ICODE-ADC/lammps
4
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update GPU container definitions for CUDA 11.7 and singularity -> apptainer

Browse Source
This commit is contained in:
Axel Kohlmeyer
2022-06-08 08:19:54 -04:00
parent 42694ba640
commit 55fdb7f12a
4 changed files with 28 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
doc/src/Tools.rst
View File

@ -95,7 +95,7 @@ Miscellaneous tools
* :ref:`LAMMPS shell <lammps_shell>`
* :ref:`LAMMPS magic patterns for file(1) <magic>`
* :ref:`Offline build tool <offline>`
* :ref:`singularity <singularity_tool>`
* :ref:`singularity/apptainer <singularity_tool>`
* :ref:`SWIG interface <swig>`
* :ref:`vim <vim>`
@ -1007,14 +1007,15 @@ Ivanov, at University of Iceland (ali5 at hi.is).
.. _singularity_tool:
singularity tool
----------------------------------------
singularity/apptainer tool
--------------------------
The singularity sub-directory contains container definitions files
that can be used to build container images for building and testing
LAMMPS on specific OS variants using the `Singularity <https://sylabs.io>`_
container software. Contributions for additional variants are welcome.
For more details please see the README.md file in that folder.
The singularity sub-directory contains container definitions files that
can be used to build container images for building and testing LAMMPS on
specific OS variants using the `Apptainer <https://apptainer.org>`_ or
`Singularity <https://sylabs.io>`_ container software. Contributions for
additional variants are welcome. For more details please see the
README.md file in that folder.
----------

10
tools/singularity/README.md
View File

@ -1,17 +1,19 @@
# Singularity container definitions for compiling/testing LAMMPS
# Apptainer (aka Singularity) container definitions for compiling/testing LAMMPS
The *.def files in this folder can be used to build container images
for [Singularity](https://sylabs.io), suitable for compiling and testing
for [Apptainer](https://apptainer.org) (previously called
[Singularity](https://sylabs.io)), suitable for compiling and testing
LAMMPS on a variety of OS variants with support for most standard
packages and - for some of them - also building/spellchecking the manual
in all supported formats. This allows to test and debug LAMMPS code on
in all supported formats. This allows to test and debug LAMMPS code on
different OS variants without doing a full installation on your development
workstation, e.g. when bugs are reported that can only be reproduced on
a specific OS or with specific (mostly older) versions of tools,
compilers, or libraries.
Here is a workflow for testing a compilation of LAMMPS with a locally
built CentOS 7.x singularity container.
built CentOS 7.x Singularity container. For Apptainer replace the
`singularity` command with `apptainer`.
```
cd some/work/directory

14
tools/singularity/ubuntu18.04_gpu.def
View File

@ -2,11 +2,11 @@ BootStrap: docker
From: ubuntu:18.04
%environment
export PATH=/usr/lib/ccache:/usr/local/cuda-11.5/bin:${PATH}:/opt/rocm/bin:/opt/rocm/profiler/bin:/opt/rocm/opencl/bin/x86_64
export CUDADIR=/usr/local/cuda-11.5
export CUDA_PATH=/usr/local/cuda-11.5
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-11.5/lib64:/opt/rocm/lib:/opt/rocm-4.5.0/llvm/lib
export LIBRARY_PATH=/usr/local/cuda-11.5/lib64/stubs
export PATH=/usr/lib/ccache:/usr/local/cuda-11.7/bin:${PATH}:/opt/rocm/bin:/opt/rocm/profiler/bin:/opt/rocm/opencl/bin/x86_64
export CUDADIR=/usr/local/cuda-11.7
export CUDA_PATH=/usr/local/cuda-11.7
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-11.7/lib64:/opt/rocm/lib:/opt/rocm-4.5.0/llvm/lib
export LIBRARY_PATH=/usr/local/cuda-11.7/lib64/stubs
%post
export DEBIAN_FRONTEND=noninteractive
apt-get update
@ -122,11 +122,11 @@ From: ubuntu:18.04
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-ubuntu1804.pin
mv cuda-ubuntu1804.pin /etc/apt/preferences.d/cuda-repository-pin-600
apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub
apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/3bf863cc.pub
add-apt-repository "deb http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/ /"
apt-get update
export CUDA_PKG_VERSION=11.5
export CUDA_PKG_VERSION=11.7
apt-get install -y --no-install-recommends \
cuda-libraries-${CUDA_PKG_VERSION} \

12
tools/singularity/ubuntu20.04_gpu.def
View File

@ -2,11 +2,11 @@ BootStrap: docker
From: ubuntu:20.04
%environment
export PATH=/usr/lib/ccache:/usr/local/cuda-11.5/bin:${PATH}:/opt/rocm/bin:/opt/rocm/profiler/bin:/opt/rocm/opencl/bin/x86_64
export CUDADIR=/usr/local/cuda-11.5
export CUDA_PATH=/usr/local/cuda-11.5
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-11.5/lib64:/opt/rocm/lib:/opt/rocm-4.5.0/llvm/lib
export LIBRARY_PATH=/usr/local/cuda-11.5/lib64/stubs
export PATH=/usr/lib/ccache:/usr/local/cuda-11.7/bin:${PATH}:/opt/rocm/bin:/opt/rocm/profiler/bin:/opt/rocm/opencl/bin/x86_64
export CUDADIR=/usr/local/cuda-11.7
export CUDA_PATH=/usr/local/cuda-11.7
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-11.7/lib64:/opt/rocm/lib:/opt/rocm-4.5.0/llvm/lib
export LIBRARY_PATH=/usr/local/cuda-11.7/lib64/stubs
%post
export DEBIAN_FRONTEND=noninteractive
apt-get update
@ -109,7 +109,7 @@ From: ubuntu:20.04
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/7fa2af80.pub
apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/3bf863cc.pub
add-apt-repository "deb http://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/ /"
apt-get update