Dockerfile.jp6

#
# build cuDF from source for JetPack 6 (see config.py for package configuration)
#
ARG BASE_IMAGE
FROM ${BASE_IMAGE}

ARG CUDF_REPO
ARG CUDF_VERSION
ARG CUDF_CMAKE_CUDA_ARCHITECTURES

ARG INSTALL_PREFIX=/usr/local
ARG BUILD_DIR=/opt/rapids

WORKDIR ${BUILD_DIR}

#
# cudf bundles many of it's dependencies, but some are still needed 
# libssl for cudf, boost and liblz4 for ORC extensions
#
RUN apt-get update && \    
    apt-get install -y --no-install-recommends \
		  libssl-dev \
		  libboost-system-dev \
		  libboost-filesystem-dev \
		  liblz4-dev \
    && rm -rf /var/lib/apt/lists/* \
    && apt-get clean
    
# set Python3 as default
RUN update-alternatives --install /usr/bin/python python /usr/bin/python3 1


# 
# build libcudf (C++)
#
ADD https://api.github.com/repos/${CUDF_REPO}/git/refs/heads/${CUDF_VERSION} /tmp/cudf_version.json

RUN git clone --branch ${CUDF_VERSION} --depth=1 --recursive https://github.com/${CUDF_REPO} cudf && \ 
    cd cudf && \ 
    ./build.sh libcudf -v


#
# build rmm-python
#
RUN pip3 install --no-cache-dir --verbose scikit-build ninja
RUN pip3 install --no-cache-dir --verbose 'Cython>3'

RUN cd cudf/cpp/build/_deps/rmm-src/python && \ 
    python3 setup.py bdist_wheel --verbose && \ 
    cp dist/rmm*.whl /opt && \ 
    pip3 install --no-cache-dir --verbose /opt/rmm*.whl

RUN pip3 show rmm && python3 -c 'import rmm; print(rmm.__version__)'


#
# build cudf (python)
#
RUN cp -r cudf/cpp/build/_deps/dlpack-src/include/dlpack /usr/local/include && \
    ls /usr/local/include/dlpack 

RUN cd cudf/python/cudf && \
    sed -i 's|"cubinlinker",||' pyproject.toml && \
    sed -i 's|"ptxcompiler",||' pyproject.toml && \
    sed -i 's|"cuda-python.*",||' pyproject.toml && \
    sed -i 's|"cupy.*",||' pyproject.toml && \
    cat pyproject.toml
    
RUN cd cudf/python/cudf && \
    SKBUILD_CONFIGURE_OPTIONS="-DCMAKE_PREFIX_PATH=${INSTALL_PREFIX} -DCMAKE_LIBRARY_PATH=${BUILD_DIR}/cudf/cpp/build -DCMAKE_CUDA_ARCHITECTURES=${CUDF_CMAKE_CUDA_ARCHITECTURES} -DFIND_CUDF_CPP=ON" \
    SKBUILD_BUILD_OPTIONS="-j$(nproc)" \
    python3 setup.py --verbose bdist_wheel && \
    cp dist/cudf*.whl /opt && \
    pip3 install --no-cache-dir --verbose /opt/cudf*.whl 


#
# build dask_cudf
#
RUN cd cudf/python/dask_cudf && \
    sed -i 's|"cupy.*",||' pyproject.toml && \
    python3 setup.py --verbose bdist_wheel && \
    cp dist/dask_cudf*.whl /opt && \
    pip3 install --no-cache-dir --verbose /opt/dask_cudf*.whl 

# a different version of cuda-python gets installed, so restore the one that we built
RUN pip3 install --no-cache-dir /opt/cuda_python*.whl

# cudf/utils/metadata/orc_column_statistics_pb2.py - your generated code is out of date and must be regenerated with protoc >= 3.19.0
RUN pip3 install --no-cache-dir --verbose 'protobuf<3.20'

# requests package needed for test_csv.py
RUN pip3 install --no-cache-dir --verbose requests

# make sure that the cudf modules load okay
RUN pip3 show cudf && python3 -c 'import cudf; print(cudf.__version__)'
RUN pip3 show dask_cudf && python3 -c 'import dask_cudf; print(dask_cudf.__version__)'  

WORKDIR /