ci: up pangea-3 build

.github/workflows/docker_build_tpls.yml

@@ -61,10 +61,11 @@ jobs:
             DOCKER_REPOSITORY: geosx/rockylinux8-clang17-cuda12.5
             TPL_DOCKERFILE: docker/tpl-rockylinux-clang-cuda-12.Dockerfile
             RUNS_ON: Runner_4core_16GB
-          - name: Pangea 3 (AlmaLinux 8.8, gcc 9.4.0, open-mpi 4.1.2, cuda 11.5.0, openblas 0.3.10)
-            DOCKER_REPOSITORY: geosx/pangea3-almalinux8-gcc9.4-openmpi4.1.2-cuda11.5.0-openblas0.3.18
-            TPL_DOCKERFILE: docker/TotalEnergies/Pangea3.Dockerfile
-            DOCKER_ROOT_IMAGE: 7g8efcehpff/pangea-almalinux8-gcc9.4-openmpi4.1.2-cuda11.5.0-openblas0.3.18:4
+          - name: Pangea 3 (AlmaLinux 8.8, gcc 11.4.0, open-mpi 4.1.6, cuda 11.8.0, openblas 0.3.18)
+            DOCKER_REPOSITORY: geosx/pangea3-gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0
+            TPL_DOCKERFILE: docker/TotalEnergies/pangea-3/pangea3-tpl.Dockerfile
+            DOCKER_ROOT_IMAGE: onetechssc/pangea3:gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0_v1.0
+            INSTALL_DIR_ROOT: /workrd/SCR/NUM/GEOS_environment/p3/install/tpl
             HOST_ARCH: ppc64le
             RUNS_ON: streak2
           - name: Sherlock CPU (centos 7.9.2009, gcc 10.1.0, open-mpi 4.1.2, openblas 0.3.10, zlib 1.2.11)

docker/TotalEnergies/pangea-3/pangea3-gcc-openmpi-openblas.Dockerfile

@@ -0,0 +1,82 @@
+#######################################
+# Pangea 3 image : gcc - openmpi - openblas
+#######################################
+#
+# Installs :
+#   - openmpi  = 4.1.6
+#   - openblas = 0.3.18
+#   - cuda     = 11.8.0
+#
+#######################################
+#
+# Description :
+#   - the image is based on onetechssc/pangea3:gcc11.4.0_v1.0 built from pangea3-gcc.Dockerfile
+#   - this image is deployed as onetechssc/pangea3:gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0_v1.0
+#   - openblas is installed via spack (as done for Pangea 3)
+#   - openmpi and cuda are installed via spack (part of system image on Pangea 3)
+#
+# Usage :
+#   build the image:
+#   - podman build --format docker -t onetechssc/pangea3:gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0_v1.0 -f pangea3-gcc-openmpi-openblas.Dockerfile -v /etc/pki/ca-trust/extracted/pem/:/etc/pki/ca-trust/extracted/pem  .
+#   run the image:
+#   - podman run -it --detach --privileged --name pangea3_gcc_openmpi_openblas -v /etc/pki/ca-trust/extracted/pem/:/etc/pki/ca-trust/extracted/pem localhost/onetechssc/pangea3:gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0_v1.0
+#   - podman exec -it pangea3_gcc_openmpi_openblas /bin/bash
+#   push the image (requires to be part of the onetechssc docker organization):
+#   - podman login docker.io
+#   - podman push localhost/onetechssc/pangea3:gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0_v1.0 docker://docker.io/onetechssc/pangea3:gcc11.4.0-openmpi4.1.6-openblas0.3.18-cuda11.8.0_v1.0
+#
+#######################################
+
+# -------------------------------------
+# PANGEA3 - GCC-HPCX-MKL
+FROM docker.io/onetechssc/pangea3:gcc11.4.0_v1.0 AS pangea3
+# ------
+# LABELS
+LABEL description="Pangea 3 image : gcc - cmake - python - openmpi - openblas - cuda"
+LABEL version="1.0"
+LABEL maintainer="TotalEnergies HPC Team"
+# ------
+# ARGS
+ARG OPENBLAS_VERSION=0.3.18
+ARG OPENMPI_VERSION=4.1.6
+ARG CUDA_VERSION=11.8.0
+# ------
+# INSTALL
+# spack config (make sure you survive our wonderfull proxy)
+RUN spack config --scope defaults add config:connect_timeout:120
+RUN spack config --scope defaults add config:url_fetch_method:curl
+# openblas (with openmp)
+RUN spack install openblas@$OPENBLAS_VERSION threads=openmp %gcc@$GCC_VERSION
+# cuda
+RUN spack install cuda@$CUDA_VERSION %gcc@$GCC_VERSION
+# openmpi
+RUN spack install openmpi@$OPENMPI_VERSION +cuda cuda_arch=70 %gcc@$GCC_VERSION
+# numpy with openblas
+RUN spack install py-numpy %gcc@$GCC_VERSION ^python@$PYTHON_VERSION ^openblas@$OPENBLAS_VERSION
+# virtualenv (openssl error when using curl)
+RUN spack config --scope defaults add config:url_fetch_method:urllib &&\
+    spack install py-virtualenv %gcc@$GCC_VERSION ^python@$PYTHON_VERSION &&\
+    spack config --scope defaults add config:url_fetch_method:curl
+# sccache
+RUN dnf makecache --refresh && dnf -y install cargo openssl-devel  &&\
+    cargo install sccache --locked && mkdir -p /opt/sccache/ && cp -r /root/.cargo/bin /opt/sccache/  &&\
+    dnf remove -y cargo openssl-devel && dnf clean all
+# flex
+RUN spack install flex %gcc@$GCC_VERSION
+# zlib
+RUN spack install zlib %gcc@$GCC_VERSION
+# pugixml
+RUN spack install pugixml %gcc@$GCC_VERSION
+# ------
+# ENV
+# sccache
+ENV SCCACHE=/opt/sccache/bin/sccache
+# create env script
+RUN <<EOF cat > /root/.setup_env.sh
+. /opt/spack/share/spack/setup-env.sh
+spack load gcc@$GCC_VERSION python@$PYTHON_VERSION cmake@$CMAKE_VERSION \
+           openblas@$OPENBLAS_VERSION openmpi@$OPENMPI_VERSION cuda@$CUDA_VERSION \
+           py-numpy py-virtualenv \
+           flex zlib
+EOF
+RUN chmod +x /root/.setup_env.sh