Merge branch 'master' into atc-doc-to-rst

This commit is contained in:
Axel Kohlmeyer
2020-03-19 15:55:08 -04:00
10 changed files with 108 additions and 33 deletions

View File

@ -151,9 +151,9 @@ one of them as a starting point and customize it to your needs.
cmake -C ../cmake/presets/nolib.cmake [OPTIONS] ../cmake # disable packages that do require extra libraries or tools
cmake -C ../cmake/presets/clang.cmake [OPTIONS] ../cmake # change settings to use the Clang compilers by default
cmake -C ../cmake/presets/intel.cmake [OPTIONS] ../cmake # change settings to use the Intel compilers by default
cmake -C ../cmake/presets/mingw.cmake [OPTIONS] ../cmake # enable packages compatible with MinGW compilers for Windows
cmake -C ../cmake/presets/all_on.cmake [OPTIONS] ../cmake # enable all packages
cmake -C ../cmake/presets/all_off.cmake [OPTIONS] ../cmake # disable all packages
mingw64-cmake -C ../cmake/presets/mingw-cross.cmake [OPTIONS] ../cmake # compile with MinGW cross compilers
.. note::

View File

@ -83,10 +83,16 @@ traditional build system, but CMake has also been successfully tested
using the mingw32-cmake and mingw64-cmake wrappers that are bundled
with the cross-compiler environment on Fedora machines. A CMake preset
selecting all packages compatible with this cross-compilation build
is provided. You will likely need to disable the GPU package unless you
download and install the contents of the pre-compiled `OpenCL ICD loader library <https://download.lammps.org/thirdparty/opencl-win-devel.tar.gz>`_
into your MinGW64 cross-compiler environment. The cross-compilation
currently will only produce non-MPI serial binaries.
is provided. The GPU package can only be compiled with OpenCL support
and you need to download and install the pre-compiled
`OpenCL ICD loader library <https://download.lammps.org/thirdparty/opencl-win-devel.tar.gz>`_
into your MinGW64 cross-compiler environment. With CMake this will be
done transparently. To compile with MPI support, a pre-compiled
library and the corresponding header files are required. There is
`one package for 32-bit Windows <https://download.lammps.org/thirdparty/mpich2-win32-devel.tar.gz>`_
and `a second package for 64-bit Windows <https://download.lammps.org/thirdparty/mpich2-win64-devel.tar.gz>`_.
When building with CMake, the matching package will be downloaded
automatically, but MPI support has to be explicitly enabled with ``-DBUILD_MPI=on``.
Please keep in mind, though, that this only applies to **compiling** LAMMPS.
Whether the resulting binaries do work correctly is not tested by the

View File

@ -140,6 +140,8 @@ accelerated styles exist.
* :doc:`coul/long/cs <pair_cs>` - long-range Coulombic potential and core/shell
* :doc:`coul/long/soft <pair_fep_soft>` - long-range Coulombic potential with a soft core
* :doc:`coul/msm <pair_coul>` - long-range MSM Coulombics
* :doc:`coul/slater/cut <pair_coul>` - smeared out Coulombics
* :doc:`coul/slater/long <pair_coul>` - long-range smeared out Coulombics
* :doc:`coul/shield <pair_coul_shield>` - Coulombics for boron nitride for use with :doc:`ilp/graphene/hbn <pair_ilp_graphene_hbn>` potential
* :doc:`coul/streitz <pair_coul>` - Coulombics via Streitz/Mintmire Slater orbitals
* :doc:`coul/wolf <pair_coul>` - Coulombics via Wolf potential