Dockerfile.rhel9.2

ARG BASE_IMAGE
ARG BASE_TAG
FROM ${BASE_IMAGE}:${BASE_TAG} AS gaudi-base
ARG ARTIFACTORY_URL
ARG VERSION
ARG REVISION

LABEL vendor="Intel Corporation"
LABEL release="${VERSION}-${REVISION}"

ENV HOME="/opt/app-root/src"
WORKDIR /opt/app-root/src

RUN dnf install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-9.noarch.rpm && \
    dnf clean all && rm -rf /var/cache/yum

RUN echo "[BaseOS]" > /etc/yum.repos.d/CentOS-Linux-BaseOS.repo && \
    echo "name=CentOS Linux 9 - BaseOS" >> /etc/yum.repos.d/CentOS-Linux-BaseOS.repo && \
    echo "baseurl=https://mirror.stream.centos.org/9-stream/BaseOS/x86_64/os" >> /etc/yum.repos.d/CentOS-Linux-BaseOS.repo && \
    echo "gpgkey=https://www.centos.org/keys/RPM-GPG-KEY-CentOS-Official-SHA256" >> /etc/yum.repos.d/CentOS-Linux-BaseOS.repo && \
    echo "gpgcheck=1" >> /etc/yum.repos.d/CentOS-Linux-BaseOS.repo

RUN echo "[centos9]" > /etc/yum.repos.d/CentOS-Linux-AppStream.repo && \
    echo "name=CentOS Linux 9 - AppStream" >> /etc/yum.repos.d/CentOS-Linux-AppStream.repo && \
    echo "baseurl=https://mirror.stream.centos.org/9-stream/AppStream/x86_64/os" >> /etc/yum.repos.d/CentOS-Linux-AppStream.repo && \
    echo "gpgkey=https://www.centos.org/keys/RPM-GPG-KEY-CentOS-Official-SHA256" >> /etc/yum.repos.d/CentOS-Linux-AppStream.repo && \
    echo "gpgcheck=1" >> /etc/yum.repos.d/CentOS-Linux-AppStream.repo

RUN dnf install -y \
    clang \
    cmake3 \
    cpp \
    gcc \
    gcc-c++ \
    glibc \
    glibc-headers \
    glibc-devel \
    jemalloc \
    libarchive \
    libksba \
    unzip \
    llvm \
    lsof \
    python3-devel \
    openssh-clients \
    openssl \
    openssl-devel \
    libjpeg-devel \
    openssh-server \
    lsb_release \
    wget \
    git \
    libffi-devel \
    bzip2-devel \
    zlib-devel \
    mesa-libGL \
    iproute \
    python3-dnf-plugin-versionlock && \
    # update pkgs (except OS version) for resolving potentials CVEs
    dnf versionlock add redhat-release* && \
    dnf update -y && \
    dnf clean all && rm -rf /var/cache/yum

RUN mkdir -p /licenses && \
    wget -O /licenses/LICENSE https://raw.githubusercontent.com/intel/ai-containers/main/LICENSE

ENV PYTHON_VERSION=3.10
COPY install-python310.sh .
RUN ./install-python310.sh rhel9.2 && rm install-python310.sh
ENV LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

COPY install_efa.sh .
RUN ./install_efa.sh && rm install_efa.sh && rm -rf /etc/ld.so.conf.d/efa.conf /etc/profile.d/efa.sh

ENV LIBFABRIC_VERSION="1.20.0"
ENV LIBFABRIC_ROOT="/opt/habanalabs/libfabric-${LIBFABRIC_VERSION}"
ENV MPI_ROOT=/opt/amazon/openmpi
ENV LD_LIBRARY_PATH=$LIBFABRIC_ROOT/lib:${MPI_ROOT}/lib:/usr/lib/habanalabs:$LD_LIBRARY_PATH
ENV PATH=${LIBFABRIC_ROOT}/bin:${MPI_ROOT}/bin:$PATH
ENV OPAL_PREFIX=${MPI_ROOT}
ENV MPICC=${MPI_ROOT}/bin/mpicc
ENV RDMAV_FORK_SAFE=1
ENV FI_EFA_USE_DEVICE_RDMA=1

RUN echo "[habanalabs]" > /etc/yum.repos.d/habanalabs.repo && \
    echo "name=Habana RH9 Linux repo" >> /etc/yum.repos.d/habanalabs.repo && \
    echo "baseurl=https://${ARTIFACTORY_URL}/artifactory/rhel/9/9.2" >> /etc/yum.repos.d/habanalabs.repo && \
    echo "gpgkey=https://${ARTIFACTORY_URL}/artifactory/rhel/9/9.2/repodata/repomd.xml.key" >> /etc/yum.repos.d/habanalabs.repo && \
    echo "gpgcheck=1" >> /etc/yum.repos.d/habanalabs.repo

# for Habana GPG key with SHA-1 signature
RUN update-crypto-policies --set DEFAULT:SHA1

RUN dnf install -y habanalabs-rdma-core-"$VERSION"-"$REVISION".el9 \
    habanalabs-thunk-"$VERSION"-"$REVISION".el9 \
    habanalabs-firmware-tools-"$VERSION"-"$REVISION".el9 \
    habanalabs-graph-"$VERSION"-"$REVISION".el9 && \
    rm -f /etc/yum.repos.d/habanalabs.repo && rm -f /etc/yum.repos.d/habana.repo && rm -rf /tmp/* && \
    dnf clean all && rm -rf /var/cache/yum

RUN rpm -V habanalabs-rdma-core && rpm -V habanalabs-thunk && rpm -V habanalabs-firmware-tools && rpm -V habanalabs-graph

# There is no need to store pip installation files inside docker image
ENV PIP_NO_CACHE_DIR=on
ENV PIP_DISABLE_PIP_VERSION_CHECK=1
ENV RDMA_CORE_ROOT=/opt/habanalabs/rdma-core/src
ENV RDMA_CORE_LIB=${RDMA_CORE_ROOT}/build/lib

RUN wget -nv -O /tmp/libfabric-${LIBFABRIC_VERSION}.tar.bz2 https://github.com/ofiwg/libfabric/releases/download/v${LIBFABRIC_VERSION}/libfabric-${LIBFABRIC_VERSION}.tar.bz2 && \
    cd /tmp/ && tar xf /tmp/libfabric-${LIBFABRIC_VERSION}.tar.bz2 && \
    cd /tmp/libfabric-${LIBFABRIC_VERSION} && \
    ./configure --prefix=$LIBFABRIC_ROOT --enable-psm3-verbs --enable-verbs=yes --with-synapseai=/usr && \
    make && make install && cd / && rm -rf /tmp/libfabric-${LIBFABRIC_VERSION}.tar.bz2 /tmp/libfabric-${LIBFABRIC_VERSION}

RUN wget -nv -O /tmp/main.zip https://github.com/HabanaAI/hccl_ofi_wrapper/archive/refs/heads/main.zip && \
    unzip /tmp/main.zip -d /tmp && \
    cd /tmp/hccl_ofi_wrapper-main && \
    make && cp -f libhccl_ofi_wrapper.so /usr/lib/habanalabs/libhccl_ofi_wrapper.so && \
    cd / && \
    rm -rf /tmp/main.zip /tmp/hccl_ofi_wrapper-main

ENV GC_KERNEL_PATH=/usr/lib/habanalabs/libtpc_kernels.so
ENV HABANA_LOGS=/opt/app-root/log/habana_logs/
ENV HABANA_SCAL_BIN_PATH=/opt/habanalabs/engines_fw
ENV HABANA_PLUGINS_LIB_PATH=/opt/habanalabs/habana_plugins

ENV APP_ROOT="/opt/app-root"

RUN python3.10 -m pip install "pip>=23.3" "setuptools>=70.0.0" "wheel==0.38.4"

WORKDIR ${APP_ROOT}

RUN python3.10 -m venv ${APP_ROOT} && \
    wget -O ${APP_ROOT}/bin/fix-permissions \
        https://raw.githubusercontent.com/sclorg/s2i-python-container/master/3.9-minimal/root/usr/bin/fix-permissions && \
    chown -R 1001:0 ${APP_ROOT} && \
    chmod +x ${APP_ROOT}/bin/fix-permissions && \
    ${APP_ROOT}/bin/fix-permissions ${APP_ROOT} -P && \
    echo "unset BASH_ENV PROMPT_COMMAND ENV" >> ${APP_ROOT}/bin/activate

USER 1001

ENV BASH_ENV="${APP_ROOT}/bin/activate"
ENV ENV="${APP_ROOT}/bin/activate"
ENV PROMPT_COMMAND=". ${APP_ROOT}/bin/activate"

SHELL ["/bin/bash", "-c"]

RUN python -m pip install habana_media_loader=="${VERSION}"."${REVISION}"


FROM gaudi-base AS gaudi-pytorch

ARG PT_VERSION
ARG VERSION
ARG REVISION
ARG ARTIFACTORY_URL
ENV BASE_NAME=rhel9.2

LABEL name="PyTorch Installer"
LABEL summary="Habanalabs PyTorch installer layer for RHEL9.2"
LABEL description="Image with pre installed Habanalabs packages for PyTorch"

RUN echo "/usr/lib/habanalabs" > $(python -c "import sysconfig; print(sysconfig.get_path('platlib'))")/habanalabs-graph.pth

USER root

RUN echo "[CRB]" > /etc/yum.repos.d/CentOS-Linux-CRB.repo && \
    echo "name=CentOS Linux 9 - CRB" >> /etc/yum.repos.d/CentOS-Linux-CRB.repo && \
    echo "baseurl=https://mirror.stream.centos.org/9-stream/CRB/x86_64/os" >> /etc/yum.repos.d/CentOS-Linux-CRB.repo && \
    echo "gpgkey=https://www.centos.org/keys/RPM-GPG-KEY-CentOS-Official-SHA256" >> /etc/yum.repos.d/CentOS-Linux-CRB.repo && \
    echo "gpgcheck=1" >> /etc/yum.repos.d/CentOS-Linux-CRB.repo

RUN dnf install --allowerasing -y \
    curl \
    cairo-devel \
    numactl-devel \
    iproute \
    which \
    zlib-devel \
    lapack-devel \
    openblas-devel \
    numactl \
    gperftools-devel && \
    dnf clean all && rm -rf /var/cache/yum

RUN dnf config-manager --add-repo https://yum.repos.intel.com/mkl/setup/intel-mkl.repo -y && \
    dnf install --allowerasing -y intel-mkl-64bit-2020.4-912 && \
    dnf clean all && rm -rf /var/cache/yum

# Set LD_PRELOAD after all required installations to
# avoid warnings during docker creation
ENV LD_PRELOAD=/lib64/libtcmalloc.so.4
ENV TCMALLOC_LARGE_ALLOC_REPORT_THRESHOLD=7516192768

RUN rm -rf /tmp/*

USER 1001

COPY --chown=1001:0 install_packages.sh .
RUN ./install_packages.sh && rm -f install_packages.sh

USER root

RUN /sbin/ldconfig && echo "source /etc/profile.d/habanalabs.sh" >> ~/.bashrc && \
    chown 1001:0 ~/.bashrc

USER 1001

FROM gaudi-pytorch AS gaudi-notebooks

WORKDIR ${APP_ROOT}/src

COPY --chown=1001:0 requirements.txt requirements.txt
COPY --chown=1001:0 start-notebook.sh /opt/app-root/bin
COPY --chown=1001:0 builder /opt/app-root/builder
COPY --chown=1001:0 utils /opt/app-root/bin/utils

USER 1001

RUN python -m pip install -r requirements.txt && \
    chmod -R g+w ${APP_ROOT}/lib/python3.10/site-packages && \
    fix-permissions ${APP_ROOT} -P && \
    chmod -R g+w /opt/app-root/src && \
    sed -i -e "s/Python.*/$(python --version | cut -d '.' -f-2)\",/" /opt/app-root/share/jupyter/kernels/python3/kernel.json && \
    jupyter labextension disable "@jupyterlab/apputils-extension:announcements"

RUN cd ${APP_ROOT}/ && \
    git clone https://github.com/HabanaAI/vllm-fork.git && \
    cd vllm-fork && \
    VLLM_TARGET_DEVICE=hpu pip install -e .

WORKDIR ${APP_ROOT}/src
ENV NOTEBOOK_SAMPLES_LINK="https://raw.githubusercontent.com/intel/ai-containers/refs/heads/main/enterprise/redhat/openshift-ai/gaudi/demo/oneapi-sample.ipynb"

ENTRYPOINT ["bash", "-c", "/opt/app-root/builder/run"]