ERROR";
+
+ case NPP_SIZE_ERROR:
+ return "NPP_SIZE_ERROR";
+
+ case NPP_NULL_POINTER_ERROR:
+ return "NPP_NULL_POINTER_ERROR";
+
+ case NPP_CUDA_KERNEL_EXECUTION_ERROR:
+ return "NPP_CUDA_KERNEL_EXECUTION_ERROR";
+
+ case NPP_NOT_IMPLEMENTED_ERROR:
+ return "NPP_NOT_IMPLEMENTED_ERROR";
+
+ case NPP_ERROR:
+ return "NPP_ERROR";
+
+ case NPP_SUCCESS:
+ return "NPP_SUCCESS";
+
+ case NPP_WRONG_INTERSECTION_QUAD_WARNING:
+ return "NPP_WRONG_INTERSECTION_QUAD_WARNING";
+
+ case NPP_MISALIGNED_DST_ROI_WARNING:
+ return "NPP_MISALIGNED_DST_ROI_WARNING";
+
+ case NPP_AFFINE_QUAD_INCORRECT_WARNING:
+ return "NPP_AFFINE_QUAD_INCORRECT_WARNING";
+
+ case NPP_DOUBLE_SIZE_WARNING:
+ return "NPP_DOUBLE_SIZE_WARNING";
+
+ case NPP_WRONG_INTERSECTION_ROI_WARNING:
+ return "NPP_WRONG_INTERSECTION_ROI_WARNING";
+
+#if ((NPP_VERSION_MAJOR << 12) + (NPP_VERSION_MINOR << 4)) >= 0x6000
+ /* These are 6.0 or higher */
+ case NPP_LUT_PALETTE_BITSIZE_ERROR:
+ return "NPP_LUT_PALETTE_BITSIZE_ERROR";
+
+ case NPP_ZC_MODE_NOT_SUPPORTED_ERROR:
+ return "NPP_ZC_MODE_NOT_SUPPORTED_ERROR";
+
+ case NPP_QUALITY_INDEX_ERROR:
+ return "NPP_QUALITY_INDEX_ERROR";
+
+ case NPP_CHANNEL_ORDER_ERROR:
+ return "NPP_CHANNEL_ORDER_ERROR";
+
+ case NPP_ZERO_MASK_VALUE_ERROR:
+ return "NPP_ZERO_MASK_VALUE_ERROR";
+
+ case NPP_NUMBER_OF_CHANNELS_ERROR:
+ return "NPP_NUMBER_OF_CHANNELS_ERROR";
+
+ case NPP_COI_ERROR:
+ return "NPP_COI_ERROR";
+
+ case NPP_DIVISOR_ERROR:
+ return "NPP_DIVISOR_ERROR";
+
+ case NPP_CHANNEL_ERROR:
+ return "NPP_CHANNEL_ERROR";
+
+ case NPP_STRIDE_ERROR:
+ return "NPP_STRIDE_ERROR";
+
+ case NPP_ANCHOR_ERROR:
+ return "NPP_ANCHOR_ERROR";
+
+ case NPP_MASK_SIZE_ERROR:
+ return "NPP_MASK_SIZE_ERROR";
+
+ case NPP_MOMENT_00_ZERO_ERROR:
+ return "NPP_MOMENT_00_ZERO_ERROR";
+
+ case NPP_THRESHOLD_NEGATIVE_LEVEL_ERROR:
+ return "NPP_THRESHOLD_NEGATIVE_LEVEL_ERROR";
+
+ case NPP_THRESHOLD_ERROR:
+ return "NPP_THRESHOLD_ERROR";
+
+ case NPP_CONTEXT_MATCH_ERROR:
+ return "NPP_CONTEXT_MATCH_ERROR";
+
+ case NPP_FFT_FLAG_ERROR:
+ return "NPP_FFT_FLAG_ERROR";
+
+ case NPP_FFT_ORDER_ERROR:
+ return "NPP_FFT_ORDER_ERROR";
+
+ case NPP_SCALE_RANGE_ERROR:
+ return "NPP_SCALE_RANGE_ERROR";
+
+ case NPP_DATA_TYPE_ERROR:
+ return "NPP_DATA_TYPE_ERROR";
+
+ case NPP_OUT_OFF_RANGE_ERROR:
+ return "NPP_OUT_OFF_RANGE_ERROR";
+
+ case NPP_DIVIDE_BY_ZERO_ERROR:
+ return "NPP_DIVIDE_BY_ZERO_ERROR";
+
+ case NPP_RANGE_ERROR:
+ return "NPP_RANGE_ERROR";
+
+ case NPP_NO_MEMORY_ERROR:
+ return "NPP_NO_MEMORY_ERROR";
+
+ case NPP_ERROR_RESERVED:
+ return "NPP_ERROR_RESERVED";
+
+ case NPP_NO_OPERATION_WARNING:
+ return "NPP_NO_OPERATION_WARNING";
+
+ case NPP_DIVIDE_BY_ZERO_WARNING:
+ return "NPP_DIVIDE_BY_ZERO_WARNING";
+#endif
+
+#if ((NPP_VERSION_MAJOR << 12) + (NPP_VERSION_MINOR << 4)) >= 0x7000
+ /* These are 7.0 or higher */
+ case NPP_OVERFLOW_ERROR:
+ return "NPP_OVERFLOW_ERROR";
+
+ case NPP_CORRUPTED_DATA_ERROR:
+ return "NPP_CORRUPTED_DATA_ERROR";
+#endif
+ }
+
+ return "<unknown>";
+}
+#endif
+
+template <typename T>
+void check(T result, char const *const func, const char *const file,
+ int const line) {
+ if (result) {
+ fprintf(stderr, "CUDA error at %s:%d code=%d(%s) \"%s\" \n", file, line,
+ static_cast<unsigned int>(result), _cudaGetErrorEnum(result), func);
+ exit(EXIT_FAILURE);
+ }
+}
+
+#ifdef __DRIVER_TYPES_H__
+// This will output the proper CUDA error strings in the event
+// that a CUDA host call returns an error
+#define checkCudaErrors(val) check((val), #val, __FILE__, __LINE__)
+
+// This will output the proper error string when calling cudaGetLastError
+#define getLastCudaError(msg) __getLastCudaError(msg, __FILE__, __LINE__)
+
+inline void __getLastCudaError(const char *errorMessage, const char *file,
+ const int line) {
+ cudaError_t err = cudaGetLastError();
+
+ if (cudaSuccess != err) {
+ fprintf(stderr,
+ "%s(%i) : getLastCudaError() CUDA error :"
+ " %s : (%d) %s.\n",
+ file, line, errorMessage, static_cast<int>(err),
+ cudaGetErrorString(err));
+ exit(EXIT_FAILURE);
+ }
+}
+
+// This will only print the proper error string when calling cudaGetLastError
+// but not exit program incase error detected.
+#define printLastCudaError(msg) __printLastCudaError(msg, __FILE__, __LINE__)
+
+inline void __printLastCudaError(const char *errorMessage, const char *file,
+ const int line) {
+ cudaError_t err = cudaGetLastError();
+
+ if (cudaSuccess != err) {
+ fprintf(stderr,
+ "%s(%i) : getLastCudaError() CUDA error :"
+ " %s : (%d) %s.\n",
+ file, line, errorMessage, static_cast<int>(err),
+ cudaGetErrorString(err));
+ }
+}
+#endif
+
+#ifndef MAX
+#define MAX(a, b) (a > b ? a : b)
+#endif
+
+// Float To Int conversion
+inline int ftoi(float value) {
+ return (value >= 0 ? static_cast<int>(value + 0.5)
+ : static_cast<int>(value - 0.5));
+}
+
+// Beginning of GPU Architecture definitions
+inline int _ConvertSMVer2Cores(int major, int minor) {
+ // Defines for GPU Architecture types (using the SM version to determine
+ // the # of cores per SM
+ typedef struct {
+ int SM; // 0xMm (hexidecimal notation), M = SM Major version,
+ // and m = SM minor version
+ int Cores;
+ } sSMtoCores;
+
+ sSMtoCores nGpuArchCoresPerSM[] = {
+ {0x30, 192},
+ {0x32, 192},
+ {0x35, 192},
+ {0x37, 192},
+ {0x50, 128},
+ {0x52, 128},
+ {0x53, 128},
+ {0x60, 64},
+ {0x61, 128},
+ {0x62, 128},
+ {0x70, 64},
+ {0x72, 64},
+ {0x75, 64},
+ {0x80, 64},
+ {0x86, 128},
+ {-1, -1}};
+
+ int index = 0;
+
+ while (nGpuArchCoresPerSM[index].SM != -1) {
+ if (nGpuArchCoresPerSM[index].SM == ((major << 4) + minor)) {
+ return nGpuArchCoresPerSM[index].Cores;
+ }
+
+ index++;
+ }
+
+ // If we don't find the values, we default use the previous one
+ // to run properly
+ printf(
+ "MapSMtoCores for SM %d.%d is undefined."
+ " Default to use %d Cores/SM\n",
+ major, minor, nGpuArchCoresPerSM[index - 1].Cores);
+ return nGpuArchCoresPerSM[index - 1].Cores;
+}
+
+inline const char* _ConvertSMVer2ArchName(int major, int minor) {
+ // Defines for GPU Architecture types (using the SM version to determine
+ // the GPU Arch name)
+ typedef struct {
+ int SM; // 0xMm (hexidecimal notation), M = SM Major version,
+ // and m = SM minor version
+ const char* name;
+ } sSMtoArchName;
+
+ sSMtoArchName nGpuArchNameSM[] = {
+ {0x30, "Kepler"},
+ {0x32, "Kepler"},
+ {0x35, "Kepler"},
+ {0x37, "Kepler"},
+ {0x50, "Maxwell"},
+ {0x52, "Maxwell"},
+ {0x53, "Maxwell"},
+ {0x60, "Pascal"},
+ {0x61, "Pascal"},
+ {0x62, "Pascal"},
+ {0x70, "Volta"},
+ {0x72, "Xavier"},
+ {0x75, "Turing"},
+ {0x80, "Ampere"},
+ {0x86, "Ampere"},
+ {-1, "Graphics Device"}};
+
+ int index = 0;
+
+ while (nGpuArchNameSM[index].SM != -1) {
+ if (nGpuArchNameSM[index].SM == ((major << 4) + minor)) {
+ return nGpuArchNameSM[index].name;
+ }
+
+ index++;
+ }
+
+ // If we don't find the values, we default use the previous one
+ // to run properly
+ printf(
+ "MapSMtoArchName for SM %d.%d is undefined."
+ " Default to use %s\n",
+ major, minor, nGpuArchNameSM[index - 1].name);
+ return nGpuArchNameSM[index - 1].name;
+}
+// end of GPU Architecture definitions
+
+#ifdef __CUDA_RUNTIME_H__
+// General GPU Device CUDA Initialization
+inline int gpuDeviceInit(int devID) {
+ int device_count;
+ checkCudaErrors(cudaGetDeviceCount(&device_count));
+
+ if (device_count == 0) {
+ fprintf(stderr,
+ "gpuDeviceInit() CUDA error: "
+ "no devices supporting CUDA.\n");
+ exit(EXIT_FAILURE);
+ }
+
+ if (devID < 0) {
+ devID = 0;
+ }
+
+ if (devID > device_count - 1) {
+ fprintf(stderr, "\n");
+ fprintf(stderr, ">> %d CUDA capable GPU device(s) detected. <<\n",
+ device_count);
+ fprintf(stderr,
+ ">> gpuDeviceInit (-device=%d) is not a valid"
+ " GPU device. <<\n",
+ devID);
+ fprintf(stderr, "\n");
+ return -devID;
+ }
+
+ int computeMode = -1, major = 0, minor = 0;
+ checkCudaErrors(cudaDeviceGetAttribute(&computeMode, cudaDevAttrComputeMode, devID));
+ checkCudaErrors(cudaDeviceGetAttribute(&major, cudaDevAttrComputeCapabilityMajor, devID));
+ checkCudaErrors(cudaDeviceGetAttribute(&minor, cudaDevAttrComputeCapabilityMinor, devID));
+ if (computeMode == cudaComputeModeProhibited) {
+ fprintf(stderr,
+ "Error: device is running in <Compute Mode "
+ "Prohibited>, no threads can use cudaSetDevice().\n");
+ return -1;
+ }
+
+ if (major < 1) {
+ fprintf(stderr, "gpuDeviceInit(): GPU device does not support CUDA.\n");
+ exit(EXIT_FAILURE);
+ }
+
+ checkCudaErrors(cudaSetDevice(devID));
+ printf("gpuDeviceInit() CUDA Device [%d]: \"%s\n", devID, _ConvertSMVer2ArchName(major, minor));
+
+ return devID;
+}
+
+// This function returns the best GPU (with maximum GFLOPS)
+inline int gpuGetMaxGflopsDeviceId() {
+ int current_device = 0, sm_per_multiproc = 0;
+ int max_perf_device = 0;
+ int device_count = 0;
+ int devices_prohibited = 0;
+
+ uint64_t max_compute_perf = 0;
+ checkCudaErrors(cudaGetDeviceCount(&device_count));
+
+ if (device_count == 0) {
+ fprintf(stderr,
+ "gpuGetMaxGflopsDeviceId() CUDA error:"
+ " no devices supporting CUDA.\n");
+ exit(EXIT_FAILURE);
+ }
+
+ // Find the best CUDA capable GPU device
+ current_device = 0;
+
+ while (current_device < device_count) {
+ int computeMode = -1, major = 0, minor = 0;
+ checkCudaErrors(cudaDeviceGetAttribute(&computeMode, cudaDevAttrComputeMode, current_device));
+ checkCudaErrors(cudaDeviceGetAttribute(&major, cudaDevAttrComputeCapabilityMajor, current_device));
+ checkCudaErrors(cudaDeviceGetAttribute(&minor, cudaDevAttrComputeCapabilityMinor, current_device));
+
+ // If this GPU is not running on Compute Mode prohibited,
+ // then we can add it to the list
+ if (computeMode != cudaComputeModeProhibited) {
+ if (major == 9999 && minor == 9999) {
+ sm_per_multiproc = 1;
+ } else {
+ sm_per_multiproc =
+ _ConvertSMVer2Cores(major, minor);
+ }
+ int multiProcessorCount = 0, clockRate = 0;
+ checkCudaErrors(cudaDeviceGetAttribute(&multiProcessorCount, cudaDevAttrMultiProcessorCount, current_device));
+ cudaError_t result = cudaDeviceGetAttribute(&clockRate, cudaDevAttrClockRate, current_device);
+ if (result != cudaSuccess) {
+ // If cudaDevAttrClockRate attribute is not supported we
+ // set clockRate as 1, to consider GPU with most SMs and CUDA Cores.
+ if(result == cudaErrorInvalidValue) {
+ clockRate = 1;
+ }
+ else {
+ fprintf(stderr, "CUDA error at %s:%d code=%d(%s) \n", __FILE__, __LINE__,
+ static_cast<unsigned int>(result), _cudaGetErrorEnum(result));
+ exit(EXIT_FAILURE);
+ }
+ }
+ uint64_t compute_perf = (uint64_t)multiProcessorCount * sm_per_multiproc * clockRate;
+
+ if (compute_perf > max_compute_perf) {
+ max_compute_perf = compute_perf;
+ max_perf_device = current_device;
+ }
+ } else {
+ devices_prohibited++;
+ }
+
+ ++current_device;
+ }
+
+ if (devices_prohibited == device_count) {
+ fprintf(stderr,
+ "gpuGetMaxGflopsDeviceId() CUDA error:"
+ " all devices have compute mode prohibited.\n");
+ exit(EXIT_FAILURE);
+ }
+
+ return max_perf_device;
+}
+
+// Initialization code to find the best CUDA Device
+inline int findCudaDevice(int argc, const char **argv) {
+ int devID = 0;
+
+ // If the command-line has a device number specified, use it
+ if (checkCmdLineFlag(argc, argv, "device")) {
+ devID = getCmdLineArgumentInt(argc, argv, "device=");
+
+ if (devID < 0) {
+ printf("Invalid command line parameter\n ");
+ exit(EXIT_FAILURE);
+ } else {
+ devID = gpuDeviceInit(devID);
+
+ if (devID < 0) {
+ printf("exiting...\n");
+ exit(EXIT_FAILURE);
+ }
+ }
+ } else {
+ // Otherwise pick the device with highest Gflops/s
+ devID = gpuGetMaxGflopsDeviceId();
+ checkCudaErrors(cudaSetDevice(devID));
+ int major = 0, minor = 0;
+ checkCudaErrors(cudaDeviceGetAttribute(&major, cudaDevAttrComputeCapabilityMajor, devID));
+ checkCudaErrors(cudaDeviceGetAttribute(&minor, cudaDevAttrComputeCapabilityMinor, devID));
+ printf("GPU Device %d: \"%s\" with compute capability %d.%d\n\n",
+ devID, _ConvertSMVer2ArchName(major, minor), major, minor);
+
+ }
+
+ return devID;
+}
+
+inline int findIntegratedGPU() {
+ int current_device = 0;
+ int device_count = 0;
+ int devices_prohibited = 0;
+
+ checkCudaErrors(cudaGetDeviceCount(&device_count));
+
+ if (device_count == 0) {
+ fprintf(stderr, "CUDA error: no devices supporting CUDA.\n");
+ exit(EXIT_FAILURE);
+ }
+
+ // Find the integrated GPU which is compute capable
+ while (current_device < device_count) {
+ int computeMode = -1, integrated = -1;
+ checkCudaErrors(cudaDeviceGetAttribute(&computeMode, cudaDevAttrComputeMode, current_device));
+ checkCudaErrors(cudaDeviceGetAttribute(&integrated, cudaDevAttrIntegrated, current_device));
+ // If GPU is integrated and is not running on Compute Mode prohibited,
+ // then cuda can map to GLES resource
+ if (integrated && (computeMode != cudaComputeModeProhibited)) {
+ checkCudaErrors(cudaSetDevice(current_device));
+
+ int major = 0, minor = 0;
+ checkCudaErrors(cudaDeviceGetAttribute(&major, cudaDevAttrComputeCapabilityMajor, current_device));
+ checkCudaErrors(cudaDeviceGetAttribute(&minor, cudaDevAttrComputeCapabilityMinor, current_device));
+ printf("GPU Device %d: \"%s\" with compute capability %d.%d\n\n",
+ current_device, _ConvertSMVer2ArchName(major, minor), major, minor);
+
+ return current_device;
+ } else {
+ devices_prohibited++;
+ }
+
+ current_device++;
+ }
+
+ if (devices_prohibited == device_count) {
+ fprintf(stderr,
+ "CUDA error:"
+ " No GLES-CUDA Interop capable GPU found.\n");
+ exit(EXIT_FAILURE);
+ }
+
+ return -1;
+}
+
+// General check for CUDA GPU SM Capabilities
+inline bool checkCudaCapabilities(int major_version, int minor_version) {
+ int dev;
+ int major = 0, minor = 0;
+
+ checkCudaErrors(cudaGetDevice(&dev));
+ checkCudaErrors(cudaDeviceGetAttribute(&major, cudaDevAttrComputeCapabilityMajor, dev));
+ checkCudaErrors(cudaDeviceGetAttribute(&minor, cudaDevAttrComputeCapabilityMinor, dev));
+
+ if ((major > major_version) ||
+ (major == major_version &&
+ minor >= minor_version)) {
+ printf(" Device %d: <%16s >, Compute SM %d.%d detected\n", dev,
+ _ConvertSMVer2ArchName(major, minor), major, minor);
+ return true;
+ } else {
+ printf(
+ " No GPU device was found that can support "
+ "CUDA compute capability %d.%d.\n",
+ major_version, minor_version);
+ return false;
+ }
+}
+#endif
+
+// end of CUDA Helper Functions
+
+#endif // COMMON_HELPER_CUDA_H_
\ No newline at end of file
diff --git a/3rdParty/cuda_samples/helper_functions.h b/3rdParty/cuda_samples/helper_functions.h
new file mode 100644
index 0000000000000000000000000000000000000000..3fc2ea47ba7d39a4bf6a882f65b16779de6de0ac
--- /dev/null
+++ b/3rdParty/cuda_samples/helper_functions.h
@@ -0,0 +1,59 @@
+/* Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
+ *
+ * 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 NVIDIA CORPORATION 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 THE COPYRIGHT HOLDERS ``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 THE COPYRIGHT OWNER 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.
+ */
+
+// These are helper functions for the SDK samples (string parsing,
+// timers, image helpers, etc)
+#ifndef COMMON_HELPER_FUNCTIONS_H_
+#define COMMON_HELPER_FUNCTIONS_H_
+
+#ifdef WIN32
+#pragma warning(disable : 4996)
+#endif
+
+// includes, project
+#include <assert.h>
+#include <exception.h>
+#include <math.h>
+#include <stdio.h>
+#include <stdlib.h>
+
+#include <algorithm>
+#include <fstream>
+#include <iostream>
+#include <string>
+#include <vector>
+
+// includes, timer, string parsing, image helpers
+#include <helper_image.h> // helper functions for image compare, dump, data comparisons
+#include <helper_string.h> // helper functions for string parsing
+#include <helper_timer.h> // helper functions for timers
+
+#ifndef EXIT_WAIVED
+#define EXIT_WAIVED 2
+#endif
+
+#endif // COMMON_HELPER_FUNCTIONS_H_
\ No newline at end of file
diff --git a/3rdParty/cuda_samples/helper_image.h b/3rdParty/cuda_samples/helper_image.h
new file mode 100644
index 0000000000000000000000000000000000000000..eb7190c21b39463222beb579337d14c1a54b0000
--- /dev/null
+++ b/3rdParty/cuda_samples/helper_image.h
@@ -0,0 +1,1009 @@
+//
+// Created by Soeren Peters on 05.02.21.
+//
+
+#ifndef VIRTUALFLUIDS_HELPER_IMAGE_H
+#define VIRTUALFLUIDS_HELPER_IMAGE_H
+
+#endif // VIRTUALFLUIDS_HELPER_IMAGE_H
+/* Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
+ *
+ * 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 NVIDIA CORPORATION 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 THE COPYRIGHT HOLDERS ``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 THE COPYRIGHT OWNER 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.
+ */
+
+// These are helper functions for the SDK samples (image,bitmap)
+#ifndef COMMON_HELPER_IMAGE_H_
+#define COMMON_HELPER_IMAGE_H_
+
+#include <assert.h>
+#include <exception.h>
+#include <math.h>
+#include <stdint.h>
+
+#include <algorithm>
+#include <fstream>
+#include <iostream>
+#include <string>
+#include <vector>
+
+#ifndef MIN
+#define MIN(a, b) ((a < b) ? a : b)
+#endif
+#ifndef MAX
+#define MAX(a, b) ((a > b) ? a : b)
+#endif
+
+#ifndef EXIT_WAIVED
+#define EXIT_WAIVED 2
+#endif
+
+#include <helper_string.h>
+
+// namespace unnamed (internal)
+namespace helper_image_internal {
+//! size of PGM file header
+const unsigned int PGMHeaderSize = 0x40;
+
+// types
+
+//! Data converter from unsigned char / unsigned byte to type T
+template <class T>
+struct ConverterFromUByte;
+
+//! Data converter from unsigned char / unsigned byte
+template <>
+struct ConverterFromUByte<unsigned char> {
+ //! Conversion operator
+ //! @return converted value
+ //! @param val value to convert
+ float operator()(const unsigned char &val) {
+ return static_cast<unsigned char>(val);
+ }
+};
+
+//! Data converter from unsigned char / unsigned byte to float
+template <>
+struct ConverterFromUByte<float> {
+ //! Conversion operator
+ //! @return converted value
+ //! @param val value to convert
+ float operator()(const unsigned char &val) {
+ return static_cast<float>(val) / 255.0f;
+ }
+};
+
+//! Data converter from unsigned char / unsigned byte to type T
+template <class T>
+struct ConverterToUByte;
+
+//! Data converter from unsigned char / unsigned byte to unsigned int
+template <>
+struct ConverterToUByte<unsigned char> {
+ //! Conversion operator (essentially a passthru
+ //! @return converted value
+ //! @param val value to convert
+ unsigned char operator()(const unsigned char &val) { return val; }
+};
+
+//! Data converter from unsigned char / unsigned byte to unsigned int
+template <>
+struct ConverterToUByte<float> {
+ //! Conversion operator
+ //! @return converted value
+ //! @param val value to convert
+ unsigned char operator()(const float &val) {
+ return static_cast<unsigned char>(val * 255.0f);
+ }
+};
+} // namespace helper_image_internal
+
+#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
+#ifndef FOPEN
+#define FOPEN(fHandle, filename, mode) fopen_s(&fHandle, filename, mode)
+#endif
+#ifndef FOPEN_FAIL
+#define FOPEN_FAIL(result) (result != 0)
+#endif
+#ifndef SSCANF
+#define SSCANF sscanf_s
+#endif
+#else
+#ifndef FOPEN
+#define FOPEN(fHandle, filename, mode) (fHandle = fopen(filename, mode))
+#endif
+#ifndef FOPEN_FAIL
+#define FOPEN_FAIL(result) (result == NULL)
+#endif
+#ifndef SSCANF
+#define SSCANF sscanf
+#endif
+#endif
+
+inline bool __loadPPM(const char *file, unsigned char **data, unsigned int *w,
+ unsigned int *h, unsigned int *channels) {
+ FILE *fp = NULL;
+
+ if (FOPEN_FAIL(FOPEN(fp, file, "rb"))) {
+ std::cerr << "__LoadPPM() : Failed to open file: " << file << std::endl;
+ return false;
+ }
+
+ // check header
+ char header[helper_image_internal::PGMHeaderSize];
+
+ if (fgets(header, helper_image_internal::PGMHeaderSize, fp) == NULL) {
+ std::cerr << "__LoadPPM() : reading PGM header returned NULL" << std::endl;
+ return false;
+ }
+
+ if (strncmp(header, "P5", 2) == 0) {
+ *channels = 1;
+ } else if (strncmp(header, "P6", 2) == 0) {
+ *channels = 3;
+ } else {
+ std::cerr << "__LoadPPM() : File is not a PPM or PGM image" << std::endl;
+ *channels = 0;
+ return false;
+ }
+
+ // parse header, read maxval, width and height
+ unsigned int width = 0;
+ unsigned int height = 0;
+ unsigned int maxval = 0;
+ unsigned int i = 0;
+
+ while (i < 3) {
+ if (fgets(header, helper_image_internal::PGMHeaderSize, fp) == NULL) {
+ std::cerr << "__LoadPPM() : reading PGM header returned NULL"
+ << std::endl;
+ return false;
+ }
+
+ if (header[0] == '#') {
+ continue;
+ }
+
+ if (i == 0) {
+ i += SSCANF(header, "%u %u %u", &width, &height, &maxval);
+ } else if (i == 1) {
+ i += SSCANF(header, "%u %u", &height, &maxval);
+ } else if (i == 2) {
+ i += SSCANF(header, "%u", &maxval);
+ }
+ }
+
+ // check if given handle for the data is initialized
+ if (NULL != *data) {
+ if (*w != width || *h != height) {
+ std::cerr << "__LoadPPM() : Invalid image dimensions." << std::endl;
+ }
+ } else {
+ *data = (unsigned char *)malloc(sizeof(unsigned char) * width * height *
+ *channels);
+ *w = width;
+ *h = height;
+ }
+
+ // read and close file
+ if (fread(*data, sizeof(unsigned char), width * height * *channels, fp) ==
+ 0) {
+ std::cerr << "__LoadPPM() read data returned error." << std::endl;
+ }
+
+ fclose(fp);
+
+ return true;
+}
+
+template <class T>
+inline bool sdkLoadPGM(const char *file, T **data, unsigned int *w,
+ unsigned int *h) {
+ unsigned char *idata = NULL;
+ unsigned int channels;
+
+ if (true != __loadPPM(file, &idata, w, h, &channels)) {
+ return false;
+ }
+
+ unsigned int size = *w * *h * channels;
+
+ // initialize mem if necessary
+ // the correct size is checked / set in loadPGMc()
+ if (NULL == *data) {
+ *data = reinterpret_cast<T *>(malloc(sizeof(T) * size));
+ }
+
+ // copy and cast data
+ std::transform(idata, idata + size, *data,
+ helper_image_internal::ConverterFromUByte<T>());
+
+ free(idata);
+
+ return true;
+}
+
+template <class T>
+inline bool sdkLoadPPM4(const char *file, T **data, unsigned int *w,
+ unsigned int *h) {
+ unsigned char *idata = 0;
+ unsigned int channels;
+
+ if (__loadPPM(file, &idata, w, h, &channels)) {
+ // pad 4th component
+ int size = *w * *h;
+ // keep the original pointer
+ unsigned char *idata_orig = idata;
+ *data = reinterpret_cast<T *>(malloc(sizeof(T) * size * 4));
+ unsigned char *ptr = *data;
+
+ for (int i = 0; i < size; i++) {
+ *ptr++ = *idata++;
+ *ptr++ = *idata++;
+ *ptr++ = *idata++;
+ *ptr++ = 0;
+ }
+
+ free(idata_orig);
+ return true;
+ } else {
+ free(idata);
+ return false;
+ }
+}
+
+inline bool __savePPM(const char *file, unsigned char *data, unsigned int w,
+ unsigned int h, unsigned int channels) {
+ assert(NULL != data);
+ assert(w > 0);
+ assert(h > 0);
+
+ std::fstream fh(file, std::fstream::out | std::fstream::binary);
+
+ if (fh.bad()) {
+ std::cerr << "__savePPM() : Opening file failed." << std::endl;
+ return false;
+ }
+
+ if (channels == 1) {
+ fh << "P5\n";
+ } else if (channels == 3) {
+ fh << "P6\n";
+ } else {
+ std::cerr << "__savePPM() : Invalid number of channels." << std::endl;
+ return false;
+ }
+
+ fh << w << "\n" << h << "\n" << 0xff << std::endl;
+
+ for (unsigned int i = 0; (i < (w * h * channels)) && fh.good(); ++i) {
+ fh << data[i];
+ }
+
+ fh.flush();
+
+ if (fh.bad()) {
+ std::cerr << "__savePPM() : Writing data failed." << std::endl;
+ return false;
+ }
+
+ fh.close();
+
+ return true;
+}
+
+template <class T>
+inline bool sdkSavePGM(const char *file, T *data, unsigned int w,
+ unsigned int h) {
+ unsigned int size = w * h;
+ unsigned char *idata = (unsigned char *)malloc(sizeof(unsigned char) * size);
+
+ std::transform(data, data + size, idata,
+ helper_image_internal::ConverterToUByte<T>());
+
+ // write file
+ bool result = __savePPM(file, idata, w, h, 1);
+
+ // cleanup
+ free(idata);
+
+ return result;
+}
+
+inline bool sdkSavePPM4ub(const char *file, unsigned char *data, unsigned int w,
+ unsigned int h) {
+ // strip 4th component
+ int size = w * h;
+ unsigned char *ndata =
+ (unsigned char *)malloc(sizeof(unsigned char) * size * 3);
+ unsigned char *ptr = ndata;
+
+ for (int i = 0; i < size; i++) {
+ *ptr++ = *data++;
+ *ptr++ = *data++;
+ *ptr++ = *data++;
+ data++;
+ }
+
+ bool result = __savePPM(file, ndata, w, h, 3);
+ free(ndata);
+ return result;
+}
+
+//////////////////////////////////////////////////////////////////////////////
+//! Read file \filename and return the data
+//! @return bool if reading the file succeeded, otherwise false
+//! @param filename name of the source file
+//! @param data uninitialized pointer, returned initialized and pointing to
+//! the data read
+//! @param len number of data elements in data, -1 on error
+//////////////////////////////////////////////////////////////////////////////
+template <class T>
+inline bool sdkReadFile(const char *filename, T **data, unsigned int *len,
+ bool verbose) {
+ // check input arguments
+ assert(NULL != filename);
+ assert(NULL != len);
+
+ // intermediate storage for the data read
+ std::vector<T> data_read;
+
+ // open file for reading
+ FILE *fh = NULL;
+
+ // check if filestream is valid
+ if (FOPEN_FAIL(FOPEN(fh, filename, "r"))) {
+ printf("Unable to open input file: %s\n", filename);
+ return false;
+ }
+
+ // read all data elements
+ T token;
+
+ while (!feof(fh)) {
+ fscanf(fh, "%f", &token);
+ data_read.push_back(token);
+ }
+
+ // the last element is read twice
+ data_read.pop_back();
+ fclose(fh);
+
+ // check if the given handle is already initialized
+ if (NULL != *data) {
+ if (*len != data_read.size()) {
+ std::cerr << "sdkReadFile() : Initialized memory given but "
+ << "size mismatch with signal read "
+ << "(data read / data init = " << (unsigned int)data_read.size()
+ << " / " << *len << ")" << std::endl;
+
+ return false;
+ }
+ } else {
+ // allocate storage for the data read
+ *data = reinterpret_cast<T *>(malloc(sizeof(T) * data_read.size()));
+ // store signal size
+ *len = static_cast<unsigned int>(data_read.size());
+ }
+
+ // copy data
+ memcpy(*data, &data_read.front(), sizeof(T) * data_read.size());
+
+ return true;
+}
+
+//////////////////////////////////////////////////////////////////////////////
+//! Read file \filename and return the data
+//! @return bool if reading the file succeeded, otherwise false
+//! @param filename name of the source file
+//! @param data uninitialized pointer, returned initialized and pointing to
+//! the data read
+//! @param len number of data elements in data, -1 on error
+//////////////////////////////////////////////////////////////////////////////
+template <class T>
+inline bool sdkReadFileBlocks(const char *filename, T **data, unsigned int *len,
+ unsigned int block_num, unsigned int block_size,
+ bool verbose) {
+ // check input arguments
+ assert(NULL != filename);
+ assert(NULL != len);
+
+ // open file for reading
+ FILE *fh = fopen(filename, "rb");
+
+ if (fh == NULL && verbose) {
+ std::cerr << "sdkReadFile() : Opening file failed." << std::endl;
+ return false;
+ }
+
+ // check if the given handle is already initialized
+ // allocate storage for the data read
+ data[block_num] = reinterpret_cast<T *>(malloc(block_size));
+
+ // read all data elements
+ fseek(fh, block_num * block_size, SEEK_SET);
+ *len = fread(data[block_num], sizeof(T), block_size / sizeof(T), fh);
+
+ fclose(fh);
+
+ return true;
+}
+
+//////////////////////////////////////////////////////////////////////////////
+//! Write a data file \filename
+//! @return true if writing the file succeeded, otherwise false
+//! @param filename name of the source file
+//! @param data data to write
+//! @param len number of data elements in data, -1 on error
+//! @param epsilon epsilon for comparison
+//////////////////////////////////////////////////////////////////////////////
+template <class T, class S>
+inline bool sdkWriteFile(const char *filename, const T *data, unsigned int len,
+ const S epsilon, bool verbose, bool append = false) {
+ assert(NULL != filename);
+ assert(NULL != data);
+
+ // open file for writing
+ // if (append) {
+ std::fstream fh(filename, std::fstream::out | std::fstream::ate);
+
+ if (verbose) {
+ std::cerr << "sdkWriteFile() : Open file " << filename
+ << " for write/append." << std::endl;
+ }
+
+ /* } else {
+ std::fstream fh(filename, std::fstream::out);
+ if (verbose) {
+ std::cerr << "sdkWriteFile() : Open file " << filename << " for
+ write." << std::endl;
+ }
+ }
+ */
+
+ // check if filestream is valid
+ if (!fh.good()) {
+ if (verbose) {
+ std::cerr << "sdkWriteFile() : Opening file failed." << std::endl;
+ }
+
+ return false;
+ }
+
+ // first write epsilon
+ fh << "# " << epsilon << "\n";
+
+ // write data
+ for (unsigned int i = 0; (i < len) && (fh.good()); ++i) {
+ fh << data[i] << ' ';
+ }
+
+ // Check if writing succeeded
+ if (!fh.good()) {
+ if (verbose) {
+ std::cerr << "sdkWriteFile() : Writing file failed." << std::endl;
+ }
+
+ return false;
+ }
+
+ // file ends with nl
+ fh << std::endl;
+
+ return true;
+}
+
+//////////////////////////////////////////////////////////////////////////////
+//! Compare two arrays of arbitrary type
+//! @return true if \a reference and \a data are identical, otherwise false
+//! @param reference timer_interface to the reference data / gold image
+//! @param data handle to the computed data
+//! @param len number of elements in reference and data
+//! @param epsilon epsilon to use for the comparison
+//////////////////////////////////////////////////////////////////////////////
+template <class T, class S>
+inline bool compareData(const T *reference, const T *data,
+ const unsigned int len, const S epsilon,
+ const float threshold) {
+ assert(epsilon >= 0);
+
+ bool result = true;
+ unsigned int error_count = 0;
+
+ for (unsigned int i = 0; i < len; ++i) {
+ float diff = static_cast<float>(reference[i]) - static_cast<float>(data[i]);
+ bool comp = (diff <= epsilon) && (diff >= -epsilon);
+ result &= comp;
+
+ error_count += !comp;
+
+#if 0
+
+ if (!comp) {
+ std::cerr << "ERROR, i = " << i << ",\t "
+ << reference[i] << " / "
+ << data[i]
+ << " (reference / data)\n";
+ }
+
+#endif
+ }
+
+ if (threshold == 0.0f) {
+ return (result) ? true : false;
+ } else {
+ if (error_count) {
+ printf("%4.2f(%%) of bytes mismatched (count=%d)\n",
+ static_cast<float>(error_count) * 100 / static_cast<float>(len),
+ error_count);
+ }
+
+ return (len * threshold > error_count) ? true : false;
+ }
+}
+
+#ifndef __MIN_EPSILON_ERROR
+#define __MIN_EPSILON_ERROR 1e-3f
+#endif
+
+//////////////////////////////////////////////////////////////////////////////
+//! Compare two arrays of arbitrary type
+//! @return true if \a reference and \a data are identical, otherwise false
+//! @param reference handle to the reference data / gold image
+//! @param data handle to the computed data
+//! @param len number of elements in reference and data
+//! @param epsilon epsilon to use for the comparison
+//! @param epsilon threshold % of (# of bytes) for pass/fail
+//////////////////////////////////////////////////////////////////////////////
+template <class T, class S>
+inline bool compareDataAsFloatThreshold(const T *reference, const T *data,
+ const unsigned int len, const S epsilon,
+ const float threshold) {
+ assert(epsilon >= 0);
+
+ // If we set epsilon to be 0, let's set a minimum threshold
+ float max_error = MAX((float)epsilon, __MIN_EPSILON_ERROR);
+ int error_count = 0;
+ bool result = true;
+
+ for (unsigned int i = 0; i < len; ++i) {
+ float diff =
+ fabs(static_cast<float>(reference[i]) - static_cast<float>(data[i]));
+ bool comp = (diff < max_error);
+ result &= comp;
+
+ if (!comp) {
+ error_count++;
+ }
+ }
+
+ if (threshold == 0.0f) {
+ if (error_count) {
+ printf("total # of errors = %d\n", error_count);
+ }
+
+ return (error_count == 0) ? true : false;
+ } else {
+ if (error_count) {
+ printf("%4.2f(%%) of bytes mismatched (count=%d)\n",
+ static_cast<float>(error_count) * 100 / static_cast<float>(len),
+ error_count);
+ }
+
+ return ((len * threshold > error_count) ? true : false);
+ }
+}
+
+inline void sdkDumpBin(void *data, unsigned int bytes, const char *filename) {
+ printf("sdkDumpBin: <%s>\n", filename);
+ FILE *fp;
+ FOPEN(fp, filename, "wb");
+ fwrite(data, bytes, 1, fp);
+ fflush(fp);
+ fclose(fp);
+}
+
+inline bool sdkCompareBin2BinUint(const char *src_file, const char *ref_file,
+ unsigned int nelements, const float epsilon,
+ const float threshold, char *exec_path) {
+ unsigned int *src_buffer, *ref_buffer;
+ FILE *src_fp = NULL, *ref_fp = NULL;
+
+ uint64_t error_count = 0;
+ size_t fsize = 0;
+
+ if (FOPEN_FAIL(FOPEN(src_fp, src_file, "rb"))) {
+ printf("compareBin2Bin <unsigned int> unable to open src_file: %s\n",
+ src_file);
+ error_count++;
+ }
+
+ char *ref_file_path = sdkFindFilePath(ref_file, exec_path);
+
+ if (ref_file_path == NULL) {
+ printf("compareBin2Bin <unsigned int> unable to find <%s> in <%s>\n",
+ ref_file, exec_path);
+ printf(">>> Check info.xml and [project//data] folder <%s> <<<\n",
+ ref_file);
+ printf("Aborting comparison!\n");
+ printf(" FAILED\n");
+ error_count++;
+
+ if (src_fp) {
+ fclose(src_fp);
+ }
+
+ if (ref_fp) {
+ fclose(ref_fp);
+ }
+ } else {
+ if (FOPEN_FAIL(FOPEN(ref_fp, ref_file_path, "rb"))) {
+ printf(
+ "compareBin2Bin <unsigned int>"
+ " unable to open ref_file: %s\n",
+ ref_file_path);
+ error_count++;
+ }
+
+ if (src_fp && ref_fp) {
+ src_buffer = (unsigned int *)malloc(nelements * sizeof(unsigned int));
+ ref_buffer = (unsigned int *)malloc(nelements * sizeof(unsigned int));
+
+ fsize = fread(src_buffer, nelements, sizeof(unsigned int), src_fp);
+ fsize = fread(ref_buffer, nelements, sizeof(unsigned int), ref_fp);
+
+ printf(
+ "> compareBin2Bin <unsigned int> nelements=%d,"
+ " epsilon=%4.2f, threshold=%4.2f\n",
+ nelements, epsilon, threshold);
+ printf(" src_file <%s>, size=%d bytes\n", src_file,
+ static_cast<int>(fsize));
+ printf(" ref_file <%s>, size=%d bytes\n", ref_file_path,
+ static_cast<int>(fsize));
+
+ if (!compareData<unsigned int, float>(ref_buffer, src_buffer, nelements,
+ epsilon, threshold)) {
+ error_count++;
+ }
+
+ fclose(src_fp);
+ fclose(ref_fp);
+
+ free(src_buffer);
+ free(ref_buffer);
+ } else {
+ if (src_fp) {
+ fclose(src_fp);
+ }
+
+ if (ref_fp) {
+ fclose(ref_fp);
+ }
+ }
+ }
+
+ if (error_count == 0) {
+ printf(" OK\n");
+ } else {
+ printf(" FAILURE: %d errors...\n", (unsigned int)error_count);
+ }
+
+ return (error_count == 0); // returns true if all pixels pass
+}
+
+inline bool sdkCompareBin2BinFloat(const char *src_file, const char *ref_file,
+ unsigned int nelements, const float epsilon,
+ const float threshold, char *exec_path) {
+ float *src_buffer = NULL, *ref_buffer = NULL;
+ FILE *src_fp = NULL, *ref_fp = NULL;
+ size_t fsize = 0;
+
+ uint64_t error_count = 0;
+
+ if (FOPEN_FAIL(FOPEN(src_fp, src_file, "rb"))) {
+ printf("compareBin2Bin <float> unable to open src_file: %s\n", src_file);
+ error_count = 1;
+ }
+
+ char *ref_file_path = sdkFindFilePath(ref_file, exec_path);
+
+ if (ref_file_path == NULL) {
+ printf("compareBin2Bin <float> unable to find <%s> in <%s>\n", ref_file,
+ exec_path);
+ printf(">>> Check info.xml and [project//data] folder <%s> <<<\n",
+ exec_path);
+ printf("Aborting comparison!\n");
+ printf(" FAILED\n");
+ error_count++;
+
+ if (src_fp) {
+ fclose(src_fp);
+ }
+
+ if (ref_fp) {
+ fclose(ref_fp);
+ }
+ } else {
+ if (FOPEN_FAIL(FOPEN(ref_fp, ref_file_path, "rb"))) {
+ printf("compareBin2Bin <float> unable to open ref_file: %s\n",
+ ref_file_path);
+ error_count = 1;
+ }
+
+ if (src_fp && ref_fp) {
+ src_buffer = reinterpret_cast<float *>(malloc(nelements * sizeof(float)));
+ ref_buffer = reinterpret_cast<float *>(malloc(nelements * sizeof(float)));
+
+ printf(
+ "> compareBin2Bin <float> nelements=%d, epsilon=%4.2f,"
+ " threshold=%4.2f\n",
+ nelements, epsilon, threshold);
+ fsize = fread(src_buffer, sizeof(float), nelements, src_fp);
+ printf(" src_file <%s>, size=%d bytes\n", src_file,
+ static_cast<int>(fsize * sizeof(float)));
+ fsize = fread(ref_buffer, sizeof(float), nelements, ref_fp);
+ printf(" ref_file <%s>, size=%d bytes\n", ref_file_path,
+ static_cast<int>(fsize * sizeof(float)));
+
+ if (!compareDataAsFloatThreshold<float, float>(
+ ref_buffer, src_buffer, nelements, epsilon, threshold)) {
+ error_count++;
+ }
+
+ fclose(src_fp);
+ fclose(ref_fp);
+
+ free(src_buffer);
+ free(ref_buffer);
+ } else {
+ if (src_fp) {
+ fclose(src_fp);
+ }
+
+ if (ref_fp) {
+ fclose(ref_fp);
+ }
+ }
+ }
+
+ if (error_count == 0) {
+ printf(" OK\n");
+ } else {
+ printf(" FAILURE: %d errors...\n", (unsigned int)error_count);
+ }
+
+ return (error_count == 0); // returns true if all pixels pass
+}
+
+inline bool sdkCompareL2fe(const float *reference, const float *data,
+ const unsigned int len, const float epsilon) {
+ assert(epsilon >= 0);
+
+ float error = 0;
+ float ref = 0;
+
+ for (unsigned int i = 0; i < len; ++i) {
+ float diff = reference[i] - data[i];
+ error += diff * diff;
+ ref += reference[i] * reference[i];
+ }
+
+ float normRef = sqrtf(ref);
+
+ if (fabs(ref) < 1e-7) {
+#ifdef _DEBUG
+ std::cerr << "ERROR, reference l2-norm is 0\n";
+#endif
+ return false;
+ }
+
+ float normError = sqrtf(error);
+ error = normError / normRef;
+ bool result = error < epsilon;
+#ifdef _DEBUG
+
+ if (!result) {
+ std::cerr << "ERROR, l2-norm error " << error << " is greater than epsilon "
+ << epsilon << "\n";
+ }
+
+#endif
+
+ return result;
+}
+
+inline bool sdkLoadPPMub(const char *file, unsigned char **data,
+ unsigned int *w, unsigned int *h) {
+ unsigned int channels;
+ return __loadPPM(file, data, w, h, &channels);
+}
+
+inline bool sdkLoadPPM4ub(const char *file, unsigned char **data,
+ unsigned int *w, unsigned int *h) {
+ unsigned char *idata = 0;
+ unsigned int channels;
+
+ if (__loadPPM(file, &idata, w, h, &channels)) {
+ // pad 4th component
+ int size = *w * *h;
+ // keep the original pointer
+ unsigned char *idata_orig = idata;
+ *data = (unsigned char *)malloc(sizeof(unsigned char) * size * 4);
+ unsigned char *ptr = *data;
+
+ for (int i = 0; i < size; i++) {
+ *ptr++ = *idata++;
+ *ptr++ = *idata++;
+ *ptr++ = *idata++;
+ *ptr++ = 0;
+ }
+
+ free(idata_orig);
+ return true;
+ } else {
+ free(idata);
+ return false;
+ }
+}
+
+inline bool sdkComparePPM(const char *src_file, const char *ref_file,
+ const float epsilon, const float threshold,
+ bool verboseErrors) {
+ unsigned char *src_data, *ref_data;
+ uint64_t error_count = 0;
+ unsigned int ref_width, ref_height;
+ unsigned int src_width, src_height;
+
+ if (src_file == NULL || ref_file == NULL) {
+ if (verboseErrors) {
+ std::cerr << "PPMvsPPM: src_file or ref_file is NULL."
+ " Aborting comparison\n";
+ }
+
+ return false;
+ }
+
+ if (verboseErrors) {
+ std::cerr << "> Compare (a)rendered: <" << src_file << ">\n";
+ std::cerr << "> (b)reference: <" << ref_file << ">\n";
+ }
+
+ if (sdkLoadPPM4ub(ref_file, &ref_data, &ref_width, &ref_height) != true) {
+ if (verboseErrors) {
+ std::cerr << "PPMvsPPM: unable to load ref image file: " << ref_file
+ << "\n";
+ }
+
+ return false;
+ }
+
+ if (sdkLoadPPM4ub(src_file, &src_data, &src_width, &src_height) != true) {
+ std::cerr << "PPMvsPPM: unable to load src image file: " << src_file
+ << "\n";
+ return false;
+ }
+
+ if (src_height != ref_height || src_width != ref_width) {
+ if (verboseErrors) {
+ std::cerr << "PPMvsPPM: source and ref size mismatch (" << src_width
+ << "," << src_height << ")vs(" << ref_width << "," << ref_height
+ << ")\n";
+ }
+ }
+
+ if (verboseErrors) {
+ std::cerr << "PPMvsPPM: comparing images size (" << src_width << ","
+ << src_height << ") epsilon(" << epsilon << "), threshold("
+ << threshold * 100 << "%)\n";
+ }
+
+ if (compareData(ref_data, src_data, src_width * src_height * 4, epsilon,
+ threshold) == false) {
+ error_count = 1;
+ }
+
+ if (error_count == 0) {
+ if (verboseErrors) {
+ std::cerr << " OK\n\n";
+ }
+ } else {
+ if (verboseErrors) {
+ std::cerr << " FAILURE! " << error_count << " errors...\n\n";
+ }
+ }
+
+ // returns true if all pixels pass
+ return (error_count == 0) ? true : false;
+}
+
+inline bool sdkComparePGM(const char *src_file, const char *ref_file,
+ const float epsilon, const float threshold,
+ bool verboseErrors) {
+ unsigned char *src_data = 0, *ref_data = 0;
+ uint64_t error_count = 0;
+ unsigned int ref_width, ref_height;
+ unsigned int src_width, src_height;
+
+ if (src_file == NULL || ref_file == NULL) {
+ if (verboseErrors) {
+ std::cerr << "PGMvsPGM: src_file or ref_file is NULL."
+ " Aborting comparison\n";
+ }
+
+ return false;
+ }
+
+ if (verboseErrors) {
+ std::cerr << "> Compare (a)rendered: <" << src_file << ">\n";
+ std::cerr << "> (b)reference: <" << ref_file << ">\n";
+ }
+
+ if (sdkLoadPPMub(ref_file, &ref_data, &ref_width, &ref_height) != true) {
+ if (verboseErrors) {
+ std::cerr << "PGMvsPGM: unable to load ref image file: " << ref_file
+ << "\n";
+ }
+
+ return false;
+ }
+
+ if (sdkLoadPPMub(src_file, &src_data, &src_width, &src_height) != true) {
+ std::cerr << "PGMvsPGM: unable to load src image file: " << src_file
+ << "\n";
+ return false;
+ }
+
+ if (src_height != ref_height || src_width != ref_width) {
+ if (verboseErrors) {
+ std::cerr << "PGMvsPGM: source and ref size mismatch (" << src_width
+ << "," << src_height << ")vs(" << ref_width << "," << ref_height
+ << ")\n";
+ }
+ }
+
+ if (verboseErrors)
+ std::cerr << "PGMvsPGM: comparing images size (" << src_width << ","
+ << src_height << ") epsilon(" << epsilon << "), threshold("
+ << threshold * 100 << "%)\n";
+
+ if (compareData(ref_data, src_data, src_width * src_height, epsilon,
+ threshold) == false) {
+ error_count = 1;
+ }
+
+ if (error_count == 0) {
+ if (verboseErrors) {
+ std::cerr << " OK\n\n";
+ }
+ } else {
+ if (verboseErrors) {
+ std::cerr << " FAILURE! " << error_count << " errors...\n\n";
+ }
+ }
+
+ // returns true if all pixels pass
+ return (error_count == 0) ? true : false;
+}
+
+#endif // COMMON_HELPER_IMAGE_H_
\ No newline at end of file
diff --git a/3rdParty/cuda_samples/helper_string.h b/3rdParty/cuda_samples/helper_string.h
new file mode 100644
index 0000000000000000000000000000000000000000..c09935174ec057757c20263df080cb8d77b53f52
--- /dev/null
+++ b/3rdParty/cuda_samples/helper_string.h
@@ -0,0 +1,368 @@
+/* Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
+ *
+ * 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 NVIDIA CORPORATION 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 THE COPYRIGHT HOLDERS ``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 THE COPYRIGHT OWNER 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.
+ */
+
+// These are helper functions for the SDK samples (string parsing, timers, etc)
+#ifndef COMMON_HELPER_STRING_H_
+#define COMMON_HELPER_STRING_H_
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <fstream>
+#include <string>
+
+#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
+#ifndef _CRT_SECURE_NO_DEPRECATE
+#define _CRT_SECURE_NO_DEPRECATE
+#endif
+#ifndef STRCASECMP
+#define STRCASECMP _stricmp
+#endif
+#ifndef STRNCASECMP
+#define STRNCASECMP _strnicmp
+#endif
+#ifndef STRCPY
+#define STRCPY(sFilePath, nLength, sPath) strcpy_s(sFilePath, nLength, sPath)
+#endif
+
+#ifndef FOPEN
+#define FOPEN(fHandle, filename, mode) fopen_s(&fHandle, filename, mode)
+#endif
+#ifndef FOPEN_FAIL
+#define FOPEN_FAIL(result) (result != 0)
+#endif
+#ifndef SSCANF
+#define SSCANF sscanf_s
+#endif
+#ifndef SPRINTF
+#define SPRINTF sprintf_s
+#endif
+#else // Linux Includes
+#include <string.h>
+#include <strings.h>
+
+#ifndef STRCASECMP
+#define STRCASECMP strcasecmp
+#endif
+#ifndef STRNCASECMP
+#define STRNCASECMP strncasecmp
+#endif
+#ifndef STRCPY
+#define STRCPY(sFilePath, nLength, sPath) strcpy(sFilePath, sPath)
+#endif
+
+#ifndef FOPEN
+#define FOPEN(fHandle, filename, mode) (fHandle = fopen(filename, mode))
+#endif
+#ifndef FOPEN_FAIL
+#define FOPEN_FAIL(result) (result == NULL)
+#endif
+#ifndef SSCANF
+#define SSCANF sscanf
+#endif
+#ifndef SPRINTF
+#define SPRINTF sprintf
+#endif
+#endif
+
+#ifndef EXIT_WAIVED
+#define EXIT_WAIVED 2
+#endif
+
+// CUDA Utility Helper Functions
+inline int stringRemoveDelimiter(char delimiter, const char *string) {
+ int string_start = 0;
+
+ while (string[string_start] == delimiter) {
+ string_start++;
+ }
+
+ if (string_start >= static_cast<int>(strlen(string) - 1)) {
+ return 0;
+ }
+
+ return string_start;
+}
+
+inline int getFileExtension(char *filename, char **extension) {
+ int string_length = static_cast<int>(strlen(filename));
+
+ while (filename[string_length--] != '.') {
+ if (string_length == 0) break;
+ }
+
+ if (string_length > 0) string_length += 2;
+
+ if (string_length == 0)
+ *extension = NULL;
+ else
+ *extension = &filename[string_length];
+
+ return string_length;
+}
+
+inline bool checkCmdLineFlag(const int argc, const char **argv,
+ const char *string_ref) {
+ bool bFound = false;
+
+ if (argc >= 1) {
+ for (int i = 1; i < argc; i++) {
+ int string_start = stringRemoveDelimiter('-', argv[i]);
+ const char *string_argv = &argv[i][string_start];
+
+ const char *equal_pos = strchr(string_argv, '=');
+ int argv_length = static_cast<int>(
+ equal_pos == 0 ? strlen(string_argv) : equal_pos - string_argv);
+
+ int length = static_cast<int>(strlen(string_ref));
+
+ if (length == argv_length &&
+ !STRNCASECMP(string_argv, string_ref, length)) {
+ bFound = true;
+ continue;
+ }
+ }
+ }
+
+ return bFound;
+}
+
+// This function wraps the CUDA Driver API into a template function
+template <class T>
+inline bool getCmdLineArgumentValue(const int argc, const char **argv,
+ const char *string_ref, T *value) {
+ bool bFound = false;
+
+ if (argc >= 1) {
+ for (int i = 1; i < argc; i++) {
+ int string_start = stringRemoveDelimiter('-', argv[i]);
+ const char *string_argv = &argv[i][string_start];
+ int length = static_cast<int>(strlen(string_ref));
+
+ if (!STRNCASECMP(string_argv, string_ref, length)) {
+ if (length + 1 <= static_cast<int>(strlen(string_argv))) {
+ int auto_inc = (string_argv[length] == '=') ? 1 : 0;
+ *value = (T)atoi(&string_argv[length + auto_inc]);
+ }
+
+ bFound = true;
+ i = argc;
+ }
+ }
+ }
+
+ return bFound;
+}
+
+inline int getCmdLineArgumentInt(const int argc, const char **argv,
+ const char *string_ref) {
+ bool bFound = false;
+ int value = -1;
+
+ if (argc >= 1) {
+ for (int i = 1; i < argc; i++) {
+ int string_start = stringRemoveDelimiter('-', argv[i]);
+ const char *string_argv = &argv[i][string_start];
+ int length = static_cast<int>(strlen(string_ref));
+
+ if (!STRNCASECMP(string_argv, string_ref, length)) {
+ if (length + 1 <= static_cast<int>(strlen(string_argv))) {
+ int auto_inc = (string_argv[length] == '=') ? 1 : 0;
+ value = atoi(&string_argv[length + auto_inc]);
+ } else {
+ value = 0;
+ }
+
+ bFound = true;
+ continue;
+ }
+ }
+ }
+
+ if (bFound) {
+ return value;
+ } else {
+ return 0;
+ }
+}
+
+inline float getCmdLineArgumentFloat(const int argc, const char **argv,
+ const char *string_ref) {
+ bool bFound = false;
+ float value = -1;
+
+ if (argc >= 1) {
+ for (int i = 1; i < argc; i++) {
+ int string_start = stringRemoveDelimiter('-', argv[i]);
+ const char *string_argv = &argv[i][string_start];
+ int length = static_cast<int>(strlen(string_ref));
+
+ if (!STRNCASECMP(string_argv, string_ref, length)) {
+ if (length + 1 <= static_cast<int>(strlen(string_argv))) {
+ int auto_inc = (string_argv[length] == '=') ? 1 : 0;
+ value = static_cast<float>(atof(&string_argv[length + auto_inc]));
+ } else {
+ value = 0.f;
+ }
+
+ bFound = true;
+ continue;
+ }
+ }
+ }
+
+ if (bFound) {
+ return value;
+ } else {
+ return 0;
+ }
+}
+
+inline bool getCmdLineArgumentString(const int argc, const char **argv,
+ const char *string_ref,
+ char **string_retval) {
+ bool bFound = false;
+
+ if (argc >= 1) {
+ for (int i = 1; i < argc; i++) {
+ int string_start = stringRemoveDelimiter('-', argv[i]);
+ char *string_argv = const_cast<char *>(&argv[i][string_start]);
+ int length = static_cast<int>(strlen(string_ref));
+
+ if (!STRNCASECMP(string_argv, string_ref, length)) {
+ *string_retval = &string_argv[length + 1];
+ bFound = true;
+ continue;
+ }
+ }
+ }
+
+ if (!bFound) {
+ *string_retval = NULL;
+ }
+
+ return bFound;
+}
+
+//////////////////////////////////////////////////////////////////////////////
+//! Find the path for a file assuming that
+//! files are found in the searchPath.
+//!
+//! @return the path if succeeded, otherwise 0
+//! @param filename name of the file
+//! @param executable_path optional absolute path of the executable
+//////////////////////////////////////////////////////////////////////////////
+inline char *sdkFindFilePath(const char *filename,
+ const char *executable_path) {
+ // <executable_name> defines a variable that is replaced with the name of the
+ // executable
+
+ // Typical relative search paths to locate needed companion files (e.g. sample
+ // input data, or JIT source files) The origin for the relative search may be
+ // the .exe file, a .bat file launching an .exe, a browser .exe launching the
+ // .exe or .bat, etc
+ const char *searchPath[] = {
+ "./", // same dir
+ "./data/", // same dir
+ "../../../../Samples/<executable_name>/", // up 4 in tree
+ "../../../Samples/<executable_name>/", // up 3 in tree
+ "../../Samples/<executable_name>/", // up 2 in tree
+ "../../../../Samples/<executable_name>/data/", // up 4 in tree
+ "../../../Samples/<executable_name>/data/", // up 3 in tree
+ "../../Samples/<executable_name>/data/", // up 2 in tree
+ "../../../../Common/data/", // up 4 in tree
+ "../../../Common/data/", // up 3 in tree
+ "../../Common/data/" // up 2 in tree
+ };
+
+ // Extract the executable name
+ std::string executable_name;
+
+ if (executable_path != 0) {
+ executable_name = std::string(executable_path);
+
+#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
+ // Windows path delimiter
+ size_t delimiter_pos = executable_name.find_last_of('\\');
+ executable_name.erase(0, delimiter_pos + 1);
+
+ if (executable_name.rfind(".exe") != std::string::npos) {
+ // we strip .exe, only if the .exe is found
+ executable_name.resize(executable_name.size() - 4);
+ }
+
+#else
+ // Linux & OSX path delimiter
+ size_t delimiter_pos = executable_name.find_last_of('/');
+ executable_name.erase(0, delimiter_pos + 1);
+#endif
+ }
+
+ // Loop over all search paths and return the first hit
+ for (unsigned int i = 0; i < sizeof(searchPath) / sizeof(char *); ++i) {
+ std::string path(searchPath[i]);
+ size_t executable_name_pos = path.find("<executable_name>");
+
+ // If there is executable_name variable in the searchPath
+ // replace it with the value
+ if (executable_name_pos != std::string::npos) {
+ if (executable_path != 0) {
+ path.replace(executable_name_pos, strlen("<executable_name>"),
+ executable_name);
+ } else {
+ // Skip this path entry if no executable argument is given
+ continue;
+ }
+ }
+
+#ifdef _DEBUG
+ printf("sdkFindFilePath <%s> in %s\n", filename, path.c_str());
+#endif
+
+ // Test if the file exists
+ path.append(filename);
+ FILE *fp;
+ FOPEN(fp, path.c_str(), "rb");
+
+ if (fp != NULL) {
+ fclose(fp);
+ // File found
+ // returning an allocated array here for backwards compatibility reasons
+ char *file_path = reinterpret_cast<char *>(malloc(path.length() + 1));
+ STRCPY(file_path, path.length() + 1, path.c_str());
+ return file_path;
+ }
+
+ if (fp) {
+ fclose(fp);
+ }
+ }
+
+ // File not found
+ return 0;
+}
+
+#endif // COMMON_HELPER_STRING_H_
\ No newline at end of file
diff --git a/3rdParty/cuda_samples/helper_timer.h b/3rdParty/cuda_samples/helper_timer.h
new file mode 100644
index 0000000000000000000000000000000000000000..51efd720993057092323ea224377e87c95d70ff2
--- /dev/null
+++ b/3rdParty/cuda_samples/helper_timer.h
@@ -0,0 +1,465 @@
+/* Copyright (c) 2019, NVIDIA CORPORATION. All rights reserved.
+ *
+ * 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 NVIDIA CORPORATION 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 THE COPYRIGHT HOLDERS ``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 THE COPYRIGHT OWNER 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.
+ */
+
+// Helper Timing Functions
+#ifndef COMMON_HELPER_TIMER_H_
+#define COMMON_HELPER_TIMER_H_
+
+#ifndef EXIT_WAIVED
+#define EXIT_WAIVED 2
+#endif
+
+// includes, system
+#include <vector>
+
+// includes, project
+#include <exception.h>
+
+// Definition of the StopWatch Interface, this is used if we don't want to use
+// the CUT functions But rather in a self contained class interface
+class StopWatchInterface {
+public:
+ StopWatchInterface() {}
+ virtual ~StopWatchInterface() {}
+
+public:
+ //! Start time measurement
+ virtual void start() = 0;
+
+ //! Stop time measurement
+ virtual void stop() = 0;
+
+ //! Reset time counters to zero
+ virtual void reset() = 0;
+
+ //! Time in msec. after start. If the stop watch is still running (i.e. there
+ //! was no call to stop()) then the elapsed time is returned, otherwise the
+ //! time between the last start() and stop call is returned
+ virtual float getTime() = 0;
+
+ //! Mean time to date based on the number of times the stopwatch has been
+ //! _stopped_ (ie finished sessions) and the current total time
+ virtual float getAverageTime() = 0;
+};
+
+//////////////////////////////////////////////////////////////////
+// Begin Stopwatch timer class definitions for all OS platforms //
+//////////////////////////////////////////////////////////////////
+#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
+// includes, system
+#define WINDOWS_LEAN_AND_MEAN
+#include <windows.h>
+#undef min
+#undef max
+
+//! Windows specific implementation of StopWatch
+class StopWatchWin : public StopWatchInterface {
+ public:
+ //! Constructor, default
+ StopWatchWin()
+ : start_time(),
+ end_time(),
+ diff_time(0.0f),
+ total_time(0.0f),
+ running(false),
+ clock_sessions(0),
+ freq(0),
+ freq_set(false) {
+ if (!freq_set) {
+ // helper variable
+ LARGE_INTEGER temp;
+
+ // get the tick frequency from the OS
+ QueryPerformanceFrequency(reinterpret_cast<LARGE_INTEGER *>(&temp));
+
+ // convert to type in which it is needed
+ freq = (static_cast<double>(temp.QuadPart)) / 1000.0;
+
+ // rememeber query
+ freq_set = true;
+ }
+ }
+
+ // Destructor
+ ~StopWatchWin() {}
+
+ public:
+ //! Start time measurement
+ inline void start();
+
+ //! Stop time measurement
+ inline void stop();
+
+ //! Reset time counters to zero
+ inline void reset();
+
+ //! Time in msec. after start. If the stop watch is still running (i.e. there
+ //! was no call to stop()) then the elapsed time is returned, otherwise the
+ //! time between the last start() and stop call is returned
+ inline float getTime();
+
+ //! Mean time to date based on the number of times the stopwatch has been
+ //! _stopped_ (ie finished sessions) and the current total time
+ inline float getAverageTime();
+
+ private:
+ // member variables
+
+ //! Start of measurement
+ LARGE_INTEGER start_time;
+ //! End of measurement
+ LARGE_INTEGER end_time;
+
+ //! Time difference between the last start and stop
+ float diff_time;
+
+ //! TOTAL time difference between starts and stops
+ float total_time;
+
+ //! flag if the stop watch is running
+ bool running;
+
+ //! Number of times clock has been started
+ //! and stopped to allow averaging
+ int clock_sessions;
+
+ //! tick frequency
+ double freq;
+
+ //! flag if the frequency has been set
+ bool freq_set;
+};
+
+// functions, inlined
+
+////////////////////////////////////////////////////////////////////////////////
+//! Start time measurement
+////////////////////////////////////////////////////////////////////////////////
+inline void StopWatchWin::start() {
+ QueryPerformanceCounter(reinterpret_cast<LARGE_INTEGER *>(&start_time));
+ running = true;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Stop time measurement and increment add to the current diff_time summation
+//! variable. Also increment the number of times this clock has been run.
+////////////////////////////////////////////////////////////////////////////////
+inline void StopWatchWin::stop() {
+ QueryPerformanceCounter(reinterpret_cast<LARGE_INTEGER *>(&end_time));
+ diff_time = static_cast<float>(((static_cast<double>(end_time.QuadPart) -
+ static_cast<double>(start_time.QuadPart)) /
+ freq));
+
+ total_time += diff_time;
+ clock_sessions++;
+ running = false;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Reset the timer to 0. Does not change the timer running state but does
+//! recapture this point in time as the current start time if it is running.
+////////////////////////////////////////////////////////////////////////////////
+inline void StopWatchWin::reset() {
+ diff_time = 0;
+ total_time = 0;
+ clock_sessions = 0;
+
+ if (running) {
+ QueryPerformanceCounter(reinterpret_cast<LARGE_INTEGER *>(&start_time));
+ }
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Time in msec. after start. If the stop watch is still running (i.e. there
+//! was no call to stop()) then the elapsed time is returned added to the
+//! current diff_time sum, otherwise the current summed time difference alone
+//! is returned.
+////////////////////////////////////////////////////////////////////////////////
+inline float StopWatchWin::getTime() {
+ // Return the TOTAL time to date
+ float retval = total_time;
+
+ if (running) {
+ LARGE_INTEGER temp;
+ QueryPerformanceCounter(reinterpret_cast<LARGE_INTEGER *>(&temp));
+ retval += static_cast<float>(((static_cast<double>(temp.QuadPart) -
+ static_cast<double>(start_time.QuadPart)) /
+ freq));
+ }
+
+ return retval;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Time in msec. for a single run based on the total number of COMPLETED runs
+//! and the total time.
+////////////////////////////////////////////////////////////////////////////////
+inline float StopWatchWin::getAverageTime() {
+ return (clock_sessions > 0) ? (total_time / clock_sessions) : 0.0f;
+}
+#else
+// Declarations for Stopwatch on Linux and Mac OSX
+// includes, system
+#include <sys/time.h>
+#include <ctime>
+
+//! Windows specific implementation of StopWatch
+class StopWatchLinux : public StopWatchInterface {
+public:
+ //! Constructor, default
+ StopWatchLinux()
+ : start_time(),
+ diff_time(0.0),
+ total_time(0.0),
+ running(false),
+ clock_sessions(0) {}
+
+ // Destructor
+ virtual ~StopWatchLinux() {}
+
+public:
+ //! Start time measurement
+ inline void start();
+
+ //! Stop time measurement
+ inline void stop();
+
+ //! Reset time counters to zero
+ inline void reset();
+
+ //! Time in msec. after start. If the stop watch is still running (i.e. there
+ //! was no call to stop()) then the elapsed time is returned, otherwise the
+ //! time between the last start() and stop call is returned
+ inline float getTime();
+
+ //! Mean time to date based on the number of times the stopwatch has been
+ //! _stopped_ (ie finished sessions) and the current total time
+ inline float getAverageTime();
+
+private:
+ // helper functions
+
+ //! Get difference between start time and current time
+ inline float getDiffTime();
+
+private:
+ // member variables
+
+ //! Start of measurement
+ struct timeval start_time;
+
+ //! Time difference between the last start and stop
+ float diff_time;
+
+ //! TOTAL time difference between starts and stops
+ float total_time;
+
+ //! flag if the stop watch is running
+ bool running;
+
+ //! Number of times clock has been started
+ //! and stopped to allow averaging
+ int clock_sessions;
+};
+
+// functions, inlined
+
+////////////////////////////////////////////////////////////////////////////////
+//! Start time measurement
+////////////////////////////////////////////////////////////////////////////////
+inline void StopWatchLinux::start() {
+ gettimeofday(&start_time, 0);
+ running = true;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Stop time measurement and increment add to the current diff_time summation
+//! variable. Also increment the number of times this clock has been run.
+////////////////////////////////////////////////////////////////////////////////
+inline void StopWatchLinux::stop() {
+ diff_time = getDiffTime();
+ total_time += diff_time;
+ running = false;
+ clock_sessions++;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Reset the timer to 0. Does not change the timer running state but does
+//! recapture this point in time as the current start time if it is running.
+////////////////////////////////////////////////////////////////////////////////
+inline void StopWatchLinux::reset() {
+ diff_time = 0;
+ total_time = 0;
+ clock_sessions = 0;
+
+ if (running) {
+ gettimeofday(&start_time, 0);
+ }
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Time in msec. after start. If the stop watch is still running (i.e. there
+//! was no call to stop()) then the elapsed time is returned added to the
+//! current diff_time sum, otherwise the current summed time difference alone
+//! is returned.
+////////////////////////////////////////////////////////////////////////////////
+inline float StopWatchLinux::getTime() {
+ // Return the TOTAL time to date
+ float retval = total_time;
+
+ if (running) {
+ retval += getDiffTime();
+ }
+
+ return retval;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Time in msec. for a single run based on the total number of COMPLETED runs
+//! and the total time.
+////////////////////////////////////////////////////////////////////////////////
+inline float StopWatchLinux::getAverageTime() {
+ return (clock_sessions > 0) ? (total_time / clock_sessions) : 0.0f;
+}
+////////////////////////////////////////////////////////////////////////////////
+
+////////////////////////////////////////////////////////////////////////////////
+inline float StopWatchLinux::getDiffTime() {
+ struct timeval t_time;
+ gettimeofday(&t_time, 0);
+
+ // time difference in milli-seconds
+ return static_cast<float>(1000.0 * (t_time.tv_sec - start_time.tv_sec) +
+ (0.001 * (t_time.tv_usec - start_time.tv_usec)));
+}
+#endif // WIN32
+
+////////////////////////////////////////////////////////////////////////////////
+//! Timer functionality exported
+
+////////////////////////////////////////////////////////////////////////////////
+//! Create a new timer
+//! @return true if a time has been created, otherwise false
+//! @param name of the new timer, 0 if the creation failed
+////////////////////////////////////////////////////////////////////////////////
+inline bool sdkCreateTimer(StopWatchInterface **timer_interface) {
+// printf("sdkCreateTimer called object %08x\n", (void *)*timer_interface);
+#if defined(WIN32) || defined(_WIN32) || defined(WIN64) || defined(_WIN64)
+ *timer_interface = reinterpret_cast<StopWatchInterface *>(new StopWatchWin());
+#else
+ *timer_interface =
+ reinterpret_cast<StopWatchInterface *>(new StopWatchLinux());
+#endif
+ return (*timer_interface != NULL) ? true : false;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Delete a timer
+//! @return true if a time has been deleted, otherwise false
+//! @param name of the timer to delete
+////////////////////////////////////////////////////////////////////////////////
+inline bool sdkDeleteTimer(StopWatchInterface **timer_interface) {
+ // printf("sdkDeleteTimer called object %08x\n", (void *)*timer_interface);
+ if (*timer_interface) {
+ delete *timer_interface;
+ *timer_interface = NULL;
+ }
+
+ return true;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Start the time with name \a name
+//! @param name name of the timer to start
+////////////////////////////////////////////////////////////////////////////////
+inline bool sdkStartTimer(StopWatchInterface **timer_interface) {
+ // printf("sdkStartTimer called object %08x\n", (void *)*timer_interface);
+ if (*timer_interface) {
+ (*timer_interface)->start();
+ }
+
+ return true;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Stop the time with name \a name. Does not reset.
+//! @param name name of the timer to stop
+////////////////////////////////////////////////////////////////////////////////
+inline bool sdkStopTimer(StopWatchInterface **timer_interface) {
+ // printf("sdkStopTimer called object %08x\n", (void *)*timer_interface);
+ if (*timer_interface) {
+ (*timer_interface)->stop();
+ }
+
+ return true;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Resets the timer's counter.
+//! @param name name of the timer to reset.
+////////////////////////////////////////////////////////////////////////////////
+inline bool sdkResetTimer(StopWatchInterface **timer_interface) {
+ // printf("sdkResetTimer called object %08x\n", (void *)*timer_interface);
+ if (*timer_interface) {
+ (*timer_interface)->reset();
+ }
+
+ return true;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Return the average time for timer execution as the total time
+//! for the timer dividied by the number of completed (stopped) runs the timer
+//! has made.
+//! Excludes the current running time if the timer is currently running.
+//! @param name name of the timer to return the time of
+////////////////////////////////////////////////////////////////////////////////
+inline float sdkGetAverageTimerValue(StopWatchInterface **timer_interface) {
+ // printf("sdkGetAverageTimerValue called object %08x\n", (void
+ // *)*timer_interface);
+ if (*timer_interface) {
+ return (*timer_interface)->getAverageTime();
+ } else {
+ return 0.0f;
+ }
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! Total execution time for the timer over all runs since the last reset
+//! or timer creation.
+//! @param name name of the timer to obtain the value of.
+////////////////////////////////////////////////////////////////////////////////
+inline float sdkGetTimerValue(StopWatchInterface **timer_interface) {
+ // printf("sdkGetTimerValue called object %08x\n", (void *)*timer_interface);
+ if (*timer_interface) {
+ return (*timer_interface)->getTime();
+ } else {
+ return 0.0f;
+ }
+}
+
+#endif // COMMON_HELPER_TIMER_H_
\ No newline at end of file
diff --git a/3rdParty/googletest/CMakeLists.txt b/3rdParty/googletest/CMakeLists.txt
index acc5fb1ed8d37bcc89e9e60aceb28c7400e7cfca..781b60be44fc76ec640fea0444a527c10125e141 100644
--- a/3rdParty/googletest/CMakeLists.txt
+++ b/3rdParty/googletest/CMakeLists.txt
@@ -1,7 +1,7 @@
# Note: CMake support is community-based. The maintainers do not use CMake
# internally.
-cmake_minimum_required(VERSION 2.8.8)
+cmake_minimum_required(VERSION 2.8.12)
if (POLICY CMP0048)
cmake_policy(SET CMP0048 NEW)
diff --git a/3rdParty/googletest/googlemock/CMakeLists.txt b/3rdParty/googletest/googlemock/CMakeLists.txt
index d32b70b5be0e0ae74f5376fb03a2226065ad599a..63cd3c61441ee2ba9177b6b366c069013b0ff4ad 100644
--- a/3rdParty/googletest/googlemock/CMakeLists.txt
+++ b/3rdParty/googletest/googlemock/CMakeLists.txt
@@ -42,7 +42,7 @@ else()
cmake_policy(SET CMP0048 NEW)
project(gmock VERSION ${GOOGLETEST_VERSION} LANGUAGES CXX C)
endif()
-cmake_minimum_required(VERSION 2.6.4)
+cmake_minimum_required(VERSION 2.8.12)
if (COMMAND set_up_hermetic_build)
set_up_hermetic_build()
diff --git a/3rdParty/googletest/googletest/CMakeLists.txt b/3rdParty/googletest/googletest/CMakeLists.txt
index db292946a59453e09929229c1fbdb3701f2bd6ab..0ef01d22e7216988d86f728a27309d6494e36b1c 100644
--- a/3rdParty/googletest/googletest/CMakeLists.txt
+++ b/3rdParty/googletest/googletest/CMakeLists.txt
@@ -53,7 +53,7 @@ else()
cmake_policy(SET CMP0048 NEW)
project(gtest VERSION ${GOOGLETEST_VERSION} LANGUAGES CXX C)
endif()
-cmake_minimum_required(VERSION 2.6.4)
+cmake_minimum_required(VERSION 2.8.12)
if (POLICY CMP0063) # Visibility
cmake_policy(SET CMP0063 NEW)
diff --git a/3rdParty/metis/metis-5.1.1/GKlib/GKlibSystem.cmake b/3rdParty/metis/metis-5.1.1/GKlib/GKlibSystem.cmake
index d83b2083c176a3addb3ddb951fd0e44923b18aa6..b8478c5d06fd5f7f1347ef8da75073a90fa5faa1 100644
--- a/3rdParty/metis/metis-5.1.1/GKlib/GKlibSystem.cmake
+++ b/3rdParty/metis/metis-5.1.1/GKlib/GKlibSystem.cmake
@@ -61,7 +61,9 @@ if(GDB)
set(GKlib_COPTS "${GKlib_COPTS} -g")
set(GKlib_COPTIONS "${GKlib_COPTIONS} -Werror")
else()
- set(GKlib_COPTS "-O3")
+ if(NOT MSVC)
+ set(GKlib_COPTS "-O3")
+ endif()
endif(GDB)
diff --git a/3rdParty/metis/metis-5.1.1/libmetis/CMakeLists.txt b/3rdParty/metis/metis-5.1.1/libmetis/CMakeLists.txt
index 4732b645ea354ada4a61540e12f73bff90540cb5..802241df51949274173b3f647cf6a54615947eb5 100644
--- a/3rdParty/metis/metis-5.1.1/libmetis/CMakeLists.txt
+++ b/3rdParty/metis/metis-5.1.1/libmetis/CMakeLists.txt
@@ -6,10 +6,12 @@ file(GLOB metis_sources *.c)
add_library(metis ${METIS_LIBRARY_TYPE} ${GKlib_sources} ${metis_sources})
if(UNIX)
target_link_libraries(metis m)
+
+ target_compile_options(metis PRIVATE "-Wno-format")
endif()
if(MSVC)
- target_compile_options(metis PRIVATE "/w")
+ target_compile_options(metis PRIVATE "/W0")
endif()
if(METIS_INSTALL)
diff --git a/CMake/3rd.cmake b/CMake/3rd.cmake
index 6cc488f94716f7cec973874b0930ed8f9b719d08..781146111d48739671b35c98bb96ebff358809b4 100644
--- a/CMake/3rd.cmake
+++ b/CMake/3rd.cmake
@@ -1,5 +1,2 @@
include(${VF_CMAKE_DIR}/3rd/boost.cmake)
-include(${VF_CMAKE_DIR}/3rd/cuda.cmake)
include(${VF_CMAKE_DIR}/3rd/gmock.cmake)
-include(${VF_CMAKE_DIR}/3rd/mpi.cmake)
-include(${VF_CMAKE_DIR}/3rd/OpenMP.cmake)
\ No newline at end of file
diff --git a/CMake/3rd/OpenMP.cmake b/CMake/3rd/OpenMP.cmake
deleted file mode 100644
index 45465ba4bf43dd5cf7687b83ac1d368332614582..0000000000000000000000000000000000000000
--- a/CMake/3rd/OpenMP.cmake
+++ /dev/null
@@ -1,13 +0,0 @@
-function (linkOpenMP targetName)
-
- if(NOT USE_OPENMP)
- return()
- endif()
-
- find_package(OpenMP REQUIRED)
-
- if(OpenMP_CXX_FOUND)
- target_link_libraries(${targetName} PUBLIC OpenMP::OpenMP_CXX)
- endif()
-
-endfunction()
\ No newline at end of file
diff --git a/CMake/3rd/boost.cmake b/CMake/3rd/boost.cmake
index 544ae2d97b3b8ef0277445cf252149986c8dfb3a..74f6f165cda1ef93f64dd326c961b2663bc6ea67 100644
--- a/CMake/3rd/boost.cmake
+++ b/CMake/3rd/boost.cmake
@@ -1,4 +1,10 @@
-function(linkBoost components)
+function(linkBoost)
+
+ set( options )
+ set( oneValueArgs )
+ set( multiValueArgs COMPONENTS)
+ cmake_parse_arguments( ARG "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN} )
+
if(BUILD_SHARED_LIBS)
if (WIN32)
set(Boost_USE_STATIC_LIBS ON)
@@ -8,7 +14,9 @@ function(linkBoost components)
set(Boost_USE_STATIC_RUNTIME OFF)
else()
set(Boost_USE_STATIC_LIBS ON)
- set(Boost_USE_STATIC_RUNTIME ON)
+ if(WIN32)
+ set(Boost_USE_STATIC_RUNTIME ON)
+ endif()
endif()
set(Boost_USE_MULTITHREADED ON)
@@ -18,8 +26,16 @@ function(linkBoost components)
# add_definitions( -DBOOST_ALL_DYN_LINK )
endif()
- vf_get_library_name(library_name)
- find_package( Boost REQUIRED COMPONENTS ${components})
+ vf_get_library_name(library_name)
+ if(DEFINED ARG_COMPONENTS)
+ find_package( Boost REQUIRED COMPONENTS ${ARG_COMPONENTS})
+ target_link_libraries(${library_name} PRIVATE ${Boost_LIBRARIES})
+ message("here")
+ else()
+ find_package( Boost REQUIRED)
+ message("or here")
+ endif()
+
+
target_include_directories(${library_name} PRIVATE ${Boost_INCLUDE_DIR})
- target_link_libraries(${library_name} PRIVATE ${Boost_LIBRARIES})
endfunction()
diff --git a/CMake/3rd/cuda.cmake b/CMake/3rd/cuda.cmake
deleted file mode 100644
index 83acbdcc858e87e04d46d5a33e9b64b11c0e20e4..0000000000000000000000000000000000000000
--- a/CMake/3rd/cuda.cmake
+++ /dev/null
@@ -1,21 +0,0 @@
-
-function(linkCUDA)
-
- find_path(CUDA_CUT_INCLUDE_DIR
- helper_cuda.h
- PATHS "$ENV{NVCUDASAMPLES_ROOT}" "${NVCUDASAMPLES_ROOT}"
- PATH_SUFFIXES "common/inc" "Common"
- DOC "Location of helper_cuda.h"
- NO_DEFAULT_PATH
- )
-
- vf_get_library_name(library_name)
- target_include_directories(${library_name} PRIVATE ${CUDA_CUT_INCLUDE_DIR})
- target_include_directories(${library_name} PRIVATE ${CMAKE_CUDA_TOOLKIT_INCLUDE_DIRECTORIES})
-
- message(${CMAKE_CUDA_TOOLKIT_INCLUDE_DIRECTORIES})
-
- # set the following properties only for specific targets
- # set_property(TARGET ${targetName} PROPERTY CUDA_SEPARABLE_COMPILATION ON)
- # set_property(TARGET ${targetName} PROPERTY CUDA_64_BIT_DEVICE_CODE ON)
-endfunction()
\ No newline at end of file
diff --git a/CMake/3rd/mpi.cmake b/CMake/3rd/mpi.cmake
deleted file mode 100644
index 93b3ee386cc46623ed02bbbffb88996328f03e52..0000000000000000000000000000000000000000
--- a/CMake/3rd/mpi.cmake
+++ /dev/null
@@ -1,9 +0,0 @@
-function (linkMPI)
-
- find_package(MPI REQUIRED)
-
- vf_get_library_name(library_name)
- target_include_directories(${library_name} PUBLIC ${MPI_CXX_INCLUDE_PATH})
- target_link_libraries(${library_name} PRIVATE MPI::MPI_CXX)
-
-endfunction()
\ No newline at end of file
diff --git a/CMake/CMakeSetCompilerFlags.cmake b/CMake/CMakeSetCompilerFlags.cmake
index 4165eeff8b8cce1e89d19beffefe7507496fee49..784f3f24a7cde518e113363aa92ce90fec0e9c2d 100644
--- a/CMake/CMakeSetCompilerFlags.cmake
+++ b/CMake/CMakeSetCompilerFlags.cmake
@@ -1,29 +1,21 @@
-###############################################################
-# set hostname -> CAB_MACHINE and load an optional config file
-###############################################################
+#########################################################################################
+## Access the hostname and loads a optional machine file hostname.cmake
+#########################################################################################
macro(loadMachineFile)
- IF(NOT CAB_MACHINE)
- SET(CAB_MACHINE $ENV{CAB_MACHINE})
+ site_name(MACHINE_NAME)
+ string(TOUPPER "${MACHINE_NAME}" MACHINE_NAME)
- IF( CAB_MACHINE )
- STRING(TOUPPER "${CAB_MACHINE}" CAB_MACHINE)
- ELSE()
- EXECUTE_PROCESS( COMMAND hostname OUTPUT_VARIABLE CAB_MACHINE)
- STRING(REGEX REPLACE "[ ]*([A-Za-z0-9]+).*[\\\\n]*" "\\1" CAB_MACHINE "${CAB_MACHINE}" )
- STRING(TOUPPER "${CAB_MACHINE}" CAB_MACHINE)
- ENDIF()
- ENDIF()
+ set(BUILD_MACHINE_FILE_PATH "${VF_CMAKE_DIR}/cmake_config_files")
- LIST(APPEND VF_COMPILER_DEFINITION CAB_MACHINE=${CAB_MACHINE})
- SET(CMAKE_CONFIG_FILE "${VF_CMAKE_DIR}/cmake_config_files/${CAB_MACHINE}.config.cmake")
+ set(MACHINE_FILE "${BUILD_MACHINE_FILE_PATH}/${MACHINE_NAME}.config.cmake")
- IF(NOT EXISTS ${CMAKE_CONFIG_FILE})
- status("No configuration file found for machine: ${CAB_MACHINE}.")
+ IF(NOT EXISTS ${MACHINE_FILE})
+ status("No configuration file found: ${MACHINE_FILE}.")
ELSE()
- status("Load configuration file ${CAB_MACHINE}.config.cmake")
- include(${CMAKE_CONFIG_FILE})
+ status("Load configuration file: ${MACHINE_FILE}")
+ include(${MACHINE_FILE})
ENDIF()
endmacro()
@@ -35,9 +27,9 @@ endmacro()
################################################################
macro(loadCompilerFlags)
- SET(CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "")
- SET(CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "")
- SET(CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_RELEASE "")
+ SET(CS_COMPILER_FLAGS_CXX "")
+ SET(CS_COMPILER_FLAGS_CXX_DEBUG "")
+ SET(CS_COMPILER_FLAGS_CXX_RELEASE "")
# https://cmake.org/cmake/help/latest/variable/CMAKE_LANG_COMPILER_ID.html#variable:CMAKE_<LANG>_COMPILER_ID
@@ -57,9 +49,9 @@ endmacro()
################################################################
function(addAdditionalFlags project_name)
- status_lib("additional compiler flags CXX: ${CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS}")
- status_lib("additional compiler flags CXX debug: ${CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG}")
- status_lib("additional compiler flags CXX release: ${CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_RELEASE}")
+ status_lib("additional compiler flags CXX: ${CS_COMPILER_FLAGS_CXX}")
+ status_lib("additional compiler flags CXX debug: ${CS_COMPILER_FLAGS_CXX_DEBUG}")
+ status_lib("additional compiler flags CXX release: ${CS_COMPILER_FLAGS_CXX_RELEASE}")
status_lib("additional compiler definitions: ${VF_COMPILER_DEFINITION}")
status_lib("additional linker flags: ${VF_LINK_OPTIONS}")
@@ -74,15 +66,15 @@ function(addAdditionalFlags project_name)
endforeach()
# compile options
- foreach(flag IN LISTS CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS)
+ foreach(flag IN LISTS CS_COMPILER_FLAGS_CXX)
target_compile_options(${project_name} PRIVATE "$<$<COMPILE_LANGUAGE:CXX>:${flag}>")
endforeach()
- foreach(flag IN LISTS CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG)
+ foreach(flag IN LISTS CS_COMPILER_FLAGS_CXX_DEBUG)
target_compile_options(${project_name} PRIVATE "$<$<AND:$<COMPILE_LANGUAGE:CXX>,$<CONFIG:DEBUG>>:${flag}>")
endforeach()
- foreach(flag IN LISTS CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_RELEASE)
+ foreach(flag IN LISTS CS_COMPILER_FLAGS_CXX_RELEASE)
target_compile_options(${project_name} PRIVATE "$<$<AND:$<COMPILE_LANGUAGE:CXX>,$<CONFIG:RELEASE>>:${flag}>")
endforeach()
diff --git a/CMake/VirtualFluidsMacros.cmake b/CMake/VirtualFluidsMacros.cmake
index 69e56fa63d4f25adb84e539678055e7f46b62d3d..debb5ee7826d7dc3d6499b3813070dcab4b94bd4 100644
--- a/CMake/VirtualFluidsMacros.cmake
+++ b/CMake/VirtualFluidsMacros.cmake
@@ -112,6 +112,15 @@ function(vf_add_library)
else()
vf_get_library_name (library_name)
endif()
+
+ if(NOT DEFINED ARG_BUILDTYPE)
+ if(BUILD_SHARED_LIBS)
+ set(ARG_BUILDTYPE "shared")
+ else()
+ set(ARG_BUILDTYPE "static")
+ endif()
+ endif()
+
status("Configuring the target: ${library_name} (type=${ARG_BUILDTYPE})...")
@@ -216,6 +225,10 @@ function(vf_add_library)
target_include_directories(${library_name} PRIVATE ${VF_SRC_DIR}/gpu)
target_include_directories(${library_name} PRIVATE ${VF_SRC_DIR}/cpu)
+ if(BUILD_VF_GPU)
+ target_include_directories(${library_name} PRIVATE "${VF_THIRD_DIR}/cuda_samples/")
+ target_include_directories(${library_name} PRIVATE ${CMAKE_CUDA_TOOLKIT_INCLUDE_DIRECTORIES})
+ endif()
status("... configuring target: ${library_name} (type=${ARG_BUILDTYPE}) done")
diff --git a/CMake/cmake_config_files/PHOENIX.config.cmake b/CMake/cmake_config_files/PHOENIX.config.cmake
index 1d69df88bc174c206fb3639b7d67e703dd41d5eb..2f576538c106a6a4d83509a49a1408a8d63efbdb 100644
--- a/CMake/cmake_config_files/PHOENIX.config.cmake
+++ b/CMake/cmake_config_files/PHOENIX.config.cmake
@@ -43,4 +43,5 @@ SET(BOOST_LIBRARYDIR "/cluster/lib/boost/1.63.0/gcc/lib" CACHE PATH "BOOST_LIB
#SET(VTK_DIR "/home/irmb/tools/VTK/build/VTK-8.2.0" CACHE PATH "VTK directory override" FORCE)
#SET(VTK_DIR "/home/stelenz/software/vtk/VTK-8.1.0/build" CACHE PATH "VTK directory override" FORCE)
-set(NVCUDASAMPLES_ROOT "/cluster/cuda/11.0/samples")
+## nvidia
+set(CMAKE_CUDA_ARCHITECTURES 60) # NVIDIA Tesla P100
\ No newline at end of file
diff --git a/CMake/compilerflags/AppleClang.cmake b/CMake/compilerflags/AppleClang.cmake
index aecc48fa5398c6418c2d47f5af243659128050bc..6f52ad35956a967b3892aed26257fa1c41cb0c46 100644
--- a/CMake/compilerflags/AppleClang.cmake
+++ b/CMake/compilerflags/AppleClang.cmake
@@ -3,20 +3,26 @@
#############################################################################################################
# debug
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-g") # generates debug information. Works best with -O0.
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-O0")
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-g") # generates debug information. Works best with -O0.
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-O0")
# release
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_RELEASE "-O3") # optimization level (-O3: most optimization which also could result in larger binaries)
+list(APPEND CS_COMPILER_FLAGS_CXX_RELEASE "-O3") # optimization level (-O3: most optimization which also could result in larger binaries)
# all
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-fPIC") # position independent code for shared libraries
+list(APPEND CS_COMPILER_FLAGS_CXX "-fPIC") # position independent code for shared libraries
#############################################################################################################
# warnings
#############################################################################################################
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wall")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wunreachable-code")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wall")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wextra")
+list(APPEND CS_COMPILER_FLAGS_CXX "-pedantic")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-unused-function")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-reorder")
+if(BUILD_WARNINGS_AS_ERRORS)
+ list(APPEND CS_COMPILER_FLAGS_CXX -Werror)
+endif()
+
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unused-function")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unused-parameter")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-reorder")
diff --git a/CMake/compilerflags/Clang.cmake b/CMake/compilerflags/Clang.cmake
index 434be42697165ea347b2c2728e9cf836b66650af..2eb4eec5ee89715a1668ef078c84f4e720bd04e6 100644
--- a/CMake/compilerflags/Clang.cmake
+++ b/CMake/compilerflags/Clang.cmake
@@ -3,22 +3,31 @@
#############################################################################################################
# debug
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-g") # generates debug information. Works best with -O0.
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-O0")
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-g") # generates debug information. Works best with -O0.
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-O0")
# release
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_RELEASE "-O3") # optimization level (-O3: most optimization which also could result in larger binaries)
+list(APPEND CS_COMPILER_FLAGS_CXX_RELEASE "-O3") # optimization level (-O3: most optimization which also could result in larger binaries)
# all
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-fPIC") # position independent code for shared libraries
+list(APPEND CS_COMPILER_FLAGS_CXX "-fPIC") # position independent code for shared libraries
#############################################################################################################
# warnings
#############################################################################################################
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wall")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-unused-function")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-reorder-ctor")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wall")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wextra")
+list(APPEND CS_COMPILER_FLAGS_CXX "-pedantic")
+
+if(BUILD_WARNINGS_AS_ERRORS)
+ list(APPEND CS_COMPILER_FLAGS_CXX -Werror)
+endif()
+
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unused-function")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unused-parameter")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-reorder-ctor")
+
#############################################################################################################
diff --git a/CMake/compilerflags/GNU.cmake b/CMake/compilerflags/GNU.cmake
index 211f2a040b9c6c71d632fbee8621377d1df63350..3e67b79e9c84d29b51b2881b17e2b74f5510bbd2 100644
--- a/CMake/compilerflags/GNU.cmake
+++ b/CMake/compilerflags/GNU.cmake
@@ -3,33 +3,41 @@
#############################################################################################################
# debug
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-g") # generates debug information. Works best with -O0.
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-O0") # no optimization
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-g") # generates debug information. Works best with -O0.
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-O0") # no optimization
# release
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_RELEASE "-O3") # optimization level (-O3: most optimization which also could result in larger binaries)
+list(APPEND CS_COMPILER_FLAGS_CXX_RELEASE "-O3") # optimization level (-O3: most optimization which also could result in larger binaries)
# all
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-fPIC") # position independent code for shared libraries
+list(APPEND CS_COMPILER_FLAGS_CXX "-fPIC") # position independent code for shared libraries
if(NOT BUILD_VF_INCLUDE_WHAT_YOU_USE) # optimization flag '-funroll-all-loops' is not supported for IWYU
- LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-funroll-all-loops")
+ LIST(APPEND CS_COMPILER_FLAGS_CXX "-funroll-all-loops")
endif()
# gcov
if (BUILD_VF_COVERAGE)
- list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "--coverage")
+ list(APPEND CS_COMPILER_FLAGS_CXX "--coverage")
set(CMAKE_EXE_LINKER_FLAGS ${CMAKE_EXE_LINKER_FLAGS} " --coverage")
endif()
#############################################################################################################
# warnings
#############################################################################################################
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wall")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-unused-function")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-reorder")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-sign-compare")
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-Wno-unknown-pragmas")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wall")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wextra")
+list(APPEND CS_COMPILER_FLAGS_CXX "-pedantic")
+
+if(BUILD_WARNINGS_AS_ERRORS)
+ list(APPEND CS_COMPILER_FLAGS_CXX -Werror)
+endif()
+
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unused-function")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unused-parameter")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-reorder")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-unknown-pragmas")
+list(APPEND CS_COMPILER_FLAGS_CXX "-Wno-cast-function-type")
#############################################################################################################
# linker options
diff --git a/CMake/compilerflags/Intel.cmake b/CMake/compilerflags/Intel.cmake
index a53998d93eec146de6158d80bad2302536b2c252..c7177d4c246402d8b865e7059273cb18d0938e44 100644
--- a/CMake/compilerflags/Intel.cmake
+++ b/CMake/compilerflags/Intel.cmake
@@ -2,22 +2,22 @@
# compiler flags
#############################################################################################################
-#LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-O")
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-wd654")
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-wd1125") #virtual function override intended
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-wd1224") #warning directive: This file includes at least one deprecated or antiquated header
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-wd377") #class "std::auto_ptr<RCF::I_ClientTransport>" has no suitable copy constructor
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-wd327") #class "std::auto_ptr<RCF::I_ClientTransport>" has no suitable copy constructor
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-wd327") #class "std::auto_ptr<RCF::I_ClientTransport>" has no suitable copy constructor
-#~
-#~ LIST(APPEND CAB_COMPILER_ADDTIONAL_C_COMPILER_FLAGS "-wd266") #function "__GKfree" declared implicitly
-#LIST(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-xHOST -O3 -ip -ipo -fno-alias -mcmodel=medium -qopt-streaming-stores=always")
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-O")
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-wd654")
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-wd1125") #virtual function override intended
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-wd1224") #warning directive: This file includes at least one deprecated or antiquated header
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-wd377") #class "std::auto_ptr<RCF::I_ClientTransport>" has no suitable copy constructor
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-wd327") #class "std::auto_ptr<RCF::I_ClientTransport>" has no suitable copy constructor
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-wd327") #class "std::auto_ptr<RCF::I_ClientTransport>" has no suitable copy constructor
+#
+#LIST(APPEND CAB_COMPILER_ADDTIONAL_C_COMPILER_FLAGS "-wd266") #function "__GKfree" declared implicitly
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-xHOST -O3 -ip -ipo -fno-alias -mcmodel=medium -qopt-streaming-stores=always")
# all
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-xHOST;-O3;-ip;-fno-alias;-mcmodel=medium;-qopt-streaming-stores=always;-xCORE-AVX512;-qopt-zmm-usage=high")
+list(APPEND CS_COMPILER_FLAGS_CXX "-xHOST;-O3;-ip;-fno-alias;-mcmodel=medium;-qopt-streaming-stores=always;-xCORE-AVX512;-qopt-zmm-usage=high")
# debug
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS_DEBUG "-g -traceback")
+list(APPEND CS_COMPILER_FLAGS_CXX_DEBUG "-g -traceback")
#############################################################################################################
diff --git a/CMake/compilerflags/MSVC.cmake b/CMake/compilerflags/MSVC.cmake
index 937d9f4d5d1fead78f35e2442845b27f13bc9a54..2af38d98e63cf04c4da476fb02754ce47510e4f6 100644
--- a/CMake/compilerflags/MSVC.cmake
+++ b/CMake/compilerflags/MSVC.cmake
@@ -1,29 +1,29 @@
#############################################################################################################
# compiler flags
#############################################################################################################
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/bigobj") # increases that address capacity to 4,294,967,296 (2^32).
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "-MP") # enable multi-threaded compiling
+list(APPEND CS_COMPILER_FLAGS_CXX "/bigobj") # increases that address capacity to 4,294,967,296 (2^32).
+list(APPEND CS_COMPILER_FLAGS_CXX "-MP") # enable multi-threaded compiling
#############################################################################################################
# warnings
#############################################################################################################
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/W4") # highest warning level
+list(APPEND CS_COMPILER_FLAGS_CXX "/W4") # highest warning level
# With W4 the following warnings appear many times. As long they are not eliminated they are suppressed:
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4458") # C4458: declaration of 'XXX' hides class member
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4100") # C4100: 'XXX': unreferenced formal parameter
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4505") # C4505: 'XXX': unreferenced local function has been removed
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4244") # C4244: '=': conversion from 'int' to 'char', possible loss of data, triggered by algorithm(2216,24)
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4310") # C4310: cast truncates constant value, triggerd by muParserbase.h
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4127") # C4127: conditional expression is constant: e.g. sizeof(int)
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4458") # C4458: declaration of 'XXX' hides class member
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4100") # C4100: 'XXX': unreferenced formal parameter
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4505") # C4505: 'XXX': unreferenced local function has been removed
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4244") # C4244: '=': conversion from 'int' to 'char', possible loss of data, triggered by algorithm(2216,24)
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4310") # C4310: cast truncates constant value, triggerd by muParserbase.h
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4127") # C4127: conditional expression is constant: e.g. sizeof(int)
# Urgent FIXME: This warning should be activated and fixed:
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4701") # C4701: potentially uninitialized local variable 'lMaxX3' used
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4701") # C4701: potentially uninitialized local variable 'lMaxX3' used
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4251") # disable needs to have dll interface
-list(APPEND CAB_COMPILER_ADDTIONAL_CXX_COMPILER_FLAGS "/wd4005") # disable macro redefinition (triggered by metis.h)
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4251") # disable needs to have dll interface
+list(APPEND CS_COMPILER_FLAGS_CXX "/wd4005") # disable macro redefinition (triggered by metis.h)
#############################################################################################################
# preprocessor definitions
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 50073d8a0016c0345e9ae7ae3cfc5507afb6ad14..efce16caaf8d40ae3f5a9ca9bcedd5c5a6767bf5 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -8,17 +8,15 @@
#################################################################################
# required cmake versions
# CMAKE 3.13: target_link_options
+# CMAKE 3.15: CMAKE_MSVC_RUNTIME_LIBRARY
#################################################################################
-cmake_minimum_required(VERSION 3.13..3.18 FATAL_ERROR)
+cmake_minimum_required(VERSION 3.15..3.19 FATAL_ERROR)
project(VirtualFluids CXX)
set(CMAKE_CXX_STANDARD 14)
set(CMAKE_CXX_STANDARD_REQUIRED ON)
-set(CMAKE_CUDA_STANDARD 14)
-set(CMAKE_CUDA_STANDARD_REQUIRED TRUE)
-
set_property(GLOBAL PROPERTY USE_FOLDERS ON)
set_property(GLOBAL PROPERTY PREDEFINED_TARGETS_FOLDER ".cmake")
set(libraryFolder "libs")
@@ -37,30 +35,36 @@ set (VF_ROOT_DIR ${CMAKE_CURRENT_SOURCE_DIR})
option(BUILD_VF_CPU "Build VirtualFluids cpu variant" OFF)
option(BUILD_VF_GPU "Build VirtualFluids gpu variant" OFF)
+option(BUILD_USE_OPENMP "Build VirtualFluids with openmp" ON)
+
+
+# vf gpu
+option(BUILD_VF_GPU "Build VirtualFluids GPU" ON )
+option(BUILD_VF_GKS "Build VirtualFluids GKS" OFF )
+option(BUILD_VF_TRAFFIC "Build VirtualFluids Traffic" OFF)
+option(BUILD_JSONCPP "Builds json cpp " OFF)
+option(BUILD_NUMERIC_TESTS "Build numeric tests" OFF)
+
option(BUILD_VF_UNIT_TESTS "Build VirtualFluids unit tests" OFF)
option(BUILD_VF_CLANG_TIDY "Add the clang tidy checks to the targets" OFF)
option(BUILD_VF_INCLUDE_WHAT_YOU_USE "Add IWYU to the targets" OFF)
option(BUILD_VF_CPPCHECK "Add cppcheck to the targets" OFF)
option(BUILD_VF_COVERAGE "Add the -coverage compiler flag." OFF)
-option(BUILD_SHARED_LIBS "" ON)
+option(BUILD_SHARED_LIBS "" OFF)
+option(BUILD_WARNINGS_AS_ERRORS "" OFF)
-option(USE_OPENMP "Include OpenMP support" ON)
+# windows: use multi-threaded dynamically-linked runtime library
+if(BUILD_SHARED_LIBS)
+ set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreaded$<$<CONFIG:Debug>:Debug>DLL")
+else()
+ set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreaded$<$<CONFIG:Debug>:Debug>")
+endif()
option(BUILD_VF_PYTHON_BINDINGS "" OFF)
option(BUILD_VF_DOUBLE_ACCURACY "Use double accuracy" OFF)
-#################################################################################
-# CMAKE POLICIES
-#################################################################################
-# CMAKE_CUDA_ARCHITECTURES
-# https://cmake.org/cmake/help/git-stage/policy/CMP0104.htmls
-if(POLICY CMP0104)
- cmake_policy(SET CMP0104 NEW)
- set(CMAKE_CUDA_ARCHITECTURES 30)
- # with cuda 11 the minimum architecture is 52
-endif()
#################################################################################
# MACROS
@@ -75,6 +79,13 @@ ENDIF()
#################################################################################
# COMMON LIBRARIES
#################################################################################
+if(BUILD_USE_OPENMP)
+ find_package(OpenMP REQUIRED)
+endif()
+
+find_package(MPI REQUIRED)
+
+
add_subdirectory(src/basics)
#################################################################################
@@ -84,6 +95,24 @@ if (BUILD_VF_CPU)
include (cpu.cmake)
endif()
if(BUILD_VF_GPU)
+
+ include(CheckLanguage)
+ check_language(CUDA)
+
+ if(NOT CMAKE_CUDA_COMPILER)
+ message(FATAL_ERROR "CUDA Compiler was requested but is not found on the system.")
+ endif()
+
+ set(CMAKE_CUDA_STANDARD 11)
+ set(CMAKE_CUDA_STANDARD_REQUIRED TRUE)
+
+ if(NOT DEFINED CMAKE_CUDA_ARCHITECTURES)
+ message(WARNING "CMAKE_CUDA_ARCHITECTURES was not defined and is set to 30 (CUDA support until 10.1 only).")
+ set(CMAKE_CUDA_ARCHITECTURES 30)
+ endif()
+
+ message("CUDA Architecture: ${CMAKE_CUDA_ARCHITECTURES}")
+
include (gpu.cmake)
endif()
@@ -91,5 +120,7 @@ endif()
# 3rd Party Libraries
#################################################################################
if(BUILD_VF_UNIT_TESTS)
- add_subdirectory(${VF_THIRD_DIR}/googletest)
+ if(NOT BUILD_NUMERIC_TESTS) # in this case googletest is already included.
+ add_subdirectory(${VF_THIRD_DIR}/googletest)
+ endif()
endif()
diff --git a/CMakePresets.json b/CMakePresets.json
new file mode 100644
index 0000000000000000000000000000000000000000..6863446af85b177bb4dc99eed475aa52f4d50269
--- /dev/null
+++ b/CMakePresets.json
@@ -0,0 +1,133 @@
+{
+ "version": 1,
+ "cmakeMinimumRequired": {
+ "major": 3,
+ "minor": 19,
+ "patch": 0
+ },
+ "configurePresets": [
+ {
+ "name": "default",
+ "hidden": true,
+ "binaryDir": "${sourceDir}/build/",
+ "cacheVariables": {
+ "BUILD_VF_UNIT_TESTS": "ON"
+ }
+ },
+ {
+ "name": "default_make",
+ "inherits": "default",
+ "hidden": true,
+ "generator": "Unix Makefiles"
+ },
+ {
+ "name": "default_ccache_make",
+ "inherits": "default_make",
+ "hidden": true,
+ "cacheVariables": {
+ "CMAKE_CXX_COMPILER_LAUNCHER": "ccache",
+ "CMAKE_CUDA_COMPILER_LAUNCHER": "ccache",
+ "CMAKE_C_COMPILER_LAUNCHER": "ccache"
+ }
+ },
+ {
+ "name": "default_msvc",
+ "inherits": "default",
+ "hidden": true,
+ "generator": "Visual Studio 16 2019",
+ "architecture": "x64"
+ },
+ {
+ "name": "default_cpu",
+ "hidden": true,
+ "description": "CPU build of VirtualFluids",
+ "cacheVariables": {
+ "BUILD_VF_CPU": "ON"
+ }
+ },
+ {
+ "name": "default_gpu",
+ "hidden": true,
+ "description": "GPU build of VirtualFluids",
+ "cacheVariables": {
+ "BUILD_VF_GPU": "ON"
+ }
+ },
+ {
+ "name": "default_gpu_numerical_tests",
+ "inherits": ["default_gpu"],
+ "hidden": true,
+ "description": "GPU numerical tests of VirtualFluids",
+ "cacheVariables": {
+ "BUILD_VF_DOUBLE_ACCURACY": "ON",
+ "BUILD_NUMERIC_TESTS": "ON"
+ }
+ },
+ {
+ "name": "default_all",
+ "hidden": true,
+ "description": "All build of VirtualFluids",
+ "inherits": ["default_cpu", "default_gpu"]
+ },
+ {
+ "name": "cpu_make",
+ "inherits": ["default_make", "default_cpu"],
+ "displayName": "cpu make configuration"
+ },
+ {
+ "name": "cpu_make_ccache",
+ "inherits": ["default_ccache_make", "default_cpu"],
+ "displayName": "cpu ccache make configuration"
+ },
+ {
+ "name": "cpu_msvc",
+ "inherits": ["default_msvc", "default_cpu"],
+ "displayName": "cpu msvc configuration"
+ },
+ {
+ "name": "gpu_make",
+ "inherits": ["default_make", "default_gpu"],
+ "displayName": "gpu make configuration"
+ },
+ {
+ "name": "gpu_make_ccache",
+ "inherits": ["default_ccache_make", "default_gpu"],
+ "displayName": "gpu ccache make configuration"
+ },
+ {
+ "name": "gpu_msvc",
+ "inherits": ["default_msvc", "default_gpu"],
+ "displayName": "gpu msvc configuration"
+ },
+ {
+ "name": "all_make",
+ "inherits": ["default_make", "default_all"],
+ "displayName": "all make configuration"
+ },
+ {
+ "name": "all_make_ccache",
+ "inherits": ["default_ccache_make", "default_all"],
+ "displayName": "all ccache make configuration"
+ },
+ {
+ "name": "all_msvc",
+ "inherits": ["default_msvc", "default_all"],
+ "displayName": "all msvc configuration"
+ },
+ {
+ "name": "gpu_numerical_tests_make",
+ "inherits": ["default_make", "default_gpu_numerical_tests"],
+ "displayName": "gpu numerical tests make configuration"
+ },
+ {
+ "name": "gpu_numerical_tests_ccache_make",
+ "inherits": ["default_ccache_make", "default_gpu_numerical_tests"],
+ "displayName": "gpu numerical tests ccache make configuration"
+ },
+ {
+ "name": "gpu_numerical_tests_msvc",
+ "inherits": ["default_msvc", "default_gpu_numerical_tests"],
+ "displayName": "gpu numerical tests msvc configuration"
+ }
+ ]
+}
diff --git a/apps/gpu/LidDrivenCavityGPU/CMakeLists.txt b/apps/gpu/LidDrivenCavityGPU/CMakeLists.txt
index f4e91979beaf3a63f1691803abe0feaf09a2d6cb..4a8f54ceea63c6efee33f38814d4b8db48695a2e 100644
--- a/apps/gpu/LidDrivenCavityGPU/CMakeLists.txt
+++ b/apps/gpu/LidDrivenCavityGPU/CMakeLists.txt
@@ -1,8 +1,3 @@
-
-
-PROJECT(LidDrivenCavityGPU)
-
+project(LidDrivenCavityGPU LANGUAGES CUDA CXX)
vf_add_library(BUILDTYPE binary PRIVATE_LINK basics GridGenerator VirtualFluids_GPU GksMeshAdapter GksGpu FILES LidDrivenCavity.cpp)
-
-linkCUDA()
diff --git a/gpu.cmake b/gpu.cmake
index 8ccd015747d94b4724317da10f0dce6eba00750e..ff11b0a60ef949089130f8a8b2885d267ce91050 100644
--- a/gpu.cmake
+++ b/gpu.cmake
@@ -1,31 +1,7 @@
-#############################################################
-### CUDAPATH ###
-#############################################################
-
-# if CMake cannot find CUDA by itself, set the correct paths manually:
-#SET(CUDA_CUT_INCLUDE_DIR "/cluster/cuda/9.0/include;/cluster/cuda/9.0/samples/common/inc" CACHE PATH "CUDA_CUT_INCLUDE_DIR")
-#SET(CUDA_SAMPLE_INCLUDE_DIR "/cluster/cuda/9.0/samples/common/inc" CACHE PATH "CUDA_CUT_INCLUDE_DIR")
-
-#############################################################
-### OPTIONS ###
-#############################################################
-
-option(VF_DOUBLE_ACCURACY "Use double accuracy" ON )
-
-
-#############################################################
-
-enable_language(CUDA)
-
-#############################################################
-
-
# only use this with device of CC larger than 6.0
IF(VF_DOUBLE_ACCURACY)
- set(CMAKE_CUDA_FLAGS " -arch=sm_60" CACHE STRING "" FORCE)
-ELSE(VF_DOUBLE_ACCURACY)
- set(CMAKE_CUDA_FLAGS "" CACHE STRING "" FORCE)
+ set(CMAKE_CUDA_ARCHITECTURES 60)
ENDIF(VF_DOUBLE_ACCURACY)
set(CMAKE_CUDA_FLAGS_DEBUG " -G" CACHE STRING "" FORCE)
diff --git a/sonar-project.properties b/sonar-project.properties
index 51f5908ff5e93027d8d0a7a02224edb7a08f7142..79510bce96750061e92d6d0bf720eb1c6427790a 100644
--- a/sonar-project.properties
+++ b/sonar-project.properties
@@ -1,12 +1,9 @@
# must be unique in a given SonarQube instance
sonar.projectKey=vf:project:open_source
-# --- optional properties ---
-
-# defaults to project key
sonar.projectName=VirtualFluids
-# defaults to 'not provided'
-#sonar.projectVersion=1.0
+
+sonar.projectVersion=1.0
sonar.language=cxx
@@ -36,4 +33,4 @@ sonar.cxx.gcc.reportPath=build/gcc_warnings.txt
sonar.cxx.funccomplexity.threshold=10
-sonar.cxx.funcsize.threshold=20
\ No newline at end of file
+sonar.cxx.funcsize.threshold=20
diff --git a/src/basics/CMakeLists.txt b/src/basics/CMakeLists.txt
index 0b643212ad9c5da74e711d72da085c1fe670704e..1703d6269c1cebe36005226373da85b14c5515f6 100644
--- a/src/basics/CMakeLists.txt
+++ b/src/basics/CMakeLists.txt
@@ -1,7 +1,7 @@
include(Core/buildInfo.cmake)
-vf_add_library(BUILDTYPE static EXCLUDE buildInfo.in.cpp)
+vf_add_library(PUBLIC_LINK MPI::MPI_CXX EXCLUDE buildInfo.in.cpp)
vf_get_library_name (library_name)
target_include_directories(${library_name} PUBLIC ${CMAKE_CURRENT_SOURCE_DIR}/Core)
@@ -21,7 +21,4 @@ IF(MSVC)
target_compile_definitions(${library_name} PUBLIC NOMINMAX) # Disable Min/Max-Macros
ENDIF(MSVC)
-
-linkMPI()
-
-vf_add_tests()
\ No newline at end of file
+vf_add_tests()
diff --git a/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.cpp b/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.cpp
index dce44b5551d2aedcf0c0d70a21dbcd9e8b071c8d..8fdb9fe1bc7f5dfe10dbfb522a87a4a3187dd60b 100644
--- a/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.cpp
+++ b/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.cpp
@@ -6,20 +6,17 @@
#include <fstream>
#include <iostream>
-BASICS_EXPORT std::shared_ptr<ConfigFileReader> ConfigFileReader::getNewInstance()
+std::shared_ptr<ConfigFileReader> ConfigFileReader::getNewInstance()
{
return std::shared_ptr<ConfigFileReader>(new ConfigFileReader());
}
-ConfigFileReader::ConfigFileReader() = default;
-
-BASICS_EXPORT ConfigFileReader::~ConfigFileReader() = default;
-
-BASICS_EXPORT std::shared_ptr<ConfigData> ConfigFileReader::readConfigFile(const std::string &filePath) const
+std::shared_ptr<ConfigData> ConfigFileReader::readConfigFile(const char* filePath) const
{
+ std::cout << filePath << std::endl;
std::shared_ptr<ConfigDataImp> data = ConfigDataImp::getNewInstance();
std::ifstream stream;
- stream.open(filePath.c_str(), std::ios::in);
+ stream.open(filePath, std::ios::in);
if (stream.fail())
throw std::runtime_error("can not open config file!");
std::unique_ptr<input::Input> input = input::Input::makeInput(stream, "config");
diff --git a/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.h b/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.h
index 9d88dfce6a7337d8ffaa7c4ffe43a4fd63949914..77c93ebfa4ba8564188d8e4a5442963382cf91e3 100644
--- a/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.h
+++ b/src/basics/Core/Input/ConfigFileReader/ConfigFileReader.h
@@ -1,9 +1,11 @@
#ifndef CONFIGFILEREADER_H
#define CONFIGFILEREADER_H
-#include "../Input.h"
#include <memory>
+#include <string>
+
+#include "basics_export.h"
class ConfigData;
@@ -11,11 +13,10 @@ class ConfigFileReader
{
public:
BASICS_EXPORT static std::shared_ptr<ConfigFileReader> getNewInstance();
- BASICS_EXPORT virtual ~ConfigFileReader();
- BASICS_EXPORT std::shared_ptr<ConfigData> readConfigFile(const std::string &filePath) const;
+ BASICS_EXPORT std::shared_ptr<ConfigData> readConfigFile(const char* filePath) const;
private:
- ConfigFileReader();
+ ConfigFileReader() = default;
};
#endif
diff --git a/src/basics/Core/StringUtilities/StringUtil.cpp b/src/basics/Core/StringUtilities/StringUtil.cpp
index 31496beae5953159c8d9a08a25fb5e808d9efea3..327a62346a91073834c8b710e90968a524ee2d28 100644
--- a/src/basics/Core/StringUtilities/StringUtil.cpp
+++ b/src/basics/Core/StringUtilities/StringUtil.cpp
@@ -1,5 +1,6 @@
#include "StringUtil.h"
+#include <string.h>
#include <regex>
#include <sstream>
diff --git a/src/basics/Core/StringUtilities/StringUtil.h b/src/basics/Core/StringUtilities/StringUtil.h
index cdf8dce290110848d5e6c50eb2fac35822dddf76..1927a69bc60bf2467fb09893463b3c9363191890 100644
--- a/src/basics/Core/StringUtilities/StringUtil.h
+++ b/src/basics/Core/StringUtilities/StringUtil.h
@@ -36,13 +36,23 @@ public:
static BASICS_EXPORT bool endsWith(const std::string &input, const std::string &end);
+
+ template<class T>
+ static T fromString(const std::string& s)
+ {
+ std::istringstream stream (s);
+ T t;
+ stream >> t;
+ return t;
+ }
+
private:
StringUtil() = default;
- ;
+
StringUtil(const StringUtil &) = default;
- ;
+
virtual ~StringUtil() = default;
- ;
+
static bool toBool(bool &t, const std::string &input, std::ios_base &(*f)(std::ios_base &));
};
diff --git a/src/basics/Core/buildInfo.in.cpp b/src/basics/Core/buildInfo.in.cpp
index 56f302208256ba86d7f814bfa6711f704aea3479..482f4592a41bfed0615858869ec4bb297764b0b1 100644
--- a/src/basics/Core/buildInfo.in.cpp
+++ b/src/basics/Core/buildInfo.in.cpp
@@ -1,5 +1,3 @@
-#include "buildInfo.h"
-
#include "basics_export.h"
namespace buildInfo
diff --git a/src/basics/basics/utilities/UbComparators.h b/src/basics/basics/utilities/UbComparators.h
index bc507e456b9603d85d42f8597fe4748a135e51d2..d461e27cb8d5f6026f13b2aea363d563ebed5c5c 100644
--- a/src/basics/basics/utilities/UbComparators.h
+++ b/src/basics/basics/utilities/UbComparators.h
@@ -170,7 +170,7 @@ struct compareMember {
// l.sort( compareConstMethods<Klasse, double, &Klasse::getVal1 >() );
//}
-}; // namespace UbComparators
+} // namespace UbComparators
#endif // UBCOMPARATOR_H
diff --git a/src/basics/basics/utilities/UbEqual.h b/src/basics/basics/utilities/UbEqual.h
index b3ca9102d585faeac0b2a4e413434d4d0d759282..87955b181ea3efaad4b7f3d2ebc746271ec95bb7 100644
--- a/src/basics/basics/utilities/UbEqual.h
+++ b/src/basics/basics/utilities/UbEqual.h
@@ -259,7 +259,7 @@ inline bool isUbEqual(const T1 &a, const T2 &b)
{
using Low = typename UbEqualTrait<T1, T2>::Low;
return specific_equal<Low, Low>(static_cast<Low>(a), static_cast<Low>(b));
-};
+}
//////////////////////////////////////////////////////////////////////////
// UbEqual-Functor
diff --git a/src/basics/basics/utilities/UbLogger.h b/src/basics/basics/utilities/UbLogger.h
index d350a763721b4b5fc1c2736fa56e2fff66a3eee0..fc2b118715a0f4afc0251b8ed8e2373a7d488153 100644
--- a/src/basics/basics/utilities/UbLogger.h
+++ b/src/basics/basics/utilities/UbLogger.h
@@ -228,7 +228,7 @@ inline std::string UbLogger<OutputPolicy>::logTimeString()
char buffer[11];
time_t t;
time(&t);
- tm r = { 0 };
+ tm r; // = { 0 };
strftime(buffer, sizeof(buffer), "%X", localtime_r(&t, &r));
struct timeval tv;
gettimeofday(&tv, 0);
diff --git a/src/basics/basics/utilities/UbSystem.h b/src/basics/basics/utilities/UbSystem.h
index 8e676811a152b176efd617a944f4e22ea538721e..0436a360c2b595115824e1b7906214621bd76314 100644
--- a/src/basics/basics/utilities/UbSystem.h
+++ b/src/basics/basics/utilities/UbSystem.h
@@ -530,7 +530,7 @@ struct select2nd {
const result_type &operator()(const argument_type &p) const { return p.second; }
};
-}; // namespace UbSystem
+} // namespace UbSystem
#define UB_STATIC_ASSERT(expr) static_cast<void>(sizeof(UbSystem::ub_static_assert<expr>));
// zum ueberpruefen von STATISCHEN ausdruecken waehrend der compile-zeit
diff --git a/src/basics/basics/writer/WbWriterVtkXmlBinary.h b/src/basics/basics/writer/WbWriterVtkXmlBinary.h
index 393c6bb13b268805a3fc2ca0850dd75d3fdd5616..421148d90497e3628ed274439c0b2fd7636b7fd2 100644
--- a/src/basics/basics/writer/WbWriterVtkXmlBinary.h
+++ b/src/basics/basics/writer/WbWriterVtkXmlBinary.h
@@ -37,7 +37,9 @@
#include <basics/writer/WbWriter.h>
-class WbWriterVtkXmlBinary : public WbWriter
+#include "basics_export.h"
+
+class BASICS_EXPORT WbWriterVtkXmlBinary : public WbWriter
{
public:
static WbWriterVtkXmlBinary *getInstance()
diff --git a/src/basics/geometry3d/GbCuboid3D.h b/src/basics/geometry3d/GbCuboid3D.h
index f0a0b0f9884999050495156f3c547b26d6398e61..762a08c2696dbf58ea4b726528c89571fc21ab62 100644
--- a/src/basics/geometry3d/GbCuboid3D.h
+++ b/src/basics/geometry3d/GbCuboid3D.h
@@ -80,7 +80,7 @@ public:
void setCenterCoordinates(const double &x1, const double &x2, const double &x3) override;
void translate(const double &x1, const double &x2, const double &x3) override;
- void rotate(const double &rx1, const double &rx2, const double &rx3) override {}
+ void rotate(const double &rx1, const double &rx2, const double &rx3) override { (void)rx1; (void)rx2; (void)rx3; }
void scale(const double &sx1, const double &sx2, const double &sx3) override;
double getLengthX1();
diff --git a/src/basics/geometry3d/GbObject3D.h b/src/basics/geometry3d/GbObject3D.h
index f60e64c8f95dfde84a389c422bbc6b9377e68187..c562f42412c8ab75b0ebcd303a5785ab4939cdfb 100644
--- a/src/basics/geometry3d/GbObject3D.h
+++ b/src/basics/geometry3d/GbObject3D.h
@@ -49,6 +49,8 @@ class GbObject3DCreator;
#include <PointerDefinitions.h>
+#include "basics_export.h"
+
//////////////////////////////////////////////////////////////////////////
//!
//! \class GbObject3D
@@ -57,7 +59,7 @@ class GbObject3DCreator;
//!
//////////////////////////////////////////////////////////////////////////
-class GbObject3D : public ObObject
+class BASICS_EXPORT GbObject3D : public ObObject
{
public:
// abstract Methods
diff --git a/src/cpu/VirtualFluidsCore/BoundaryConditions/BCAlgorithm.h b/src/cpu/VirtualFluidsCore/BoundaryConditions/BCAlgorithm.h
index 7cef8205ce13d89802c9bbd7eebb5d6eb3759b3f..734aa9342df011846ce345ba7c37a181383e8828 100644
--- a/src/cpu/VirtualFluidsCore/BoundaryConditions/BCAlgorithm.h
+++ b/src/cpu/VirtualFluidsCore/BoundaryConditions/BCAlgorithm.h
@@ -47,18 +47,30 @@ class BoundaryConditions;
class BCAlgorithm
{
public:
- static const char VelocityBCAlgorithm = 0;
- static const char EqDensityBCAlgorithm = 1;
- static const char NonEqDensityBCAlgorithm = 2;
- static const char NoSlipBCAlgorithm = 3;
- static const char SlipBCAlgorithm = 4;
- static const char HighViscosityNoSlipBCAlgorithm = 5;
- static const char ThinWallNoSlipBCAlgorithm = 6;
- static const char VelocityWithDensityBCAlgorithm = 7;
+ static const char VelocityBCAlgorithm = 0;
+ static const char EqDensityBCAlgorithm = 1;
+ static const char NonEqDensityBCAlgorithm = 2;
+ static const char NoSlipBCAlgorithm = 3;
+ static const char SlipBCAlgorithm = 4;
+ static const char HighViscosityNoSlipBCAlgorithm = 5;
+ static const char ThinWallNoSlipBCAlgorithm = 6;
+ static const char VelocityWithDensityBCAlgorithm = 7;
static const char NonReflectingOutflowBCAlgorithm = 8;
+ static const char VelocityAndThixotropyBCAlgorithm = 9;
+ static const char DensityAndThixotropyBCAlgorithm = 10;
+ static const char NoSlipAndThixotropyBCAlgorithm = 11;
+ static const char NonReflectingOutflowAndThixotropyBCAlgorithm = 12;
+ static const char VelocityWithDensityAndThixotropyBCAlgorithm = 13;
+ static const char BinghamModelNoSlipBCAlgorithm = 14;
+ static const char HerschelBulkleyModelNoSlipBCAlgorithm = 15;
+ static const char SimpleVelocityBCAlgorithm = 16;
+ static const char SimpleSlipBCAlgorithm = 17;
+ static const char PowellEyringModelNoSlipBCAlgorithm = 18;
+ static const char BinghamModelVelocityBCAlgorithm = 19;
+
public:
- BCAlgorithm() = default;
+ BCAlgorithm() = default;
virtual ~BCAlgorithm() = default;
virtual void addDistributions(SPtr<DistributionArray3D> distributions) = 0;
@@ -72,11 +84,13 @@ public:
SPtr<BCArray3D> getBcArray();
void setBcArray(SPtr<BCArray3D> bcarray);
virtual void applyBC() = 0;
+ bool getThixotropy(){ return thixotropy; };
protected:
- bool compressible{ false };
+ bool compressible { false };
char type;
bool preCollision;
+ bool thixotropy { false };
SPtr<BoundaryConditions> bcPtr;
SPtr<DistributionArray3D> distributions;
diff --git a/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.cpp b/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.cpp
index 87606eecf03943259dfec89a805336d2a3190bfa..88f4a52b2ff0445838af5aade25d9c78ce6809a8 100644
--- a/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.cpp
+++ b/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.cpp
@@ -63,11 +63,11 @@ void BCArray3D::resize(std::size_t nx1, std::size_t nx2, std::size_t nx3, int va
//////////////////////////////////////////////////////////////////////////
bool BCArray3D::validIndices(std::size_t x1, std::size_t x2, std::size_t x3) const
{
- if (x1 < 0 || x1 >= this->bcindexmatrix.getNX1())
+ if (x1 >= this->bcindexmatrix.getNX1())
return false;
- if (x2 < 0 || x2 >= this->bcindexmatrix.getNX2())
+ if (x2 >= this->bcindexmatrix.getNX2())
return false;
- if (x3 < 0 || x3 >= this->bcindexmatrix.getNX3())
+ if (x3 >= this->bcindexmatrix.getNX3())
return false;
return true;
}
diff --git a/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.h b/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.h
index 835e5b1c95454a9fbe8186d6942c3936a5e0e2cc..b9d08f7117d9dc41c008c9d92a5780aceedad21c 100644
--- a/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.h
+++ b/src/cpu/VirtualFluidsCore/BoundaryConditions/BCArray3D.h
@@ -35,7 +35,7 @@
#define BCArray_H
#include "BoundaryConditions.h"
-#include "basics/container/CbArray3D.h"
+#include "CbArray3D.h"
#include <typeinfo>
diff --git a/src/cpu/VirtualFluidsCore/BoundaryConditions/BoundaryConditions.h b/src/cpu/VirtualFluidsCore/BoundaryConditions/BoundaryConditions.h
index 1ab3b4e284de49fd1b80bad13d65fee98d221e57..84ba7a6041d38546da03323f35a23f7084df9809 100644
--- a/src/cpu/VirtualFluidsCore/BoundaryConditions/BoundaryConditions.h
+++ b/src/cpu/VirtualFluidsCore/BoundaryConditions/BoundaryConditions.h
@@ -48,7 +48,6 @@ class BoundaryConditions
{
public:
BoundaryConditions()
-
{
UB_STATIC_ASSERT(sizeof(long long) >= 8);
UB_STATIC_ASSERT((sizeof(long long) * 8) >= (D3Q27System::FENDDIR + 1) * BoundaryConditions::optionDigits);
@@ -150,13 +149,13 @@ public:
{
return (short)(((slipBoundaryFlags >> (optionDigits * direction)) & maxOptionVal) - 1);
}
- void setNormalVector(const LBMReal &nx1, const LBMReal &nx2, const LBMReal &nx3)
+ void setNormalVector(const float &nx1, const float &nx2, const float &nx3)
{
this->nx1 = nx1;
this->nx2 = nx2;
this->nx3 = nx3;
}
- UbTupleDouble3 getNormalVector() { return makeUbTuple(nx1, nx2, nx3); }
+ UbTupleFloat3 getNormalVector() { return makeUbTuple(nx1, nx2, nx3); }
/*============== Velocity Boundary ========================*/
void setVelocityBoundaryFlag(const int &direction, const short &secOpt = 0)
@@ -181,72 +180,73 @@ public:
void setBoundaryVelocity(const Vector3D &vx)
{
- setBoundaryVelocityX1((LBMReal)vx[0]);
- setBoundaryVelocityX2((LBMReal)vx[1]);
- setBoundaryVelocityX3((LBMReal)vx[2]);
+ setBoundaryVelocityX1((float)vx[0]);
+ setBoundaryVelocityX2((float)vx[1]);
+ setBoundaryVelocityX3((float)vx[2]);
}
- void setBoundaryVelocityX1(const LBMReal &vx1) { this->bcVelocityX1 = vx1; }
- void setBoundaryVelocityX2(const LBMReal &vx2) { this->bcVelocityX2 = vx2; }
- void setBoundaryVelocityX3(const LBMReal &vx3) { this->bcVelocityX3 = vx3; }
- LBMReal getBoundaryVelocityX1() { return this->bcVelocityX1; }
- LBMReal getBoundaryVelocityX2() { return this->bcVelocityX2; }
- LBMReal getBoundaryVelocityX3() { return this->bcVelocityX3; }
- LBMReal getBoundaryVelocity(const int &direction)
+ void setBoundaryVelocityX1(const float &vx1) { this->bcVelocityX1 = vx1; }
+ void setBoundaryVelocityX2(const float &vx2) { this->bcVelocityX2 = vx2; }
+ void setBoundaryVelocityX3(const float &vx3) { this->bcVelocityX3 = vx3; }
+ float getBoundaryVelocityX1() { return this->bcVelocityX1; }
+ float getBoundaryVelocityX2() { return this->bcVelocityX2; }
+ float getBoundaryVelocityX3() { return this->bcVelocityX3; }
+ float getBoundaryVelocity(const int &direction)
{
switch (direction) {
case D3Q27System::E:
- return (LBMReal)(UbMath::c4o9 *
- (+bcVelocityX1)); //(2/cs^2)(=6)*rho_0(=1 for incompressible)*wi*u*ei with cs=1/sqrt(3)
+ return (float)(UbMath::c4o9 *
+ (+bcVelocityX1)); //(2/cs^2)(=6)*rho_0(=1 bei inkompr)*wi*u*ei mit cs=1/sqrt(3)
case D3Q27System::W:
- return (LBMReal)(UbMath::c4o9 * (-bcVelocityX1));
+ return (float)(UbMath::c4o9 *
+ (-bcVelocityX1)); // z.B. aus paper manfred MRT LB models in three dimensions (2002)
case D3Q27System::N:
- return (LBMReal)(UbMath::c4o9 * (+bcVelocityX2));
+ return (float)(UbMath::c4o9 * (+bcVelocityX2));
case D3Q27System::S:
- return (LBMReal)(UbMath::c4o9 * (-bcVelocityX2));
+ return (float)(UbMath::c4o9 * (-bcVelocityX2));
case D3Q27System::T:
- return (LBMReal)(UbMath::c4o9 * (+bcVelocityX3));
+ return (float)(UbMath::c4o9 * (+bcVelocityX3));
case D3Q27System::B:
- return (LBMReal)(UbMath::c4o9 * (-bcVelocityX3));
+ return (float)(UbMath::c4o9 * (-bcVelocityX3));
case D3Q27System::NE:
- return (LBMReal)(UbMath::c1o9 * (+bcVelocityX1 + bcVelocityX2));
+ return (float)(UbMath::c1o9 * (+bcVelocityX1 + bcVelocityX2));
case D3Q27System::SW:
- return (LBMReal)(UbMath::c1o9 * (-bcVelocityX1 - bcVelocityX2));
+ return (float)(UbMath::c1o9 * (-bcVelocityX1 - bcVelocityX2));
case D3Q27System::SE:
- return (LBMReal)(UbMath::c1o9 * (+bcVelocityX1 - bcVelocityX2));
+ return (float)(UbMath::c1o9 * (+bcVelocityX1 - bcVelocityX2));
case D3Q27System::NW:
- return (LBMReal)(UbMath::c1o9 * (-bcVelocityX1 + bcVelocityX2));
+ return (float)(UbMath::c1o9 * (-bcVelocityX1 + bcVelocityX2));
case D3Q27System::TE:
- return (LBMReal)(UbMath::c1o9 * (+bcVelocityX1 + bcVelocityX3));
+ return (float)(UbMath::c1o9 * (+bcVelocityX1 + bcVelocityX3));
case D3Q27System::BW:
- return (LBMReal)(UbMath::c1o9 * (-bcVelocityX1 - bcVelocityX3));
+ return (float)(UbMath::c1o9 * (-bcVelocityX1 - bcVelocityX3));
case D3Q27System::BE:
- return (LBMReal)(UbMath::c1o9 * (+bcVelocityX1 - bcVelocityX3));
+ return (float)(UbMath::c1o9 * (+bcVelocityX1 - bcVelocityX3));
case D3Q27System::TW:
- return (LBMReal)(UbMath::c1o9 * (-bcVelocityX1 + bcVelocityX3));
+ return (float)(UbMath::c1o9 * (-bcVelocityX1 + bcVelocityX3));
case D3Q27System::TN:
- return (LBMReal)(UbMath::c1o9 * (+bcVelocityX2 + bcVelocityX3));
+ return (float)(UbMath::c1o9 * (+bcVelocityX2 + bcVelocityX3));
case D3Q27System::BS:
- return (LBMReal)(UbMath::c1o9 * (-bcVelocityX2 - bcVelocityX3));
+ return (float)(UbMath::c1o9 * (-bcVelocityX2 - bcVelocityX3));
case D3Q27System::BN:
- return (LBMReal)(UbMath::c1o9 * (+bcVelocityX2 - bcVelocityX3));
+ return (float)(UbMath::c1o9 * (+bcVelocityX2 - bcVelocityX3));
case D3Q27System::TS:
- return (LBMReal)(UbMath::c1o9 * (-bcVelocityX2 + bcVelocityX3));
+ return (float)(UbMath::c1o9 * (-bcVelocityX2 + bcVelocityX3));
case D3Q27System::TNE:
- return (LBMReal)(UbMath::c1o36 * (+bcVelocityX1 + bcVelocityX2 + bcVelocityX3));
+ return (float)(UbMath::c1o36 * (+bcVelocityX1 + bcVelocityX2 + bcVelocityX3));
case D3Q27System::BSW:
- return (LBMReal)(UbMath::c1o36 * (-bcVelocityX1 - bcVelocityX2 - bcVelocityX3));
+ return (float)(UbMath::c1o36 * (-bcVelocityX1 - bcVelocityX2 - bcVelocityX3));
case D3Q27System::BNE:
- return (LBMReal)(UbMath::c1o36 * (+bcVelocityX1 + bcVelocityX2 - bcVelocityX3));
+ return (float)(UbMath::c1o36 * (+bcVelocityX1 + bcVelocityX2 - bcVelocityX3));
case D3Q27System::TSW:
- return (LBMReal)(UbMath::c1o36 * (-bcVelocityX1 - bcVelocityX2 + bcVelocityX3));
+ return (float)(UbMath::c1o36 * (-bcVelocityX1 - bcVelocityX2 + bcVelocityX3));
case D3Q27System::TSE:
- return (LBMReal)(UbMath::c1o36 * (+bcVelocityX1 - bcVelocityX2 + bcVelocityX3));
+ return (float)(UbMath::c1o36 * (+bcVelocityX1 - bcVelocityX2 + bcVelocityX3));
case D3Q27System::BNW:
- return (LBMReal)(UbMath::c1o36 * (-bcVelocityX1 + bcVelocityX2 - bcVelocityX3));
+ return (float)(UbMath::c1o36 * (-bcVelocityX1 + bcVelocityX2 - bcVelocityX3));
case D3Q27System::BSE:
- return (LBMReal)(UbMath::c1o36 * (+bcVelocityX1 - bcVelocityX2 - bcVelocityX3));
+ return (float)(UbMath::c1o36 * (+bcVelocityX1 - bcVelocityX2 - bcVelocityX3));
case D3Q27System::TNW:
- return (LBMReal)(UbMath::c1o36 * (-bcVelocityX1 + bcVelocityX2 + bcVelocityX3));
+ return (float)(UbMath::c1o36 * (-bcVelocityX1 + bcVelocityX2 + bcVelocityX3));
default:
throw UbException(UB_EXARGS, "unknown error");
}
@@ -273,44 +273,44 @@ public:
return (short)(((densityBoundaryFlags >> (optionDigits * direction)) & maxOptionVal) - 1);
}
- void setBoundaryDensity(LBMReal density) { this->bcDensity = density; }
- LBMReal getBoundaryDensity() { return this->bcDensity; }
+ void setBoundaryDensity(float density) { this->bcDensity = density; }
+ float getBoundaryDensity() { return this->bcDensity; }
- // Lodi extension
- void setDensityLodiDensity(const LBMReal &bcLodiDensity) { this->bcLodiDensity = bcLodiDensity; }
- void setDensityLodiVelocityX1(const LBMReal &bcLodiVelocityX1) { this->bcLodiVelocityX1 = bcLodiVelocityX1; }
- void setDensityLodiVelocityX2(const LBMReal &bcLodiVelocityX2) { this->bcLodiVelocityX2 = bcLodiVelocityX2; }
- void setDensityLodiVelocityX3(const LBMReal &bcLodiVelocityX3) { this->bcLodiVelocityX3 = bcLodiVelocityX3; }
- void setDensityLodiLength(const LBMReal &bcLodiLentgh) { this->bcLodiLentgh = bcLodiLentgh; }
- LBMReal getDensityLodiDensity() const { return this->bcLodiDensity; }
- LBMReal getDensityLodiVelocityX1() const { return this->bcLodiVelocityX1; }
- LBMReal getDensityLodiVelocityX2() const { return this->bcLodiVelocityX2; }
- LBMReal getDensityLodiVelocityX3() const { return this->bcLodiVelocityX3; }
- LBMReal getDensityLodiLength() const { return this->bcLodiLentgh; }
+ ////Lodi extension
+ void setDensityLodiDensity(const float &bcLodiDensity) { this->bcLodiDensity = bcLodiDensity; }
+ void setDensityLodiVelocityX1(const float &bcLodiVelocityX1) { this->bcLodiVelocityX1 = bcLodiVelocityX1; }
+ void setDensityLodiVelocityX2(const float &bcLodiVelocityX2) { this->bcLodiVelocityX2 = bcLodiVelocityX2; }
+ void setDensityLodiVelocityX3(const float &bcLodiVelocityX3) { this->bcLodiVelocityX3 = bcLodiVelocityX3; }
+ void setDensityLodiLength(const float &bcLodiLentgh) { this->bcLodiLentgh = bcLodiLentgh; }
+ float getDensityLodiDensity() const { return this->bcLodiDensity; }
+ float getDensityLodiVelocityX1() const { return this->bcLodiVelocityX1; }
+ float getDensityLodiVelocityX2() const { return this->bcLodiVelocityX2; }
+ float getDensityLodiVelocityX3() const { return this->bcLodiVelocityX3; }
+ float getDensityLodiLength() const { return this->bcLodiLentgh; }
- LBMReal &densityLodiDensity() { return this->bcLodiDensity; }
- LBMReal &densityLodiVelocityX1() { return this->bcLodiVelocityX1; }
- LBMReal &densityLodiVelocityX2() { return this->bcLodiVelocityX2; }
- LBMReal &densityLodiVelocityX3() { return this->bcLodiVelocityX3; }
- LBMReal &densityLodiLentgh() { return this->bcLodiLentgh; }
+ float &densityLodiDensity() { return this->bcLodiDensity; }
+ float &densityLodiVelocityX1() { return this->bcLodiVelocityX1; }
+ float &densityLodiVelocityX2() { return this->bcLodiVelocityX2; }
+ float &densityLodiVelocityX3() { return this->bcLodiVelocityX3; }
+ float &densityLodiLentgh() { return this->bcLodiLentgh; }
- const LBMReal &densityLodiDensity() const { return this->bcLodiDensity; }
- const LBMReal &densityLodiVelocityX1() const { return this->bcLodiVelocityX1; }
- const LBMReal &densityLodiVelocityX2() const { return this->bcLodiVelocityX2; }
- const LBMReal &densityLodiVelocityX3() const { return this->bcLodiVelocityX3; }
- const LBMReal &densityLodiLentgh() const { return this->bcLodiLentgh; }
+ const float &densityLodiDensity() const { return this->bcLodiDensity; }
+ const float &densityLodiVelocityX1() const { return this->bcLodiVelocityX1; }
+ const float &densityLodiVelocityX2() const { return this->bcLodiVelocityX2; }
+ const float &densityLodiVelocityX3() const { return this->bcLodiVelocityX3; }
+ const float &densityLodiLentgh() const { return this->bcLodiLentgh; }
/*======================= Qs =============================*/
- void setQ(const LBMReal &val, const int &direction) { q[direction] = val; }
- LBMReal getQ(const int &direction) { return q[direction]; }
+ void setQ(const float &val, const int &direction) { q[direction] = val; }
+ float getQ(const int &direction) { return q[direction]; }
virtual std::vector<std::string> getBCNames()
{
std::vector<std::string> tmp;
- tmp.emplace_back("NoSlipBC");
- tmp.emplace_back("SlipBC");
- tmp.emplace_back("VelocityBC");
- tmp.emplace_back("DensityBC");
+ tmp.push_back("NoSlipBC");
+ tmp.push_back("SlipBC");
+ tmp.push_back("VelocityBC");
+ tmp.push_back("DensityBC");
return tmp;
}
virtual std::vector<long long> getBCFlags()
@@ -336,7 +336,7 @@ public:
static const long long maxOptionVal; // = ( 1<<optionDigits ) - 1; //2^3-1 -> 7
protected:
- LBMReal q[D3Q27System::FENDDIR + 1];
+ float q[D3Q27System::FENDDIR + 1];
long long noslipBoundaryFlags{ 0 };
long long slipBoundaryFlags{ 0 };
@@ -344,20 +344,26 @@ protected:
long long densityBoundaryFlags{ 0 };
long long wallModelBoundaryFlags{ 0 };
- LBMReal bcVelocityX1{ 0.0f };
- LBMReal bcVelocityX2{ 0.0f };
- LBMReal bcVelocityX3{ 0.0f };
- LBMReal bcDensity{ 0.0f };
+ float bcVelocityX1{ 0.0f };
+ float bcVelocityX2{ 0.0f };
+ float bcVelocityX3{ 0.0f };
+ float bcDensity{ 0.0f };
+ // float bcThixotropy{ 0.0f };
+
+ float bcLodiDensity{ 0.0f };
+ float bcLodiVelocityX1{ 0.0f };
+ float bcLodiVelocityX2{ 0.0f };
+ float bcLodiVelocityX3{ 0.0f };
+ float bcLodiLentgh{ 0.0f };
- LBMReal bcLodiDensity{ 0.0f };
- LBMReal bcLodiVelocityX1{ 0.0f };
- LBMReal bcLodiVelocityX2{ 0.0f };
- LBMReal bcLodiVelocityX3{ 0.0f };
- LBMReal bcLodiLentgh{ 0.0f };
+ float nx1{ 0.0f }, nx2{ 0.0f }, nx3{ 0.0f };
- LBMReal nx1{ 0.0f }, nx2{ 0.0f }, nx3{ 0.0f };
+ char algorithmType { -1 };
- char algorithmType{ -1 };
+private:
+ friend class MPIIORestartCoProcessor;
+ friend class MPIIOMigrationCoProcessor;
+ friend class MPIIOMigrationBECoProcessor;
};
#endif
diff --git a/src/cpu/VirtualFluidsCore/CMakeLists.txt b/src/cpu/VirtualFluidsCore/CMakeLists.txt
index 871aa123b3922d109d521eedcf83adfc87493e16..ac9da38b2c28d52a9be8d8489abf42247f65506e 100644
--- a/src/cpu/VirtualFluidsCore/CMakeLists.txt
+++ b/src/cpu/VirtualFluidsCore/CMakeLists.txt
@@ -1,6 +1,11 @@
-vf_add_library(BUILDTYPE static PUBLIC_LINK basics muparser)
+set(CAB_ADDITIONAL_LINK_LIBRARIES "")
+if(BUILD_USE_OPENMP)
+list(APPEND CAB_ADDITIONAL_LINK_LIBRARIES OpenMP::OpenMP_CXX)
+endif()
+
+vf_add_library(BUILDTYPE static PUBLIC_LINK basics muparser MPI::MPI_CXX ${CAB_ADDITIONAL_LINK_LIBRARIES})
vf_get_library_name(library_name)
diff --git a/src/cpu/VirtualFluidsCore/CoProcessors/WriteBoundaryConditionsCoProcessor.cpp b/src/cpu/VirtualFluidsCore/CoProcessors/WriteBoundaryConditionsCoProcessor.cpp
index a7ba3a84505f1695a0c59175d2d2d8adf3d4dbdb..e85f2806df40e11c7f30cca1c86bcb5dc639ee73 100644
--- a/src/cpu/VirtualFluidsCore/CoProcessors/WriteBoundaryConditionsCoProcessor.cpp
+++ b/src/cpu/VirtualFluidsCore/CoProcessors/WriteBoundaryConditionsCoProcessor.cpp
@@ -151,8 +151,6 @@ void WriteBoundaryConditionsCoProcessor::addDataGeo(SPtr<Block3D> block)
SPtr<ILBMKernel> kernel = block->getKernel();
SPtr<BCArray3D> bcArray = kernel->getBCProcessor()->getBCArray();
- // knotennummerierung faengt immer bei 0 an!
- unsigned int SWB, SEB, NEB, NWB, SWT, SET, NET, NWT;
int minX1 = 0;
int minX2 = 0;
@@ -171,9 +169,9 @@ void WriteBoundaryConditionsCoProcessor::addDataGeo(SPtr<Block3D> block)
maxX2 -= 1;
maxX3 -= 1;
- for (size_t ix3 = minX3; ix3 <= maxX3; ix3++) {
- for (size_t ix2 = minX2; ix2 <= maxX2; ix2++) {
- for (size_t ix1 = minX1; ix1 <= maxX1; ix1++) {
+ for (int ix3 = minX3; ix3 <= maxX3; ix3++) {
+ for (int ix2 = minX2; ix2 <= maxX2; ix2++) {
+ for (int ix1 = minX1; ix1 <= maxX1; ix1++) {
if (!bcArray->isUndefined(ix1, ix2, ix3)) {
int index = 0;
nodeNumbers(ix1, ix2, ix3) = nr++;
@@ -239,6 +237,9 @@ void WriteBoundaryConditionsCoProcessor::addDataGeo(SPtr<Block3D> block)
maxX2 -= 1;
maxX3 -= 1;
+ // knotennummerierung faengt immer bei 0 an!
+ int SWB = 0, SEB = 0, NEB = 0, NWB = 0, SWT = 0, SET = 0, NET = 0, NWT = 0;
+
// cell vector erstellen
for (int ix3 = minX3; ix3 <= maxX3; ix3++) {
for (int ix2 = minX2; ix2 <= maxX2; ix2++) {
@@ -248,7 +249,9 @@ void WriteBoundaryConditionsCoProcessor::addDataGeo(SPtr<Block3D> block)
(SWT = nodeNumbers(ix1, ix2, ix3 + 1)) >= 0 && (SET = nodeNumbers(ix1 + 1, ix2, ix3 + 1)) >= 0 &&
(NET = nodeNumbers(ix1 + 1, ix2 + 1, ix3 + 1)) >= 0 &&
(NWT = nodeNumbers(ix1, ix2 + 1, ix3 + 1)) >= 0) {
- cells.push_back(makeUbTuple(SWB, SEB, NEB, NWB, SWT, SET, NET, NWT));
+ cells.push_back(makeUbTuple((unsigned int)SWB, (unsigned int)SEB, (unsigned int)NEB,
+ (unsigned int)NWB, (unsigned int)SWT, (unsigned int)SET,
+ (unsigned int)NET, (unsigned int)NWT));
}
}
}
diff --git a/src/cpu/VirtualFluidsCore/CoProcessors/WriteMacroscopicQuantitiesCoProcessor.cpp b/src/cpu/VirtualFluidsCore/CoProcessors/WriteMacroscopicQuantitiesCoProcessor.cpp
index 27cf92056c88f1481f256ced4080003ed511f241..eef7d3bf6a0d0ad80aa1bdd7d83dfb469b044584 100644
--- a/src/cpu/VirtualFluidsCore/CoProcessors/WriteMacroscopicQuantitiesCoProcessor.cpp
+++ b/src/cpu/VirtualFluidsCore/CoProcessors/WriteMacroscopicQuantitiesCoProcessor.cpp
@@ -53,20 +53,24 @@ WriteMacroscopicQuantitiesCoProcessor::WriteMacroscopicQuantitiesCoProcessor(SPt
WbWriter *const writer,
SPtr<LBMUnitConverter> conv,
SPtr<Communicator> comm)
- : CoProcessor(grid, s), path(path), writer(writer), conv(conv), comm(comm)
+ : CoProcessor(grid, s), path(path), writer(writer), conv(conv), comm(comm)
{
- gridRank = comm->getProcessID();
+ gridRank = comm->getProcessID();
minInitLevel = this->grid->getCoarsestInitializedLevel();
maxInitLevel = this->grid->getFinestInitializedLevel();
blockVector.resize(maxInitLevel + 1);
- for (int level = minInitLevel; level <= maxInitLevel; level++) {
+ for (int level = minInitLevel; level <= maxInitLevel; level++)
+ {
grid->getBlocks(level, gridRank, true, blockVector[level]);
}
}
+
//////////////////////////////////////////////////////////////////////////
-void WriteMacroscopicQuantitiesCoProcessor::init() {}
+void WriteMacroscopicQuantitiesCoProcessor::init()
+{}
+
//////////////////////////////////////////////////////////////////////////
void WriteMacroscopicQuantitiesCoProcessor::process(double step)
{
@@ -75,14 +79,18 @@ void WriteMacroscopicQuantitiesCoProcessor::process(double step)
UBLOG(logDEBUG3, "WriteMacroscopicQuantitiesCoProcessor::update:" << step);
}
+
//////////////////////////////////////////////////////////////////////////
void WriteMacroscopicQuantitiesCoProcessor::collectData(double step)
{
int istep = static_cast<int>(step);
- for (int level = minInitLevel; level <= maxInitLevel; level++) {
- for (SPtr<Block3D> block : blockVector[level]) {
- if (block) {
+ for (int level = minInitLevel; level <= maxInitLevel; level++)
+ {
+ for (SPtr<Block3D> block : blockVector[level])
+ {
+ if (block)
+ {
addDataMQ(block);
}
}
@@ -93,27 +101,29 @@ void WriteMacroscopicQuantitiesCoProcessor::collectData(double step)
subfolder = "mq" + UbSystem::toString(istep);
pfilePath = path + "/mq/" + subfolder;
cfilePath = path + "/mq/mq_collection";
- partPath = pfilePath + "/mq" + UbSystem::toString(gridRank) + "_" + UbSystem::toString(istep);
+ partPath = pfilePath + "/mq" + UbSystem::toString(gridRank) + "_" + UbSystem::toString(istep);
std::string partName = writer->writeOctsWithNodeData(partPath, nodes, cells, datanames, data);
- size_t found = partName.find_last_of("/");
- std::string piece = partName.substr(found + 1);
- piece = subfolder + "/" + piece;
+ size_t found = partName.find_last_of("/");
+ std::string piece = partName.substr(found + 1);
+ piece = subfolder + "/" + piece;
std::vector<std::string> cellDataNames;
std::vector<std::string> pieces;
pieces.push_back(piece);
if (comm->getProcessID() == comm->getRoot()) {
std::string pname =
- WbWriterVtkXmlASCII::getInstance()->writeParallelFile(pfilePath, pieces, datanames, cellDataNames);
+ WbWriterVtkXmlASCII::getInstance()->writeParallelFile(pfilePath, pieces, datanames, cellDataNames);
found = pname.find_last_of("/");
piece = pname.substr(found + 1);
std::vector<std::string> filenames;
filenames.push_back(piece);
- if (step == CoProcessor::scheduler->getMinBegin()) {
+ if (step == CoProcessor::scheduler->getMinBegin())
+ {
WbWriterVtkXmlASCII::getInstance()->writeCollection(cfilePath, filenames, istep, false);
- } else {
+ } else
+ {
WbWriterVtkXmlASCII::getInstance()->addFilesToCollection(cfilePath, filenames, istep, false);
}
UBLOG(logINFO, "WriteMacroscopicQuantitiesCoProcessor step: " << istep);
@@ -121,6 +131,7 @@ void WriteMacroscopicQuantitiesCoProcessor::collectData(double step)
clearData();
}
+
//////////////////////////////////////////////////////////////////////////
void WriteMacroscopicQuantitiesCoProcessor::clearData()
{
@@ -129,16 +140,23 @@ void WriteMacroscopicQuantitiesCoProcessor::clearData()
datanames.clear();
data.clear();
}
+
//////////////////////////////////////////////////////////////////////////
void WriteMacroscopicQuantitiesCoProcessor::addDataMQ(SPtr<Block3D> block)
{
- // This data is written:
+ double level = (double)block->getLevel();
+
+ // Diese Daten werden geschrieben:
datanames.resize(0);
- datanames.emplace_back("DRho");
- datanames.emplace_back("Press");
- datanames.emplace_back("Vx");
- datanames.emplace_back("Vy");
- datanames.emplace_back("Vz");
+ datanames.push_back("Rho");
+ datanames.push_back("Vx");
+ datanames.push_back("Vy");
+ datanames.push_back("Vz");
+ // datanames.push_back("Press");
+ datanames.push_back("Level");
+ // datanames.push_back("BlockID");
+ // datanames.push_back("gamma");
+ // datanames.push_back("collFactor");
data.resize(datanames.size());
@@ -146,10 +164,10 @@ void WriteMacroscopicQuantitiesCoProcessor::addDataMQ(SPtr<Block3D> block)
SPtr<BCArray3D> bcArray = kernel->getBCProcessor()->getBCArray();
SPtr<DistributionArray3D> distributions = kernel->getDataSet()->getFdistributions();
LBMReal f[D3Q27System::ENDF + 1];
- LBMReal vx1, vx2, vx3, drho, press;
+ LBMReal vx1, vx2, vx3, rho;
- // node numbering always starts at 0!
- unsigned int SWB, SEB, NEB, NWB, SWT, SET, NET, NWT;
+ // knotennummerierung faengt immer bei 0 an!
+ int SWB, SEB, NEB, NWB, SWT, SET, NET, NWT;
if (block->getKernel()->getCompressible()) {
calcMacros = &D3Q27System::calcCompMacroscopicValues;
@@ -165,12 +183,21 @@ void WriteMacroscopicQuantitiesCoProcessor::addDataMQ(SPtr<Block3D> block)
int maxX2 = (int)(distributions->getNX2());
int maxX3 = (int)(distributions->getNX3());
- // assign numbers and create node vector + collect data
+ // int minX1 = 1;
+ // int minX2 = 1;
+ // int minX3 = 1;
+
+ // int maxX1 = (int)(distributions->getNX1());
+ // int maxX2 = (int)(distributions->getNX2());
+ // int maxX3 = (int)(distributions->getNX3());
+
+ // nummern vergeben und node vector erstellen + daten sammeln
CbArray3D<int> nodeNumbers((int)maxX1, (int)maxX2, (int)maxX3, -1);
maxX1 -= 2;
maxX2 -= 2;
maxX3 -= 2;
+ // D3Q27BoundaryConditionPtr bcPtr;
int nr = (int)nodes.size();
for (int ix3 = minX3; ix3 <= maxX3; ix3++) {
@@ -180,44 +207,60 @@ void WriteMacroscopicQuantitiesCoProcessor::addDataMQ(SPtr<Block3D> block)
int index = 0;
nodeNumbers(ix1, ix2, ix3) = nr++;
Vector3D worldCoordinates = grid->getNodeCoordinates(block, ix1, ix2, ix3);
- nodes.emplace_back(float(worldCoordinates[0]), float(worldCoordinates[1]),
- float(worldCoordinates[2]));
+ nodes.push_back(UbTupleFloat3(float(worldCoordinates[0]), float(worldCoordinates[1]),
+ float(worldCoordinates[2])));
distributions->getDistribution(f, ix1, ix2, ix3);
- calcMacros(f, drho, vx1, vx2, vx3);
- press = D3Q27System::calcPress(f, drho, vx1, vx2, vx3);
-
- if (UbMath::isNaN(drho) || UbMath::isInfinity(drho))
- UB_THROW(UbException(
- UB_EXARGS, "drho is not a number (nan or -1.#IND) or infinity number -1.#INF in block=" +
- block->toString() + ", node=" + UbSystem::toString(ix1) + "," +
- UbSystem::toString(ix2) + "," + UbSystem::toString(ix3)));
+ calcMacros(f, rho, vx1, vx2, vx3);
+ double press = D3Q27System::getPressure(f); // D3Q27System::calcPress(f,rho,vx1,vx2,vx3);
+
+ if (UbMath::isNaN(rho) || UbMath::isInfinity(rho))
+ // UB_THROW( UbException(UB_EXARGS,"rho is not a number (nan or -1.#IND) or infinity number
+ // -1.#INF in block="+block->toString()+",
+ // node="+UbSystem::toString(ix1)+","+UbSystem::toString(ix2)+","+UbSystem::toString(ix3)));
+ rho = 999.0;
if (UbMath::isNaN(press) || UbMath::isInfinity(press))
- UB_THROW(UbException(
- UB_EXARGS, "press is not a number (nan or -1.#IND) or infinity number -1.#INF in block=" +
- block->toString() + ", node=" + UbSystem::toString(ix1) + "," +
- UbSystem::toString(ix2) + "," + UbSystem::toString(ix3)));
+ // UB_THROW( UbException(UB_EXARGS,"press is not a number (nan or -1.#IND) or infinity number
+ // -1.#INF in block="+block->toString()+",
+ // node="+UbSystem::toString(ix1)+","+UbSystem::toString(ix2)+","+UbSystem::toString(ix3)));
+ press = 999.0;
if (UbMath::isNaN(vx1) || UbMath::isInfinity(vx1))
- UB_THROW(UbException(
- UB_EXARGS, "vx1 is not a number (nan or -1.#IND) or infinity number -1.#INF in block=" +
- block->toString() + ", node=" + UbSystem::toString(ix1) + "," +
- UbSystem::toString(ix2) + "," + UbSystem::toString(ix3)));
+ // UB_THROW( UbException(UB_EXARGS,"vx1 is not a number (nan or -1.#IND) or infinity number
+ // -1.#INF in block="+block->toString()+",
+ // node="+UbSystem::toString(ix1)+","+UbSystem::toString(ix2)+","+UbSystem::toString(ix3)));
+ vx1 = 999.0;
if (UbMath::isNaN(vx2) || UbMath::isInfinity(vx2))
- UB_THROW(UbException(
- UB_EXARGS, "vx2 is not a number (nan or -1.#IND) or infinity number -1.#INF in block=" +
- block->toString() + ", node=" + UbSystem::toString(ix1) + "," +
- UbSystem::toString(ix2) + "," + UbSystem::toString(ix3)));
+ // UB_THROW( UbException(UB_EXARGS,"vx2 is not a number (nan or -1.#IND) or infinity number
+ // -1.#INF in block="+block->toString()+",
+ // node="+UbSystem::toString(ix1)+","+UbSystem::toString(ix2)+","+UbSystem::toString(ix3)));
+ vx2 = 999.0;
if (UbMath::isNaN(vx3) || UbMath::isInfinity(vx3))
- UB_THROW(UbException(
- UB_EXARGS, "vx3 is not a number (nan or -1.#IND) or infinity number -1.#INF in block=" +
- block->toString() + ", node=" + UbSystem::toString(ix1) + "," +
- UbSystem::toString(ix2) + "," + UbSystem::toString(ix3)));
-
- data[index++].push_back(drho * conv->getFactorDensityLbToW());
- data[index++].push_back(press * conv->getFactorPressureLbToW());
- data[index++].push_back(vx1 * conv->getFactorVelocityLbToW());
- data[index++].push_back(vx2 * conv->getFactorVelocityLbToW());
- data[index++].push_back(vx3 * conv->getFactorVelocityLbToW());
+ // UB_THROW( UbException(UB_EXARGS,"vx3 is not a number (nan or -1.#IND) or infinity number
+ // -1.#INF in block="+block->toString()+",
+ // node="+UbSystem::toString(ix1)+","+UbSystem::toString(ix2)+","+UbSystem::toString(ix3)));
+ vx3 = 999.0;
+
+ data[index++].push_back(rho);
+ data[index++].push_back(vx1);
+ data[index++].push_back(vx2);
+ data[index++].push_back(vx3);
+
+ // shearRate = D3Q27System::getShearRate(f, collFactor);
+
+ // LBMReal collFactorF = BinghamModelLBMKernel::getBinghamCollFactor(collFactor, yieldStress,
+ // shearRate, rho);
+
+ // data[index++].push_back(shearRate);
+ // data[index++].push_back(collFactorF);
+
+ // data[index++].push_back((rho+1.0) * conv->getFactorDensityLbToW() );
+ // data[index++].push_back(vx1 * conv->getFactorVelocityLbToW());
+ // data[index++].push_back(vx2 * conv->getFactorVelocityLbToW());
+ // data[index++].push_back(vx3 * conv->getFactorVelocityLbToW());
+ // data[index++].push_back((press * conv->getFactorPressureLbToW()) / ((rho+1.0) *
+ // conv->getFactorDensityLbToW()));
+ data[index++].push_back(level);
+ // data[index++].push_back(blockID);
}
}
}
@@ -234,7 +277,9 @@ void WriteMacroscopicQuantitiesCoProcessor::addDataMQ(SPtr<Block3D> block)
(SWT = nodeNumbers(ix1, ix2, ix3 + 1)) >= 0 && (SET = nodeNumbers(ix1 + 1, ix2, ix3 + 1)) >= 0 &&
(NET = nodeNumbers(ix1 + 1, ix2 + 1, ix3 + 1)) >= 0 &&
(NWT = nodeNumbers(ix1, ix2 + 1, ix3 + 1)) >= 0) {
- cells.push_back(makeUbTuple(SWB, SEB, NEB, NWB, SWT, SET, NET, NWT));
+ cells.push_back(makeUbTuple((unsigned int)SWB, (unsigned int)SEB, (unsigned int)NEB,
+ (unsigned int)NWB, (unsigned int)SWT, (unsigned int)SET,
+ (unsigned int)NET, (unsigned int)NWT));
}
}
}
diff --git a/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.cpp b/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.cpp
index d7e2286d4599aabf4e7b6e4c9d6a824b38e6d873..6f8a6e74664cf82a550b9000071d4f6beb9ebac2 100644
--- a/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.cpp
+++ b/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.cpp
@@ -47,7 +47,7 @@ D3Q27EsoTwist3DSplittedVector::D3Q27EsoTwist3DSplittedVector(size_t nx1, size_t
this->nonLocalDistributions =
std::make_shared<CbArray4D<LBMReal, IndexerX4X3X2X1>>(13, nx1 + 1, nx2 + 1, nx3 + 1, value);
- this->restDistributions = std::make_shared<CbArray3D<LBMReal, IndexerX3X2X1>>(nx1, nx2, nx3, value);
+ this->zeroDistributions = std::make_shared<CbArray3D<LBMReal, IndexerX3X2X1>>(nx1, nx2, nx3, value);
}
//////////////////////////////////////////////////////////////////////////
D3Q27EsoTwist3DSplittedVector::~D3Q27EsoTwist3DSplittedVector() = default;
@@ -84,7 +84,7 @@ void D3Q27EsoTwist3DSplittedVector::getDistribution(LBMReal *const f, size_t x1,
f[D3Q27System::BNW] = (*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1 + 1, x2, x3 + 1);
f[D3Q27System::BNE] = (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1);
- f[D3Q27System::REST] = (*this->restDistributions)(x1, x2, x3);
+ f[D3Q27System::ZERO] = (*this->zeroDistributions)(x1, x2, x3);
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setDistribution(const LBMReal *const f, size_t x1, size_t x2, size_t x3)
@@ -117,7 +117,7 @@ void D3Q27EsoTwist3DSplittedVector::setDistribution(const LBMReal *const f, size
(*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1 + 1, x2, x3 + 1) = f[D3Q27System::INV_BNW];
(*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1) = f[D3Q27System::INV_BNE];
- (*this->restDistributions)(x1, x2, x3) = f[D3Q27System::REST];
+ (*this->zeroDistributions)(x1, x2, x3) = f[D3Q27System::ZERO];
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::getDistributionInv(LBMReal *const f, size_t x1, size_t x2, size_t x3)
@@ -150,7 +150,7 @@ void D3Q27EsoTwist3DSplittedVector::getDistributionInv(LBMReal *const f, size_t
f[D3Q27System::INV_BNW] = (*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1 + 1, x2, x3 + 1);
f[D3Q27System::INV_BNE] = (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1);
- f[D3Q27System::REST] = (*this->restDistributions)(x1, x2, x3);
+ f[D3Q27System::ZERO] = (*this->zeroDistributions)(x1, x2, x3);
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setDistributionInv(const LBMReal *const f, size_t x1, size_t x2, size_t x3)
@@ -183,7 +183,7 @@ void D3Q27EsoTwist3DSplittedVector::setDistributionInv(const LBMReal *const f, s
(*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1 + 1, x2, x3 + 1) = f[D3Q27System::BNW];
(*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1) = f[D3Q27System::BNE];
- (*this->restDistributions)(x1, x2, x3) = f[D3Q27System::REST];
+ (*this->zeroDistributions)(x1, x2, x3) = f[D3Q27System::ZERO];
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setDistributionForDirection(const LBMReal *const f, size_t x1, size_t x2, size_t x3,
@@ -241,8 +241,8 @@ void D3Q27EsoTwist3DSplittedVector::setDistributionForDirection(const LBMReal *c
(*this->localDistributions)(D3Q27System::ET_TSW, x1 + 1, x2 + 1, x3) = f[D3Q27System::BNE];
if ((direction & EsoTwistD3Q27System::etTSW) == EsoTwistD3Q27System::etTSW)
(*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1) = f[D3Q27System::TSW];
- if ((direction & EsoTwistD3Q27System::REST) == EsoTwistD3Q27System::REST)
- (*this->restDistributions)(x1, x2, x3) = f[D3Q27System::REST];
+ if ((direction & EsoTwistD3Q27System::ZERO) == EsoTwistD3Q27System::ZERO)
+ (*this->zeroDistributions)(x1, x2, x3) = f[D3Q27System::ZERO];
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setDistributionForDirection(LBMReal f, size_t x1, size_t x2, size_t x3,
@@ -327,8 +327,8 @@ void D3Q27EsoTwist3DSplittedVector::setDistributionForDirection(LBMReal f, size_
case D3Q27System::TSW:
(*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1) = f;
break;
- case D3Q27System::REST:
- (*this->restDistributions)(x1, x2, x3) = f;
+ case D3Q27System::ZERO:
+ (*this->zeroDistributions)(x1, x2, x3) = f;
break;
default:
UB_THROW(UbException(UB_EXARGS, "Direction didn't find"));
@@ -390,8 +390,8 @@ void D3Q27EsoTwist3DSplittedVector::setDistributionInvForDirection(const LBMReal
(*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1) = f[D3Q27System::BNE];
if ((direction & EsoTwistD3Q27System::etTSW) == EsoTwistD3Q27System::etTSW)
(*this->localDistributions)(D3Q27System::ET_TSW, x1 + 1, x2 + 1, x3) = f[D3Q27System::TSW];
- if ((direction & EsoTwistD3Q27System::REST) == EsoTwistD3Q27System::REST)
- (*this->restDistributions)(x1, x2, x3) = f[D3Q27System::REST];
+ if ((direction & EsoTwistD3Q27System::ZERO) == EsoTwistD3Q27System::ZERO)
+ (*this->zeroDistributions)(x1, x2, x3) = f[D3Q27System::ZERO];
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setDistributionInvForDirection(LBMReal f, size_t x1, size_t x2, size_t x3,
@@ -476,8 +476,8 @@ void D3Q27EsoTwist3DSplittedVector::setDistributionInvForDirection(LBMReal f, si
case D3Q27System::TSW:
(*this->localDistributions)(D3Q27System::ET_TSW, x1 + 1, x2 + 1, x3) = f;
break;
- case D3Q27System::REST:
- (*this->restDistributions)(x1, x2, x3) = f;
+ case D3Q27System::ZERO:
+ (*this->zeroDistributions)(x1, x2, x3) = f;
break;
default:
UB_THROW(UbException(UB_EXARGS, "Direction didn't find"));
@@ -539,8 +539,8 @@ LBMReal D3Q27EsoTwist3DSplittedVector::getDistributionForDirection(size_t x1, si
return (*this->localDistributions)(D3Q27System::ET_TSW, x1 + 1, x2 + 1, x3);
case D3Q27System::BNE:
return (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1);
- case D3Q27System::REST:
- return (*this->restDistributions)(x1, x2, x3);
+ case D3Q27System::ZERO:
+ return (*this->zeroDistributions)(x1, x2, x3);
default:
UB_THROW(UbException(UB_EXARGS, "Direction didn't find"));
}
@@ -601,8 +601,8 @@ LBMReal D3Q27EsoTwist3DSplittedVector::getDistributionInvForDirection(size_t x1,
return (*this->localDistributions)(D3Q27System::ET_TSW, x1 + 1, x2 + 1, x3);
case D3Q27System::TSW:
return (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3 + 1);
- case D3Q27System::REST:
- return (*this->restDistributions)(x1, x2, x3);
+ case D3Q27System::ZERO:
+ return (*this->zeroDistributions)(x1, x2, x3);
default:
UB_THROW(UbException(UB_EXARGS, "Direction didn't find"));
}
@@ -626,7 +626,7 @@ CbArray4D<LBMReal, IndexerX4X3X2X1>::CbArray4DPtr D3Q27EsoTwist3DSplittedVector:
//////////////////////////////////////////////////////////////////////////
CbArray3D<LBMReal, IndexerX3X2X1>::CbArray3DPtr D3Q27EsoTwist3DSplittedVector::getZeroDistributions()
{
- return this->restDistributions;
+ return this->zeroDistributions;
}
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setNX1(size_t newNX1) { NX1 = newNX1; }
@@ -647,7 +647,7 @@ void D3Q27EsoTwist3DSplittedVector::setNonLocalDistributions(CbArray4D<LBMReal,
//////////////////////////////////////////////////////////////////////////
void D3Q27EsoTwist3DSplittedVector::setZeroDistributions(CbArray3D<LBMReal, IndexerX3X2X1>::CbArray3DPtr array)
{
- restDistributions = array;
+ zeroDistributions = array;
}
//////////////////////////////////////////////////////////////////////////
diff --git a/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.h b/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.h
index df44457b04a36918643400d60ec5f514e32982ea..1c0d7d05f1392c8c116863e9e0b41000c90ed15e 100644
--- a/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.h
+++ b/src/cpu/VirtualFluidsCore/Data/D3Q27EsoTwist3DSplittedVector.h
@@ -100,7 +100,7 @@ public:
protected:
CbArray4D<LBMReal, IndexerX4X3X2X1>::CbArray4DPtr localDistributions;
CbArray4D<LBMReal, IndexerX4X3X2X1>::CbArray4DPtr nonLocalDistributions;
- CbArray3D<LBMReal, IndexerX3X2X1>::CbArray3DPtr restDistributions;
+ CbArray3D<LBMReal, IndexerX3X2X1>::CbArray3DPtr zeroDistributions;
size_t NX1, NX2, NX3;
friend class MPIIORestartCoProcessor;
diff --git a/src/cpu/VirtualFluidsCore/Data/DataSet3D.h b/src/cpu/VirtualFluidsCore/Data/DataSet3D.h
index c0171588bcd4a74680326c44db062dba63d4c41c..4930beeab491caf541c436a2e60b326b7cd54c64 100644
--- a/src/cpu/VirtualFluidsCore/Data/DataSet3D.h
+++ b/src/cpu/VirtualFluidsCore/Data/DataSet3D.h
@@ -51,6 +51,9 @@ public:
SPtr<DistributionArray3D> getFdistributions() const;
void setFdistributions(SPtr<DistributionArray3D> distributions);
+ SPtr<DistributionArray3D> getHdistributions() const;
+ void setHdistributions(SPtr<DistributionArray3D> distributions);
+
SPtr<AverageValuesArray3D> getAverageDensity() const;
void setAverageDensity(SPtr<AverageValuesArray3D> values);
@@ -71,10 +74,12 @@ public:
SPtr<RelaxationFactorArray3D> getRelaxationFactor() const;
void setRelaxationFactor(SPtr<RelaxationFactorArray3D> values);
-
protected:
private:
SPtr<DistributionArray3D> fdistributions;
+
+ SPtr<DistributionArray3D> hdistributions;
+
SPtr<AverageValuesArray3D> averageValues;
SPtr<AverageValuesArray3D> averageDensity;
@@ -85,40 +90,96 @@ private:
SPtr<ShearStressValuesArray3D> shearStressValues;
SPtr<RelaxationFactorArray3D> relaxationFactor;
+
};
-inline SPtr<DistributionArray3D> DataSet3D::getFdistributions() const { return fdistributions; }
+inline SPtr<DistributionArray3D> DataSet3D::getFdistributions() const
+{
+ return fdistributions;
+}
-inline void DataSet3D::setFdistributions(SPtr<DistributionArray3D> distributions) { fdistributions = distributions; }
+inline void DataSet3D::setFdistributions(SPtr<DistributionArray3D> distributions)
+{
+ fdistributions = distributions;
+}
-inline SPtr<AverageValuesArray3D> DataSet3D::getAverageValues() const { return averageValues; }
+inline SPtr<DistributionArray3D> DataSet3D::getHdistributions() const
+{
+ return hdistributions;
+}
-inline void DataSet3D::setAverageValues(SPtr<AverageValuesArray3D> values) { averageValues = values; }
+inline void DataSet3D::setHdistributions(SPtr<DistributionArray3D> distributions)
+{
+ hdistributions = distributions;
+}
-inline SPtr<AverageValuesArray3D> DataSet3D::getAverageDensity() const { return averageDensity; }
+inline SPtr<AverageValuesArray3D> DataSet3D::getAverageValues() const
+{
+ return averageValues;
+}
-inline void DataSet3D::setAverageDensity(SPtr<AverageValuesArray3D> values) { averageDensity = values; }
+inline void DataSet3D::setAverageValues(SPtr<AverageValuesArray3D> values)
+{
+ averageValues = values;
+}
-inline SPtr<AverageValuesArray3D> DataSet3D::getAverageVelocity() const { return averageVelocity; }
+inline SPtr<AverageValuesArray3D> DataSet3D::getAverageDensity() const
+{
+ return averageDensity;
+}
-inline void DataSet3D::setAverageVelocity(SPtr<AverageValuesArray3D> values) { averageVelocity = values; }
+inline void DataSet3D::setAverageDensity(SPtr<AverageValuesArray3D> values)
+{
+ averageDensity = values;
+}
-inline SPtr<AverageValuesArray3D> DataSet3D::getAverageFluctuations() const { return averageFluktuations; }
+inline SPtr<AverageValuesArray3D> DataSet3D::getAverageVelocity() const
+{
+ return averageVelocity;
+}
+
+inline void DataSet3D::setAverageVelocity(SPtr<AverageValuesArray3D> values)
+{
+ averageVelocity = values;
+}
-inline void DataSet3D::setAverageFluctuations(SPtr<AverageValuesArray3D> values) { averageFluktuations = values; }
+inline SPtr<AverageValuesArray3D> DataSet3D::getAverageFluctuations() const
+{
+ return averageFluktuations;
+}
+
+inline void DataSet3D::setAverageFluctuations(SPtr<AverageValuesArray3D> values)
+{
+ averageFluktuations = values;
+}
-inline SPtr<AverageValuesArray3D> DataSet3D::getAverageTriplecorrelations() const { return averageTriplecorrelations; }
+inline SPtr<AverageValuesArray3D> DataSet3D::getAverageTriplecorrelations() const
+{
+ return averageTriplecorrelations;
+}
inline void DataSet3D::setAverageTriplecorrelations(SPtr<AverageValuesArray3D> values)
{
averageTriplecorrelations = values;
}
-inline SPtr<ShearStressValuesArray3D> DataSet3D::getShearStressValues() const { return shearStressValues; }
+inline SPtr<ShearStressValuesArray3D> DataSet3D::getShearStressValues() const
+{
+ return shearStressValues;
+}
-inline void DataSet3D::setShearStressValues(SPtr<ShearStressValuesArray3D> values) { shearStressValues = values; }
+inline void DataSet3D::setShearStressValues(SPtr<ShearStressValuesArray3D> values)
+{
+ shearStressValues = values;
+}
-inline SPtr<RelaxationFactorArray3D> DataSet3D::getRelaxationFactor() const { return relaxationFactor; }
+inline SPtr<RelaxationFactorArray3D> DataSet3D::getRelaxationFactor() const
+{
+ return relaxationFactor;
+}
-inline void DataSet3D::setRelaxationFactor(SPtr<RelaxationFactorArray3D> values) { relaxationFactor = values; }
+inline void DataSet3D::setRelaxationFactor(SPtr<RelaxationFactorArray3D> values)
+{
+ relaxationFactor = values;
+}
#endif
diff --git a/src/cpu/VirtualFluidsCore/Data/DistributionArray3D.h b/src/cpu/VirtualFluidsCore/Data/DistributionArray3D.h
index 242c6b7a1c156216a88d4c06c4945040383e39af..8fe4dccea1b53da0513a093e8a741cd0071caf48 100644
--- a/src/cpu/VirtualFluidsCore/Data/DistributionArray3D.h
+++ b/src/cpu/VirtualFluidsCore/Data/DistributionArray3D.h
@@ -41,9 +41,9 @@ class DistributionArray3D
{
public:
DistributionArray3D() = default;
- ;
+
virtual ~DistributionArray3D() = default;
- ;
+
//! get number of nodes for x1 direction
virtual size_t getNX1() const = 0;
//! get number of nodes for x2 direction
diff --git a/src/cpu/VirtualFluidsCore/Data/EsoTwist3D.h b/src/cpu/VirtualFluidsCore/Data/EsoTwist3D.h
index 689589222634edfc52dcfbf358ffc8d32dd1186b..319a9200cc204b0f9b869b2e52353e717a89d783 100644
--- a/src/cpu/VirtualFluidsCore/Data/EsoTwist3D.h
+++ b/src/cpu/VirtualFluidsCore/Data/EsoTwist3D.h
@@ -43,13 +43,23 @@
// Geier, M., & Schönherr, M. (2017). Esoteric twist: an efficient in-place streaming algorithmus for the lattice
// Boltzmann method on massively parallel hardware. Computation, 5(2), 19.
+class EsoTwistD3Q27UnrollArray
+{
+};
+class EsoTwistPlusD3Q27UnrollArray
+{
+};
+class EsoTwistPlusD3Q19UnrollArray
+{
+};
+
class EsoTwist3D : public DistributionArray3D
{
public:
EsoTwist3D() = default;
- ;
+
~EsoTwist3D() override = default;
- ;
+
//////////////////////////////////////////////////////////////////////////
void swap() override = 0;
//////////////////////////////////////////////////////////////////////////
diff --git a/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.cpp b/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.cpp
index c456be678449744475a0ac6932850dceb0ee6f1c..1a13aa008ab49a48f1d16c7a2a71ea39dfb191ab 100644
--- a/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.cpp
+++ b/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.cpp
@@ -35,7 +35,7 @@
// index 0 1 2 3 4 5 6 7 8 9 10 11 12
// 13 14 15 16 17 18 19 20 21 22 23 24 25 26 f: E, W, N, S, T, B, NE, SW, SE, NW, TE, BW, BE, TW, TN,
-// BS, BN, TS, TNE TNW TSE TSW BNE BNW BSE BSW REST
+// BS, BN, TS, TNE TNW TSE TSW BNE BNW BSE BSW ZERO
const int EsoTwistD3Q27System::ETX1[EsoTwistD3Q27System::ENDF + 1] = { 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1,
0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1, 0 };
const int EsoTwistD3Q27System::ETX2[EsoTwistD3Q27System::ENDF + 1] = { 0, 0, 0, 1, 0, 0, 0, 1, 0, -1, 0, 0, 0, 0,
@@ -49,7 +49,7 @@ const int EsoTwistD3Q27System::etINVDIR[EsoTwistD3Q27System::ENDF + 1] = {
D3Q27System::INV_TE, D3Q27System::INV_BW, D3Q27System::INV_BE, D3Q27System::INV_TW, D3Q27System::INV_TN,
D3Q27System::INV_BS, D3Q27System::INV_BN, D3Q27System::INV_TS, D3Q27System::INV_TNE, D3Q27System::INV_TNW,
D3Q27System::INV_TSE, D3Q27System::INV_TSW, D3Q27System::INV_BNE, D3Q27System::INV_BNW, D3Q27System::INV_BSE,
- D3Q27System::INV_BSW, D3Q27System::REST
+ D3Q27System::INV_BSW, D3Q27System::ZERO
};
const unsigned long int EsoTwistD3Q27System::etDIR[EsoTwistD3Q27System::ENDF + 1] = {
diff --git a/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.h b/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.h
index 21752cc48a84b02bc24cb7efe9e3c5912f476dfd..a9214673ec4b4a66a52fa53f9b625ead0180768b 100644
--- a/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.h
+++ b/src/cpu/VirtualFluidsCore/Data/EsoTwistD3Q27System.h
@@ -47,7 +47,7 @@ struct EsoTwistD3Q27System {
const static int STARTDIR = D3Q27System::STARTDIR;
const static int ENDDIR = D3Q27System::ENDDIR;
- static const int REST = D3Q27System::REST; /*f0 */
+ static const int ZERO = D3Q27System::ZERO; /*f0 */
static const int E = D3Q27System::E; /*f1 */
static const int W = D3Q27System::W; /*f2 */
static const int N = D3Q27System::N; /*f3 */
diff --git a/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.cpp b/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.cpp
index 724d855e3dbfecbe3388ad4071a2a1b1666c1010..f1b2e5ad8c62babaea92cea50a646434f9757cd9 100644
--- a/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.cpp
+++ b/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.cpp
@@ -127,7 +127,7 @@ void BasicCalculator::calculate()
if (refinement) {
if (straightStartLevel < maxInitLevel)
exchangeBlockData(straightStartLevel, maxInitLevel);
- //////////////////////////////////////////////////////////////////////////
+ //////////////////////////////////////////////////////////////////////////
#ifdef TIMING
time[4] = timer.stop();
UBLOG(logINFO, "refinement exchangeBlockData time = " << time[4]);
@@ -155,14 +155,16 @@ void BasicCalculator::calculate()
} catch (std::exception &e) {
UBLOG(logERROR, e.what());
UBLOG(logERROR, " step = " << calcStep);
- // throw;
- exit(EXIT_FAILURE);
+ // throw e;
+ // exit(EXIT_FAILURE);
} catch (std::string &s) {
UBLOG(logERROR, s);
- exit(EXIT_FAILURE);
+ // exit(EXIT_FAILURE);
+ // throw s;
} catch (...) {
UBLOG(logERROR, "unknown exception");
- exit(EXIT_FAILURE);
+ // exit(EXIT_FAILURE);
+ // throw;
}
}
//////////////////////////////////////////////////////////////////////////
@@ -173,28 +175,27 @@ void BasicCalculator::calculateBlocks(int startLevel, int maxInitLevel, int calc
#endif
{
SPtr<Block3D> blockTemp;
- try {
- // startLevel bis maxInitLevel
- for (int level = startLevel; level <= maxInitLevel; level++) {
- // timer.resetAndStart();
- // call LBM kernel
- int size = (int)blocks[level].size();
+ // startLevel bis maxInitLevel
+ for (int level = startLevel; level <= maxInitLevel; level++) {
+ // timer.resetAndStart();
+ // call LBM kernel
+ int size = (int)blocks[level].size();
#ifdef _OPENMP
#pragma omp for schedule(OMP_SCHEDULE)
#endif
- for (int i = 0; i < size; i++) {
+ for (int i = 0; i < size; i++) {
+ try {
blockTemp = blocks[level][i];
blockTemp->getKernel()->calculate(calcStep);
+ } catch (std::exception &e) {
+ UBLOG(logERROR, e.what());
+ UBLOG(logERROR, blockTemp->toString() << " step = " << calcStep);
+ std::exit(EXIT_FAILURE);
}
- // timer.stop();
- // UBLOG(logINFO, "level = " << level << " blocks = " << blocks[level].size() << " collision time = " <<
- // timer.getTotalTime());
}
- } catch (std::exception &e) {
- UBLOG(logERROR, e.what());
- // UBLOG(logERROR, blockTemp->toString()<<" step = "<<calcStep);
- // throw;
- exit(EXIT_FAILURE);
+ // timer.stop();
+ // UBLOG(logINFO, "level = " << level << " blocks = " << blocks[level].size() << " collision time = " <<
+ // timer.getTotalTime());
}
}
}
@@ -239,8 +240,13 @@ void BasicCalculator::connectorsPrepareLocal(std::vector<SPtr<Block3DConnector>>
#pragma omp parallel for schedule(OMP_SCHEDULE)
#endif
for (int i = 0; i < size; i++) {
- connectors[i]->prepareForReceive();
- connectors[i]->prepareForSend();
+ try {
+ connectors[i]->prepareForReceive();
+ connectors[i]->prepareForSend();
+ } catch (std::exception &e) {
+ UBLOG(logERROR, e.what());
+ std::exit(EXIT_FAILURE);
+ }
}
}
//////////////////////////////////////////////////////////////////////////
@@ -251,8 +257,13 @@ void BasicCalculator::connectorsSendLocal(std::vector<SPtr<Block3DConnector>> &c
#pragma omp parallel for schedule(OMP_SCHEDULE)
#endif
for (int i = 0; i < size; i++) {
- connectors[i]->fillSendVectors();
- connectors[i]->sendVectors();
+ try {
+ connectors[i]->fillSendVectors();
+ connectors[i]->sendVectors();
+ } catch (std::exception &e) {
+ UBLOG(logERROR, e.what());
+ std::exit(EXIT_FAILURE);
+ }
}
}
//////////////////////////////////////////////////////////////////////////
@@ -321,36 +332,43 @@ void BasicCalculator::applyPreCollisionBC(int startLevel, int maxInitLevel)
#pragma omp parallel for schedule(OMP_SCHEDULE)
#endif
for (int i = 0; i < size; i++) {
- blocks[level][i]->getKernel()->getBCProcessor()->applyPreCollisionBC();
+ try {
+ blocks[level][i]->getKernel()->getBCProcessor()->applyPreCollisionBC();
+ } catch (std::exception &e) {
+ UBLOG(logERROR, e.what());
+ exit(EXIT_FAILURE);
+ } catch (std::string &s) {
+ UBLOG(logERROR, s);
+ exit(EXIT_FAILURE);
+ } catch (...) {
+ UBLOG(logERROR, "unknown exception");
+ exit(EXIT_FAILURE);
+ }
}
}
}
//////////////////////////////////////////////////////////////////////////
void BasicCalculator::applyPostCollisionBC(int startLevel, int maxInitLevel)
{
- try {
- // from startLevel to maxInitLevel
- for (int level = startLevel; level <= maxInitLevel; level++) {
- int size = (int)blocks[level].size();
+ // from startLevel to maxInitLevel
+ for (int level = startLevel; level <= maxInitLevel; level++) {
+ int size = (int)blocks[level].size();
#ifdef _OPENMP
#pragma omp parallel for schedule(OMP_SCHEDULE)
#endif
- for (int i = 0; i < size; i++) {
+ for (int i = 0; i < size; i++) {
+ try {
blocks[level][i]->getKernel()->getBCProcessor()->applyPostCollisionBC();
+ } catch (std::exception &e) {
+ UBLOG(logERROR, e.what());
+ exit(EXIT_FAILURE);
+ } catch (std::string &s) {
+ UBLOG(logERROR, s);
+ exit(EXIT_FAILURE);
+ } catch (...) {
+ UBLOG(logERROR, "unknown exception");
+ exit(EXIT_FAILURE);
}
}
- } catch (std::exception &e) {
- UBLOG(logERROR, e.what());
- // UBLOG(logERROR, " step = "<<calcStep);
- // throw;
- exit(EXIT_FAILURE);
- } catch (std::string &s) {
- UBLOG(logERROR, s);
- // throw;
- exit(EXIT_FAILURE);
- } catch (...) {
- UBLOG(logERROR, "unknown exception");
- // throw;
- exit(EXIT_FAILURE);
}
}
diff --git a/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.h b/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.h
index fb6e68ce0a56ed7e407ab23605a1facd83eace52..3ef1f4c712e552ea5d5b5e82306e2bd94d74d7ab 100644
--- a/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.h
+++ b/src/cpu/VirtualFluidsCore/Grid/BasicCalculator.h
@@ -39,7 +39,8 @@
class Block3DConnector;
//! \class BasicCalculator
-//! \brief Class implements basic functionality with OpenMP parallelization for main calculation LBM loop
+//! \brief Class implements basic functionality with MPI + OpenMP parallelization for main calculation loop
+//! \author Konstantin Kutscher
class BasicCalculator : public Calculator
{
diff --git a/src/cpu/VirtualFluidsCore/Grid/Block3D.cpp b/src/cpu/VirtualFluidsCore/Grid/Block3D.cpp
index 5e921c4f20390da51724f8e53caaab23e0270549..79753c144f5cfff831f1d0415e9434c50b11bcea 100644
--- a/src/cpu/VirtualFluidsCore/Grid/Block3D.cpp
+++ b/src/cpu/VirtualFluidsCore/Grid/Block3D.cpp
@@ -318,6 +318,13 @@ void Block3D::deleteInterpolationFlag()
interpolationFlagCF = 0;
}
//////////////////////////////////////////////////////////////////////////
+double Block3D::getWorkLoad()
+{
+ double l = kernel->getCalculationTime();
+ l *= static_cast<double>(1 << level);
+ return l;
+}
+//////////////////////////////////////////////////////////////////////////
std::string Block3D::toString()
{
std::stringstream ss;
diff --git a/src/cpu/VirtualFluidsCore/Grid/Block3D.h b/src/cpu/VirtualFluidsCore/Grid/Block3D.h
index 7a4a2aad75825d6e2aabd252f9c480ca3a96c5a5..b2279b069e6ee322023d30419f8eed5c587f95e8 100644
--- a/src/cpu/VirtualFluidsCore/Grid/Block3D.h
+++ b/src/cpu/VirtualFluidsCore/Grid/Block3D.h
@@ -134,6 +134,8 @@ public:
bool hasInterpolationFlagFC(int dir);
bool hasInterpolationFlagFC();
+ double getWorkLoad();
+
std::string toString();
static int getMaxGlobalID() { return counter; }
diff --git a/src/cpu/VirtualFluidsCore/Grid/Calculator.cpp b/src/cpu/VirtualFluidsCore/Grid/Calculator.cpp
index 083b30b9a4080183b2543dfe14f33d3551a093ec..fbeb2de979bb31dfb87441b5cfcfdf3393f0043c 100644
--- a/src/cpu/VirtualFluidsCore/Grid/Calculator.cpp
+++ b/src/cpu/VirtualFluidsCore/Grid/Calculator.cpp
@@ -124,11 +124,11 @@ void Calculator::initRemoteConnectors()
// grid->getBlocks(level, gridRank, true, blockVector);
grid->getBlocks(l, blockVector);
for (SPtr<Block3D> block : blockVector) {
- int l = block->getLevel();
- block->pushBackRemoteSameLevelConnectors(remoteConns[l]);
+ int block_level = block->getLevel();
+ block->pushBackRemoteSameLevelConnectors(remoteConns[block_level]);
- block->pushBackRemoteInterpolationConnectorsCF(remoteInterConnsCF[l]);
- block->pushBackRemoteInterpolationConnectorsFC(remoteInterConnsFC[l]);
+ block->pushBackRemoteInterpolationConnectorsCF(remoteInterConnsCF[block_level]);
+ block->pushBackRemoteInterpolationConnectorsFC(remoteInterConnsFC[block_level]);
}
}
diff --git a/src/cpu/VirtualFluidsCore/Grid/Grid3D.cpp b/src/cpu/VirtualFluidsCore/Grid/Grid3D.cpp
index c78fda22d5f1b2fe95fcac5a94435f72cea1c6a5..92be5ed5a06e1909a34144cdd0d1b31000309281 100644
--- a/src/cpu/VirtualFluidsCore/Grid/Grid3D.cpp
+++ b/src/cpu/VirtualFluidsCore/Grid/Grid3D.cpp
@@ -153,6 +153,17 @@ bool Grid3D::deleteBlock(int ix1, int ix2, int ix3, int level)
return false;
}
}
+void Grid3D::deleteBlocks()
+{
+ std::vector<std::vector<SPtr<Block3D>>> blocksVector(25);
+ int minInitLevel = Grid3DSystem::MINLEVEL;
+ int maxInitLevel = Grid3DSystem::MAXLEVEL;
+ for (int level = minInitLevel; level < maxInitLevel; level++) {
+ getBlocks(level, blocksVector[level]);
+ for (SPtr<Block3D> block : blocksVector[level]) // blocks of the current level
+ deleteBlock(block);
+ }
+}
//////////////////////////////////////////////////////////////////////////
void Grid3D::replaceBlock(SPtr<Block3D> block)
{
@@ -1328,7 +1339,7 @@ void Grid3D::getNeighborBlocksForDirectionWithDirZero(int dir, int ix1, int ix2,
case Grid3DSystem::BSW:
this->getNeighborsBottomSouthWest(ix1, ix2, ix3, level, levelDepth, blocks);
break;
- case Grid3DSystem::REST:
+ case Grid3DSystem::ZERO:
this->getNeighborsZero(ix1, ix2, ix3, level, levelDepth, blocks);
break;
default:
diff --git a/src/cpu/VirtualFluidsCore/Grid/Grid3D.h b/src/cpu/VirtualFluidsCore/Grid/Grid3D.h
index 69c5106847bba69b651ee2c9ed84c0616798b1c3..84c821e84b8c98f17e39814a211de3262e75f804 100644
--- a/src/cpu/VirtualFluidsCore/Grid/Grid3D.h
+++ b/src/cpu/VirtualFluidsCore/Grid/Grid3D.h
@@ -74,6 +74,7 @@ public:
void addBlock(SPtr<Block3D> block);
bool deleteBlock(SPtr<Block3D> block);
bool deleteBlock(int ix1, int ix2, int ix3, int level);
+ void deleteBlocks();
void deleteBlocks(const std::vector<int> &ids);
void replaceBlock(SPtr<Block3D> block);
SPtr<Block3D> getBlock(int ix1, int ix2, int ix3, int level) const;
diff --git a/src/cpu/VirtualFluidsCore/Grid/Grid3DSystem.h b/src/cpu/VirtualFluidsCore/Grid/Grid3DSystem.h
index 008e38b88aec6be8695411d51c263084816fd2e1..ee61b8f7327e76a9393d4d3caa13c3a796470c08 100644
--- a/src/cpu/VirtualFluidsCore/Grid/Grid3DSystem.h
+++ b/src/cpu/VirtualFluidsCore/Grid/Grid3DSystem.h
@@ -71,7 +71,7 @@ static const int BNE = 22;
static const int BNW = 23;
static const int BSE = 24;
static const int BSW = 25;
-static const int REST /*f0 */ = 26;
+static const int ZERO /*f0 */ = 26;
static const int ENDDIR = 25;
@@ -104,6 +104,7 @@ static const int INV_BSW = TNE;
extern const int INVDIR[ENDDIR + 1];
+static const int MINLEVEL = 0;
static const int MAXLEVEL = 25;
extern const int EX1[ENDDIR + 1];
diff --git a/src/cpu/VirtualFluidsCore/Interactors/D3Q27Interactor.cpp b/src/cpu/VirtualFluidsCore/Interactors/D3Q27Interactor.cpp
index 55ce8d1dd332c51d1acd8e1285ea0e81ec9ab00f..bf1895b930f1c61d36d537319b53fe4b0abcd960 100644
--- a/src/cpu/VirtualFluidsCore/Interactors/D3Q27Interactor.cpp
+++ b/src/cpu/VirtualFluidsCore/Interactors/D3Q27Interactor.cpp
@@ -219,7 +219,7 @@ void D3Q27Interactor::initInteractor(const double &timeStep)
else
this->unsetTimeDependent();
- Interactor3D::initInteractor(timeStep);
+ updateBlocks();
}
//////////////////////////////////////////////////////////////////////////
void D3Q27Interactor::updateInteractor(const double ×tep)
diff --git a/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.cpp b/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.cpp
index 0127c9c880f03d574f657ebc43e53ccaa4b67c7e..84526c62598b1d718b1f179228ae2a3f51839856 100644
--- a/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.cpp
+++ b/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.cpp
@@ -290,11 +290,10 @@ void Interactor3D::setInactive() { active = false; }
//////////////////////////////////////////////////////////////////////////
bool Interactor3D::isActive() { return active; }
//////////////////////////////////////////////////////////////////////////
-void Interactor3D::initInteractor(const double & /*timeStep*/)
+void Interactor3D::updateBlocks()
{
- // UBLOG(logINFO, "transBlocks.size = "<<transBlocks.size());
-
- for (SPtr<Block3D> block : bcBlocks) {
+ for (SPtr<Block3D> block : bcBlocks)
+ {
this->setDifferencesToGbObject3D(block);
}
}
diff --git a/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.h b/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.h
index 9bf3a03ba179a1da7fa932a209d78e39b6622bcf..74627b76addaf6badaea678d1c4a20b274234b3a 100644
--- a/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.h
+++ b/src/cpu/VirtualFluidsCore/Interactors/Interactor3D.h
@@ -57,7 +57,7 @@ public:
Interactor3D(SPtr<GbObject3D> geoObject3D, SPtr<Grid3D> grid, int type, Interactor3D::Accuracy a);
virtual ~Interactor3D();
- virtual void initInteractor(const double ×tep = 0);
+ virtual void initInteractor(const double ×tep = 0) = 0;
virtual void updateInteractor(const double ×tep = 0) = 0;
void setSolidBlock(SPtr<Block3D> block);
@@ -76,7 +76,7 @@ public:
SPtr<Grid3D> getGrid3D() const { return grid.lock(); }
void setGrid3D(SPtr<Grid3D> grid) { this->grid = grid; }
virtual SPtr<GbObject3D> getGbObject3D() const { return geoObject3D; }
- virtual bool setDifferencesToGbObject3D(const SPtr<Block3D> /*block*//*, const double& x1, const double& x2, const double& x3, const double& blockLengthX1, const double& blockLengthX2, const double& blockLengthX3, const double& timestep=0*/)
+ virtual bool setDifferencesToGbObject3D(const SPtr<Block3D>)
{
// UBLOG(logINFO, "Interactor3D::setDifferencesToGbObject3D()");
return false;
@@ -123,6 +123,8 @@ protected:
bool isBlockCuttingGeoObject(double minX1, double minX2, double minX3, double maxX1, double maxX2, double maxX3,
double delta);
+ void updateBlocks();
+
SPtr<GbObject3D> geoObject3D;
WPtr<Grid3D> grid;
int type;
diff --git a/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.cpp b/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.cpp
index 9cecabe4653b78df759db47fe3f48a43a090dad5..daed493b9cc1afddbd92acabcd551da0f463ea26 100644
--- a/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.cpp
+++ b/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.cpp
@@ -30,24 +30,24 @@
//! \ingroup LBM
//! \author Konstantin Kutscher, Martin Geier
//=======================================================================================
-
#include "CumulantK17LBMKernel.h"
-#include "BCArray3D.h"
-#include "Block3D.h"
-#include "D3Q27EsoTwist3DSplittedVector.h"
#include "D3Q27System.h"
+#include "D3Q27EsoTwist3DSplittedVector.h"
+#include <cmath>
#include "DataSet3D.h"
#include "LBMKernel.h"
-#include <cmath>
+#include "Block3D.h"
+#include "BCArray3D.h"
#define PROOF_CORRECTNESS
using namespace UbMath;
//////////////////////////////////////////////////////////////////////////
-CumulantK17LBMKernel::CumulantK17LBMKernel() { this->compressible = true; }
-//////////////////////////////////////////////////////////////////////////
-CumulantK17LBMKernel::~CumulantK17LBMKernel(void) = default;
+CumulantK17LBMKernel::CumulantK17LBMKernel()
+{
+ this->compressible = true;
+}
//////////////////////////////////////////////////////////////////////////
void CumulantK17LBMKernel::initDataSet()
{
@@ -75,561 +75,561 @@ SPtr<LBMKernel> CumulantK17LBMKernel::clone()
//////////////////////////////////////////////////////////////////////////
void CumulantK17LBMKernel::calculate(int step)
{
- //////////////////////////////////////////////////////////////////////////
- //! Cumulant K17 Kernel is based on
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //! and
- //! <a href="https://doi.org/10.1016/j.jcp.2017.07.004"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.07.004 ]</b></a>
- //!
- //! The cumulant kernel is executed in the following steps
- //!
- ////////////////////////////////////////////////////////////////////////////////
- //! - Get node index coordinates from thredIdx, blockIdx, blockDim and gridDim.
- //!
- using namespace D3Q27System;
- using namespace std;
-
- //initializing of forcing stuff
- if (withForcing)
- {
- muForcingX1.DefineVar("x1", &muX1); muForcingX1.DefineVar("x2", &muX2); muForcingX1.DefineVar("x3", &muX3);
- muForcingX2.DefineVar("x1", &muX1); muForcingX2.DefineVar("x2", &muX2); muForcingX2.DefineVar("x3", &muX3);
- muForcingX3.DefineVar("x1", &muX1); muForcingX3.DefineVar("x2", &muX2); muForcingX3.DefineVar("x3", &muX3);
-
- muDeltaT = deltaT;
-
- muForcingX1.DefineVar("dt", &muDeltaT);
- muForcingX2.DefineVar("dt", &muDeltaT);
- muForcingX3.DefineVar("dt", &muDeltaT);
-
- muNu = (1.0 / 3.0) * (1.0 / collFactor - 1.0 / 2.0);
-
- muForcingX1.DefineVar("nu", &muNu);
- muForcingX2.DefineVar("nu", &muNu);
- muForcingX3.DefineVar("nu", &muNu);
- }
- /////////////////////////////////////
-
- localDistributions = dynamic_pointer_cast<D3Q27EsoTwist3DSplittedVector>(dataSet->getFdistributions())->getLocalDistributions();
- nonLocalDistributions = dynamic_pointer_cast<D3Q27EsoTwist3DSplittedVector>(dataSet->getFdistributions())->getNonLocalDistributions();
- restDistributions = dynamic_pointer_cast<D3Q27EsoTwist3DSplittedVector>(dataSet->getFdistributions())->getZeroDistributions();
-
- SPtr<BCArray3D> bcArray = this->getBCProcessor()->getBCArray();
-
- const int bcArrayMaxX1 = (int)bcArray->getNX1();
- const int bcArrayMaxX2 = (int)bcArray->getNX2();
- const int bcArrayMaxX3 = (int)bcArray->getNX3();
-
- int minX1 = ghostLayerWidth;
- int minX2 = ghostLayerWidth;
- int minX3 = ghostLayerWidth;
- int maxX1 = bcArrayMaxX1 - ghostLayerWidth;
- int maxX2 = bcArrayMaxX2 - ghostLayerWidth;
- int maxX3 = bcArrayMaxX3 - ghostLayerWidth;
-
- LBMReal omega = collFactor;
-
- for (int x3 = minX3; x3 < maxX3; x3++)
- {
- for (int x2 = minX2; x2 < maxX2; x2++)
- {
- for (int x1 = minX1; x1 < maxX1; x1++)
- {
- if (!bcArray->isSolid(x1, x2, x3) && !bcArray->isUndefined(x1, x2, x3))
+ //////////////////////////////////////////////////////////////////////////
+ //! Cumulant K17 Kernel is based on
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //! and
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.07.004"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.07.004 ]</b></a>
+ //!
+ //! The cumulant kernel is executed in the following steps
+ //!
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from thredIdx, blockIdx, blockDim and gridDim.
+ //!
+
+ using namespace std;
+
+ //initializing of forcing stuff
+ if (withForcing)
+ {
+ muForcingX1.DefineVar("x1", &muX1); muForcingX1.DefineVar("x2", &muX2); muForcingX1.DefineVar("x3", &muX3);
+ muForcingX2.DefineVar("x1", &muX1); muForcingX2.DefineVar("x2", &muX2); muForcingX2.DefineVar("x3", &muX3);
+ muForcingX3.DefineVar("x1", &muX1); muForcingX3.DefineVar("x2", &muX2); muForcingX3.DefineVar("x3", &muX3);
+
+ muDeltaT = deltaT;
+
+ muForcingX1.DefineVar("dt", &muDeltaT);
+ muForcingX2.DefineVar("dt", &muDeltaT);
+ muForcingX3.DefineVar("dt", &muDeltaT);
+
+ muNu = (1.0 / 3.0) * (1.0 / collFactor - 1.0 / 2.0);
+
+ muForcingX1.DefineVar("nu", &muNu);
+ muForcingX2.DefineVar("nu", &muNu);
+ muForcingX3.DefineVar("nu", &muNu);
+ }
+ /////////////////////////////////////
+
+ localDistributions = dynamic_pointer_cast<D3Q27EsoTwist3DSplittedVector>(dataSet->getFdistributions())->getLocalDistributions();
+ nonLocalDistributions = dynamic_pointer_cast<D3Q27EsoTwist3DSplittedVector>(dataSet->getFdistributions())->getNonLocalDistributions();
+ restDistributions = dynamic_pointer_cast<D3Q27EsoTwist3DSplittedVector>(dataSet->getFdistributions())->getZeroDistributions();
+
+ SPtr<BCArray3D> bcArray = this->getBCProcessor()->getBCArray();
+
+ const int bcArrayMaxX1 = (int)bcArray->getNX1();
+ const int bcArrayMaxX2 = (int)bcArray->getNX2();
+ const int bcArrayMaxX3 = (int)bcArray->getNX3();
+
+ int minX1 = ghostLayerWidth;
+ int minX2 = ghostLayerWidth;
+ int minX3 = ghostLayerWidth;
+ int maxX1 = bcArrayMaxX1 - ghostLayerWidth;
+ int maxX2 = bcArrayMaxX2 - ghostLayerWidth;
+ int maxX3 = bcArrayMaxX3 - ghostLayerWidth;
+
+ LBMReal omega = collFactor;
+
+ for (int x3 = minX3; x3 < maxX3; x3++)
+ {
+ for (int x2 = minX2; x2 < maxX2; x2++)
+ {
+ for (int x1 = minX1; x1 < maxX1; x1++)
{
- int x1p = x1 + 1;
- int x2p = x2 + 1;
- int x3p = x3 + 1;
- //////////////////////////////////////////////////////////////////////////
- //////////////////////////////////////////////////////////////////////////
- //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm
- //! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////
- //////////////////////////////////////////////////////////////////////////
-
- //E N T
- //c c c
- //////////
- //W S B
- //a a a
-
- //Rest is b
-
- //mfxyz
- //a - negative
- //b - null
- //c - positive
-
- // a b c
- //-1 0 1
-
- LBMReal mfcbb = (*this->localDistributions)(D3Q27System::ET_E, x1, x2, x3);
- LBMReal mfbcb = (*this->localDistributions)(D3Q27System::ET_N, x1, x2, x3);
- LBMReal mfbbc = (*this->localDistributions)(D3Q27System::ET_T, x1, x2, x3);
- LBMReal mfccb = (*this->localDistributions)(D3Q27System::ET_NE, x1, x2, x3);
- LBMReal mfacb = (*this->localDistributions)(D3Q27System::ET_NW, x1p, x2, x3);
- LBMReal mfcbc = (*this->localDistributions)(D3Q27System::ET_TE, x1, x2, x3);
- LBMReal mfabc = (*this->localDistributions)(D3Q27System::ET_TW, x1p, x2, x3);
- LBMReal mfbcc = (*this->localDistributions)(D3Q27System::ET_TN, x1, x2, x3);
- LBMReal mfbac = (*this->localDistributions)(D3Q27System::ET_TS, x1, x2p, x3);
- LBMReal mfccc = (*this->localDistributions)(D3Q27System::ET_TNE, x1, x2, x3);
- LBMReal mfacc = (*this->localDistributions)(D3Q27System::ET_TNW, x1p, x2, x3);
- LBMReal mfcac = (*this->localDistributions)(D3Q27System::ET_TSE, x1, x2p, x3);
- LBMReal mfaac = (*this->localDistributions)(D3Q27System::ET_TSW, x1p, x2p, x3);
-
- LBMReal mfabb = (*this->nonLocalDistributions)(D3Q27System::ET_W, x1p, x2, x3);
- LBMReal mfbab = (*this->nonLocalDistributions)(D3Q27System::ET_S, x1, x2p, x3);
- LBMReal mfbba = (*this->nonLocalDistributions)(D3Q27System::ET_B, x1, x2, x3p);
- LBMReal mfaab = (*this->nonLocalDistributions)(D3Q27System::ET_SW, x1p, x2p, x3);
- LBMReal mfcab = (*this->nonLocalDistributions)(D3Q27System::ET_SE, x1, x2p, x3);
- LBMReal mfaba = (*this->nonLocalDistributions)(D3Q27System::ET_BW, x1p, x2, x3p);
- LBMReal mfcba = (*this->nonLocalDistributions)(D3Q27System::ET_BE, x1, x2, x3p);
- LBMReal mfbaa = (*this->nonLocalDistributions)(D3Q27System::ET_BS, x1, x2p, x3p);
- LBMReal mfbca = (*this->nonLocalDistributions)(D3Q27System::ET_BN, x1, x2, x3p);
- LBMReal mfaaa = (*this->nonLocalDistributions)(D3Q27System::ET_BSW, x1p, x2p, x3p);
- LBMReal mfcaa = (*this->nonLocalDistributions)(D3Q27System::ET_BSE, x1, x2p, x3p);
- LBMReal mfaca = (*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1p, x2, x3p);
- LBMReal mfcca = (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3p);
-
- LBMReal mfbbb = (*this->restDistributions)(x1, x2, x3);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3)
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- LBMReal drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
- (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
- ((mfabb + mfcbb) + (mfbab + mfbcb)) + (mfbba + mfbbc)) + mfbbb;
-
- LBMReal rho = c1 + drho;
- LBMReal OOrho = c1 / rho;
- ////////////////////////////////////////////////////////////////////////////////////
- LBMReal vvx = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
- (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
- (mfcbb - mfabb)) / rho;
- LBMReal vvy = ((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
- (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
- (mfbcb - mfbab)) / rho;
- LBMReal vvz = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
- (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
- (mfbbc - mfbba)) / rho;
- ////////////////////////////////////////////////////////////////////////////////////
- //forcing
- ///////////////////////////////////////////////////////////////////////////////////////////
- if (withForcing)
- {
- muX1 = static_cast<double>(x1 - 1 + ix1 * maxX1);
- muX2 = static_cast<double>(x2 - 1 + ix2 * maxX2);
- muX3 = static_cast<double>(x3 - 1 + ix3 * maxX3);
-
- forcingX1 = muForcingX1.Eval();
- forcingX2 = muForcingX2.Eval();
- forcingX3 = muForcingX3.Eval();
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Add half of the acceleration (body force) to the velocity as in Eq. (42)
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- vvx += forcingX1 * deltaT * c1o2; // X
- vvy += forcingX2 * deltaT * c1o2; // Y
- vvz += forcingX3 * deltaT * c1o2; // Z
- }
- ////////////////////////////////////////////////////////////////////////////////////
- // calculate the square of velocities for this lattice node
- LBMReal vx2 = vvx * vvx;
- LBMReal vy2 = vvy * vvy;
- LBMReal vz2 = vvz * vvz;
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Set relaxation limiters for third order cumulants to default value \f$ \lambda=0.001 \f$ according to section 6 in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- LBMReal wadjust;
- LBMReal qudricLimitP = c1o100;
- LBMReal qudricLimitM = c1o100;
- LBMReal qudricLimitD = c1o100;
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Chimera transform from well conditioned distributions to central moments as defined in Appendix J in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //! see also Eq. (6)-(14) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- forwardInverseChimeraWithK(mfaaa, mfaab, mfaac, vvz, vz2, c36, c1o36);
- forwardInverseChimeraWithK(mfaba, mfabb, mfabc, vvz, vz2, c9, c1o9);
- forwardInverseChimeraWithK(mfaca, mfacb, mfacc, vvz, vz2, c36, c1o36);
- forwardInverseChimeraWithK(mfbaa, mfbab, mfbac, vvz, vz2, c9, c1o9);
- forwardInverseChimeraWithK(mfbba, mfbbb, mfbbc, vvz, vz2, c9o4, c4o9);
- forwardInverseChimeraWithK(mfbca, mfbcb, mfbcc, vvz, vz2, c9, c1o9);
- forwardInverseChimeraWithK(mfcaa, mfcab, mfcac, vvz, vz2, c36, c1o36);
- forwardInverseChimeraWithK(mfcba, mfcbb, mfcbc, vvz, vz2, c9, c1o9);
- forwardInverseChimeraWithK(mfcca, mfccb, mfccc, vvz, vz2, c36, c1o36);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- forwardInverseChimeraWithK(mfaaa, mfaba, mfaca, vvy, vy2, c6, c1o6);
- forwardChimera(mfaab, mfabb, mfacb, vvy, vy2);
- forwardInverseChimeraWithK(mfaac, mfabc, mfacc, vvy, vy2, c18, c1o18);
- forwardInverseChimeraWithK(mfbaa, mfbba, mfbca, vvy, vy2, c3o2, c2o3);
- forwardChimera(mfbab, mfbbb, mfbcb, vvy, vy2);
- forwardInverseChimeraWithK(mfbac, mfbbc, mfbcc, vvy, vy2, c9o2, c2o9);
- forwardInverseChimeraWithK(mfcaa, mfcba, mfcca, vvy, vy2, c6, c1o6);
- forwardChimera(mfcab, mfcbb, mfccb, vvy, vy2);
- forwardInverseChimeraWithK(mfcac, mfcbc, mfccc, vvy, vy2, c18, c1o18);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- forwardInverseChimeraWithK(mfaaa, mfbaa, mfcaa, vvx, vx2, c1, c1);
- forwardChimera(mfaba, mfbba, mfcba, vvx, vx2);
- forwardInverseChimeraWithK(mfaca, mfbca, mfcca, vvx, vx2, c3, c1o3);
- forwardChimera(mfaab, mfbab, mfcab, vvx, vx2);
- forwardChimera(mfabb, mfbbb, mfcbb, vvx, vx2);
- forwardChimera(mfacb, mfbcb, mfccb, vvx, vx2);
- forwardInverseChimeraWithK(mfaac, mfbac, mfcac, vvx, vx2, c3, c1o3);
- forwardChimera(mfabc, mfbbc, mfcbc, vvx, vx2);
- forwardInverseChimeraWithK(mfacc, mfbcc, mfccc, vvx, vx2, c9, c1o9);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Setting relaxation rates for non-hydrodynamic cumulants (default values). Variable names and equations according to
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //! => [NAME IN PAPER]=[NAME IN CODE]=[DEFAULT VALUE].
- //! - Trace of second order cumulants \f$ C_{200}+C_{020}+C_{002} \f$ used to adjust bulk viscosity:\f$\omega_2=OxxPyyPzz=1.0 \f$.
- //! - Third order cumulants \f$ C_{120}+C_{102} \f$, \f$ C_{210}+C_{012} \f$, \f$ C_{201}+C_{021} \f$: \f$\omega_3=OxyyPxzz\f$ set according to Eq. (111) with simplifications assuming \f$\omega_2=1.0\f$.
- //! - Third order cumulants \f$ C_{120}-C_{102} \f$, \f$ C_{210}-C_{012} \f$, \f$ C_{201}-C_{021} \f$: \f$\omega_4 = OxyyMxzz\f$ set according to Eq. (112) with simplifications assuming \f$\omega_2 = 1.0\f$.
- //! - Third order cumulants \f$ C_{111} \f$: \f$\omega_5 = Oxyz\f$ set according to Eq. (113) with simplifications assuming \f$\omega_2 = 1.0\f$ (modify for different bulk viscosity).
- //! - Fourth order cumulants \f$ C_{220} \f$, \f$ C_{202} \f$, \f$ C_{022} \f$, \f$ C_{211} \f$, \f$ C_{121} \f$, \f$ C_{112} \f$: for simplification all set to the same default value \f$ \omega_6=\omega_7=\omega_8=O4=1.0 \f$.
- //! - Fifth order cumulants \f$ C_{221}\f$, \f$C_{212}\f$, \f$C_{122}\f$: \f$\omega_9=O5=1.0\f$.
- //! - Sixth order cumulant \f$ C_{222}\f$: \f$\omega_{10}=O6=1.0\f$.
- //!
- ////////////////////////////////////////////////////////////
- //2.
- LBMReal OxxPyyPzz = c1;
- ////////////////////////////////////////////////////////////
- //3.
- LBMReal OxyyPxzz = c8 * (-c2 + omega) * ( c1 + c2*omega) / (-c8 - c14*omega + c7*omega*omega);
- LBMReal OxyyMxzz = c8 * (-c2 + omega) * (-c7 + c4*omega) / (c56 - c50*omega + c9*omega*omega);
- LBMReal Oxyz = c24 * (-c2 + omega) * (-c2 - c7*omega + c3*omega*omega) / (c48 + c152*omega - c130*omega*omega + c29*omega*omega*omega);
- ////////////////////////////////////////////////////////////
- //4.
- LBMReal O4 = c1;
- ////////////////////////////////////////////////////////////
- //5.
- LBMReal O5 = c1;
- ////////////////////////////////////////////////////////////
- //6.
- LBMReal O6 = c1;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - A and B: parameters for fourth order convergence of the diffusion term according to Eq. (114) and (115)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //! with simplifications assuming \f$\omega_2 = 1.0\f$ (modify for different bulk viscosity).
- //!
- LBMReal A = (c4 + c2*omega - c3*omega*omega) / (c2 - c7*omega + c5*omega*omega);
- LBMReal B = (c4 + c28*omega - c14*omega*omega) / (c6 - c21*omega + c15*omega*omega);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute cumulants from central moments according to Eq. (20)-(23) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- ////////////////////////////////////////////////////////////
- //4.
- LBMReal CUMcbb = mfcbb - ((mfcaa + c1o3) * mfabb + c2 * mfbba * mfbab) * OOrho;
- LBMReal CUMbcb = mfbcb - ((mfaca + c1o3) * mfbab + c2 * mfbba * mfabb) * OOrho;
- LBMReal CUMbbc = mfbbc - ((mfaac + c1o3) * mfbba + c2 * mfbab * mfabb) * OOrho;
-
- LBMReal CUMcca = mfcca - (((mfcaa * mfaca + c2 * mfbba * mfbba) + c1o3 * (mfcaa + mfaca)) * OOrho - c1o9 * (drho * OOrho));
- LBMReal CUMcac = mfcac - (((mfcaa * mfaac + c2 * mfbab * mfbab) + c1o3 * (mfcaa + mfaac)) * OOrho - c1o9 * (drho * OOrho));
- LBMReal CUMacc = mfacc - (((mfaac * mfaca + c2 * mfabb * mfabb) + c1o3 * (mfaac + mfaca)) * OOrho - c1o9 * (drho * OOrho));
- ////////////////////////////////////////////////////////////
- //5.
- LBMReal CUMbcc = mfbcc - ((mfaac * mfbca + mfaca * mfbac + c4 * mfabb * mfbbb + c2 * (mfbab * mfacb + mfbba * mfabc)) + c1o3 * (mfbca + mfbac)) * OOrho;
- LBMReal CUMcbc = mfcbc - ((mfaac * mfcba + mfcaa * mfabc + c4 * mfbab * mfbbb + c2 * (mfabb * mfcab + mfbba * mfbac)) + c1o3 * (mfcba + mfabc)) * OOrho;
- LBMReal CUMccb = mfccb - ((mfcaa * mfacb + mfaca * mfcab + c4 * mfbba * mfbbb + c2 * (mfbab * mfbca + mfabb * mfcba)) + c1o3 * (mfacb + mfcab)) * OOrho;
- ////////////////////////////////////////////////////////////
- //6.
- LBMReal CUMccc = mfccc + ((-c4 * mfbbb * mfbbb
- - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca)
- - c4 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc)
- - c2 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) * OOrho
- + (c4 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac)
- + c2 * (mfcaa * mfaca * mfaac)
- + c16 * mfbba * mfbab * mfabb) * OOrho * OOrho
- - c1o3 * (mfacc + mfcac + mfcca) * OOrho
- - c1o9 * (mfcaa + mfaca + mfaac) * OOrho
- + (c2 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba)
- + (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 * (mfaac + mfaca + mfcaa)) * OOrho * OOrho * c2o3
- + c1o27 * ((drho * drho - drho) * OOrho * OOrho));
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute linear combinations of second and third order cumulants
- //!
- ////////////////////////////////////////////////////////////
- //2.
- LBMReal mxxPyyPzz = mfcaa + mfaca + mfaac;
- LBMReal mxxMyy = mfcaa - mfaca;
- LBMReal mxxMzz = mfcaa - mfaac;
- ////////////////////////////////////////////////////////////
- //3.
- LBMReal mxxyPyzz = mfcba + mfabc;
- LBMReal mxxyMyzz = mfcba - mfabc;
-
- LBMReal mxxzPyyz = mfcab + mfacb;
- LBMReal mxxzMyyz = mfcab - mfacb;
-
- LBMReal mxyyPxzz = mfbca + mfbac;
- LBMReal mxyyMxzz = mfbca - mfbac;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //incl. correction
- ////////////////////////////////////////////////////////////
- //! - Compute velocity gradients from second order cumulants according to Eq. (27)-(32)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //! Further explanations of the correction in viscosity in Appendix H of
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //! Note that the division by rho is omitted here as we need rho times the gradients later.
- //!
- LBMReal Dxy = -c3 * omega * mfbba;
- LBMReal Dxz = -c3 * omega * mfbab;
- LBMReal Dyz = -c3 * omega * mfabb;
- LBMReal dxux = c1o2 * (-omega) * (mxxMyy + mxxMzz) + c1o2 * OxxPyyPzz * (mfaaa - mxxPyyPzz);
- LBMReal dyuy = dxux + omega * c3o2 * mxxMyy;
- LBMReal dzuz = dxux + omega * c3o2 * mxxMzz;
- ////////////////////////////////////////////////////////////
- //! - Relaxation of second order cumulants with correction terms according to Eq. (33)-(35) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- mxxPyyPzz += OxxPyyPzz * (mfaaa - mxxPyyPzz) - c3 * (c1 - c1o2 * OxxPyyPzz) * (vx2 * dxux + vy2 * dyuy + vz2 * dzuz);
- mxxMyy += omega * (-mxxMyy) - c3 * (c1 + c1o2 * (-omega)) * (vx2 * dxux - vy2 * dyuy);
- mxxMzz += omega * (-mxxMzz) - c3 * (c1 + c1o2 * (-omega)) * (vx2 * dxux - vz2 * dzuz);
-
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- ////no correction
- //mxxPyyPzz += OxxPyyPzz*(mfaaa - mxxPyyPzz);
- //mxxMyy += -(-omega) * (-mxxMyy);
- //mxxMzz += -(-omega) * (-mxxMzz);
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- mfabb += omega * (-mfabb);
- mfbab += omega * (-mfbab);
- mfbba += omega * (-mfbba);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //relax
- //////////////////////////////////////////////////////////////////////////
- // incl. limiter
- //! - Relaxation of third order cumulants including limiter according to Eq. (116)-(123)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- wadjust = Oxyz + (c1 - Oxyz) * abs(mfbbb) / (abs(mfbbb) + qudricLimitD);
- mfbbb += wadjust * (-mfbbb);
- wadjust = OxyyPxzz + (c1 - OxyyPxzz) * abs(mxxyPyzz) / (abs(mxxyPyzz) + qudricLimitP);
- mxxyPyzz += wadjust * (-mxxyPyzz);
- wadjust = OxyyMxzz + (c1 - OxyyMxzz) * abs(mxxyMyzz) / (abs(mxxyMyzz) + qudricLimitM);
- mxxyMyzz += wadjust * (-mxxyMyzz);
- wadjust = OxyyPxzz + (c1 - OxyyPxzz) * abs(mxxzPyyz) / (abs(mxxzPyyz) + qudricLimitP);
- mxxzPyyz += wadjust * (-mxxzPyyz);
- wadjust = OxyyMxzz + (c1 - OxyyMxzz) * abs(mxxzMyyz) / (abs(mxxzMyyz) + qudricLimitM);
- mxxzMyyz += wadjust * (-mxxzMyyz);
- wadjust = OxyyPxzz + (c1 - OxyyPxzz) * abs(mxyyPxzz) / (abs(mxyyPxzz) + qudricLimitP);
- mxyyPxzz += wadjust * (-mxyyPxzz);
- wadjust = OxyyMxzz + (c1 - OxyyMxzz) * abs(mxyyMxzz) / (abs(mxyyMxzz) + qudricLimitM);
- mxyyMxzz += wadjust * (-mxyyMxzz);
- //////////////////////////////////////////////////////////////////////////
- // no limiter
- //mfbbb += OxyyMxzz * (-mfbbb);
- //mxxyPyzz += OxyyPxzz * (-mxxyPyzz);
- //mxxyMyzz += OxyyMxzz * (-mxxyMyzz);
- //mxxzPyyz += OxyyPxzz * (-mxxzPyyz);
- //mxxzMyyz += OxyyMxzz * (-mxxzMyyz);
- //mxyyPxzz += OxyyPxzz * (-mxyyPxzz);
- //mxyyMxzz += OxyyMxzz * (-mxyyMxzz);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute inverse linear combinations of second and third order cumulants
- //!
- mfcaa = c1o3 * (mxxMyy + mxxMzz + mxxPyyPzz);
- mfaca = c1o3 * (-c2 * mxxMyy + mxxMzz + mxxPyyPzz);
- mfaac = c1o3 * (mxxMyy - c2 * mxxMzz + mxxPyyPzz);
-
- mfcba = (mxxyMyzz + mxxyPyzz) * c1o2;
- mfabc = (-mxxyMyzz + mxxyPyzz) * c1o2;
- mfcab = (mxxzMyyz + mxxzPyyz) * c1o2;
- mfacb = (-mxxzMyyz + mxxzPyyz) * c1o2;
- mfbca = (mxyyMxzz + mxyyPxzz) * c1o2;
- mfbac = (-mxyyMxzz + mxyyPxzz) * c1o2;
- //////////////////////////////////////////////////////////////////////////
-
- //////////////////////////////////////////////////////////////////////////
- //4.
- // no limiter
- //! - Relax fourth order cumulants to modified equilibrium for fourth order convergence of diffusion according to Eq. (43)-(48)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- CUMacc = -O4 * (c1 / omega - c1o2) * (dyuy + dzuz) * c2o3 * A + (c1 - O4) * (CUMacc);
- CUMcac = -O4 * (c1 / omega - c1o2) * (dxux + dzuz) * c2o3 * A + (c1 - O4) * (CUMcac);
- CUMcca = -O4 * (c1 / omega - c1o2) * (dyuy + dxux) * c2o3 * A + (c1 - O4) * (CUMcca);
- CUMbbc = -O4 * (c1 / omega - c1o2) * Dxy * c1o3 * B + (c1 - O4) * (CUMbbc);
- CUMbcb = -O4 * (c1 / omega - c1o2) * Dxz * c1o3 * B + (c1 - O4) * (CUMbcb);
- CUMcbb = -O4 * (c1 / omega - c1o2) * Dyz * c1o3 * B + (c1 - O4) * (CUMcbb);
-
- //////////////////////////////////////////////////////////////////////////
- //5.
- CUMbcc += O5 * (-CUMbcc);
- CUMcbc += O5 * (-CUMcbc);
- CUMccb += O5 * (-CUMccb);
-
- //////////////////////////////////////////////////////////////////////////
- //6.
- CUMccc += O6 * (-CUMccc);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute central moments from post collision cumulants according to Eq. (53)-(56) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
-
- //////////////////////////////////////////////////////////////////////////
- //4.
- mfcbb = CUMcbb + c1o3 * ((c3 * mfcaa + c1) * mfabb + c6 * mfbba * mfbab) * OOrho;
- mfbcb = CUMbcb + c1o3 * ((c3 * mfaca + c1) * mfbab + c6 * mfbba * mfabb) * OOrho;
- mfbbc = CUMbbc + c1o3 * ((c3 * mfaac + c1) * mfbba + c6 * mfbab * mfabb) * OOrho;
-
- mfcca = CUMcca + (((mfcaa * mfaca + c2 * mfbba * mfbba) * c9 + c3 * (mfcaa + mfaca)) * OOrho - (drho * OOrho)) * c1o9;
- mfcac = CUMcac + (((mfcaa * mfaac + c2 * mfbab * mfbab) * c9 + c3 * (mfcaa + mfaac)) * OOrho - (drho * OOrho)) * c1o9;
- mfacc = CUMacc + (((mfaac * mfaca + c2 * mfabb * mfabb) * c9 + c3 * (mfaac + mfaca)) * OOrho - (drho * OOrho)) * c1o9;
-
- //////////////////////////////////////////////////////////////////////////
- //5.
- mfbcc = CUMbcc + c1o3 * (c3 * (mfaac * mfbca + mfaca * mfbac + c4 * mfabb * mfbbb + c2 * (mfbab * mfacb + mfbba * mfabc)) + (mfbca + mfbac)) * OOrho;
- mfcbc = CUMcbc + c1o3 * (c3 * (mfaac * mfcba + mfcaa * mfabc + c4 * mfbab * mfbbb + c2 * (mfabb * mfcab + mfbba * mfbac)) + (mfcba + mfabc)) * OOrho;
- mfccb = CUMccb + c1o3 * (c3 * (mfcaa * mfacb + mfaca * mfcab + c4 * mfbba * mfbbb + c2 * (mfbab * mfbca + mfabb * mfcba)) + (mfacb + mfcab)) * OOrho;
-
- //////////////////////////////////////////////////////////////////////////
- //6.
- mfccc = CUMccc - ((-c4 * mfbbb * mfbbb
- - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca)
- - c4 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc)
- - c2 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) * OOrho
- + (c4 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac)
- + c2 * (mfcaa * mfaca * mfaac)
- + c16 * mfbba * mfbab * mfabb) * OOrho * OOrho
- - c1o3 * (mfacc + mfcac + mfcca) * OOrho
- - c1o9 * (mfcaa + mfaca + mfaac) * OOrho
- + (c2 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba)
- + (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 * (mfaac + mfaca + mfcaa)) * OOrho * OOrho * c2o3
- + c1o27 * ((drho * drho - drho) * OOrho * OOrho));
-
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Add acceleration (body force) to first order cumulants according to Eq. (85)-(87) in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- mfbaa = -mfbaa;
- mfaba = -mfaba;
- mfaab = -mfaab;
- ////////////////////////////////////////////////////////////////////////////////////
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Chimera transform from central moments to well conditioned distributions as defined in Appendix J in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //! see also Eq. (88)-(96) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- backwardInverseChimeraWithK(mfaaa, mfbaa, mfcaa, vvx, vx2, c1, c1);
- backwardChimera(mfaba, mfbba, mfcba, vvx, vx2);
- backwardInverseChimeraWithK(mfaca, mfbca, mfcca, vvx, vx2, c3, c1o3);
- backwardChimera(mfaab, mfbab, mfcab, vvx, vx2);
- backwardChimera(mfabb, mfbbb, mfcbb, vvx, vx2);
- backwardChimera(mfacb, mfbcb, mfccb, vvx, vx2);
- backwardInverseChimeraWithK(mfaac, mfbac, mfcac, vvx, vx2, c3, c1o3);
- backwardChimera(mfabc, mfbbc, mfcbc, vvx, vx2);
- backwardInverseChimeraWithK(mfacc, mfbcc, mfccc, vvx, vx2, c9, c1o9);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- backwardInverseChimeraWithK(mfaaa, mfaba, mfaca, vvy, vy2, c6, c1o6);
- backwardChimera(mfaab, mfabb, mfacb, vvy, vy2);
- backwardInverseChimeraWithK(mfaac, mfabc, mfacc, vvy, vy2, c18, c1o18);
- backwardInverseChimeraWithK(mfbaa, mfbba, mfbca, vvy, vy2, c3o2, c2o3);
- backwardChimera(mfbab, mfbbb, mfbcb, vvy, vy2);
- backwardInverseChimeraWithK(mfbac, mfbbc, mfbcc, vvy, vy2, c9o2, c2o9);
- backwardInverseChimeraWithK(mfcaa, mfcba, mfcca, vvy, vy2, c6, c1o6);
- backwardChimera(mfcab, mfcbb, mfccb, vvy, vy2);
- backwardInverseChimeraWithK(mfcac, mfcbc, mfccc, vvy, vy2, c18, c1o18);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- backwardInverseChimeraWithK(mfaaa, mfaab, mfaac, vvz, vz2, c36, c1o36);
- backwardInverseChimeraWithK(mfaba, mfabb, mfabc, vvz, vz2, c9, c1o9);
- backwardInverseChimeraWithK(mfaca, mfacb, mfacc, vvz, vz2, c36, c1o36);
- backwardInverseChimeraWithK(mfbaa, mfbab, mfbac, vvz, vz2, c9, c1o9);
- backwardInverseChimeraWithK(mfbba, mfbbb, mfbbc, vvz, vz2, c9o4, c4o9);
- backwardInverseChimeraWithK(mfbca, mfbcb, mfbcc, vvz, vz2, c9, c1o9);
- backwardInverseChimeraWithK(mfcaa, mfcab, mfcac, vvz, vz2, c36, c1o36);
- backwardInverseChimeraWithK(mfcba, mfcbb, mfcbc, vvz, vz2, c9, c1o9);
- backwardInverseChimeraWithK(mfcca, mfccb, mfccc, vvz, vz2, c36, c1o36);
- ////////////////////////////////////////////////////////////////////////////////////
-
- //////////////////////////////////////////////////////////////////////////
- //proof correctness
- //////////////////////////////////////////////////////////////////////////
+ if (!bcArray->isSolid(x1, x2, x3) && !bcArray->isUndefined(x1, x2, x3))
+ {
+ int x1p = x1 + 1;
+ int x2p = x2 + 1;
+ int x3p = x3 + 1;
+ //////////////////////////////////////////////////////////////////////////
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm
+ //! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ ////////////////////////////////////////////////////////////////////////////
+ //////////////////////////////////////////////////////////////////////////
+
+ //E N T
+ //c c c
+ //////////
+ //W S B
+ //a a a
+
+ //Rest is b
+
+ //mfxyz
+ //a - negative
+ //b - null
+ //c - positive
+
+ // a b c
+ //-1 0 1
+
+ LBMReal mfcbb = (*this->localDistributions)(D3Q27System::ET_E, x1, x2, x3);
+ LBMReal mfbcb = (*this->localDistributions)(D3Q27System::ET_N, x1, x2, x3);
+ LBMReal mfbbc = (*this->localDistributions)(D3Q27System::ET_T, x1, x2, x3);
+ LBMReal mfccb = (*this->localDistributions)(D3Q27System::ET_NE, x1, x2, x3);
+ LBMReal mfacb = (*this->localDistributions)(D3Q27System::ET_NW, x1p, x2, x3);
+ LBMReal mfcbc = (*this->localDistributions)(D3Q27System::ET_TE, x1, x2, x3);
+ LBMReal mfabc = (*this->localDistributions)(D3Q27System::ET_TW, x1p, x2, x3);
+ LBMReal mfbcc = (*this->localDistributions)(D3Q27System::ET_TN, x1, x2, x3);
+ LBMReal mfbac = (*this->localDistributions)(D3Q27System::ET_TS, x1, x2p, x3);
+ LBMReal mfccc = (*this->localDistributions)(D3Q27System::ET_TNE, x1, x2, x3);
+ LBMReal mfacc = (*this->localDistributions)(D3Q27System::ET_TNW, x1p, x2, x3);
+ LBMReal mfcac = (*this->localDistributions)(D3Q27System::ET_TSE, x1, x2p, x3);
+ LBMReal mfaac = (*this->localDistributions)(D3Q27System::ET_TSW, x1p, x2p, x3);
+
+ LBMReal mfabb = (*this->nonLocalDistributions)(D3Q27System::ET_W, x1p, x2, x3);
+ LBMReal mfbab = (*this->nonLocalDistributions)(D3Q27System::ET_S, x1, x2p, x3);
+ LBMReal mfbba = (*this->nonLocalDistributions)(D3Q27System::ET_B, x1, x2, x3p);
+ LBMReal mfaab = (*this->nonLocalDistributions)(D3Q27System::ET_SW, x1p, x2p, x3);
+ LBMReal mfcab = (*this->nonLocalDistributions)(D3Q27System::ET_SE, x1, x2p, x3);
+ LBMReal mfaba = (*this->nonLocalDistributions)(D3Q27System::ET_BW, x1p, x2, x3p);
+ LBMReal mfcba = (*this->nonLocalDistributions)(D3Q27System::ET_BE, x1, x2, x3p);
+ LBMReal mfbaa = (*this->nonLocalDistributions)(D3Q27System::ET_BS, x1, x2p, x3p);
+ LBMReal mfbca = (*this->nonLocalDistributions)(D3Q27System::ET_BN, x1, x2, x3p);
+ LBMReal mfaaa = (*this->nonLocalDistributions)(D3Q27System::ET_BSW, x1p, x2p, x3p);
+ LBMReal mfcaa = (*this->nonLocalDistributions)(D3Q27System::ET_BSE, x1, x2p, x3p);
+ LBMReal mfaca = (*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1p, x2, x3p);
+ LBMReal mfcca = (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3p);
+
+ LBMReal mfbbb = (*this->restDistributions)(x1, x2, x3);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3)
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //!
+ LBMReal drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
+ (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
+ ((mfabb + mfcbb) + (mfbab + mfbcb)) + (mfbba + mfbbc)) + mfbbb;
+
+ LBMReal rho = c1 + drho;
+ LBMReal OOrho = c1 / rho;
+ ////////////////////////////////////////////////////////////////////////////////////
+ LBMReal vvx = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
+ (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
+ (mfcbb - mfabb)) / rho;
+ LBMReal vvy = ((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
+ (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
+ (mfbcb - mfbab)) / rho;
+ LBMReal vvz = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
+ (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
+ (mfbbc - mfbba)) / rho;
+ ////////////////////////////////////////////////////////////////////////////////////
+ //forcing
+ ///////////////////////////////////////////////////////////////////////////////////////////
+ if (withForcing)
+ {
+ muX1 = static_cast<double>(x1 - 1 + ix1 * maxX1);
+ muX2 = static_cast<double>(x2 - 1 + ix2 * maxX2);
+ muX3 = static_cast<double>(x3 - 1 + ix3 * maxX3);
+
+ forcingX1 = muForcingX1.Eval();
+ forcingX2 = muForcingX2.Eval();
+ forcingX3 = muForcingX3.Eval();
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Add half of the acceleration (body force) to the velocity as in Eq. (42)
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //!
+ vvx += forcingX1 * deltaT * c1o2; // X
+ vvy += forcingX2 * deltaT * c1o2; // Y
+ vvz += forcingX3 * deltaT * c1o2; // Z
+ }
+ ////////////////////////////////////////////////////////////////////////////////////
+ // calculate the square of velocities for this lattice node
+ LBMReal vx2 = vvx * vvx;
+ LBMReal vy2 = vvy * vvy;
+ LBMReal vz2 = vvz * vvz;
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Set relaxation limiters for third order cumulants to default value \f$ \lambda=0.001 \f$ according to section 6 in
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ LBMReal wadjust;
+ LBMReal qudricLimitP = c1o100;
+ LBMReal qudricLimitM = c1o100;
+ LBMReal qudricLimitD = c1o100;
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Chimera transform from well conditioned distributions to central moments as defined in Appendix J in
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //! see also Eq. (6)-(14) in
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ ////////////////////////////////////////////////////////////////////////////////////
+ // Z - Dir
+ forwardInverseChimeraWithK(mfaaa, mfaab, mfaac, vvz, vz2, c36, c1o36);
+ forwardInverseChimeraWithK(mfaba, mfabb, mfabc, vvz, vz2, c9, c1o9);
+ forwardInverseChimeraWithK(mfaca, mfacb, mfacc, vvz, vz2, c36, c1o36);
+ forwardInverseChimeraWithK(mfbaa, mfbab, mfbac, vvz, vz2, c9, c1o9);
+ forwardInverseChimeraWithK(mfbba, mfbbb, mfbbc, vvz, vz2, c9o4, c4o9);
+ forwardInverseChimeraWithK(mfbca, mfbcb, mfbcc, vvz, vz2, c9, c1o9);
+ forwardInverseChimeraWithK(mfcaa, mfcab, mfcac, vvz, vz2, c36, c1o36);
+ forwardInverseChimeraWithK(mfcba, mfcbb, mfcbc, vvz, vz2, c9, c1o9);
+ forwardInverseChimeraWithK(mfcca, mfccb, mfccc, vvz, vz2, c36, c1o36);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ // Y - Dir
+ forwardInverseChimeraWithK(mfaaa, mfaba, mfaca, vvy, vy2, c6, c1o6);
+ forwardChimera(mfaab, mfabb, mfacb, vvy, vy2);
+ forwardInverseChimeraWithK(mfaac, mfabc, mfacc, vvy, vy2, c18, c1o18);
+ forwardInverseChimeraWithK(mfbaa, mfbba, mfbca, vvy, vy2, c3o2, c2o3);
+ forwardChimera(mfbab, mfbbb, mfbcb, vvy, vy2);
+ forwardInverseChimeraWithK(mfbac, mfbbc, mfbcc, vvy, vy2, c9o2, c2o9);
+ forwardInverseChimeraWithK(mfcaa, mfcba, mfcca, vvy, vy2, c6, c1o6);
+ forwardChimera(mfcab, mfcbb, mfccb, vvy, vy2);
+ forwardInverseChimeraWithK(mfcac, mfcbc, mfccc, vvy, vy2, c18, c1o18);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ // X - Dir
+ forwardInverseChimeraWithK(mfaaa, mfbaa, mfcaa, vvx, vx2, c1, c1);
+ forwardChimera(mfaba, mfbba, mfcba, vvx, vx2);
+ forwardInverseChimeraWithK(mfaca, mfbca, mfcca, vvx, vx2, c3, c1o3);
+ forwardChimera(mfaab, mfbab, mfcab, vvx, vx2);
+ forwardChimera(mfabb, mfbbb, mfcbb, vvx, vx2);
+ forwardChimera(mfacb, mfbcb, mfccb, vvx, vx2);
+ forwardInverseChimeraWithK(mfaac, mfbac, mfcac, vvx, vx2, c3, c1o3);
+ forwardChimera(mfabc, mfbbc, mfcbc, vvx, vx2);
+ forwardInverseChimeraWithK(mfacc, mfbcc, mfccc, vvx, vx2, c9, c1o9);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Setting relaxation rates for non-hydrodynamic cumulants (default values). Variable names and equations according to
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //! => [NAME IN PAPER]=[NAME IN CODE]=[DEFAULT VALUE].
+ //! - Trace of second order cumulants \f$ C_{200}+C_{020}+C_{002} \f$ used to adjust bulk viscosity:\f$\omega_2=OxxPyyPzz=1.0 \f$.
+ //! - Third order cumulants \f$ C_{120}+C_{102} \f$, \f$ C_{210}+C_{012} \f$, \f$ C_{201}+C_{021} \f$: \f$\omega_3=OxyyPxzz\f$ set according to Eq. (111) with simplifications assuming \f$\omega_2=1.0\f$.
+ //! - Third order cumulants \f$ C_{120}-C_{102} \f$, \f$ C_{210}-C_{012} \f$, \f$ C_{201}-C_{021} \f$: \f$\omega_4 = OxyyMxzz\f$ set according to Eq. (112) with simplifications assuming \f$\omega_2 = 1.0\f$.
+ //! - Third order cumulants \f$ C_{111} \f$: \f$\omega_5 = Oxyz\f$ set according to Eq. (113) with simplifications assuming \f$\omega_2 = 1.0\f$ (modify for different bulk viscosity).
+ //! - Fourth order cumulants \f$ C_{220} \f$, \f$ C_{202} \f$, \f$ C_{022} \f$, \f$ C_{211} \f$, \f$ C_{121} \f$, \f$ C_{112} \f$: for simplification all set to the same default value \f$ \omega_6=\omega_7=\omega_8=O4=1.0 \f$.
+ //! - Fifth order cumulants \f$ C_{221}\f$, \f$C_{212}\f$, \f$C_{122}\f$: \f$\omega_9=O5=1.0\f$.
+ //! - Sixth order cumulant \f$ C_{222}\f$: \f$\omega_{10}=O6=1.0\f$.
+ //!
+ ////////////////////////////////////////////////////////////
+ //2.
+ LBMReal OxxPyyPzz = c1;
+ ////////////////////////////////////////////////////////////
+ //3.
+ LBMReal OxyyPxzz = c8 * (-c2 + omega) * ( c1 + c2*omega) / (-c8 - c14*omega + c7*omega*omega);
+ LBMReal OxyyMxzz = c8 * (-c2 + omega) * (-c7 + c4*omega) / (c56 - c50*omega + c9*omega*omega);
+ LBMReal Oxyz = c24 * (-c2 + omega) * (-c2 - c7*omega + c3*omega*omega) / (c48 + c152*omega - c130*omega*omega + c29*omega*omega*omega);
+ ////////////////////////////////////////////////////////////
+ //4.
+ LBMReal O4 = c1;
+ ////////////////////////////////////////////////////////////
+ //5.
+ LBMReal O5 = c1;
+ ////////////////////////////////////////////////////////////
+ //6.
+ LBMReal O6 = c1;
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - A and B: parameters for fourth order convergence of the diffusion term according to Eq. (114) and (115)
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //! with simplifications assuming \f$\omega_2 = 1.0\f$ (modify for different bulk viscosity).
+ //!
+ LBMReal A = (c4 + c2*omega - c3*omega*omega) / (c2 - c7*omega + c5*omega*omega);
+ LBMReal B = (c4 + c28*omega - c14*omega*omega) / (c6 - c21*omega + c15*omega*omega);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Compute cumulants from central moments according to Eq. (20)-(23) in
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ ////////////////////////////////////////////////////////////
+ //4.
+ LBMReal CUMcbb = mfcbb - ((mfcaa + c1o3) * mfabb + c2 * mfbba * mfbab) * OOrho;
+ LBMReal CUMbcb = mfbcb - ((mfaca + c1o3) * mfbab + c2 * mfbba * mfabb) * OOrho;
+ LBMReal CUMbbc = mfbbc - ((mfaac + c1o3) * mfbba + c2 * mfbab * mfabb) * OOrho;
+
+ LBMReal CUMcca = mfcca - (((mfcaa * mfaca + c2 * mfbba * mfbba) + c1o3 * (mfcaa + mfaca)) * OOrho - c1o9 * (drho * OOrho));
+ LBMReal CUMcac = mfcac - (((mfcaa * mfaac + c2 * mfbab * mfbab) + c1o3 * (mfcaa + mfaac)) * OOrho - c1o9 * (drho * OOrho));
+ LBMReal CUMacc = mfacc - (((mfaac * mfaca + c2 * mfabb * mfabb) + c1o3 * (mfaac + mfaca)) * OOrho - c1o9 * (drho * OOrho));
+ ////////////////////////////////////////////////////////////
+ //5.
+ LBMReal CUMbcc = mfbcc - ((mfaac * mfbca + mfaca * mfbac + c4 * mfabb * mfbbb + c2 * (mfbab * mfacb + mfbba * mfabc)) + c1o3 * (mfbca + mfbac)) * OOrho;
+ LBMReal CUMcbc = mfcbc - ((mfaac * mfcba + mfcaa * mfabc + c4 * mfbab * mfbbb + c2 * (mfabb * mfcab + mfbba * mfbac)) + c1o3 * (mfcba + mfabc)) * OOrho;
+ LBMReal CUMccb = mfccb - ((mfcaa * mfacb + mfaca * mfcab + c4 * mfbba * mfbbb + c2 * (mfbab * mfbca + mfabb * mfcba)) + c1o3 * (mfacb + mfcab)) * OOrho;
+ ////////////////////////////////////////////////////////////
+ //6.
+ LBMReal CUMccc = mfccc + ((-c4 * mfbbb * mfbbb
+ - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca)
+ - c4 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc)
+ - c2 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) * OOrho
+ + (c4 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac)
+ + c2 * (mfcaa * mfaca * mfaac)
+ + c16 * mfbba * mfbab * mfabb) * OOrho * OOrho
+ - c1o3 * (mfacc + mfcac + mfcca) * OOrho
+ - c1o9 * (mfcaa + mfaca + mfaac) * OOrho
+ + (c2 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba)
+ + (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 * (mfaac + mfaca + mfcaa)) * OOrho * OOrho * c2o3
+ + c1o27 * ((drho * drho - drho) * OOrho * OOrho));
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Compute linear combinations of second and third order cumulants
+ //!
+ ////////////////////////////////////////////////////////////
+ //2.
+ LBMReal mxxPyyPzz = mfcaa + mfaca + mfaac;
+ LBMReal mxxMyy = mfcaa - mfaca;
+ LBMReal mxxMzz = mfcaa - mfaac;
+ ////////////////////////////////////////////////////////////
+ //3.
+ LBMReal mxxyPyzz = mfcba + mfabc;
+ LBMReal mxxyMyzz = mfcba - mfabc;
+
+ LBMReal mxxzPyyz = mfcab + mfacb;
+ LBMReal mxxzMyyz = mfcab - mfacb;
+
+ LBMReal mxyyPxzz = mfbca + mfbac;
+ LBMReal mxyyMxzz = mfbca - mfbac;
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //incl. correction
+ ////////////////////////////////////////////////////////////
+ //! - Compute velocity gradients from second order cumulants according to Eq. (27)-(32)
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //! Further explanations of the correction in viscosity in Appendix H of
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //! Note that the division by rho is omitted here as we need rho times the gradients later.
+ //!
+ LBMReal Dxy = -c3 * omega * mfbba;
+ LBMReal Dxz = -c3 * omega * mfbab;
+ LBMReal Dyz = -c3 * omega * mfabb;
+ LBMReal dxux = c1o2 * (-omega) * (mxxMyy + mxxMzz) + c1o2 * OxxPyyPzz * (mfaaa - mxxPyyPzz);
+ LBMReal dyuy = dxux + omega * c3o2 * mxxMyy;
+ LBMReal dzuz = dxux + omega * c3o2 * mxxMzz;
+ ////////////////////////////////////////////////////////////
+ //! - Relaxation of second order cumulants with correction terms according to Eq. (33)-(35) in
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ mxxPyyPzz += OxxPyyPzz * (mfaaa - mxxPyyPzz) - c3 * (c1 - c1o2 * OxxPyyPzz) * (vx2 * dxux + vy2 * dyuy + vz2 * dzuz);
+ mxxMyy += omega * (-mxxMyy) - c3 * (c1 + c1o2 * (-omega)) * (vx2 * dxux - vy2 * dyuy);
+ mxxMzz += omega * (-mxxMzz) - c3 * (c1 + c1o2 * (-omega)) * (vx2 * dxux - vz2 * dzuz);
+
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////no correction
+ //mxxPyyPzz += OxxPyyPzz*(mfaaa - mxxPyyPzz);
+ //mxxMyy += -(-omega) * (-mxxMyy);
+ //mxxMzz += -(-omega) * (-mxxMzz);
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ mfabb += omega * (-mfabb);
+ mfbab += omega * (-mfbab);
+ mfbba += omega * (-mfbba);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //relax
+ //////////////////////////////////////////////////////////////////////////
+ // incl. limiter
+ //! - Relaxation of third order cumulants including limiter according to Eq. (116)-(123)
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ wadjust = Oxyz + (c1 - Oxyz) * abs(mfbbb) / (abs(mfbbb) + qudricLimitD);
+ mfbbb += wadjust * (-mfbbb);
+ wadjust = OxyyPxzz + (c1 - OxyyPxzz) * abs(mxxyPyzz) / (abs(mxxyPyzz) + qudricLimitP);
+ mxxyPyzz += wadjust * (-mxxyPyzz);
+ wadjust = OxyyMxzz + (c1 - OxyyMxzz) * abs(mxxyMyzz) / (abs(mxxyMyzz) + qudricLimitM);
+ mxxyMyzz += wadjust * (-mxxyMyzz);
+ wadjust = OxyyPxzz + (c1 - OxyyPxzz) * abs(mxxzPyyz) / (abs(mxxzPyyz) + qudricLimitP);
+ mxxzPyyz += wadjust * (-mxxzPyyz);
+ wadjust = OxyyMxzz + (c1 - OxyyMxzz) * abs(mxxzMyyz) / (abs(mxxzMyyz) + qudricLimitM);
+ mxxzMyyz += wadjust * (-mxxzMyyz);
+ wadjust = OxyyPxzz + (c1 - OxyyPxzz) * abs(mxyyPxzz) / (abs(mxyyPxzz) + qudricLimitP);
+ mxyyPxzz += wadjust * (-mxyyPxzz);
+ wadjust = OxyyMxzz + (c1 - OxyyMxzz) * abs(mxyyMxzz) / (abs(mxyyMxzz) + qudricLimitM);
+ mxyyMxzz += wadjust * (-mxyyMxzz);
+ //////////////////////////////////////////////////////////////////////////
+ // no limiter
+ //mfbbb += OxyyMxzz * (-mfbbb);
+ //mxxyPyzz += OxyyPxzz * (-mxxyPyzz);
+ //mxxyMyzz += OxyyMxzz * (-mxxyMyzz);
+ //mxxzPyyz += OxyyPxzz * (-mxxzPyyz);
+ //mxxzMyyz += OxyyMxzz * (-mxxzMyyz);
+ //mxyyPxzz += OxyyPxzz * (-mxyyPxzz);
+ //mxyyMxzz += OxyyMxzz * (-mxyyMxzz);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Compute inverse linear combinations of second and third order cumulants
+ //!
+ mfcaa = c1o3 * (mxxMyy + mxxMzz + mxxPyyPzz);
+ mfaca = c1o3 * (-c2 * mxxMyy + mxxMzz + mxxPyyPzz);
+ mfaac = c1o3 * (mxxMyy - c2 * mxxMzz + mxxPyyPzz);
+
+ mfcba = (mxxyMyzz + mxxyPyzz) * c1o2;
+ mfabc = (-mxxyMyzz + mxxyPyzz) * c1o2;
+ mfcab = (mxxzMyyz + mxxzPyyz) * c1o2;
+ mfacb = (-mxxzMyyz + mxxzPyyz) * c1o2;
+ mfbca = (mxyyMxzz + mxyyPxzz) * c1o2;
+ mfbac = (-mxyyMxzz + mxyyPxzz) * c1o2;
+ //////////////////////////////////////////////////////////////////////////
+
+ //////////////////////////////////////////////////////////////////////////
+ //4.
+ // no limiter
+ //! - Relax fourth order cumulants to modified equilibrium for fourth order convergence of diffusion according to Eq. (43)-(48)
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ CUMacc = -O4 * (c1 / omega - c1o2) * (dyuy + dzuz) * c2o3 * A + (c1 - O4) * (CUMacc);
+ CUMcac = -O4 * (c1 / omega - c1o2) * (dxux + dzuz) * c2o3 * A + (c1 - O4) * (CUMcac);
+ CUMcca = -O4 * (c1 / omega - c1o2) * (dyuy + dxux) * c2o3 * A + (c1 - O4) * (CUMcca);
+ CUMbbc = -O4 * (c1 / omega - c1o2) * Dxy * c1o3 * B + (c1 - O4) * (CUMbbc);
+ CUMbcb = -O4 * (c1 / omega - c1o2) * Dxz * c1o3 * B + (c1 - O4) * (CUMbcb);
+ CUMcbb = -O4 * (c1 / omega - c1o2) * Dyz * c1o3 * B + (c1 - O4) * (CUMcbb);
+
+ //////////////////////////////////////////////////////////////////////////
+ //5.
+ CUMbcc += O5 * (-CUMbcc);
+ CUMcbc += O5 * (-CUMcbc);
+ CUMccb += O5 * (-CUMccb);
+
+ //////////////////////////////////////////////////////////////////////////
+ //6.
+ CUMccc += O6 * (-CUMccc);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Compute central moments from post collision cumulants according to Eq. (53)-(56) in
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+
+ //////////////////////////////////////////////////////////////////////////
+ //4.
+ mfcbb = CUMcbb + c1o3 * ((c3 * mfcaa + c1) * mfabb + c6 * mfbba * mfbab) * OOrho;
+ mfbcb = CUMbcb + c1o3 * ((c3 * mfaca + c1) * mfbab + c6 * mfbba * mfabb) * OOrho;
+ mfbbc = CUMbbc + c1o3 * ((c3 * mfaac + c1) * mfbba + c6 * mfbab * mfabb) * OOrho;
+
+ mfcca = CUMcca + (((mfcaa * mfaca + c2 * mfbba * mfbba) * c9 + c3 * (mfcaa + mfaca)) * OOrho - (drho * OOrho)) * c1o9;
+ mfcac = CUMcac + (((mfcaa * mfaac + c2 * mfbab * mfbab) * c9 + c3 * (mfcaa + mfaac)) * OOrho - (drho * OOrho)) * c1o9;
+ mfacc = CUMacc + (((mfaac * mfaca + c2 * mfabb * mfabb) * c9 + c3 * (mfaac + mfaca)) * OOrho - (drho * OOrho)) * c1o9;
+
+ //////////////////////////////////////////////////////////////////////////
+ //5.
+ mfbcc = CUMbcc + c1o3 * (c3 * (mfaac * mfbca + mfaca * mfbac + c4 * mfabb * mfbbb + c2 * (mfbab * mfacb + mfbba * mfabc)) + (mfbca + mfbac)) * OOrho;
+ mfcbc = CUMcbc + c1o3 * (c3 * (mfaac * mfcba + mfcaa * mfabc + c4 * mfbab * mfbbb + c2 * (mfabb * mfcab + mfbba * mfbac)) + (mfcba + mfabc)) * OOrho;
+ mfccb = CUMccb + c1o3 * (c3 * (mfcaa * mfacb + mfaca * mfcab + c4 * mfbba * mfbbb + c2 * (mfbab * mfbca + mfabb * mfcba)) + (mfacb + mfcab)) * OOrho;
+
+ //////////////////////////////////////////////////////////////////////////
+ //6.
+ mfccc = CUMccc - ((-c4 * mfbbb * mfbbb
+ - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca)
+ - c4 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc)
+ - c2 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) * OOrho
+ + (c4 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac)
+ + c2 * (mfcaa * mfaca * mfaac)
+ + c16 * mfbba * mfbab * mfabb) * OOrho * OOrho
+ - c1o3 * (mfacc + mfcac + mfcca) * OOrho
+ - c1o9 * (mfcaa + mfaca + mfaac) * OOrho
+ + (c2 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba)
+ + (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 * (mfaac + mfaca + mfcaa)) * OOrho * OOrho * c2o3
+ + c1o27 * ((drho * drho - drho) * OOrho * OOrho));
+
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Add acceleration (body force) to first order cumulants according to Eq. (85)-(87) in
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //!
+ mfbaa = -mfbaa;
+ mfaba = -mfaba;
+ mfaab = -mfaab;
+ ////////////////////////////////////////////////////////////////////////////////////
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Chimera transform from central moments to well conditioned distributions as defined in Appendix J in
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //! see also Eq. (88)-(96) in
+ //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+ //!
+ ////////////////////////////////////////////////////////////////////////////////////
+ // X - Dir
+ backwardInverseChimeraWithK(mfaaa, mfbaa, mfcaa, vvx, vx2, c1, c1);
+ backwardChimera(mfaba, mfbba, mfcba, vvx, vx2);
+ backwardInverseChimeraWithK(mfaca, mfbca, mfcca, vvx, vx2, c3, c1o3);
+ backwardChimera(mfaab, mfbab, mfcab, vvx, vx2);
+ backwardChimera(mfabb, mfbbb, mfcbb, vvx, vx2);
+ backwardChimera(mfacb, mfbcb, mfccb, vvx, vx2);
+ backwardInverseChimeraWithK(mfaac, mfbac, mfcac, vvx, vx2, c3, c1o3);
+ backwardChimera(mfabc, mfbbc, mfcbc, vvx, vx2);
+ backwardInverseChimeraWithK(mfacc, mfbcc, mfccc, vvx, vx2, c9, c1o9);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ // Y - Dir
+ backwardInverseChimeraWithK(mfaaa, mfaba, mfaca, vvy, vy2, c6, c1o6);
+ backwardChimera(mfaab, mfabb, mfacb, vvy, vy2);
+ backwardInverseChimeraWithK(mfaac, mfabc, mfacc, vvy, vy2, c18, c1o18);
+ backwardInverseChimeraWithK(mfbaa, mfbba, mfbca, vvy, vy2, c3o2, c2o3);
+ backwardChimera(mfbab, mfbbb, mfbcb, vvy, vy2);
+ backwardInverseChimeraWithK(mfbac, mfbbc, mfbcc, vvy, vy2, c9o2, c2o9);
+ backwardInverseChimeraWithK(mfcaa, mfcba, mfcca, vvy, vy2, c6, c1o6);
+ backwardChimera(mfcab, mfcbb, mfccb, vvy, vy2);
+ backwardInverseChimeraWithK(mfcac, mfcbc, mfccc, vvy, vy2, c18, c1o18);
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ // Z - Dir
+ backwardInverseChimeraWithK(mfaaa, mfaab, mfaac, vvz, vz2, c36, c1o36);
+ backwardInverseChimeraWithK(mfaba, mfabb, mfabc, vvz, vz2, c9, c1o9);
+ backwardInverseChimeraWithK(mfaca, mfacb, mfacc, vvz, vz2, c36, c1o36);
+ backwardInverseChimeraWithK(mfbaa, mfbab, mfbac, vvz, vz2, c9, c1o9);
+ backwardInverseChimeraWithK(mfbba, mfbbb, mfbbc, vvz, vz2, c9o4, c4o9);
+ backwardInverseChimeraWithK(mfbca, mfbcb, mfbcc, vvz, vz2, c9, c1o9);
+ backwardInverseChimeraWithK(mfcaa, mfcab, mfcac, vvz, vz2, c36, c1o36);
+ backwardInverseChimeraWithK(mfcba, mfcbb, mfcbc, vvz, vz2, c9, c1o9);
+ backwardInverseChimeraWithK(mfcca, mfccb, mfccc, vvz, vz2, c36, c1o36);
+ ////////////////////////////////////////////////////////////////////////////////////
+
+ //////////////////////////////////////////////////////////////////////////
+ //proof correctness
+ //////////////////////////////////////////////////////////////////////////
#ifdef PROOF_CORRECTNESS
- LBMReal drho_post = (mfaaa + mfaac + mfaca + mfcaa + mfacc + mfcac + mfccc + mfcca)
- + (mfaab + mfacb + mfcab + mfccb) + (mfaba + mfabc + mfcba + mfcbc) + (mfbaa + mfbac + mfbca + mfbcc)
- + (mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc) + mfbbb;
- LBMReal dif = drho - drho_post;
+ LBMReal drho_post = (mfaaa + mfaac + mfaca + mfcaa + mfacc + mfcac + mfccc + mfcca)
+ + (mfaab + mfacb + mfcab + mfccb) + (mfaba + mfabc + mfcba + mfcbc) + (mfbaa + mfbac + mfbca + mfbcc)
+ + (mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc) + mfbbb;
+ LBMReal dif = drho - drho_post;
#ifdef SINGLEPRECISION
- if (dif > 10.0E-7 || dif < -10.0E-7)
+ if (dif > 10.0E-7 || dif < -10.0E-7)
#else
- if (dif > 10.0E-15 || dif < -10.0E-15)
+ if (dif > 10.0E-15 || dif < -10.0E-15)
#endif
- {
- UB_THROW(UbException(UB_EXARGS, "rho=" + UbSystem::toString(drho) + ", rho_post=" + UbSystem::toString(drho_post)
- + " dif=" + UbSystem::toString(dif)
- + " rho is not correct for node " + UbSystem::toString(x1) + "," + UbSystem::toString(x2) + "," + UbSystem::toString(x3)
- + " in " + block.lock()->toString() + " step = " + UbSystem::toString(step)));
- }
+ {
+ UB_THROW(UbException(UB_EXARGS, "rho=" + UbSystem::toString(drho) + ", rho_post=" + UbSystem::toString(drho_post)
+ + " dif=" + UbSystem::toString(dif)
+ + " rho is not correct for node " + UbSystem::toString(x1) + "," + UbSystem::toString(x2) + "," + UbSystem::toString(x3)
+ + " in " + block.lock()->toString() + " step = " + UbSystem::toString(step)));
+ }
#endif
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Write distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm
- //! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
- //!
- (*this->localDistributions)(D3Q27System::ET_E, x1, x2, x3) = mfabb;
- (*this->localDistributions)(D3Q27System::ET_N, x1, x2, x3) = mfbab;
- (*this->localDistributions)(D3Q27System::ET_T, x1, x2, x3) = mfbba;
- (*this->localDistributions)(D3Q27System::ET_NE, x1, x2, x3) = mfaab;
- (*this->localDistributions)(D3Q27System::ET_NW, x1p, x2, x3) = mfcab;
- (*this->localDistributions)(D3Q27System::ET_TE, x1, x2, x3) = mfaba;
- (*this->localDistributions)(D3Q27System::ET_TW, x1p, x2, x3) = mfcba;
- (*this->localDistributions)(D3Q27System::ET_TN, x1, x2, x3) = mfbaa;
- (*this->localDistributions)(D3Q27System::ET_TS, x1, x2p, x3) = mfbca;
- (*this->localDistributions)(D3Q27System::ET_TNE, x1, x2, x3) = mfaaa;
- (*this->localDistributions)(D3Q27System::ET_TNW, x1p, x2, x3) = mfcaa;
- (*this->localDistributions)(D3Q27System::ET_TSE, x1, x2p, x3) = mfaca;
- (*this->localDistributions)(D3Q27System::ET_TSW, x1p, x2p, x3) = mfcca;
-
- (*this->nonLocalDistributions)(D3Q27System::ET_W, x1p, x2, x3) = mfcbb;
- (*this->nonLocalDistributions)(D3Q27System::ET_S, x1, x2p, x3) = mfbcb;
- (*this->nonLocalDistributions)(D3Q27System::ET_B, x1, x2, x3p) = mfbbc;
- (*this->nonLocalDistributions)(D3Q27System::ET_SW, x1p, x2p, x3) = mfccb;
- (*this->nonLocalDistributions)(D3Q27System::ET_SE, x1, x2p, x3) = mfacb;
- (*this->nonLocalDistributions)(D3Q27System::ET_BW, x1p, x2, x3p) = mfcbc;
- (*this->nonLocalDistributions)(D3Q27System::ET_BE, x1, x2, x3p) = mfabc;
- (*this->nonLocalDistributions)(D3Q27System::ET_BS, x1, x2p, x3p) = mfbcc;
- (*this->nonLocalDistributions)(D3Q27System::ET_BN, x1, x2, x3p) = mfbac;
- (*this->nonLocalDistributions)(D3Q27System::ET_BSW, x1p, x2p, x3p) = mfccc;
- (*this->nonLocalDistributions)(D3Q27System::ET_BSE, x1, x2p, x3p) = mfacc;
- (*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1p, x2, x3p) = mfcac;
- (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3p) = mfaac;
-
- (*this->restDistributions)(x1, x2, x3) = mfbbb;
- //////////////////////////////////////////////////////////////////////////
-
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Write distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm
+ //! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ (*this->localDistributions)(D3Q27System::ET_E, x1, x2, x3) = mfabb;
+ (*this->localDistributions)(D3Q27System::ET_N, x1, x2, x3) = mfbab;
+ (*this->localDistributions)(D3Q27System::ET_T, x1, x2, x3) = mfbba;
+ (*this->localDistributions)(D3Q27System::ET_NE, x1, x2, x3) = mfaab;
+ (*this->localDistributions)(D3Q27System::ET_NW, x1p, x2, x3) = mfcab;
+ (*this->localDistributions)(D3Q27System::ET_TE, x1, x2, x3) = mfaba;
+ (*this->localDistributions)(D3Q27System::ET_TW, x1p, x2, x3) = mfcba;
+ (*this->localDistributions)(D3Q27System::ET_TN, x1, x2, x3) = mfbaa;
+ (*this->localDistributions)(D3Q27System::ET_TS, x1, x2p, x3) = mfbca;
+ (*this->localDistributions)(D3Q27System::ET_TNE, x1, x2, x3) = mfaaa;
+ (*this->localDistributions)(D3Q27System::ET_TNW, x1p, x2, x3) = mfcaa;
+ (*this->localDistributions)(D3Q27System::ET_TSE, x1, x2p, x3) = mfaca;
+ (*this->localDistributions)(D3Q27System::ET_TSW, x1p, x2p, x3) = mfcca;
+
+ (*this->nonLocalDistributions)(D3Q27System::ET_W, x1p, x2, x3) = mfcbb;
+ (*this->nonLocalDistributions)(D3Q27System::ET_S, x1, x2p, x3) = mfbcb;
+ (*this->nonLocalDistributions)(D3Q27System::ET_B, x1, x2, x3p) = mfbbc;
+ (*this->nonLocalDistributions)(D3Q27System::ET_SW, x1p, x2p, x3) = mfccb;
+ (*this->nonLocalDistributions)(D3Q27System::ET_SE, x1, x2p, x3) = mfacb;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BW, x1p, x2, x3p) = mfcbc;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BE, x1, x2, x3p) = mfabc;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BS, x1, x2p, x3p) = mfbcc;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BN, x1, x2, x3p) = mfbac;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BSW, x1p, x2p, x3p) = mfccc;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BSE, x1, x2p, x3p) = mfacc;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BNW, x1p, x2, x3p) = mfcac;
+ (*this->nonLocalDistributions)(D3Q27System::ET_BNE, x1, x2, x3p) = mfaac;
+
+ (*this->restDistributions)(x1, x2, x3) = mfbbb;
+ //////////////////////////////////////////////////////////////////////////
+
+ }
}
- }
- }
- }
+ }
+ }
}
//////////////////////////////////////////////////////////////////////////
diff --git a/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.h b/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.h
index ca652cb37953f530d8a49a6917026652e4382a87..10cfd49264bb829eac1fc6b9bedeee3b6eace265 100644
--- a/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.h
+++ b/src/cpu/VirtualFluidsCore/LBM/CumulantK17LBMKernel.h
@@ -34,12 +34,12 @@
#ifndef CumulantK17LBMKernel_h__
#define CumulantK17LBMKernel_h__
+#include "LBMKernel.h"
#include "BCProcessor.h"
#include "D3Q27System.h"
-#include "LBMKernel.h"
-#include "basics/container/CbArray3D.h"
-#include "basics/container/CbArray4D.h"
#include "basics/utilities/UbTiming.h"
+#include "basics/container/CbArray4D.h"
+#include "basics/container/CbArray3D.h"
//! \brief Compressible cumulant LBM kernel.
//! \details LBM implementation that use Cascaded Cumulant Lattice Boltzmann method for D3Q27 model
@@ -52,17 +52,16 @@ class CumulantK17LBMKernel : public LBMKernel
{
public:
CumulantK17LBMKernel();
- ~CumulantK17LBMKernel() override;
+ ~CumulantK17LBMKernel() = default;
void calculate(int step) override;
SPtr<LBMKernel> clone() override;
+ double getCalculationTime() override { return .0; }
protected:
- inline void forwardInverseChimeraWithK(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv, LBMReal v2,
- LBMReal Kinverse, LBMReal K);
- inline void backwardInverseChimeraWithK(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv, LBMReal v2,
- LBMReal Kinverse, LBMReal K);
- inline void forwardChimera(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv, LBMReal v2);
- inline void backwardChimera(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv, LBMReal v2);
+ inline void forwardInverseChimeraWithK(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2, LBMReal Kinverse, LBMReal K);
+ inline void backwardInverseChimeraWithK(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2, LBMReal Kinverse, LBMReal K);
+ inline void forwardChimera(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2);
+ inline void backwardChimera(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2);
virtual void initDataSet();
LBMReal f[D3Q27System::ENDF + 1];
@@ -80,19 +79,18 @@ protected:
};
////////////////////////////////////////////////////////////////////////////////
-//! \brief forward chimera transformation \ref forwardInverseChimeraWithK
+//! \brief forward chimera transformation \ref forwardInverseChimeraWithK
//! Transformation from distributions to central moments according to Eq. (6)-(14) in
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> Modified for lower round-off errors.
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+//! Modified for lower round-off errors.
////////////////////////////////////////////////////////////////////////////////
-inline void CumulantK17LBMKernel::forwardInverseChimeraWithK(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv,
- LBMReal v2, LBMReal Kinverse, LBMReal K)
+inline void CumulantK17LBMKernel::forwardInverseChimeraWithK(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2, LBMReal Kinverse, LBMReal K)
{
using namespace UbMath;
LBMReal m2 = mfa + mfc;
LBMReal m1 = mfc - mfa;
LBMReal m0 = m2 + mfb;
- mfa = m0;
+ mfa = m0;
m0 *= Kinverse;
m0 += c1;
mfb = (m1 * Kinverse - m0 * vv) * K;
@@ -101,50 +99,49 @@ inline void CumulantK17LBMKernel::forwardInverseChimeraWithK(LBMReal &mfa, LBMRe
////////////////////////////////////////////////////////////////////////////////
//! \brief backward chimera transformation \ref backwardInverseChimeraWithK
//! Transformation from central moments to distributions according to Eq. (57)-(65) in
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> Modified for lower round-off errors.
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+//! ] Modified for lower round-off errors.
////////////////////////////////////////////////////////////////////////////////
-inline void CumulantK17LBMKernel::backwardInverseChimeraWithK(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv,
- LBMReal v2, LBMReal Kinverse, LBMReal K)
+inline void CumulantK17LBMKernel::backwardInverseChimeraWithK(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2, LBMReal Kinverse, LBMReal K)
{
using namespace UbMath;
LBMReal m0 = (((mfc - mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1) * (v2 - vv) * c1o2) * K;
LBMReal m1 = (((mfa - mfc) - c2 * mfb * vv) * Kinverse + (mfa * Kinverse + c1) * (-v2)) * K;
- mfc = (((mfc + mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1) * (v2 + vv) * c1o2) * K;
- mfa = m0;
- mfb = m1;
+ mfc = (((mfc + mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1) * (v2 + vv) * c1o2) * K;
+ mfa = m0;
+ mfb = m1;
}
////////////////////////////////////////////////////////////////////////////////
-//! \brief forward chimera transformation \ref forwardChimera
+//! \brief forward chimera transformation \ref forwardChimera
//! Transformation from distributions to central moments according to Eq. (6)-(14) in
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations. Modified for lower round-off
-//! errors.
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+//! for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations.
+//! Modified for lower round-off errors.
////////////////////////////////////////////////////////////////////////////////
-inline void CumulantK17LBMKernel::forwardChimera(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv, LBMReal v2)
+inline void CumulantK17LBMKernel::forwardChimera(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2)
{
using namespace UbMath;
LBMReal m1 = (mfa + mfc) + mfb;
LBMReal m2 = mfc - mfa;
- mfc = (mfc + mfa) + (v2 * m1 - c2 * vv * m2);
- mfb = m2 - vv * m1;
- mfa = m1;
+ mfc = (mfc + mfa) + (v2 * m1 - c2 * vv * m2);
+ mfb = m2 - vv * m1;
+ mfa = m1;
}
////////////////////////////////////////////////////////////////////////////////
-//! \brief backward chimera transformation \ref backwardChimera
+//! \brief backward chimera transformation \ref backwardChimera
//! Transformation from central moments to distributions according to Eq. (57)-(65) in
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations. Modified for lower round-off
-//! errors.
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
+//! for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations.
+//! Modified for lower round-off errors.
////////////////////////////////////////////////////////////////////////////////
-inline void CumulantK17LBMKernel::backwardChimera(LBMReal &mfa, LBMReal &mfb, LBMReal &mfc, LBMReal vv, LBMReal v2)
+inline void CumulantK17LBMKernel::backwardChimera(LBMReal& mfa, LBMReal& mfb, LBMReal& mfc, LBMReal vv, LBMReal v2)
{
using namespace UbMath;
LBMReal ma = (mfc + mfa * (v2 - vv)) * c1o2 + mfb * (vv - c1o2);
LBMReal mb = ((mfa - mfc) - mfa * v2) - c2 * mfb * vv;
- mfc = (mfc + mfa * (v2 + vv)) * c1o2 + mfb * (vv + c1o2);
- mfb = mb;
- mfa = ma;
+ mfc = (mfc + mfa * (v2 + vv)) * c1o2 + mfb * (vv + c1o2);
+ mfb = mb;
+ mfa = ma;
}
#endif // CumulantK17LBMKernel_h__
\ No newline at end of file
diff --git a/src/cpu/VirtualFluidsCore/LBM/D3Q27System.cpp b/src/cpu/VirtualFluidsCore/LBM/D3Q27System.cpp
index e6362563e31ad994ccc7b42c26c5e81a56a6e101..e4bea8735887c3ee9237e4fc368554e4e0002b55 100644
--- a/src/cpu/VirtualFluidsCore/LBM/D3Q27System.cpp
+++ b/src/cpu/VirtualFluidsCore/LBM/D3Q27System.cpp
@@ -1,9 +1,11 @@
#include "D3Q27System.h"
+
namespace D3Q27System
{
using namespace UbMath;
+
// index 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18//falsch
-// f: REST, E, W, N, S, T, B, NE, SW, SE, NW, TE, BW, BE, TW, TN, BS, BN, TS, TNE TNW TSE TSW BNE BNW
+// f: ZERO, E, W, N, S, T, B, NE, SW, SE, NW, TE, BW, BE, TW, TN, BS, BN, TS, TNE TNW TSE TSW BNE BNW
// BSE BSW const int EX1[] = { 0, 1, -1, 0, 0, 0, 0, 1, -1, 1, -1, 1, -1, 1, -1, 0, 0, 0, 0, 1, -1, 1, -1,
// 1, -1, 1, -1 }; const int EX2[] = { 0, 0, 0, 1, -1, 0, 0, 1, -1, -1, 1, 0, 0, 0, 0, 1, -1, 1, -1, 1, 1,
// -1, -1, 1, 1, -1, -1 }; const int EX3[] = { 0, 0, 0, 0, 0, 1, -1, 0, 0, 0, 0, 1, -1, -1, 1, 1, -1, -1,
@@ -152,7 +154,7 @@ const double cNorm[3][ENDDIR] = { { double(DX1[0]),
// const int BNW = 23;
// const int BSE = 24;
// const int BSW = 25;
-// const int REST /*f0 */ = 26;
+// const int ZERO /*f0 */ = 26;
// const int INV_E = W;
// const int INV_W = E;
diff --git a/src/cpu/VirtualFluidsCore/LBM/D3Q27System.h b/src/cpu/VirtualFluidsCore/LBM/D3Q27System.h
index b3da79948c60e986b1c96f6f4266012dd7bffd23..b5d88d6c3791d716cd0dca567d7aaa803e863536 100644
--- a/src/cpu/VirtualFluidsCore/LBM/D3Q27System.h
+++ b/src/cpu/VirtualFluidsCore/LBM/D3Q27System.h
@@ -35,8 +35,8 @@
#define D3Q27SYSTEM_H
#include <cmath>
-#include <iostream>
#include <string>
+#include <iostream>
#include "LBMSystem.h"
#include "UbException.h"
@@ -89,7 +89,7 @@ static const int BNE = 22;
static const int BNW = 23;
static const int BSE = 24;
static const int BSW = 25;
-static const int REST = 26;
+static const int ZERO = 26;
static const int INV_E = W;
static const int INV_W = E;
@@ -155,7 +155,7 @@ static LBMReal getDensity(const LBMReal *const &f /*[27]*/)
return ((f[TNE] + f[BSW]) + (f[TSE] + f[BNW])) + ((f[BSE] + f[TNW]) + (f[TSW] + f[BNE])) +
(((f[NE] + f[SW]) + (f[SE] + f[NW])) + ((f[TE] + f[BW]) + (f[BE] + f[TW])) +
((f[BN] + f[TS]) + (f[TN] + f[BS]))) +
- ((f[E] + f[W]) + (f[N] + f[S]) + (f[T] + f[B])) + f[REST];
+ ((f[E] + f[W]) + (f[N] + f[S]) + (f[T] + f[B])) + f[ZERO];
}
/*=====================================================================*/
// ATTENTION: does not apply to all models -> use certificate instead of static! to do
@@ -184,7 +184,7 @@ static void calcDensity(const LBMReal *const &f /*[27]*/, LBMReal &rho)
rho = ((f[TNE] + f[BSW]) + (f[TSE] + f[BNW])) + ((f[BSE] + f[TNW]) + (f[TSW] + f[BNE])) +
(((f[NE] + f[SW]) + (f[SE] + f[NW])) + ((f[TE] + f[BW]) + (f[BE] + f[TW])) +
((f[BN] + f[TS]) + (f[TN] + f[BS]))) +
- ((f[E] + f[W]) + (f[N] + f[S]) + (f[T] + f[B])) + f[REST];
+ ((f[E] + f[W]) + (f[N] + f[S]) + (f[T] + f[B])) + f[ZERO];
}
/*=====================================================================*/
static void calcIncompVelocityX1(const LBMReal *const &f /*[27]*/, LBMReal &vx1)
@@ -279,7 +279,7 @@ static LBMReal getCompFeqForDirection(const int &direction, const LBMReal &drho,
////-----
LBMReal rho = drho + c1;
switch (direction) {
- case REST:
+ case ZERO:
return REAL_CAST(c8o27 * (drho + rho * (-cu_sq)));
case E:
return REAL_CAST(c2o27 * (drho + rho * (3.0 * (vx1) + c9o2 * (vx1) * (vx1)-cu_sq)));
@@ -354,7 +354,7 @@ static void calcCompFeq(LBMReal *const &feq /*[27]*/, const LBMReal &drho, const
LBMReal cu_sq = 1.5 * (vx1 * vx1 + vx2 * vx2 + vx3 * vx3);
LBMReal rho = drho + c1;
- feq[REST] = c8o27 * (drho + rho * (-cu_sq));
+ feq[ZERO] = c8o27 * (drho + rho * (-cu_sq));
feq[E] = c2o27 * (drho + rho * (3.0 * (vx1) + c9o2 * (vx1) * (vx1)-cu_sq));
feq[W] = c2o27 * (drho + rho * (3.0 * (-vx1) + c9o2 * (-vx1) * (-vx1) - cu_sq));
feq[N] = c2o27 * (drho + rho * (3.0 * (vx2) + c9o2 * (vx2) * (vx2)-cu_sq));
@@ -395,7 +395,7 @@ static LBMReal getIncompFeqForDirection(const int &direction, const LBMReal &drh
LBMReal cu_sq = 1.5f * (vx1 * vx1 + vx2 * vx2 + vx3 * vx3);
switch (direction) {
- case REST:
+ case ZERO:
return REAL_CAST(c8o27 * (drho - cu_sq));
case E:
return REAL_CAST(c2o27 * (drho + 3.0 * (vx1) + c9o2 * (vx1) * (vx1)-cu_sq));
@@ -435,25 +435,25 @@ static LBMReal getIncompFeqForDirection(const int &direction, const LBMReal &drh
return REAL_CAST(c1o54 * (drho + 3.0 * (-vx2 + vx3) + c9o2 * (-vx2 + vx3) * (-vx2 + vx3) - cu_sq));
case TNE:
return REAL_CAST(c1o216 *
- (drho + 3.0 * (vx1 + vx2 + vx3) + c9o2 * (vx1 + vx2 + vx3) * (vx1 + vx2 + vx3) - cu_sq));
+ (drho + 3.0 * (vx1 + vx2 + vx3) + c9o2 * (vx1 + vx2 + vx3) * (vx1 + vx2 + vx3) - cu_sq));
case BSW:
return REAL_CAST(
c1o216 * (drho + 3.0 * (-vx1 - vx2 - vx3) + c9o2 * (-vx1 - vx2 - vx3) * (-vx1 - vx2 - vx3) - cu_sq));
case BNE:
return REAL_CAST(c1o216 *
- (drho + 3.0 * (vx1 + vx2 - vx3) + c9o2 * (vx1 + vx2 - vx3) * (vx1 + vx2 - vx3) - cu_sq));
+ (drho + 3.0 * (vx1 + vx2 - vx3) + c9o2 * (vx1 + vx2 - vx3) * (vx1 + vx2 - vx3) - cu_sq));
case TSW:
return REAL_CAST(
c1o216 * (drho + 3.0 * (-vx1 - vx2 + vx3) + c9o2 * (-vx1 - vx2 + vx3) * (-vx1 - vx2 + vx3) - cu_sq));
case TSE:
return REAL_CAST(c1o216 *
- (drho + 3.0 * (vx1 - vx2 + vx3) + c9o2 * (vx1 - vx2 + vx3) * (vx1 - vx2 + vx3) - cu_sq));
+ (drho + 3.0 * (vx1 - vx2 + vx3) + c9o2 * (vx1 - vx2 + vx3) * (vx1 - vx2 + vx3) - cu_sq));
case BNW:
return REAL_CAST(
c1o216 * (drho + 3.0 * (-vx1 + vx2 - vx3) + c9o2 * (-vx1 + vx2 - vx3) * (-vx1 + vx2 - vx3) - cu_sq));
case BSE:
return REAL_CAST(c1o216 *
- (drho + 3.0 * (vx1 - vx2 - vx3) + c9o2 * (vx1 - vx2 - vx3) * (vx1 - vx2 - vx3) - cu_sq));
+ (drho + 3.0 * (vx1 - vx2 - vx3) + c9o2 * (vx1 - vx2 - vx3) * (vx1 - vx2 - vx3) - cu_sq));
case TNW:
return REAL_CAST(
c1o216 * (drho + 3.0 * (-vx1 + vx2 + vx3) + c9o2 * (-vx1 + vx2 + vx3) * (-vx1 + vx2 + vx3) - cu_sq));
@@ -469,7 +469,7 @@ static void calcIncompFeq(LBMReal *const &feq /*[27]*/, const LBMReal &drho, con
LBMReal cu_sq = 1.5 * (vx1 * vx1 + vx2 * vx2 + vx3 * vx3);
- feq[REST] = c8o27 * (drho - cu_sq);
+ feq[ZERO] = c8o27 * (drho - cu_sq);
feq[E] = c2o27 * (drho + 3.0 * (vx1) + c9o2 * (vx1) * (vx1)-cu_sq);
feq[W] = c2o27 * (drho + 3.0 * (-vx1) + c9o2 * (-vx1) * (-vx1) - cu_sq);
feq[N] = c2o27 * (drho + 3.0 * (vx2) + c9o2 * (vx2) * (vx2)-cu_sq);
@@ -758,7 +758,7 @@ static inline LBMReal calcPress(const LBMReal *const f, LBMReal rho, LBMReal vx1
c2 * (f[NE] + f[SW] + f[SE] + f[NW] + f[TE] + f[BW] + f[BE] + f[TW] + f[TN] + f[BS] + f[BN] + f[TS]) +
c3 * (f[TNE] + f[TSW] + f[TSE] + f[TNW] + f[BNE] + f[BSW] + f[BSE] + f[BNW]) -
(vx1 * vx1 + vx2 * vx2 + vx3 * vx3)) *
- (c1 - c1o2 * OxxPyyPzz) +
+ (c1 - c1o2 * OxxPyyPzz) +
OxxPyyPzz * c1o2 * (rho)) *
c1o3;
}
diff --git a/src/cpu/VirtualFluidsCore/LBM/ILBMKernel.h b/src/cpu/VirtualFluidsCore/LBM/ILBMKernel.h
index 44d8d3273d6bfb555acb2c994b97dfbeb676f1c3..4dbe8eee09a37c0c220f47619b72bade2e6ec527 100644
--- a/src/cpu/VirtualFluidsCore/LBM/ILBMKernel.h
+++ b/src/cpu/VirtualFluidsCore/LBM/ILBMKernel.h
@@ -45,8 +45,9 @@ class ILBMKernel
public:
virtual ~ILBMKernel() = default;
- virtual void calculate(int step) = 0;
- virtual void swapDistributions() = 0;
+ virtual void calculate(int step) = 0;
+ virtual double getCalculationTime() = 0;
+ virtual void swapDistributions() = 0;
virtual bool getCompressible() const = 0;
virtual SPtr<BCProcessor> getBCProcessor() const = 0;
diff --git a/src/cpu/VirtualFluidsCore/LBM/LBMKernel.h b/src/cpu/VirtualFluidsCore/LBM/LBMKernel.h
index bfaf9d31275d12ac2d4795c46787af864329980d..be29589b9b7ab239bece700126ae906795c83977 100644
--- a/src/cpu/VirtualFluidsCore/LBM/LBMKernel.h
+++ b/src/cpu/VirtualFluidsCore/LBM/LBMKernel.h
@@ -56,6 +56,9 @@ public:
virtual SPtr<LBMKernel> clone() = 0;
+ void calculate(int step) override = 0;
+ double getCalculationTime() override = 0;
+
void setBCProcessor(SPtr<BCProcessor> bcp) override;
SPtr<BCProcessor> getBCProcessor() const override;
diff --git a/src/cpu/VirtualFluidsCore/LBM/LBMUnitConverter.h b/src/cpu/VirtualFluidsCore/LBM/LBMUnitConverter.h
index 376330314497224f2ac92e61de7ae4bf81589c93..40570cc3847f71a1942791afa7e95145daafb53b 100644
--- a/src/cpu/VirtualFluidsCore/LBM/LBMUnitConverter.h
+++ b/src/cpu/VirtualFluidsCore/LBM/LBMUnitConverter.h
@@ -97,6 +97,14 @@ public:
this->init(refLengthWorld, csWorld, rhoWorld, csWorld, refLengthLb, rhoLb, csLb);
}
+ LBMUnitConverter(int /*dummy*/, double uReal, double uLB, double nuReal, double nuLB)
+ {
+ factorVelocityLbToW = uReal / uLB;
+ factorViscosityLbToW = nuReal / nuLB;
+ factorDensityLbToW = factorViscosityLbToW * factorVelocityLbToW * factorVelocityLbToW;
+ factorPressureLbToW = factorDensityLbToW;
+ }
+
virtual ~LBMUnitConverter() = default;
double getRefRhoLb() { return refRhoLb; }
@@ -132,6 +140,10 @@ public:
double getFactorAccWToLb() { return 1.0 / this->getFactorAccLbToW(); }
double getFactorTimeLbToW(double deltaX) const { return factorTimeWithoutDx * deltaX; }
+ //////////////////////////////////////////////////////////////////////////
+ double getFactorVelocityLbToW2() { return factorVelocityLbToW; }
+ double getFactorDensityLbToW2() { return factorDensityLbToW; }
+ double getFactorPressureLbToW2() { return factorPressureLbToW; }
/*==========================================================*/
friend inline std::ostream &operator<<(std::ostream &os, LBMUnitConverter c)
@@ -199,7 +211,12 @@ protected:
double factorTimeLbToW{ 1.0 };
double factorMassLbToW{ 1.0 };
double refRhoLb{ 1.0 };
- double factorTimeWithoutDx;
+ double factorTimeWithoutDx{ 0.0 };
+
+ double factorVelocityLbToW{ 1.0 };
+ double factorViscosityLbToW{ 1.0 };
+ double factorDensityLbToW{ 1.0 };
+ double factorPressureLbToW{ 1.0 };
};
#endif // LBMUNITCONVERTER_H
diff --git a/src/cpu/VirtualFluidsCore/Utilities/MemoryUtil.h b/src/cpu/VirtualFluidsCore/Utilities/MemoryUtil.h
index b476fd7750fb5ef950d13b6157349607b6d4d7c9..670a597cb84bd4e98450dad2743a8100f04497ea 100644
--- a/src/cpu/VirtualFluidsCore/Utilities/MemoryUtil.h
+++ b/src/cpu/VirtualFluidsCore/Utilities/MemoryUtil.h
@@ -36,8 +36,8 @@
#if defined(_WIN32) || defined(_WIN64)
#define MEMORYUTIL_WINDOWS
-#include "psapi.h"
#include "windows.h"
+#include "psapi.h"
#pragma comment(lib, "psapi.lib")
#elif defined __APPLE__
#define MEMORYUTIL_APPLE
diff --git a/src/cpu/VirtualFluidsCore/Visitors/InitDistributionsBlockVisitor.cpp b/src/cpu/VirtualFluidsCore/Visitors/InitDistributionsBlockVisitor.cpp
index 8cbf1801f94111f26f5874753271e24873420a1e..c424a6376a62159da2e4b9f73ccf01858fbde521 100644
--- a/src/cpu/VirtualFluidsCore/Visitors/InitDistributionsBlockVisitor.cpp
+++ b/src/cpu/VirtualFluidsCore/Visitors/InitDistributionsBlockVisitor.cpp
@@ -294,7 +294,7 @@ void InitDistributionsBlockVisitor::visit(const SPtr<Grid3D> grid, SPtr<Block3D>
f[BNW] = f_TSE + feq[BNW];
f[BSE] = f_TNW + feq[BSE];
f[BSW] = f_TNE + feq[BSW];
- f[REST] = f_ZERO + feq[REST];
+ f[ZERO] = f_ZERO + feq[ZERO];
// calcFeqsFct(f,rho,vx1,vx2,vx3);
// distributions->setDistribution(f, ix1, ix2, ix3);
diff --git a/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.cpp b/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.cpp
index 9814ab036621d94833ad9b4baff61866b22eac4c..c0efdcc6135b1e06f766621dd4528f96fa32247d 100644
--- a/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.cpp
+++ b/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.cpp
@@ -37,13 +37,11 @@
#include "Grid3D.h"
#include "Grid3DSystem.h"
-SetConnectorsBlockVisitor::SetConnectorsBlockVisitor(SPtr<Communicator> comm, bool fullConnector, int dirs, LBMReal nu)
- : Block3DVisitor(0, Grid3DSystem::MAXLEVEL), comm(comm), fullConnector(fullConnector), dirs(dirs), nu(nu)
+SetConnectorsBlockVisitor::SetConnectorsBlockVisitor(SPtr<Communicator> comm, bool fullConnector, int dirs, LBMReal nue)
+ : Block3DVisitor(0, Grid3DSystem::MAXLEVEL), comm(comm), fullConnector(fullConnector), dirs(dirs), nue(nue)
{
}
//////////////////////////////////////////////////////////////////////////
-SetConnectorsBlockVisitor::~SetConnectorsBlockVisitor(void) = default;
-//////////////////////////////////////////////////////////////////////////
void SetConnectorsBlockVisitor::visit(SPtr<Grid3D> grid, SPtr<Block3D> block)
{
if (!block)
diff --git a/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.h b/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.h
index 7d209e0524a13ab89a48dfc2179203eb17b95162..f6eb15206371af2ff6106a5c82c6c71eba26fb34 100644
--- a/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.h
+++ b/src/cpu/VirtualFluidsCore/Visitors/SetConnectorsBlockVisitor.h
@@ -39,6 +39,7 @@
#include "Block3DVisitor.h"
#include "D3Q27System.h"
+
class Grid3D;
class Block3D;
class Communicator;
@@ -48,8 +49,8 @@ class InterpolationProcessor;
class SetConnectorsBlockVisitor : public Block3DVisitor
{
public:
- SetConnectorsBlockVisitor(SPtr<Communicator> comm, bool fullConnector, int dirs, LBMReal nu);
- ~SetConnectorsBlockVisitor() override;
+ SetConnectorsBlockVisitor(SPtr<Communicator> comm, bool fullConnector, int dirs, LBMReal nue);
+ ~SetConnectorsBlockVisitor() = default;
void visit(SPtr<Grid3D> grid, SPtr<Block3D> block) override;
//////////////////////////////////////////////////////////////////////////
protected:
@@ -58,7 +59,7 @@ protected:
bool fullConnector;
int dirs;
int gridRank;
- LBMReal nu;
+ LBMReal nue;
SPtr<InterpolationProcessor> iProcessor;
};
diff --git a/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.cpp b/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.cpp
index bc4b3d25701789cfb62b107880680a3287486f8b..7dde0c34edf5c16dff22361a6bbc36394a9783ed 100644
--- a/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.cpp
+++ b/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.cpp
@@ -1,3 +1,4 @@
+#include "MemoryUtil.h"
//=======================================================================================
// ____ ____ __ ______ __________ __ __ __ __
// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
@@ -43,13 +44,24 @@
#include <utility>
//////////////////////////////////////////////////////////////////////////
-SetKernelBlockVisitor::SetKernelBlockVisitor(SPtr<LBMKernel> kernel, LBMReal nue, SetKernelBlockVisitor::Action action)
+SetKernelBlockVisitor::SetKernelBlockVisitor(SPtr<LBMKernel> kernel, LBMReal nue,
+ SetKernelBlockVisitor::Action action)
: Block3DVisitor(0, Grid3DSystem::MAXLEVEL), kernel(std::move(kernel)), nue(nue), action(action), dataSetFlag(true)
{
}
+
+SetKernelBlockVisitor::SetKernelBlockVisitor(SPtr<LBMKernel> kernel, LBMReal nue, int &numberOfProcesses,
+ SetKernelBlockVisitor::Action action)
+ : Block3DVisitor(0, Grid3DSystem::MAXLEVEL), kernel(std::move(kernel)), nue(nue), action(action), dataSetFlag(true),
+ numberOfProcesses(numberOfProcesses)
+{
+}
+
//////////////////////////////////////////////////////////////////////////
void SetKernelBlockVisitor::visit(SPtr<Grid3D> grid, SPtr<Block3D> block)
{
+ throwExceptionIfNotEnoughMemory(grid);
+
if (kernel && (block->getRank() == grid->getRank())) {
LBMReal collFactor = LBMSystem::calcCollisionFactor(nue, block->getLevel());
kernel->setCollisionFactor(collFactor);
@@ -98,3 +110,27 @@ void SetKernelBlockVisitor::visit(SPtr<Grid3D> grid, SPtr<Block3D> block)
}
void SetKernelBlockVisitor::setNoDataSetFlag(bool flag) { dataSetFlag = flag; }
+
+void SetKernelBlockVisitor::throwExceptionIfNotEnoughMemory(const SPtr<Grid3D> &grid)
+{
+ auto availableMemory = Utilities::getTotalPhysMem();
+ auto requiredMemory = getRequiredPhysicalMemory(grid);
+ if (requiredMemory > availableMemory)
+ throw UbException(UB_EXARGS, "SetKernelBlockVisitor: Not enough memory!!!");
+}
+
+double SetKernelBlockVisitor::getRequiredPhysicalMemory(const SPtr<Grid3D> &grid) const
+{
+ unsigned long long numberOfNodesPerBlockWithGhostLayer;
+ auto numberOfBlocks = (unsigned long long)grid->getNumberOfBlocks();
+ auto blockNx = grid->getBlockNX();
+ int ghostLayer = 3;
+
+ numberOfNodesPerBlockWithGhostLayer = numberOfBlocks * (val<1>(blockNx) + ghostLayer) *
+ (val<2>(blockNx) + ghostLayer) * (val<3>(blockNx) + ghostLayer);
+
+ auto needMemAll =
+ double(numberOfNodesPerBlockWithGhostLayer * (27 * sizeof(double) + sizeof(int) + sizeof(float) * 4));
+
+ return needMemAll / double(numberOfProcesses);
+}
diff --git a/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.h b/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.h
index c2863555c0490b24e23ae2d99933e62ce94ac467..7ce7c852e2a815bdf0a37f3ef2960e1b5b76e4b4 100644
--- a/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.h
+++ b/src/cpu/VirtualFluidsCore/Visitors/SetKernelBlockVisitor.h
@@ -51,6 +51,10 @@ public:
SetKernelBlockVisitor(SPtr<LBMKernel> kernel, LBMReal nue,
SetKernelBlockVisitor::Action action = SetKernelBlockVisitor::NewKernel);
+
+ SetKernelBlockVisitor(SPtr<LBMKernel> kernel, LBMReal nue, int &numberOfProcesses,
+ SetKernelBlockVisitor::Action action = SetKernelBlockVisitor::NewKernel);
+
~SetKernelBlockVisitor() override = default;
void visit(SPtr<Grid3D> grid, SPtr<Block3D> block) override;
@@ -62,6 +66,12 @@ private:
LBMReal nue;
Action action;
bool dataSetFlag;
+
+ int numberOfProcesses{ 1 };
+
+ double getRequiredPhysicalMemory(const SPtr<Grid3D> &grid) const;
+
+ void throwExceptionIfNotEnoughMemory(const SPtr<Grid3D> &grid);
};
#endif
diff --git a/src/gpu/GksGpu/CMakeLists.txt b/src/gpu/GksGpu/CMakeLists.txt
index 88df905d3eb847c1ed5e9b5bc5a1ec68c549e313..da404e0209ed2c9f36ae323d2e6bd234fb6dfb96 100644
--- a/src/gpu/GksGpu/CMakeLists.txt
+++ b/src/gpu/GksGpu/CMakeLists.txt
@@ -1,8 +1,3 @@
+project(GksGpu LANGUAGES CUDA CXX)
-vf_add_library(BUILDTYPE shared PRIVATE_LINK basics GksMeshAdapter)
-
-linkCUDA()
-linkMPI()
-
-vf_get_library_name(library_name)
-linkOpenMP(${library_name})
+vf_add_library(PRIVATE_LINK basics GksMeshAdapter OpenMP::OpenMP_CXX MPI::MPI_CXX)
diff --git a/src/gpu/GksMeshAdapter/CMakeLists.txt b/src/gpu/GksMeshAdapter/CMakeLists.txt
index 4a16d6fbc0b99d1ec400944178844123eb50ffc0..cb00b3c016786c41ef5640eb362322bb0a3768f8 100644
--- a/src/gpu/GksMeshAdapter/CMakeLists.txt
+++ b/src/gpu/GksMeshAdapter/CMakeLists.txt
@@ -1,5 +1,3 @@
+project(GksMeshAdapter LANGUAGES CUDA CXX)
-
-vf_add_library(BUILDTYPE shared PRIVATE_LINK basics GridGenerator)
-
-linkCUDA()
+vf_add_library(PRIVATE_LINK basics GridGenerator)
diff --git a/src/gpu/GridGenerator/CMakeLists.txt b/src/gpu/GridGenerator/CMakeLists.txt
index aaabd6d5de07940194e8428f0d3c9a18a741ae33..1ce294bf420f657f35397c427929f9f310d04556 100644
--- a/src/gpu/GridGenerator/CMakeLists.txt
+++ b/src/gpu/GridGenerator/CMakeLists.txt
@@ -1,9 +1,13 @@
+project(GridGenerator LANGUAGES CUDA CXX)
-vf_add_library(BUILDTYPE shared PRIVATE_LINK basics)
-vf_get_library_name(library_name)
+vf_add_library(PRIVATE_LINK basics OpenMP::OpenMP_CXX)
-linkCUDA()
-linkOpenMP(${library_name})
+vf_get_library_name(library_name)
+set_target_properties(${library_name} PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
-set_target_properties(${library_name} PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
\ No newline at end of file
+# according to linker error when building static libraries.
+# https://stackoverflow.com/questions/50033435/cmake-cuda-separate-compilation-static-lib-link-error-on-windows-but-not-on-ubun
+if (NOT BUILD_SHARED_LIBRARY)
+ set_target_properties(${library_name} PROPERTIES CUDA_RESOLVE_DEVICE_SYMBOLS ON)
+endif()
\ No newline at end of file
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp
index 750550d7e79b5aba978d8052b83668958cebb837..4081aeffd9165e838959c78e39c9b51d6082c4a7 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp
@@ -37,7 +37,7 @@
#include "grid/BoundaryConditions/Side.h"
#include "grid/Grid.h"
-bool BoundaryCondition::isSide( SideType side ) const
+bool gg::BoundaryCondition::isSide( SideType side ) const
{
return this->side->whoAmI() == side;
}
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h
index e4e67af092841b426e0d7774048af78abedc79ac..b082e6a7402a606b72d08bc28e9b612fa6661974 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h
@@ -44,25 +44,30 @@ class Grid;
class Side;
enum class SideType;
+namespace gg
+{
+
class BoundaryCondition
{
public:
std::vector<uint> indices;
SPtr<Side> side;
- std::vector<std::vector<real> > qs;
+ std::vector<std::vector<real>> qs;
std::vector<uint> patches;
virtual char getType() const = 0;
- bool isSide( SideType side ) const;
+ bool isSide(SideType side) const;
- real getQ( uint index, uint dir ){ return this->qs[index][dir]; }
+ real getQ(uint index, uint dir) { return this->qs[index][dir]; }
};
+}
+
//////////////////////////////////////////////////////////////////////////
-class PressureBoundaryCondition : public BoundaryCondition
+class PressureBoundaryCondition : public gg::BoundaryCondition
{
public:
static SPtr<PressureBoundaryCondition> make(real rho)
@@ -91,7 +96,7 @@ public:
//////////////////////////////////////////////////////////////////////////
-class VelocityBoundaryCondition : public BoundaryCondition
+class VelocityBoundaryCondition : public gg ::BoundaryCondition
{
public:
static SPtr<VelocityBoundaryCondition> make(real vx, real vy, real vz)
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp
index 6937ac88c35e75e95c9e6a3d2973d170857dc028..99097a393735a31abad0dd717a2dcfc2b1d35326 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp
@@ -38,6 +38,8 @@
#include "utilities/math/Math.h"
+using namespace gg;
+
void Side::addIndices(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition, std::string coord, real constant,
real startInner, real endInner, real startOuter, real endOuter)
{
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h
index 7208ece4835e7476cdb17d236bf9bfabbba0493a..d8dc9a0e4ac8c2825d49dd19148d53538d027a8e 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h
@@ -46,7 +46,11 @@
#define NEGATIVE_DIR -1
class Grid;
+
+namespace gg
+{
class BoundaryCondition;
+}
class Side;
@@ -60,7 +64,7 @@ enum class SideType
class Side
{
public:
- virtual void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) = 0;
+ virtual void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) = 0;
virtual int getCoordinate() const = 0;
virtual int getDirection() const = 0;
@@ -68,12 +72,12 @@ public:
virtual SideType whoAmI() const = 0;
protected:
- static void addIndices(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition, std::string coord, real constant,
+ static void addIndices(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, std::string coord, real constant,
real startInner, real endInner, real startOuter, real endOuter);
- static void setPressureNeighborIndices(SPtr<BoundaryCondition> boundaryCondition, SPtr<Grid> grid, const uint index);
+ static void setPressureNeighborIndices(SPtr<gg::BoundaryCondition> boundaryCondition, SPtr<Grid> grid, const uint index);
- static void setQs(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition, uint index);
+ static void setQs(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, uint index);
private:
static uint getIndex(SPtr<Grid> grid, std::string coord, real constant, real v1, real v2);
@@ -82,7 +86,7 @@ private:
class MX : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) override;
+ void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -103,7 +107,7 @@ public:
class PX : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) override;
+ void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -125,7 +129,7 @@ public:
class MY : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) override;
+ void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -146,7 +150,7 @@ public:
class PY : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) override;
+ void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -168,7 +172,7 @@ public:
class MZ : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) override;
+ void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -189,7 +193,7 @@ public:
class PZ : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition) override;
+ void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
diff --git a/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h b/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h
index 84b6c8062d5d4cbe77a8ac13919ba9b7ca4b976e..f398e89a0936e0e03261c7b75f22e7311d96e161 100644
--- a/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h
+++ b/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h
@@ -60,7 +60,10 @@ class Grid;
enum class SideType;
+namespace gg
+{
class BoundaryCondition;
+}
class GeometryBoundaryCondition;
class GridBuilder
@@ -93,7 +96,7 @@ public:
virtual void getVelocityValues(real* vx, real* vy, real* vz, int* indices, int level) const = 0;
virtual void getVelocityQs(real* qs[27], int level) const = 0;
- virtual SPtr<BoundaryCondition> getBoundaryCondition( SideType side, uint level ) const = 0;
+ virtual SPtr<gg::BoundaryCondition> getBoundaryCondition( SideType side, uint level ) const = 0;
};
diff --git a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp
index 09debba7d011ec20c389ae9657b8a46185863b58..7cf735197b9fd69c7f85351e659fb31c4818efd6 100644
--- a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp
+++ b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp
@@ -227,7 +227,7 @@ void LevelGridBuilder::getVelocityQs(real* qs[27], int level) const
}
}
-GRIDGENERATOR_EXPORT SPtr<BoundaryCondition> LevelGridBuilder::getBoundaryCondition(SideType side, uint level) const
+GRIDGENERATOR_EXPORT SPtr<gg::BoundaryCondition> LevelGridBuilder::getBoundaryCondition(SideType side, uint level) const
{
for( auto bc : this->boundaryConditions[level]->velocityBoundaryConditions )
if( bc->isSide(side) )
diff --git a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h
index dcf491070de02ed4d700e778d0d883ce2b9b959e..100dedede48d962174b852b88ede465777bc25c2 100644
--- a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h
+++ b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h
@@ -87,7 +87,7 @@ public:
GRIDGENERATOR_EXPORT virtual void getVelocityValues(real* vx, real* vy, real* vz, int* indices, int level) const;
GRIDGENERATOR_EXPORT virtual void getVelocityQs(real* qs[27], int level) const;
- GRIDGENERATOR_EXPORT SPtr<BoundaryCondition> getBoundaryCondition( SideType side, uint level ) const override;
+ GRIDGENERATOR_EXPORT SPtr<gg::BoundaryCondition> getBoundaryCondition( SideType side, uint level ) const override;
protected:
diff --git a/src/gpu/VirtualFluids_GPU/CMakeLists.txt b/src/gpu/VirtualFluids_GPU/CMakeLists.txt
index 477a8cc73b17873dfbfc0e11173db0ff5ad5593d..95405b25864f506e5580fa6711a4389d39d8d7a7 100644
--- a/src/gpu/VirtualFluids_GPU/CMakeLists.txt
+++ b/src/gpu/VirtualFluids_GPU/CMakeLists.txt
@@ -1,21 +1,14 @@
+project(VirtualFluids_GPU LANGUAGES CUDA CXX)
+set(additional_libraries "")
+if(MSVC)
+ set(additional_libraries ws2_32 Traffic) # ws_32 throws an error on Phoenix
+endif()
-IF(MSVC)
- set(libsToLink ws2_32 GridGenerator basics) # ws_32 throws an error on Phoenix
-ELSE(MSVC)
- set(libsToLink GridGenerator basics)
-ENDIF(MSVC)
+vf_add_library(PRIVATE_LINK ${additional_libraries} GridGenerator basics MPI::MPI_CXX)
-
-vf_add_library(BUILDTYPE shared PRIVATE_LINK ${libsToLink})
-
-linkCUDA()
+linkBoost(COMPONENTS "serialization")
#SET(TPN_WIN32 "/EHsc")
#https://stackoverflow.com/questions/6832666/lnk2019-when-including-asio-headers-solution-generated-with-cmake
#https://stackoverflow.com/questions/27442885/syntax-error-with-stdnumeric-limitsmax
-
-IF(MSVC)
- vf_get_library_name(library_name)
- set_target_properties(${library_name} PROPERTIES LINK_FLAGS "/ignore:4251")
-ENDIF(MSVC)