Merge branch 'release-candidate-4.4.00' for 4.4.00

Part of Kokkos C++ Performance Portability Programming EcoSystem 4.4
ndellingwood · Aug 7, 2024 · 03a8d18 · 03a8d18
2 parents 6ecdf60 + f15e90c
commit 03a8d18
Show file tree

Hide file tree

Showing 264 changed files with 14,413 additions and 9,984 deletions.
diff --git a/.github/dependabot.yml b/.github/dependabot.yml
@@ -0,0 +1,6 @@
+version: 2
+updates:
+  - package-ecosystem: "github-actions"
+    directory: /
+    schedule:
+      interval: "weekly"
diff --git a/.github/workflows/clang-format-check.yml b/.github/workflows/clang-format-check.yml
@@ -10,6 +10,6 @@ jobs:
     steps:
     - uses: actions/checkout@v4
     - name: Run clang-format style check.
-      uses: DoozyX/clang-format-lint-action@v0.16.2
+      uses: DoozyX/clang-format-lint-action@v0.17
       with:
         clangFormatVersion: 8
diff --git a/.github/workflows/continuous-integration-workflow-32bit.yml b/.github/workflows/continuous-integration-workflow-32bit.yml
@@ -36,7 +36,7 @@ jobs:
             -DKokkos_ENABLE_DEPRECATED_CODE_4=ON \
             -DKokkos_ENABLE_DEPRECATION_WARNINGS=OFF \
             -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
-            -DCMAKE_CXX_FLAGS="-Werror -m32 -DKOKKOS_IMPL_32BIT" \
+            -DCMAKE_CXX_FLAGS="-Werror -m32" \
             -DCMAKE_CXX_COMPILER=g++ \
             -DCMAKE_BUILD_TYPE=RelWithDebInfo
       - name: Build

diff --git a/.github/workflows/continuous-integration-workflow.yml b/.github/workflows/continuous-integration-workflow.yml
@@ -20,50 +20,59 @@ jobs:
     continue-on-error: true
     strategy:
       matrix:
-        distro: ['fedora:latest', 'fedora:rawhide', 'ubuntu:latest']
+        distro: ['fedora:latest', 'ubuntu:latest']
         cxx: ['g++', 'clang++']
         cxx_extra_flags: ['']
         cmake_build_type: ['Release', 'Debug']
         backend: ['OPENMP']
         clang-tidy: ['']
+        stdcxx: [17]
         include:
           - distro: 'ubuntu:intel'
             cxx: 'icpc'
             cxx_extra_flags: '-diag-disable=177,10441'
             cmake_build_type: 'Release'
             backend: 'OPENMP'
+            stdcxx: '17'
           - distro: 'ubuntu:intel'
             cxx: 'icpc'
             cxx_extra_flags: '-diag-disable=177,10441'
             cmake_build_type: 'Debug'
             backend: 'OPENMP'
+            stdcxx: '17'
           - distro: 'ubuntu:intel'
             cxx: 'icpx'
-            cxx_extra_flags: '-fp-model=precise -Wno-pass-failed'
+            cxx_extra_flags: '-fp-model=precise -Wno-pass-failed -fsanitize=address,undefined -fno-sanitize=function -fno-sanitize-recover=all'
+            extra_linker_flags: '-fsanitize=address,undefined -fno-sanitize=function -fno-sanitize-recover=all'
             cmake_build_type: 'Release'
             backend: 'OPENMP'
+            stdcxx: '17'
           - distro: 'ubuntu:intel'
             cxx: 'icpx'
             cxx_extra_flags: '-fp-model=precise -Wno-pass-failed'
             cmake_build_type: 'Debug'
             backend: 'OPENMP'
+            stdcxx: '20'
           - distro: 'ubuntu:latest'
             cxx: 'clang++'
-            cxx_extra_flags: '-fsanitize=address'
-            extra_linker_flags: '-fsanitize=address'
+            cxx_extra_flags: '-fsanitize=address,undefined -fno-sanitize=function -fno-sanitize-recover=all'
+            extra_linker_flags: '-fsanitize=address,undefined -fno-sanitize=function -fno-sanitize-recover=all'
             cmake_build_type: 'RelWithDebInfo'
             backend: 'THREADS'
             clang-tidy: '-DCMAKE_CXX_CLANG_TIDY="clang-tidy;-warnings-as-errors=*"'
+            stdcxx: '23'
           - distro: 'ubuntu:latest'
             cxx: 'clang++'
-            cxx_extra_flags: '-fsanitize=address'
-            extra_linker_flags: '-fsanitize=address'
+            cxx_extra_flags: '-fsanitize=address,undefined -fno-sanitize=function -fno-sanitize-recover=all'
+            extra_linker_flags: '-fsanitize=address,undefined -fno-sanitize=function -fno-sanitize-recover=all'
             cmake_build_type: 'RelWithDebInfo'
             backend: 'SERIAL'
+            stdcxx: '20'
           - distro: 'ubuntu:latest'
             cxx: 'g++'
             cmake_build_type: 'RelWithDebInfo'
             backend: 'THREADS'
+            stdcxx: '23'
     runs-on: ubuntu-latest
     container:
       image: ghcr.io/kokkos/ci-containers/${{ matrix.distro }}
@@ -72,7 +81,7 @@ jobs:
         uses: actions/checkout@v4
         with:
           repository: desul/desul
-          ref: 477da9c8f40f8db369c28dd3f93a67e376d8511b
+          ref: 779d0441a778c7088a36d38c4cbf8df3cfa182cc
           path: desul
       - name: Install desul
         working-directory: desul
@@ -114,8 +123,8 @@ jobs:
             -DKokkos_ENABLE_DEPRECATED_CODE_4=ON \
             -DKokkos_ENABLE_DEPRECATION_WARNINGS=OFF \
             -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
-            -DKokkos_ENABLE_IMPL_MDSPAN=ON \
             -DCMAKE_CXX_FLAGS="-Werror ${{ matrix.cxx_extra_flags }}" \
+            -DCMAKE_CXX_STANDARD="${{ matrix.stdcxx }}" \
             -DCMAKE_EXE_LINKER_FLAGS="${{ matrix.extra_linker_flags }}" \
             -DCMAKE_CXX_COMPILER=${{ matrix.cxx }} \
             -DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
@@ -129,7 +138,7 @@ jobs:
         working-directory: builddir
         run: ctest --output-on-failure
       - name: Test linking against build dir
-        if: ${{ matrix.cxx_extra_flags != '-fsanitize=address' }}
+        if: ${{ !contains(matrix.cxx_extra_flags, '-fsanitize=address') }}
         working-directory: example/build_cmake_installed
         run: |
           cmake -B builddir_buildtree -DCMAKE_CXX_COMPILER=${{ matrix.cxx }} -DKokkos_ROOT=../../builddir
@@ -140,7 +149,7 @@ jobs:
       - name: Install
         run: sudo cmake --build builddir --target install
       - name: Test install
-        if: ${{ matrix.cxx_extra_flags != '-fsanitize=address' }}
+        if: ${{ !contains(matrix.cxx_extra_flags, '-fsanitize=address') }}
         working-directory: example/build_cmake_installed
         run: |
           cmake -B builddir -DCMAKE_CXX_COMPILER=${{ matrix.cxx }}

diff --git a/.github/workflows/scorecard.yml b/.github/workflows/scorecard.yml
@@ -33,12 +33,12 @@ jobs:
 
     steps:
       - name: "Checkout code"
-        uses: actions/checkout@8e5e7e5ab8b370d6c329ec480221332ada57f0ab # v3.5.2
+        uses: actions/checkout@a5ac7e51b41094c92402da3b24376905380afc29 # v4.1.6
         with:
           persist-credentials: false
 
       - name: "Run analysis"
-        uses: ossf/scorecard-action@80e868c13c90f172d68d1f4501dee99e2479f7af # v2.1.3
+        uses: ossf/scorecard-action@62b2cac7ed8198b15735ed49ab1e5cf35480ba46 # v2.4.0
         with:
           results_file: results.sarif
           results_format: sarif
@@ -60,14 +60,14 @@ jobs:
       # Upload the results as artifacts (optional). Commenting out will disable uploads of run results in SARIF
       # format to the repository Actions tab.
       - name: "Upload artifact"
-        uses: actions/upload-artifact@0b7f8abb1508181956e8e162db84b466c27e18ce # v3.1.2
+        uses: actions/upload-artifact@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
         with:
           name: SARIF file
           path: results.sarif
           retention-days: 5
 
       # Upload the results to GitHub's code scanning dashboard.
       - name: "Upload SARIF results to code scanning"
-        uses: github/codeql-action/upload-sarif@83f0fe6c4988d98a455712a27f0255212bba9bd4 # v2.3.6
+        uses: github/codeql-action/upload-sarif@9fdb3e49720b44c48891d036bb502feb25684276 # v3.25.6
         with:
           sarif_file: results.sarif
diff --git a/.github/workflows/windows.yml b/.github/workflows/windows.yml
@@ -17,10 +17,10 @@ jobs:
     runs-on: windows-2022
 
     steps:
-    - uses: Jimver/cuda-toolkit@v0.2.14
+    - uses: Jimver/cuda-toolkit@v0.2.16
       id: cuda-toolkit
       with:
-        cuda: '12.1.0'
+        cuda: '12.4.1'
     - uses: actions/checkout@v4
     - name: configure
       shell: bash

diff --git a/.jenkins b/.jenkins
@@ -58,7 +58,7 @@ pipeline {
                               make -j8 && ctest --verbose'''
                     }
                 }
-                stage('CUDA-12.2-NVHPC') {
+                stage('CUDA-12.2-NVHPC-AS-HOST-COMPILER') {
                     agent {
                         dockerfile {
                             filename 'Dockerfile.nvhpc'
@@ -82,16 +82,14 @@ pipeline {
                                 -DCMAKE_BUILD_TYPE=RelWithDebInfo \
                                 -DCMAKE_CXX_COMPILER=nvc++ \
                                 -DCMAKE_CXX_STANDARD=17 \
-                                -DCMAKE_CXX_FLAGS="--diag_suppress=implicit_return_from_non_void_function,no_device_stack" \
+                                -DCMAKE_CXX_FLAGS="--diag_suppress=implicit_return_from_non_void_function" \
                                 -DKokkos_ARCH_NATIVE=ON \
                                 -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
                                 -DKokkos_ENABLE_DEPRECATED_CODE_4=OFF \
                                 -DKokkos_ENABLE_TESTS=ON \
                                 -DKokkos_ENABLE_CUDA=ON \
                                 -DKokkos_ENABLE_CUDA_LAMBDA=ON \
                                 -DKokkos_ENABLE_OPENMP=ON \
-                                -DKokkos_ENABLE_IMPL_MDSPAN=ON \
-                                -DKokkos_ENABLE_IMPL_NVHPC_AS_DEVICE_COMPILER=ON \
                               .. && \
                               make -j8 && ctest --verbose'''
                     }
@@ -113,6 +111,7 @@ pipeline {
                                 -DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
                                 -DCMAKE_CXX_COMPILER=clang++ \
                                 -DCMAKE_CXX_FLAGS="-fsycl-device-code-split=per_kernel -Wno-deprecated-declarations -Werror -Wno-gnu-zero-variadic-macro-arguments -Wno-unknown-cuda-version -Wno-sycl-target" \
+                                -DCMAKE_PREFIX_PATH="$ONE_DPL_DIR" \
                                 -DKOKKOS_IMPL_SYCL_DEVICE_GLOBAL_SUPPORTED=0 \
                                 -DKokkos_ARCH_NATIVE=ON \
                                 -DKokkos_ARCH_AMPERE80=ON \
@@ -166,6 +165,7 @@ pipeline {
                                 -DKokkos_ENABLE_BENCHMARKS=ON \
                                 -DKokkos_ENABLE_HIP=ON \
                                 -DKokkos_ENABLE_OPENMP=ON \
+                                -DKokkos_ENABLE_IMPL_MDSPAN=OFF \
                                 -DKokkos_ENABLE_HIP_MULTIPLE_KERNEL_INSTANTIATIONS=ON \
                               .. && \
                               make -j8 && ctest --verbose'''
@@ -181,7 +181,7 @@ pipeline {
                         dockerfile {
                             filename 'Dockerfile.hipcc'
                             dir 'scripts/docker'
-                            additionalBuildArgs '--build-arg BASE=rocm/dev-ubuntu-20.04:5.6-complete'
+                            additionalBuildArgs '--build-arg BASE=rocm/dev-ubuntu-22.04:5.6-complete'
                             label 'rocm-docker'
                             args '-v /tmp/ccache.kokkos:/tmp/ccache --device=/dev/kfd --device=/dev/dri --security-opt seccomp=unconfined --group-add video --env HIP_VISIBLE_DEVICES=$HIP_VISIBLE_DEVICES'
                         }
@@ -190,6 +190,7 @@ pipeline {
                         sh 'ccache --zero-stats'
                         sh '''rm -rf build && mkdir -p build && cd build && \
                               cmake \
+                                -DBUILD_SHARED_LIBS=ON \
                                 -DCMAKE_BUILD_TYPE=RelWithDebInfo \
                                 -DCMAKE_CXX_COMPILER=hipcc \
                                 -DCMAKE_CXX_FLAGS="-Werror -Wno-unused-command-line-argument" \
@@ -345,7 +346,7 @@ pipeline {
                         sh '''rm -rf build && mkdir -p build && cd build && \
                               ../gnu_generate_makefile.bash \
                                 --with-options=compiler_warnings \
-                                --cxxflags="-Werror" \
+                                --cxxflags="-Werror -Werror all-warnings -Xcudafe --diag_suppress=20208" \
                                 --cxxstandard=c++17 \
                                 --with-cuda \
                                 --with-cuda-options=enable_lambda \
@@ -365,7 +366,7 @@ pipeline {
                             filename 'Dockerfile.nvcc'
                             dir 'scripts/docker'
                             additionalBuildArgs '--build-arg BASE=nvidia/cuda:11.0.3-devel-ubuntu18.04 --build-arg ADDITIONAL_PACKAGES="g++-8 gfortran clang" --build-arg CMAKE_VERSION=3.17.3'
-                            label 'nvidia-docker'
+                            label 'nvidia-docker && (volta || ampere)'
                             args '-v /tmp/ccache.kokkos:/tmp/ccache --env NVIDIA_VISIBLE_DEVICES=$NVIDIA_VISIBLE_DEVICES'
                         }
                     }
@@ -396,7 +397,6 @@ pipeline {
                                 -DKokkos_ENABLE_CUDA_RELOCATABLE_DEVICE_CODE=ON \
                                 -DKokkos_ENABLE_DEPRECATED_CODE_4=ON \
                                 -DKokkos_ENABLE_DEPRECATION_WARNINGS=OFF \
-                                -DKokkos_ENABLE_IMPL_MDSPAN=ON \
                                 -DCMAKE_INSTALL_PREFIX=${PWD}/../install \
                               .. && \
                               make -j8 install && \
@@ -437,18 +437,19 @@ pipeline {
                             filename 'Dockerfile.nvcc'
                             dir 'scripts/docker'
                             additionalBuildArgs '--build-arg BASE=nvidia/cuda:11.6.2-devel-ubuntu20.04'
-                            label 'nvidia-docker'
+                            label 'nvidia-docker && (volta || ampere)'
                             args '-v /tmp/ccache.kokkos:/tmp/ccache --env NVIDIA_VISIBLE_DEVICES=$NVIDIA_VISIBLE_DEVICES'
                         }
                     }
                     steps {
                         sh 'ccache --zero-stats'
                         sh '''rm -rf build && mkdir -p build && cd build && \
                               cmake \
+                                -DBUILD_SHARED_LIBS=ON \
                                 -DCMAKE_BUILD_TYPE=Debug \
                                 -DCMAKE_CXX_COMPILER_LAUNCHER=ccache \
                                 -DCMAKE_CXX_COMPILER=$WORKSPACE/bin/nvcc_wrapper \
-                                -DCMAKE_CXX_FLAGS=-Werror \
+                                -DCMAKE_CXX_FLAGS="-Werror -Werror all-warnings -Xcudafe --diag_suppress=20208" \
                                 -DCMAKE_CXX_STANDARD=17 \
                                 -DKokkos_ARCH_NATIVE=ON \
                                 -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
@@ -460,7 +461,7 @@ pipeline {
                                 -DKokkos_ENABLE_CUDA=ON \
                                 -DKokkos_ENABLE_CUDA_LAMBDA=ON \
                                 -DKokkos_ENABLE_LIBDL=OFF \
-                                -DKokkos_ENABLE_IMPL_MDSPAN=ON \
+                                -DKokkos_ENABLE_IMPL_MDSPAN=OFF \
                                 -DKokkos_ENABLE_IMPL_CUDA_MALLOC_ASYNC=OFF \
                               .. && \
                               make -j8 && ctest --verbose && \
@@ -497,7 +498,7 @@ pipeline {
                                 -DKokkos_ARCH_NATIVE=ON \
                                 -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
                                 -DKokkos_ENABLE_DEPRECATED_CODE_4=ON \
-                                -DKokkos_ENABLE_DEPRECATION_WARNINGS=OFF \
+                                -DKokkos_ENABLE_DEPRECATION_WARNINGS=ON \
                                 -DKokkos_ENABLE_TESTS=ON \
                                 -DKokkos_ENABLE_BENCHMARKS=ON \
                                 -DKokkos_ENABLE_OPENMP=ON \

diff --git a/.jenkins_nightly b/.jenkins_nightly
@@ -70,38 +70,74 @@ pipeline {
                           '''
                     }      
                 }   
-                stage('GCC-13') {
+                stage('GCC-14') {
                     agent {
                         docker {
-                            image 'gcc:13.1'
+                            image 'gcc:14.1'
                             label 'docker'
                         }
                     }
                     steps {
                         sh '''
-                          DEBIAN_FRONTEND=noninteractive && \
-                          apt-get update && apt-get upgrade -y && apt-get install -y \
-                          cmake \
-                          && \
-                          apt-get clean && rm -rf /var/lib/apt/lists/*
+                          wget https://github.com/Kitware/CMake/releases/download/v3.30.0/cmake-3.30.0-linux-x86_64.sh && \
+                          chmod +x cmake-3.30.0-linux-x86_64.sh && ./cmake-3.30.0-linux-x86_64.sh --skip-license --prefix=/usr
 
                           mkdir -p build && cd build && \
                           cmake \
                             -DCMAKE_BUILD_TYPE=Release \
-                            -DCMAKE_CXX_STANDARD=23 \
+                            -DCMAKE_CXX_STANDARD=26 \
                             -DCMAKE_CXX_FLAGS=-Werror \
                             -DKokkos_ARCH_NATIVE=ON \
                             -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
                             -DKokkos_ENABLE_BENCHMARKS=ON \
                             -DKokkos_ENABLE_EXAMPLES=ON \
                             -DKokkos_ENABLE_TESTS=ON \
-                            -DKokkos_ENABLE_DEPRECATION_WARNINGS=OFF \
+                            -DKokkos_ENABLE_DEPRECATED_CODE_4=ON \
+                            -DKokkos_ENABLE_DEPRECATION_WARNINGS=ON \
                             -DKokkos_ENABLE_SERIAL=ON \
                           .. && \
                           make -j8 && ctest --verbose
                           '''
                     }
                 }
+                stage('HIP-ROCM-6.1') {
+                    agent {
+                        dockerfile {
+                            filename 'Dockerfile.hipcc'
+                            dir 'scripts/docker'
+                            additionalBuildArgs '--build-arg BASE=rocm/dev-ubuntu-22.04:6.1.2-complete'
+                            label 'rocm-docker && AMD_Radeon_Instinct_MI210'
+                            args '-v /tmp/ccache.kokkos:/tmp/ccache --device=/dev/kfd --device=/dev/dri --security-opt seccomp=unconfined --group-add video --env HIP_VISIBLE_DEVICES=$HIP_VISIBLE_DEVICES'
+                        }
+                    }
+                    environment {
+                        // FIXME Test returns a wrong value
+                        GTEST_FILTER = '-hip_hostpinned.view_allocation_large_rank'
+                    }
+                    steps {
+                        sh 'ccache --zero-stats'
+                        sh '''rm -rf build && mkdir -p build && cd build && \
+                              cmake \
+                                -DCMAKE_BUILD_TYPE=RelWithDebInfo \
+                                -DCMAKE_CXX_COMPILER=hipcc \
+                                -DCMAKE_CXX_FLAGS="-Werror -Wno-unused-command-line-argument" \
+                                -DCMAKE_CXX_STANDARD=20 \
+                                -DKokkos_ARCH_NATIVE=ON \
+                                -DKokkos_ENABLE_COMPILER_WARNINGS=ON \
+                                -DKokkos_ENABLE_DEPRECATED_CODE_4=ON \
+                                -DKokkos_ENABLE_DEPRECATION_WARNINGS=ON \
+                                -DKokkos_ENABLE_TESTS=ON \
+                                -DKokkos_ENABLE_BENCHMARKS=ON \
+                                -DKokkos_ENABLE_HIP=ON \
+                              .. && \
+                              make -j8 && ctest --verbose'''
+                    }
+                    post {
+                        always {
+                            sh 'ccache --show-stats'
+                        }
+                    }
+                }
             }
         }
     }

diff --git a/.olcf-gitlab-ci.yml b/.olcf-gitlab-ci.yml
@@ -0,0 +1,12 @@
+test:
+  stage: test
+  tags: [frontier, shell]
+  id_tokens:
+    OLCF_ID_TOKEN:
+      aud: https://code.olcf.ornl.gov
+  script:
+    - module load rocm/6.0
+    - cmake -B build -DCMAKE_CXX_COMPILER=hipcc -DKokkos_ENABLE_HIP=ON -DKokkos_ENABLE_TESTS=ON
+    - cmake --build build -j48
+    - cd build
+    - ctest -E Kokkos_CoreUnitTest_DeviceAndThreads -V