Merge pull request #603 from Xiangyu-Hu/sycl_dockerfile

add docker file for sycl develop
Xiangyu-Hu · Jun 24, 2024 · 64956ed · 64956ed
2 parents 9861695 + cd2f342
commit 64956ed
Showing 1 changed file with 69 additions and 0 deletions.
diff --git a/dev.sycl.Dockerfile b/dev.sycl.Dockerfile
@@ -0,0 +1,69 @@
+FROM nvidia/cuda:12.5.0-devel-ubuntu20.04
+
+ARG build_with_dependencies_source=0
+ARG SPH_ONLY_STATIC_BUILD=0
+ARG was_build=0
+ARG build_with_visualization=off
+
+ENV TZ=Europe/Berlin
+RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
+
+RUN apt-get update && apt-get upgrade -y && apt-get install -y \ 
+    apt-utils \
+    build-essential \
+    cmake \
+    googletest \
+    pkg-config \
+    wget \
+    && apt-get clean \
+    && rm -rf /var/lib/apt/lists/*
+
+RUN wget https://registrationcenter-download.intel.com/akdlm/IRC_NAS/20f4e6a1-6b0b-4752-b8c1-e5eacba10e01/l_BaseKit_p_2024.0.0.49564_offline.sh
+# Install Intel oneAPI
+RUN sh ./l_BaseKit_p_2024.0.0.49564_offline.sh -a --silent --eula accept
+# RUN source /opt/intel/oneapi/setvars.sh --include-intel-llvm
+
+# Install the oneAPI Plugin for NVIDIA GPU
+# https://developer.codeplay.com/settings/api/download/
+# A new token and download URL needs to be generated. TODO: can we automate this completely?
+# Remember to add version=2024.0.0
+RUN wget --content-disposition 'https://developer.codeplay.com/api/v1/products/download?product=oneapi&variant=nvidia&version=2024.0.0&filters[]=linux&aat=MmZjNDU1MmNlMTVlZDE1Mzk5MGY3NTM2ZN7wAABGXejelYo4i2h1NCo3IgJJ-pf47QH1cLClaINraNLZAkfcpVTVYalyktXkInzpHrwQSpie1y23pwPkP-F5aV47'
+RUN sh oneapi-for-nvidia-gpus-2024.0.0-cuda-12.0-linux.sh -y
+
+RUN apt-get update && apt-get install -y ccache python3-dev gfortran git curl zip unzip tar \
+    autoconf automake autoconf-archive
+
+RUN cd $HOME && git clone https://www.github.com/microsoft/vcpkg.git
+RUN cd $HOME && cd vcpkg && \
+./bootstrap-vcpkg.sh
+RUN cd $HOME && cd vcpkg && ./vcpkg install --clean-after-build \
+    eigen3                          \
+    tbb                             \
+    boost-program-options           \
+    boost-geometry                  \
+    gtest                           \
+    pybind11                        \
+    simbody
+
+COPY ./ /home/SPHinXsys/
+WORKDIR /home/SPHinXsys
+RUN rm -rf build
+
+ENV TBB_HOME=/usr/lib/x86_64-linux-gnu
+ENV BOOST_HOME=/usr/lib/x86_64-linux-gnu
+ENV SIMBODY_HOME=/home/simbody
+ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$SIMBODY_HOME/lib
+ENV CPLUS_INCLUDE_PATH=$CPLUS_INCLUDE_PATH:$SIMBODY_HOME/include
+
+SHELL ["/bin/bash", "-c"]
+RUN source /opt/intel/oneapi/setvars.sh --include-intel-llvm && cmake   -G "Unix Makefiles"                                                     \
+    -D CMAKE_BUILD_TYPE=Release                                                 \
+    -D CMAKE_C_COMPILER=icx -D CMAKE_CXX_COMPILER=icpx                          \
+    -D CMAKE_TOOLCHAIN_FILE="$HOME/vcpkg/scripts/buildsystems/vcpkg.cmake"      \
+    -D CMAKE_C_COMPILER_LAUNCHER=ccache -D CMAKE_CXX_COMPILER_LAUNCHER=ccache   \
+    -D SPHINXSYS_USE_SYCL=ON                                                    \
+    -D SPHINXSYS_SYCL_TARGETS=nvptx64-nvidia-cuda                               \
+    -S .                                                                        \
+    -B ./build
+RUN cmake --build build/ --target test_2d_dambreak_sycl
+RUN mkdir build && cd build && cmake .. -DWASM_BUILD=${was_build} -DBUILD_WITH_DEPENDENCIES_SOURCE=${build_with_dependencies_source} -DSTATIC_BUILD=${SPH_ONLY_STATIC_BUILD} && make -j$(nproc)