Python dynamic library path could not be determined

[JavierYepez]

I'm trying to build the python bindings for Deepstream in the official Docker image like this:

# bring in the micromamba image so we can copy files from it
FROM mambaorg/micromamba:1.5.6 as micromamba

FROM nvcr.io/nvidia/deepstream:6.4-triton-multiarch

# Copy Micromamba
ENV MAMBA_ROOT_PREFIX="/opt/conda"
ENV MAMBA_EXE="/bin/micromamba"

COPY --from=micromamba "$MAMBA_EXE" "$MAMBA_EXE"
COPY --from=micromamba /usr/local/bin/_activate_current_env.sh /usr/local/bin/_activate_current_env.sh
COPY --from=micromamba /usr/local/bin/_dockerfile_shell.sh /usr/local/bin/_dockerfile_shell.sh
COPY --from=micromamba /usr/local/bin/_entrypoint.sh /usr/local/bin/_entrypoint.sh
COPY --from=micromamba /usr/local/bin/_dockerfile_initialize_user_accounts.sh /usr/local/bin/_dockerfile_initialize_user_accounts.sh
COPY --from=micromamba /usr/local/bin/_dockerfile_setup_root_prefix.sh /usr/local/bin/_dockerfile_setup_root_prefix.sh

RUN /usr/local/bin/_dockerfile_setup_root_prefix.sh

ENTRYPOINT ["/usr/local/bin/_entrypoint.sh"]
SHELL ["/usr/local/bin/_dockerfile_shell.sh"]
CMD ["/bin/bash"]

# Copy and install env
COPY env.yml /tmp/env.yml
RUN micromamba install --yes -f /tmp/env.yml &&\
    micromamba clean --all --yes

# # Activate environment
SHELL ["micromamba", "run", "-n", "base", "/bin/bash", "-c"]

ENV PATH="/opt/conda/bin:$PATH"

# # Clean in a separate layer as calling conda still generates some __pycache__ files
RUN find  -path ./proc -prune -name '*.a' -type f -exec rm -f '{}' '+' && \
    rm -rf /opt/conda/conda-meta && \
    rm -rf /opt/conda/include && \
    find -path ./proc -prune -name '__pycache__' -type d -exec rm -rf '{}' '+'

ENV NVIDIA_DRIVER_CAPABILITIES $NVIDIA_DRIVER_CAPABILITIES,video

# RUN wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-keyring_1.0-1_all.deb \
#     && dpkg -i cuda-keyring_1.0-1_all.deb
# RUN /opt/nvidia/deepstream/deepstream/user_additional_install.sh
# RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/3bf863cc.pub
RUN apt-get update  && apt install -y python3-gi python3-dev python3-gst-1.0 \
    python-gi-dev git \
    python3 python3-pip python3.10-dev cmake g++ build-essential libglib2.0-dev \
    libglib2.0-dev-bin libgstreamer1.0-dev libtool m4 autoconf \
    automake libgirepository1.0-dev libcairo2-dev

RUN apt-get install -y apt-transport-https ca-certificates -y
RUN update-ca-certificates

RUN cd sources && \
    git clone https://github.com/NVIDIA-AI-IOT/deepstream_python_apps && \
    cd deepstream_python_apps && \
    git checkout v1.1.10 && \
    git submodule update --init

RUN cd sources/deepstream_python_apps/3rdparty/gstreamer/subprojects/gst-python/ && \
    meson build && \
    meson configure && \
    cd build && \
    ninja && \
    ninja install


...

But meson build fails with the following error: meson.build:48:6: ERROR: Problem encountered: Python dynamic library path could not be determined. Here is the output:

The Meson build system
Version: 1.2.1
Source dir: /opt/nvidia/deepstream/deepstream-6.4/sources/deepstream_python_apps/3rdparty/gstreamer/subprojects/gst-python
Build dir: /opt/nvidia/deepstream/deepstream-6.4/sources/deepstream_python_apps/3rdparty/gstreamer/subprojects/gst-python/build
Build type: native build
Project name: gst-python
Project version: 1.20.3
C compiler for the host machine: cc (gcc 11.4.0 "cc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0")
C linker for the host machine: cc ld.bfd 2.38
Host machine cpu family: aarch64
Host machine cpu: aarch64
Found pkg-config: /usr/bin/pkg-config (0.29.2)
Run-time dependency gstreamer-1.0 found: YES 1.20.3
Run-time dependency gstreamer-base-1.0 found: YES 1.20.3
Run-time dependency gmodule-2.0 found: YES 2.72.4
Run-time dependency pygobject-3.0 found: YES 3.42.1
Program python3 found: YES (/opt/conda/bin/python3)
Run-time dependency python found: YES 3.10
Message: pylib_loc = /opt/conda/lib

meson.build:48:6: ERROR: Problem encountered: Python dynamic library path could not be determined

A full log can be found at /opt/nvidia/deepstream/deepstream-6.4/sources/deepstream_python_apps/3rdparty/gstreamer/subprojects/gst-python/build/meson-logs/meson-log.txt
WARNING: Running the setup command as `meson [options]` instead of `meson setup [options]` is ambiguous and deprecated.

I can provide the full log if needed. Thank you

[eli-schwartz]

This error message comes from the gst-python codebase here:

https://github.com/GStreamer/gstreamer/blob/ccf22e315cedf81e0075ab179ffb1b733da5206e/subprojects/gst-python/meson.build#L39

It looks like they manually look up the location of the python shared library because they need to embed its path as a define to look up at runtime... interesting.

There's a build option to set its path, see line 35.

[JavierYepez]

Sorry, it's the first time I use meson. How could you set up an option when running meson build?