tx258/lammps-gran-kokkos
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'develop' of https://www.github.com/lammps/lammps into kmc

Browse Source
This commit is contained in:
Jacob Tavenner
2024-12-19 11:29:52 -07:00
parent 82569f4448 aeb2190582
commit acb107af5e
39 changed files with 1810 additions and 6077 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
doc/doxygen/Doxyfile.in
View File

@ -2,7 +2,7 @@
DOXYFILE_ENCODING = UTF-8
PROJECT_NAME = "LAMMPS Programmer's Guide"
PROJECT_NUMBER = "4 May 2022"
PROJECT_NUMBER = "19 November 2024"
PROJECT_BRIEF = "Documentation of the LAMMPS library interface and Python wrapper"
PROJECT_LOGO = lammps-logo.png
CREATE_SUBDIRS = NO

4
doc/src/Intro_authors.rst
View File

@ -8,6 +8,8 @@ send an email to all of them at this address: "developers at
lammps.org". General questions about LAMMPS should be posted in the
`LAMMPS forum on MatSci <https://matsci.org/lammps/>`_.
.. We need to keep this file in sync with https://www.lammps.org/authors.html
.. raw:: latex
\small
@ -27,7 +29,7 @@ lammps.org". General questions about LAMMPS should be posted in the
* - `Steve Plimpton <sjp_>`_
- SNL (retired)
- sjplimp at gmail.com
- MD kernels, parallel algorithms & scalability, code structure and design
- original author, MD kernels, parallel algorithms & scalability, code structure and design
* - `Aidan Thompson <at_>`_
- SNL
- athomps at sandia.gov

28
doc/src/Packages_details.rst
View File

@ -2789,14 +2789,15 @@ implements smoothed particle hydrodynamics (SPH) for liquids. See the
related :ref:`MACHDYN package <PKG-MACHDYN>` package for smooth Mach dynamics
(SMD) for solids.
This package contains ideal gas, Lennard-Jones equation of states,
Tait, and full support for complete (i.e. internal-energy dependent)
equations of state. It allows for plain or Monaghans XSPH integration
of the equations of motion. It has options for density continuity or
density summation to propagate the density field. It has
:doc:`set <set>` command options to set the internal energy and density
of particles from the input script and allows the same quantities to
be output with thermodynamic output or to dump files via the :doc:`compute property/atom <compute_property_atom>` command.
This package contains ideal gas, Lennard-Jones equation of states, Tait,
and full support for complete (i.e. internal-energy dependent) equations
of state. It allows for plain or Monaghans XSPH integration of the
equations of motion. It has options for density continuity or density
summation to propagate the density field. It has :doc:`set <set>`
command options to set the internal energy and density of particles from
the input script and allows the same quantities to be output with
thermodynamic output or to dump files via the :doc:`compute
property/atom <compute_property_atom>` command.
**Author:** Georg Ganzenmuller (Fraunhofer-Institute for High-Speed
Dynamics, Ernst Mach Institute, Germany).
@ -2809,6 +2810,17 @@ Dynamics, Ernst Mach Institute, Germany).
* ``examples/PACKAGES/sph``
* https://www.lammps.org/movies.html#sph
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
.. note::
Please also note, that the :ref:`RHEO package <PKG-RHEO>` offers
similar functionality in a more modern and flexible implementation.
----------
.. _PKG-SPIN:

2
doc/src/bond_bpm_rotational.rst
View File

@ -155,7 +155,7 @@ page on BPMs.
If the *break* keyword is set to *no*, LAMMPS assumes bonds should not break
during a simulation run. This will prevent some unnecessary calculation.
The recommended bond communication distance no longer depends on bond failure
coefficients (which are ignored) but instead corresponds to the typical heurestic
coefficients (which are ignored) but instead corresponds to the typical heuristic
maximum strain used by typical non-bpm bond styles. Similar behavior to *break no*
can also be attained by setting arbitrarily high values for all four failure
coefficients. One cannot use *break no* with *smooth yes*.

2
doc/src/bond_bpm_spring.rst
View File

@ -119,7 +119,7 @@ If the *break* keyword is set to *no*, LAMMPS assumes bonds should not break
during a simulation run. This will prevent some unnecessary calculation.
The recommended bond communication distance no longer depends on the value of
:math:`\epsilon_c` (which is ignored) but instead corresponds to the typical
heurestic maximum strain used by typical non-bpm bond styles. Similar behavior
heuristic maximum strain used by typical non-bpm bond styles. Similar behavior
to *break no* can also be attained by setting an arbitrarily high value of
:math:`\epsilon_c`. One cannot use *break no* with *smooth yes*.

6
doc/src/compute_sph_e_atom.rst
View File

@ -33,6 +33,12 @@ particle.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The value of the internal energy will be 0.0 for atoms not in the
specified compute group.

6
doc/src/compute_sph_rho_atom.rst
View File

@ -32,6 +32,12 @@ kernel function interpolation using "pair style sph/rhosum".
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The value of the SPH density will be 0.0 for atoms not in the
specified compute group.

6
doc/src/compute_sph_t_atom.rst
View File

@ -37,6 +37,12 @@ particles, i.e. a Smooth-Particle Hydrodynamics particle.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The value of the internal energy will be 0.0 for atoms not in the
specified compute group.

10
doc/src/fix_rheo.rst
View File

@ -64,7 +64,7 @@ Description
Perform time integration for RHEO particles, updating positions, velocities,
and densities. For a detailed breakdown of the integration timestep and
numerical details, see :ref:`(Palermo) <rheo_palermo>`. For an overview
numerical details, see :ref:`(Palermo) <fix_rheo_palermo>`. For an overview
and list of other features available in the RHEO package, see
:doc:`the RHEO howto <Howto_rheo>`.
@ -101,7 +101,7 @@ A modified form of Fickian particle shifting can be enabled with the
more uniform spatial distribution. By default, shifting does not consider the
type of a particle and therefore may be inappropriate in systems consisting
of multiple atom types representing multiple fluid phases. However, two
optional subarguments can follow the *shift* keyword, *exclude/type* and
optional sub-arguments can follow the *shift* keyword, *exclude/type* and
*scale/cross/type* to adjust shifting at fluid interfaces.
The *exclude/type* option lets the user specify a list of atom types which
@ -155,7 +155,7 @@ threshold for this classification is set by the numerical value of
By default, RHEO integrates particles' densities using a mass diffusion
equation. Alternatively, one can update densities every timestep by performing
a kernel summation of the masses of neighboring particles by specifying the *rho/sum*
keyword. Following this keyword, one may include the optional *self/mass* subargument
keyword. Following this keyword, one may include the optional *self/mass* sub-argument
which modifies the behavior of the density summation. Typically, the density
:math:`\rho` of a particle is calculated as the sum over neighbors
@ -218,11 +218,11 @@ Default
----------
.. _rheo_palermo:
.. _fix_rheo_palermo:
**(Palermo)** Palermo, Wolf, Clemmer, O'Connor, Phys. Fluids, 36, 113337 (2024).
.. _rheo_yang:
.. _fix_rheo_yang:
**(Yang)** Yang, Rakhsha, Hu, Negrut, J. Comp. Physics, 458, 111079 (2022).

6
doc/src/fix_sph.rst
View File

@ -32,6 +32,12 @@ Hydrodynamics.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
Restart, fix_modify, output, run start/stop, minimize info
"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""

6
doc/src/fix_sph_stationary.rst
View File

@ -32,6 +32,12 @@ space. SPH stands for Smoothed Particle Hydrodynamics.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
Restart, fix_modify, output, run start/stop, minimize info
"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""

6
doc/src/pair_sph_heatconduction.rst
View File

@ -30,6 +30,12 @@ The transport model is the diffusion equation for the internal energy.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The following coefficients must be defined for each pair of atoms
types via the :doc:`pair_coeff <pair_coeff>` command as in the examples
above.

6
doc/src/pair_sph_idealgas.rst
View File

@ -36,6 +36,12 @@ particles from interpenetrating :ref:`(Monaghan) <ideal-Monoghan>`.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The following coefficients must be defined for each pair of atoms
types via the :doc:`pair_coeff <pair_coeff>` command as in the examples
above.

6
doc/src/pair_sph_lj.rst
View File

@ -34,6 +34,12 @@ interpenetrating :ref:`(Monaghan) <Monoghan>`.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The following coefficients must be defined for each pair of atoms
types via the :doc:`pair_coeff <pair_coeff>` command as in the examples
above.

6
doc/src/pair_sph_rhosum.rst
View File

@ -29,6 +29,12 @@ SPH particles by kernel function interpolation, every Nstep timesteps.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The following coefficients must be defined for each pair of atoms
types via the :doc:`pair_coeff <pair_coeff>` command as in the examples
above.

6
doc/src/pair_sph_taitwater.rst
View File

@ -41,6 +41,12 @@ prevent particles from interpenetrating :ref:`(Monaghan) <Monaghan>`.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The following coefficients must be defined for each pair of atoms
types via the :doc:`pair_coeff <pair_coeff>` command as in the examples
above.

6
doc/src/pair_sph_taitwater_morris.rst
View File

@ -37,6 +37,12 @@ This pair style also computes laminar viscosity :ref:`(Morris) <Morris>`.
See `this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in
LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
The following coefficients must be defined for each pair of atoms
types via the :doc:`pair_coeff <pair_coeff>` command as in the examples
above.

6
doc/src/set.rst
View File

@ -516,6 +516,12 @@ Keywords *sph/e*, *sph/cv*, and *sph/rho* set the energy, heat capacity,
and density of smoothed particle hydrodynamics (SPH) particles. See
`this PDF guide <PDF/SPH_LAMMPS_userguide.pdf>`_ to using SPH in LAMMPS.
.. note::
Please note that the SPH PDF guide file has not been updated for
many years and thus does not reflect the current *syntax* of the
SPH package commands. For that please refer to the LAMMPS manual.
Keyword *smd/mass/density* sets the mass of all selected particles, but
it is only applicable to the Smooth Mach Dynamics package MACHDYN. It
assumes that the particle volume has already been correctly set and

2
doc/utils/sphinx-config/false_positives.txt
View File

@ -2499,6 +2499,7 @@ neel
Neel
Neelov
Negre
Negrut
nelem
Nelement
Nelements
@ -3116,6 +3117,7 @@ Rafferty
rahman
Rahman
Rajamanickam
Rakhsha
Ralf
Raman
ramped

2
examples/comb/README
View File

@ -20,5 +20,3 @@ Examples:
4. in.comb.Cu2O.elastic: Cu2O crystal, qeq on, minimizes, then calculates
elastic constants
5. in.comb.HfO2: HfO2 polymorphs: Monoclinic HfO2 NVT @ 300K
6. in.comb.CuaS: Metallic Cu and amorphous silica interface, qeq on,
five step NVE run

4147
examples/comb/data.CuaS
View File

File diff suppressed because it is too large Load Diff

2
src/COLVARS/colvarproxy_lammps.cpp
View File

@ -204,7 +204,7 @@ cvm::rvector colvarproxy_lammps::position_distance(cvm::atom_pos const &pos1,
double xtmp = pos2.x - pos1.x;
double ytmp = pos2.y - pos1.y;
double ztmp = pos2.z - pos1.z;
_lmp->domain->minimum_image(xtmp,ytmp,ztmp);
_lmp->domain->minimum_image_big(xtmp,ytmp,ztmp);
return {xtmp, ytmp, ztmp};
}

173
src/KOKKOS/pair_snap_kokkos.h
View File

@ -30,29 +30,34 @@ PairStyle(snap/kk/host,PairSNAPKokkosDevice<LMPHostType>);
#include "pair_snap.h"
#include "kokkos_type.h"
#include "neigh_list_kokkos.h"
#include "sna_kokkos.h"
#include "pair_kokkos.h"
namespace LAMMPS_NS {
// pre-declare so sna_kokkos.h can refer to it
template<class DeviceType, typename real_type_, int vector_length_> class PairSNAPKokkos;
};
#include "sna_kokkos.h"
namespace LAMMPS_NS {
// Routines for both the CPU and GPU backend
struct TagPairSNAPPreUi{};
struct TagPairSNAPTransformUi{}; // re-order ulisttot from SoA to AoSoA, zero ylist
template <bool chemsnap> struct TagPairSNAPComputeZi{};
template <bool chemsnap> struct TagPairSNAPComputeBi{};
struct TagPairSNAPComputeBetaLinear{};
struct TagPairSNAPComputeBetaQuadratic{};
template <bool chemsnap> struct TagPairSNAPComputeYi{};
template <bool chemsnap> struct TagPairSNAPComputeYiWithZlist{};
template<int NEIGHFLAG, int EVFLAG>
struct TagPairSNAPComputeForce{};
// GPU backend only
struct TagPairSNAPComputeNeigh{};
struct TagPairSNAPComputeCayleyKlein{};
struct TagPairSNAPPreUi{};
struct TagPairSNAPComputeUiSmall{}; // more parallelism, more divergence
struct TagPairSNAPComputeUiLarge{}; // less parallelism, no divergence
struct TagPairSNAPTransformUi{}; // re-order ulisttot from SoA to AoSoA, zero ylist
struct TagPairSNAPComputeZi{};
struct TagPairSNAPBeta{};
struct TagPairSNAPComputeBi{};
struct TagPairSNAPTransformBi{}; // re-order blist from AoSoA to AoS
struct TagPairSNAPComputeYi{};
struct TagPairSNAPComputeYiWithZlist{};
template<int dir>
struct TagPairSNAPComputeFusedDeidrjSmall{}; // more parallelism, more divergence
template<int dir>
@ -60,14 +65,7 @@ struct TagPairSNAPComputeFusedDeidrjLarge{}; // less parallelism, no divergence
// CPU backend only
struct TagPairSNAPComputeNeighCPU{};
struct TagPairSNAPPreUiCPU{};
struct TagPairSNAPComputeUiCPU{};
struct TagPairSNAPTransformUiCPU{};
struct TagPairSNAPComputeZiCPU{};
struct TagPairSNAPBetaCPU{};
struct TagPairSNAPComputeBiCPU{};
struct TagPairSNAPZeroYiCPU{};
struct TagPairSNAPComputeYiCPU{};
struct TagPairSNAPComputeDuidrjCPU{};
struct TagPairSNAPComputeDeidrjCPU{};
@ -80,6 +78,8 @@ class PairSNAPKokkos : public PairSNAP {
typedef ArrayTypes<DeviceType> AT;
typedef EV_FLOAT value_type;
static constexpr LAMMPS_NS::ExecutionSpace execution_space = ExecutionSpaceFromDevice<DeviceType>::space;
static constexpr int host_flag = (execution_space == LAMMPS_NS::Host);
static constexpr int vector_length = vector_length_;
using real_type = real_type_;
using complex = SNAComplex<real_type>;
@ -93,9 +93,11 @@ class PairSNAPKokkos : public PairSNAP {
static constexpr int team_size_compute_ui = 2;
static constexpr int tile_size_transform_ui = 2;
static constexpr int tile_size_compute_zi = 2;
static constexpr int min_blocks_compute_zi = 0; // no minimum bound
static constexpr int tile_size_compute_bi = 2;
static constexpr int tile_size_transform_bi = 2;
static constexpr int tile_size_compute_beta = 2;
static constexpr int tile_size_compute_yi = 2;
static constexpr int min_blocks_compute_yi = 0; // no minimum bound
static constexpr int team_size_compute_fused_deidrj = 2;
#elif defined(KOKKOS_ENABLE_SYCL)
static constexpr int team_size_compute_neigh = 4;
@ -104,9 +106,11 @@ class PairSNAPKokkos : public PairSNAP {
static constexpr int team_size_compute_ui = 8;
static constexpr int tile_size_transform_ui = 8;
static constexpr int tile_size_compute_zi = 4;
static constexpr int min_blocks_compute_zi = 0; // no minimum bound
static constexpr int tile_size_compute_bi = 4;
static constexpr int tile_size_transform_bi = 4;
static constexpr int tile_size_compute_beta = 8;
static constexpr int tile_size_compute_yi = 8;
static constexpr int min_blocks_compute_yi = 0; // no minimum bound
static constexpr int team_size_compute_fused_deidrj = 4;
#else
static constexpr int team_size_compute_neigh = 4;
@ -116,17 +120,21 @@ class PairSNAPKokkos : public PairSNAP {
static constexpr int tile_size_transform_ui = 4;
static constexpr int tile_size_compute_zi = 8;
static constexpr int tile_size_compute_bi = 4;
static constexpr int tile_size_transform_bi = 4;
static constexpr int tile_size_compute_beta = 4;
static constexpr int tile_size_compute_yi = 8;
static constexpr int team_size_compute_fused_deidrj = sizeof(real_type) == 4 ? 4 : 2;
// this empirically reduces perf fluctuations from compiler version to compiler version
static constexpr int min_blocks_compute_zi = 4;
static constexpr int min_blocks_compute_yi = 4;
#endif
// Custom MDRangePolicy, Rank3, to reduce verbosity of kernel launches
// This hides the Kokkos::IndexType<int> and Kokkos::Rank<3...>
// and reduces the verbosity of the LaunchBound by hiding the explicit
// multiplication by vector_length
template <class Device, int num_tiles, class TagPairSNAP>
using Snap3DRangePolicy = typename Kokkos::MDRangePolicy<Device, Kokkos::IndexType<int>, Kokkos::Rank<3, Kokkos::Iterate::Left, Kokkos::Iterate::Left>, Kokkos::LaunchBounds<vector_length * num_tiles>, TagPairSNAP>;
template <class Device, int num_tiles, class TagPairSNAP, int min_blocks = 0>
using Snap3DRangePolicy = typename Kokkos::MDRangePolicy<Device, Kokkos::IndexType<int>, Kokkos::Rank<3, Kokkos::Iterate::Left, Kokkos::Iterate::Left>, Kokkos::LaunchBounds<vector_length * num_tiles, min_blocks>, TagPairSNAP>;
// Custom SnapAoSoATeamPolicy to reduce the verbosity of kernel launches
// This hides the LaunchBounds abstraction by hiding the explicit
@ -134,6 +142,29 @@ class PairSNAPKokkos : public PairSNAP {
template <class Device, int num_teams, class TagPairSNAP>
using SnapAoSoATeamPolicy = typename Kokkos::TeamPolicy<Device, Kokkos::LaunchBounds<vector_length * num_teams>, TagPairSNAP>;
// Custom MDRangePolicy, Rank2, on the host, to reduce verbosity of kernel launches. The striding of this launch is intentionally
// different from the tiled 3D range policy on the device.
template <class Device, class TagPairSNAP>
using Snap2DHostRangePolicy = typename Kokkos::MDRangePolicy<Device, Kokkos::Schedule<Kokkos::Dynamic>, Kokkos::IndexType<int>, Kokkos::Rank<2, Kokkos::Iterate::Right, Kokkos::Iterate::Right>, TagPairSNAP>;
// Custom RangePolicy, Rank2, on the host, to reduce verbosity of kernel launches
template <class Device, class TagPairSNAP>
using Snap1DHostRangePolicy = typename Kokkos::RangePolicy<Device, Kokkos::Schedule<Kokkos::Dynamic>, TagPairSNAP>;
// Helper routine that returns a CPU or a GPU policy as appropriate
template <class Device, int num_tiles, class TagPairSNAP, int min_blocks = 0>
auto snap_get_policy(const int& chunk_size_div, const int& second_loop) {
if constexpr (host_flag) {
return Snap1DHostRangePolicy<Device, TagPairSNAP>(0, chunk_size_div * vector_length);
// the 2-d policy is still correct but it has atomics so it's slower on the CPU
//return Snap2DHostRangePolicy<Device, TagPairSNAP>({0, 0}, {chunk_size_div * vector_length, second_loop});
} else
return Snap3DRangePolicy<Device, num_tiles, TagPairSNAP, min_blocks>({0, 0, 0},
{vector_length, second_loop, chunk_size_div},
{vector_length, num_tiles, 1});
}
PairSNAPKokkos(class LAMMPS *);
~PairSNAPKokkos() override;
@ -149,6 +180,7 @@ class PairSNAPKokkos : public PairSNAP {
template<class TagStyle>
void check_team_size_reduce(int, int&);
// CPU and GPU backend
template<int NEIGHFLAG, int EVFLAG>
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeForce<NEIGHFLAG,EVFLAG>,const int& ii) const;
@ -157,18 +189,23 @@ class PairSNAPKokkos : public PairSNAP {
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeForce<NEIGHFLAG,EVFLAG>,const int& ii, EV_FLOAT&) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPBetaCPU,const int& ii) const;
// GPU backend only
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeNeigh,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeNeigh>::member_type& team) const;
// GPU backend only
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeCayleyKlein, const int iatom_mod, const int jnbor, const int iatom_div) const;
// CPU and GPU
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPPreUi,const int iatom_mod, const int j, const int iatom_div) const;
void operator() (TagPairSNAPPreUi, const int& iatom_mod, const int& j, const int& iatom_div) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPPreUi, const int& iatom, const int& j) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPPreUi, const int& iatom) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeUiSmall,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeUiSmall>::member_type& team) const;
@ -177,25 +214,67 @@ class PairSNAPKokkos : public PairSNAP {
void operator() (TagPairSNAPComputeUiLarge,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeUiLarge>::member_type& team) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPTransformUi,const int iatom_mod, const int j, const int iatom_div) const;
void operator() (TagPairSNAPTransformUi, const int& iatom_mod, const int& idxu, const int& iatom_div) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeZi,const int iatom_mod, const int idxz, const int iatom_div) const;
void operator() (TagPairSNAPTransformUi, const int& iatom, const int& idxu) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPBeta, const int& ii) const;
void operator() (TagPairSNAPTransformUi, const int& iatom) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeZi<chemsnap>, const int& iatom_mod, const int& idxz, const int& iatom_div) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeZi<chemsnap>, const int& iatom, const int& idxz) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeZi<chemsnap>, const int& iatom) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBi<chemsnap>, const int& iatom_mod, const int& idxb, const int& iatom_div) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBi<chemsnap>, const int& iatom, const int& idxb) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBi<chemsnap>, const int& iatom) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBi,const int iatom_mod, const int idxb, const int iatom_div) const;
void operator() (TagPairSNAPComputeBetaLinear, const int& iatom_mod, const int& idxb, const int& iatom_div) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPTransformBi,const int iatom_mod, const int idxb, const int iatom_div) const;
void operator() (TagPairSNAPComputeBetaLinear, const int& iatom, const int& idxb) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYi,const int iatom_mod, const int idxz, const int iatom_div) const;
void operator() (TagPairSNAPComputeBetaLinear, const int& iatom) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYiWithZlist,const int iatom_mod, const int idxz, const int iatom_div) const;
void operator() (TagPairSNAPComputeBetaQuadratic, const int& iatom_mod, const int& idxb, const int& iatom_div) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBetaQuadratic, const int& iatom, const int& idxb) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBetaQuadratic, const int& iatom) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYi<chemsnap>, const int& iatom_mod, const int& idxz, const int& iatom_div) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYi<chemsnap>, const int& iatom, const int& idxz) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYi<chemsnap>, const int& iatom) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYiWithZlist<chemsnap>, const int& iatom_mod, const int& idxz, const int& iatom_div) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYiWithZlist<chemsnap>, const int& iatom, const int& idxz) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYiWithZlist<chemsnap>, const int& iatom) const;
template<int dir>
KOKKOS_INLINE_FUNCTION
@ -210,28 +289,22 @@ class PairSNAPKokkos : public PairSNAP {
void operator() (TagPairSNAPComputeNeighCPU,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeNeighCPU>::member_type& team) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPPreUiCPU,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPPreUiCPU>::member_type& team) const;
void operator() (TagPairSNAPComputeUiCPU, const int& iatom, const int& jnbor) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeUiCPU,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeUiCPU>::member_type& team) const;
void operator() (TagPairSNAPComputeUiCPU, const int& iatom) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPTransformUiCPU, const int j, const int iatom) const;
void operator() (TagPairSNAPComputeDuidrjCPU, const int& iatom, const int& jnbor) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeZiCPU,const int& ii) const;
void operator() (TagPairSNAPComputeDuidrjCPU, const int& iatom) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeBiCPU,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeBiCPU>::member_type& team) const;
void operator() (TagPairSNAPComputeDeidrjCPU, const int& iatom, const int& jnbor) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeYiCPU,const int& ii) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeDuidrjCPU,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeDuidrjCPU>::member_type& team) const;
KOKKOS_INLINE_FUNCTION
void operator() (TagPairSNAPComputeDeidrjCPU,const typename Kokkos::TeamPolicy<DeviceType, TagPairSNAPComputeDeidrjCPU>::member_type& team) const;
void operator() (TagPairSNAPComputeDeidrjCPU, const int& iatom) const;
template<int NEIGHFLAG>
KOKKOS_INLINE_FUNCTION
@ -252,7 +325,7 @@ class PairSNAPKokkos : public PairSNAP {
SNAKokkos<DeviceType, real_type, vector_length> snaKK;
int inum,max_neighs,chunk_size,chunk_offset;
int host_flag,neighflag;
int neighflag;
int eflag,vflag;
@ -260,13 +333,12 @@ class PairSNAPKokkos : public PairSNAP {
Kokkos::View<real_type*, DeviceType> d_radelem; // element radii
Kokkos::View<real_type*, DeviceType> d_wjelem; // elements weights
Kokkos::View<real_type**, Kokkos::LayoutRight, DeviceType> d_coeffelem; // element bispectrum coefficients
typename SNAKokkos<DeviceType, real_type, vector_length>::t_sna_2d_lr d_coeffelem; // element bispectrum coefficients
Kokkos::View<real_type*, DeviceType> d_sinnerelem; // element inner cutoff midpoint
Kokkos::View<real_type*, DeviceType> d_dinnerelem; // element inner cutoff half-width
Kokkos::View<T_INT*, DeviceType> d_map; // mapping from atom types to elements
Kokkos::View<T_INT*, DeviceType> d_ninside; // ninside for all atoms in list
Kokkos::View<real_type**, DeviceType> d_beta; // betas for all atoms in list
Kokkos::View<real_type***, Kokkos::LayoutLeft, DeviceType> d_beta_pack; // betas for all atoms in list, GPU
typename SNAKokkos<DeviceType, real_type, vector_length>::t_sna_2d d_beta; // betas for all atoms in list
typedef Kokkos::DualView<F_FLOAT**, DeviceType> tdual_fparams;
tdual_fparams k_cutsq;
@ -301,6 +373,9 @@ class PairSNAPKokkos : public PairSNAP {
template <typename scratch_type>
int scratch_size_helper(int values_per_team);
// Make SNAKokkos a friend
friend class SNAKokkos<DeviceType, real_type, vector_length>;
};

1410
src/KOKKOS/pair_snap_kokkos_impl.h
View File

File diff suppressed because it is too large Load Diff

145
src/KOKKOS/sna_kokkos.h
View File

@ -134,6 +134,8 @@ class SNAKokkos {
static constexpr int vector_length = vector_length_;
using KKDeviceType = typename KKDevice<DeviceType>::value;
static constexpr LAMMPS_NS::ExecutionSpace execution_space = ExecutionSpaceFromDevice<DeviceType>::space;
static constexpr int host_flag = (execution_space == LAMMPS_NS::Host);
typedef Kokkos::View<int*, DeviceType> t_sna_1i;
typedef Kokkos::View<real_type*, DeviceType> t_sna_1d;
@ -141,6 +143,7 @@ class SNAKokkos {
typedef Kokkos::View<int**, DeviceType> t_sna_2i;
typedef Kokkos::View<real_type**, DeviceType> t_sna_2d;
typedef Kokkos::View<real_type**, Kokkos::LayoutLeft, DeviceType> t_sna_2d_ll;
typedef Kokkos::View<real_type**, Kokkos::LayoutRight, DeviceType> t_sna_2d_lr;
typedef Kokkos::View<real_type***, DeviceType> t_sna_3d;
typedef Kokkos::View<real_type***, Kokkos::LayoutLeft, DeviceType> t_sna_3d_ll;
typedef Kokkos::View<real_type***[3], DeviceType> t_sna_4d;
@ -156,7 +159,7 @@ class SNAKokkos {
typedef Kokkos::View<complex***, DeviceType> t_sna_3c;
typedef Kokkos::View<complex***, Kokkos::LayoutLeft, DeviceType> t_sna_3c_ll;
typedef Kokkos::View<complex***[3], DeviceType> t_sna_4c;
typedef Kokkos::View<complex***[3], Kokkos::LayoutLeft, DeviceType> t_sna_4c3_ll;
typedef Kokkos::View<complex***[3], DeviceType> t_sna_4c3;
typedef Kokkos::View<complex****, Kokkos::LayoutLeft, DeviceType> t_sna_4c_ll;
typedef Kokkos::View<complex**[3], DeviceType> t_sna_3c3;
typedef Kokkos::View<complex*****, DeviceType> t_sna_5c;
@ -168,7 +171,8 @@ class SNAKokkos {
SNAKokkos(const SNAKokkos<DeviceType,real_type,vector_length>& sna, const typename Kokkos::TeamPolicy<DeviceType>::member_type& team);
inline
SNAKokkos(real_type, int, real_type, int, int, int, int, int, int, int);
//SNAKokkos(real_type, int, real_type, int, int, int, int, int, int, int);
SNAKokkos(const PairSNAPKokkos<DeviceType, real_type, vector_length>&);
KOKKOS_INLINE_FUNCTION
~SNAKokkos();
@ -182,88 +186,87 @@ class SNAKokkos {
double memory_usage();
int ncoeff;
int host_flag;
// functions for bispectrum coefficients, GPU only
KOKKOS_INLINE_FUNCTION
void compute_cayley_klein(const int&, const int&, const int&);
void compute_cayley_klein(const int&, const int&) const;
KOKKOS_INLINE_FUNCTION
void pre_ui(const int&, const int&, const int&, const int&); // ForceSNAP
void pre_ui(const int&, const int&, const int&) const; // ForceSNAP
// version of the code with parallelism over j_bend
KOKKOS_INLINE_FUNCTION
void compute_ui_small(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int, const int); // ForceSNAP
void compute_ui_small(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int, const int) const; // ForceSNAP
// version of the code without parallelism over j_bend
KOKKOS_INLINE_FUNCTION
void compute_ui_large(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int); // ForceSNAP
void compute_ui_large(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int) const; // ForceSNAP
// desymmetrize ulisttot
KOKKOS_INLINE_FUNCTION
void compute_zi(const int&, const int&, const int&); // ForceSNAP
void transform_ui(const int&, const int&) const;
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void compute_zi(const int&, const int&) const; // ForceSNAP
template <bool chemsnap, bool need_atomics> KOKKOS_INLINE_FUNCTION
void compute_yi(const int&, const int&) const; // ForceSNAP
template <bool chemsnap, bool need_atomics> KOKKOS_INLINE_FUNCTION
void compute_yi_with_zlist(const int&, const int&) const; // ForceSNAP
template <bool chemsnap> KOKKOS_INLINE_FUNCTION
void compute_bi(const int&, const int&) const; // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_yi(int,int,int,
const Kokkos::View<real_type***, Kokkos::LayoutLeft, DeviceType> &beta_pack); // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_yi_with_zlist(int,int,int,
const Kokkos::View<real_type***, Kokkos::LayoutLeft, DeviceType> &beta_pack); // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_bi(const int&, const int&, const int&); // ForceSNAP
void compute_beta_linear(const int&, const int&, const int&) const;
template <bool need_atomics> KOKKOS_INLINE_FUNCTION
void compute_beta_quadratic(const int&, const int&, const int&) const;
// functions for derivatives, GPU only
// version of the code with parallelism over j_bend
template<int dir>
KOKKOS_INLINE_FUNCTION
void compute_fused_deidrj_small(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int, const int); //ForceSNAP
void compute_fused_deidrj_small(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int, const int) const; //ForceSNAP
// version of the code without parallelism over j_bend
template<int dir>
KOKKOS_INLINE_FUNCTION
void compute_fused_deidrj_large(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int); //ForceSNAP
void compute_fused_deidrj_large(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, const int, const int, const int) const; //ForceSNAP
// core "evaluation" functions that get plugged into "compute" functions
// plugged into compute_ui_small, compute_ui_large
KOKKOS_FORCEINLINE_FUNCTION
void evaluate_ui_jbend(const WignerWrapper<real_type, vector_length>&, const complex&, const complex&, const real_type&, const int&,
const int&, const int&, const int&);
const int&, const int&) const;
// plugged into compute_zi, compute_yi
KOKKOS_FORCEINLINE_FUNCTION
complex evaluate_zi(const int&, const int&, const int&, const int&, const int&, const int&, const int&, const int&, const int&,
const int&, const int&, const int&, const int&, const real_type*);
// plugged into compute_yi, compute_yi_with_zlist
const int&, const int&, const int&, const real_type*) const;
// plugged into compute_bi
KOKKOS_FORCEINLINE_FUNCTION
real_type evaluate_beta_scaled(const int&, const int&, const int&, const int&, const int&, const int&, const int&, const int&,
const Kokkos::View<real_type***, Kokkos::LayoutLeft, DeviceType> &);
real_type evaluate_bi(const int&, const int&, const int&, const int&,
const int&, const int&, const int&) const;
// plugged into compute_yi, compute_yi_with_zlist
template <bool chemsnap> KOKKOS_FORCEINLINE_FUNCTION
real_type evaluate_beta_scaled(const int&, const int&, const int&, const int&, const int&, const int&, const int&) const;
// plugged into compute_fused_deidrj_small, compute_fused_deidrj_large
KOKKOS_FORCEINLINE_FUNCTION
real_type evaluate_duidrj_jbend(const WignerWrapper<real_type, vector_length>&, const complex&, const complex&, const real_type&,
const WignerWrapper<real_type, vector_length>&, const complex&, const complex&, const real_type&,
const int&, const int&, const int&, const int&);
const int&, const int&, const int&) const;
// functions for bispectrum coefficients, CPU only
KOKKOS_INLINE_FUNCTION
void pre_ui_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team,const int&,const int&); // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_ui_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int, int); // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_zi_cpu(const int&); // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_yi_cpu(int,
const Kokkos::View<real_type**, DeviceType> &beta); // ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_bi_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int); // ForceSNAP
template <bool need_atomics> KOKKOS_INLINE_FUNCTION
void compute_ui_cpu(const int&, const int&) const; // ForceSNAP
// functions for derivatives, CPU only
KOKKOS_INLINE_FUNCTION
void compute_duidrj_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int, int); //ForceSNAP
void compute_duidrj_cpu(const int&, const int&) const; //ForceSNAP
KOKKOS_INLINE_FUNCTION
void compute_deidrj_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int, int); // ForceSNAP
void compute_deidrj_cpu(const int&, const int&) const; // ForceSNAP
KOKKOS_INLINE_FUNCTION
real_type compute_sfac(real_type, real_type, real_type, real_type); // add_uarraytot, compute_duarray
real_type compute_sfac(real_type, real_type, real_type, real_type) const; // add_uarraytot, compute_duarray
KOKKOS_INLINE_FUNCTION
real_type compute_dsfac(real_type, real_type, real_type, real_type); // compute_duarray
real_type compute_dsfac(real_type, real_type, real_type, real_type) const; // compute_duarray
KOKKOS_INLINE_FUNCTION
void compute_s_dsfac(const real_type, const real_type, const real_type, const real_type, real_type&, real_type&); // compute_cayley_klein
void compute_s_dsfac(const real_type, const real_type, const real_type, const real_type, real_type&, real_type&) const; // compute_cayley_klein
#ifdef TIMING_INFO
double* timers;
@ -283,37 +286,41 @@ class SNAKokkos {
t_sna_2d dinnerij;
t_sna_2i element;
t_sna_3d dedr;
int natom, nmax;
int natom, natom_pad, nmax;
void grow_rij(int, int);
int twojmax, diagonalstyle;
// Input beta coefficients; aliases the object in PairSnapKokkos
t_sna_2d_lr d_coeffelem;
// Beta for all atoms in list; aliases the object in PairSnapKokkos
// for qSNAP the quadratic terms get accumulated into it
// in compute_bi
t_sna_2d d_beta;
// Structures for both the CPU, GPU backend
t_sna_3d ulisttot_re;
t_sna_3d ulisttot_im;
t_sna_3c ulisttot; // un-folded ulisttot
t_sna_3c zlist;
t_sna_3d blist;
t_sna_3c_ll ulisttot;
t_sna_3c_ll ulisttot_full; // un-folded ulisttot, cpu only
t_sna_3c_ll zlist;
t_sna_3c_ll ulist;
t_sna_3c_ll ylist;
t_sna_3d ylist_re;
t_sna_3d ylist_im;
// derivatives of data
t_sna_4c3_ll dulist;
// Structures for the CPU backend only
t_sna_3c ulist_cpu;
t_sna_4c3 dulist_cpu;
// Modified structures for GPU backend
t_sna_3c_ll a_pack; // Cayley-Klein `a`
t_sna_3c_ll b_pack; // `b`
t_sna_4c_ll da_pack; // `da`
t_sna_4c_ll db_pack; // `db`
t_sna_4d_ll sfac_pack; // sfac, dsfac_{x,y,z}
t_sna_4d_ll ulisttot_re_pack; // split real,
t_sna_4d_ll ulisttot_im_pack; // imag, AoSoA, flattened
t_sna_4c_ll ulisttot_pack; // AoSoA layout
t_sna_4c_ll zlist_pack; // AoSoA layout
t_sna_4d_ll blist_pack;
t_sna_4d_ll ylist_pack_re; // split real,
t_sna_4d_ll ylist_pack_im; // imag AoSoA layout
t_sna_2c a_gpu; // Cayley-Klein `a`
t_sna_2c b_gpu; // `b`
t_sna_3c da_gpu; // `da`
t_sna_3c db_gpu; // `db`
t_sna_3d sfac_gpu; // sfac, dsfac_{x,y,z}
int idxcg_max, idxu_max, idxu_half_max, idxu_cache_max, idxz_max, idxb_max;
@ -363,25 +370,11 @@ class SNAKokkos {
inline
void init_rootpqarray(); // init()
KOKKOS_INLINE_FUNCTION
void add_uarraytot(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int, int, const real_type&, const real_type&, const real_type&, const real_type&, const real_type&, int); // compute_ui
KOKKOS_INLINE_FUNCTION
void compute_uarray_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int, int,
const real_type&, const real_type&, const real_type&,
const real_type&, const real_type&); // compute_ui_cpu
inline
double deltacg(int, int, int); // init_clebsch_gordan
inline
int compute_ncoeff(); // SNAKokkos()
KOKKOS_INLINE_FUNCTION
void compute_duarray_cpu(const typename Kokkos::TeamPolicy<DeviceType>::member_type& team, int, int,
const real_type&, const real_type&, const real_type&, // compute_duidrj_cpu
const real_type&, const real_type&, const real_type&, const real_type&, const real_type&,
const real_type&, const real_type&);
// Sets the style for the switching function
// 0 = none
@ -401,6 +394,9 @@ class SNAKokkos {
real_type wself;
int wselfall_flag;
// quadratic flag
int quadratic_flag;
int bzero_flag; // 1 if bzero subtracted from barray
Kokkos::View<real_type*, DeviceType> bzero; // array of B values for isolated atoms
};
@ -409,4 +405,3 @@ class SNAKokkos {
#include "sna_kokkos_impl.h"
#endif

1833
src/KOKKOS/sna_kokkos_impl.h
View File

File diff suppressed because it is too large Load Diff

2
src/PHONON/fix_phonon.cpp
View File

@ -400,7 +400,7 @@ void FixPhonon::end_of_step()
ndim = sysdim;
for (i = 1; i < nucell; ++i) {
for (idim = 0; idim < sysdim; ++idim) dist2orig[idim] = Rnow[idx][ndim++] - Rnow[idx][idim];
domain->minimum_image(dist2orig);
domain->minimum_image_big(dist2orig);
for (idim = 0; idim < sysdim; ++idim) basis[i][idim] += dist2orig[idim];
}
}

2
src/REAXFF/fix_qtpie_reaxff.cpp
View File

@ -49,7 +49,6 @@ using namespace LAMMPS_NS;
using namespace FixConst;
static constexpr double CONV_TO_EV = 14.4;
static constexpr double SMALL = 1.0e-14;
static constexpr double QSUMSMALL = 0.00001;
static constexpr double ANGSTROM_TO_BOHRRADIUS = 1.8897261259;
@ -1101,7 +1100,6 @@ void FixQtpieReaxFF::calc_chi_eff()
memset(&chi_eff[0],0,atom->nmax*sizeof(double));
const auto x = (const double * const *)atom->x;
const int ntypes = atom->ntypes;
const int *type = atom->type;
double dist,overlap,sum_n,sum_d,expa,expb,chia,chib,phia,phib,p,m;

2
src/RIGID/fix_ehex.cpp
View File

@ -434,7 +434,7 @@ bool FixEHEX::check_cluster(tagint *shake_atom, int n, Region *region)
// take into account pbc
domain->minimum_image(xtemp);
domain->minimum_image_big(xtemp);
for (int k = 0; k < 3; k++) xcom[k] += mi * (x[lid[0]][k] + xtemp[k]);
}

4
src/angle_zero.cpp
View File

@ -54,13 +54,13 @@ void AngleZero::compute(int eflag, int vflag)
void AngleZero::settings(int narg, char **arg)
{
if ((narg != 0) && (narg != 1)) error->all(FLERR, "Illegal angle_style command");
if (narg > 1) error->all(FLERR, "Too many angle_style zero keywords");
if (narg == 1) {
if (strcmp("nocoeff", arg[0]) == 0)
coeffflag = 0;
else
error->all(FLERR, "Illegal angle_style command");
error->all(FLERR, "Unknown angle_style zero keyword {}", arg[0]);
}
}

1
src/dump.cpp
View File

@ -73,6 +73,7 @@ Dump::Dump(LAMMPS *lmp, int /*narg*/, char **arg) :
clearstep = 0;
sort_flag = 0;
sortcol = 0;
balance_flag = 0;
append_flag = 0;
buffer_allow = 0;

4
src/fix_vector.cpp
View File

@ -114,9 +114,11 @@ FixVector::FixVector(LAMMPS *lmp, int narg, char **arg) :
error->all(FLERR, "Fix for fix {} vector not computed at compatible time", val.id);
if (val.argindex == 0)
value = ifix->extscalar;
else if (ifix->extvector >= 0)
value = ifix->extvector;
else
value = ifix->extarray;
value = ifix->extlist[val.argindex - 1];
val.val.f = ifix;
} else if (val.which == ArgInfo::VARIABLE) {

2
src/library.cpp
View File

@ -519,7 +519,7 @@ must be freed with :cpp:func:`lammps_free` after use to avoid a memory leak.
\endverbatim
*
* \param handle pointer to a previously created LAMMPS instance
* \param cmd string with a single LAMMPS input line
* \param line string with a single LAMMPS input line
* \return string with expanded line */
char *lammps_expand(void *handle, const char *line)

2
src/lmptype.h
View File

@ -287,7 +287,7 @@ struct multitype {
int64_t b;
} data;
multitype() : type(LAMMPS_NONE) { data.d = 0.0; }
multitype() noexcept : type(LAMMPS_NONE) { data.d = 0.0; }
multitype(const multitype &) = default;
multitype(multitype &&) = default;
~multitype() = default;

12
src/tokenizer.h
View File

@ -59,17 +59,21 @@ class TokenizerException : public std::exception {
std::string message;
public:
// remove unused default constructor
/** The default constructor is disabled */
TokenizerException() = delete;
/** Thrown during retrieving or skipping tokens
*
* \param msg String with error message
* \param token String of the token/word that caused the error */
* \param msg String with error message
* \param token String of the token or word that caused the error */
explicit TokenizerException(const std::string &msg, const std::string &token);
/** Retrieve message describing the thrown exception
* \return string with error message */
*
* This function provides the message that can be retrieved when the corresponding
* exception is caught.
*
* \return String with error message */
const char *what() const noexcept override { return message.c_str(); }
};

16
src/utils.h
View File

@ -59,7 +59,7 @@ namespace utils {
void missing_cmd_args(const std::string &file, int line, const std::string &cmd, Error *error);
/* Internal function handling the argument list for logmesg(). */
/*! Internal function handling the argument list for logmesg(). */
void fmtargs_logmesg(LAMMPS *lmp, fmt::string_view format, fmt::format_args args);
@ -426,12 +426,11 @@ This functions adds the following case to :cpp:func:`utils::bounds() <LAMMPS_NS:
* \param ref per-grid reference from input script, e.g. "c_10:grid:data[2]"
* \param nevery frequency at which caller will access fix for per-grid info,
* ignored when reference is to a compute
* \param id ID of Compute or Fix
* \param igrid which grid is referenced (0 to N-1)
* \param idata which data on grid is referenced (0 to N-1)
* \param index which column of data is referenced (0 for vec, 1-N for array)
* \param lmp pointer to top-level LAMMPS class instance
* \param verify check bounds for interaction type
* \return id ID of Compute or Fix
* \return igrid which grid is referenced (0 to N-1)
* \return idata which data on grid is referenced (0 to N-1)
* \return index which column of data is referenced (0 for vec, 1-N for array)
* \return ArgINFO::COMPUTE or FIX or UNKNOWN or NONE */
int check_grid_reference(char *errstr, char *ref, int nevery, char *&id, int &igrid, int &idata,
@ -442,7 +441,10 @@ This functions adds the following case to :cpp:func:`utils::bounds() <LAMMPS_NS:
* Format of grid ID reference = id:gname:dname.
* Return vector with the 3 sub-strings.
*
* \param name = complete grid ID
* \param file name of source file for error message
* \param line line number in source file for error message
* \param name complete grid ID
* \param error pointer to Error class
* \return std::vector<std::string> containing the 3 sub-strings */
std::vector<std::string> parse_grid_id(const char *file, int line, const std::string &name,

2
tools/lammps-gui/CMakeLists.txt
View File

@ -205,7 +205,7 @@ if(FLATPAK_COMMAND AND FLATPAK_BUILDER)
file(STRINGS ${LAMMPS_DIR}/src/version.h line REGEX LAMMPS_VERSION)
string(REGEX REPLACE "#define LAMMPS_VERSION \"([0-9]+) ([A-Za-z][A-Za-z][A-Za-z])[A-Za-z]* ([0-9]+)\""
"\\1\\2\\3" LAMMPS_RELEASE "${line}")
set(FLATPAK_BUNDLE "LAMMPS_GUI-Linux-amd64-${LAMMPS_RELEASE}.flatpak")
set(FLATPAK_BUNDLE "LAMMPS-Linux-x86_64-GUI-${LAMMPS_RELEASE}.flatpak")
add_custom_target(flatpak
COMMAND ${FLATPAK_COMMAND} --user remote-add --if-not-exists flathub https://dl.flathub.org/repo/flathub.flatpakrepo
COMMAND ${FLATPAK_BUILDER} --force-clean --verbose --repo=${CMAKE_CURRENT_BINARY_DIR}/flatpak-repo

3
tools/lammps-gui/lammps-gui.appdata.xml
View File

@ -56,6 +56,9 @@
<releases>
<release version="1.6.11" timestamp="1725080055">
<description>
move cursor to end of log buffer before inserting new text
remove empirical filter to remove outliers from corrupted data
change tutorial download URL to tutorial website
</description>
</release>
<release version="1.6.10" timestamp="1724585189">

1
tools/lammps-gui/org.lammps.lammps-gui.yml
View File

@ -88,6 +88,7 @@ modules:
- -D PKG_QTB=yes
- -D PKG_REACTION=yes
- -D PKG_REAXFF=yes
- -D PKG_RHEO=yes
- -D PKG_RIGID=yes
- -D PKG_SHOCK=yes
- -D PKG_SMTBQ=yes