forked from lijiext/lammps
Merge pull request #1567 from akohlmey/collected-small-fixes
Collected small fixes and update for the next patch
This commit is contained in:
Axel Kohlmeyer
GitHub
commit
981b9a1f7a
|
|
@ -37,7 +37,7 @@ include(PreventInSourceBuilds)
|
|||
|
||||
if(NOT CMAKE_BUILD_TYPE AND NOT CMAKE_CXX_FLAGS)
|
||||
#release comes with -O3 by default
|
||||
set(CMAKE_BUILD_TYPE Release CACHE STRING "Choose the type of build, options are: None Debug Release RelWithDebInfo MinSizeRel." FORCE)
|
||||
set(CMAKE_BUILD_TYPE RelWithDebInfo CACHE STRING "Choose the type of build, options are: None Debug Release RelWithDebInfo MinSizeRel." FORCE)
|
||||
endif(NOT CMAKE_BUILD_TYPE AND NOT CMAKE_CXX_FLAGS)
|
||||
string(TOUPPER "${CMAKE_BUILD_TYPE}" BTYPE)
|
||||
|
||||
|
|
@ -673,5 +673,10 @@ if(PKG_KOKKOS)
|
|||
message(STATUS "Kokkos Arch: ${KOKKOS_ARCH}")
|
||||
endif()
|
||||
if(PKG_KSPACE)
|
||||
message(STATUS "Using ${FFT} as FFT")
|
||||
message(STATUS "Using ${FFT} as primary FFT library")
|
||||
if(FFT_SINGLE)
|
||||
message(STATUS "Using single precision FFTs")
|
||||
else()
|
||||
message(STATUS "Using double precision FFTs")
|
||||
endif()
|
||||
endif()
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
if(PKG_KSPACE)
|
||||
option(FFT_SINGLE "Use single precision FFT instead of double" OFF)
|
||||
option(FFT_SINGLE "Use single precision FFTs instead of double precision FFTs" OFF)
|
||||
set(FFTW "FFTW3")
|
||||
if(FFT_SINGLE)
|
||||
set(FFTW "FFTW3F")
|
||||
|
|
@ -7,26 +7,30 @@ if(PKG_KSPACE)
|
|||
endif()
|
||||
find_package(${FFTW} QUIET)
|
||||
if(${FFTW}_FOUND)
|
||||
set(FFT "${FFTW}" CACHE STRING "FFT library for KSPACE package")
|
||||
set(FFT "FFTW3" CACHE STRING "FFT library for KSPACE package")
|
||||
else()
|
||||
set(FFT "KISS" CACHE STRING "FFT library for KSPACE package")
|
||||
endif()
|
||||
set(FFT_VALUES KISS ${FFTW} MKL)
|
||||
set(FFT_VALUES KISS FFTW3 MKL)
|
||||
set_property(CACHE FFT PROPERTY STRINGS ${FFT_VALUES})
|
||||
validate_option(FFT FFT_VALUES)
|
||||
string(TOUPPER ${FFT} FFT)
|
||||
if(NOT FFT STREQUAL "KISS")
|
||||
find_package(${FFT} REQUIRED)
|
||||
if(NOT FFT STREQUAL "FFTW3F")
|
||||
add_definitions(-DFFT_FFTW)
|
||||
else()
|
||||
add_definitions(-DFFT_${FFT})
|
||||
endif()
|
||||
include_directories(${${FFT}_INCLUDE_DIRS})
|
||||
list(APPEND LAMMPS_LINK_LIBS ${${FFT}_LIBRARIES})
|
||||
|
||||
if(FFT STREQUAL "FFTW3")
|
||||
find_package(${FFTW} REQUIRED)
|
||||
add_definitions(-DFFT_FFTW3)
|
||||
include_directories(${${FFTW}_INCLUDE_DIRS})
|
||||
list(APPEND LAMMPS_LINK_LIBS ${${FFTW}_LIBRARIES})
|
||||
elseif(FFT STREQUAL "MKL")
|
||||
find_package(MKL REQUIRED)
|
||||
add_definitions(-DFFT_MKL)
|
||||
include_directories(${MKL_INCLUDE_DIRS})
|
||||
list(APPEND LAMMPS_LINK_LIBS ${MKL_LIBRARIES})
|
||||
else()
|
||||
# last option is KISSFFT
|
||||
add_definitions(-DFFT_KISS)
|
||||
endif()
|
||||
|
||||
set(FFT_PACK "array" CACHE STRING "Optimization for FFT")
|
||||
set(FFT_PACK_VALUES array pointer memcpy)
|
||||
set_property(CACHE FFT_PACK PROPERTY STRINGS ${FFT_PACK_VALUES})
|
||||
|
|
|
|||
|
|
@ -171,42 +171,40 @@ The individual style names on the "Commands
|
|||
compute"_Commands_compute.html doc page are followed by one or more of
|
||||
(g,i,k,o,t) to indicate which accelerated styles exist.
|
||||
|
||||
"ackland/atom"_compute_ackland_atom.html -
|
||||
"ackland/atom"_compute_ackland_atom.html - determines the local lattice structure based on the Ackland formulation
|
||||
"adf"_compute_adf.html - angular distribution function of triples of atoms
|
||||
"aggregate/atom"_compute_cluster_atom.html - aggregate ID for each atom
|
||||
"angle"_compute_angle.html -
|
||||
"angle/local"_compute_angle_local.html -
|
||||
"angle/local"_compute_bond_local.html - theta and energy of each angle
|
||||
"angle"_compute_angle.html - energy of each angle sub-style
|
||||
"angle/local"_compute_angle_local.html - theta and energy of each angle
|
||||
"angmom/chunk"_compute_angmom_chunk.html - angular momentum for each chunk
|
||||
"basal/atom"_compute_basal_atom.html -
|
||||
"basal/atom"_compute_basal_atom.html - calculates the hexagonal close-packed “c” lattice vector of each atom
|
||||
"body/local"_compute_body_local.html - attributes of body sub-particles
|
||||
"bond"_compute_bond.html - values computed by a bond style
|
||||
"bond"_compute_bond.html - energy of each bond sub-style
|
||||
"bond/local"_compute_bond_local.html - distance and energy of each bond
|
||||
"centro/atom"_compute_centro_atom.html - centro-symmetry parameter for each atom
|
||||
"chunk/atom"_compute_chunk_atom.html - assign chunk IDs to each atom
|
||||
"chunk/spread/atom"_compute_chunk_spread_atom.html - spreads chunk values to each atom in chunk
|
||||
"cluster/atom"_compute_cluster_atom.html - cluster ID for each atom
|
||||
"cna/atom"_compute_cna_atom.html - common neighbor analysis (CNA) for each atom
|
||||
"cnp/atom"_compute_cnp_atom.html -
|
||||
"cnp/atom"_compute_cnp_atom.html - common neighborhood parameter (CNP) for each atom
|
||||
"com"_compute_com.html - center-of-mass of group of atoms
|
||||
"com/chunk"_compute_com_chunk.html - center-of-mass for each chunk
|
||||
"contact/atom"_compute_contact_atom.html - contact count for each spherical particle
|
||||
"coord/atom"_compute_coord_atom.html - coordination number for each atom
|
||||
"damage/atom"_compute_damage_atom.html - Peridynamic damage for each atom
|
||||
"dihedral"_compute_dihedral.html -
|
||||
"dihedral"_compute_dihedral.html - energy of each dihedral sub-style
|
||||
"dihedral/local"_compute_dihedral_local.html - angle of each dihedral
|
||||
"dilatation/atom"_compute_dilatation_atom.html - Peridynamic dilatation for each atom
|
||||
"dipole/chunk"_compute_dipole_chunk.html -
|
||||
"dipole/chunk"_compute_dipole_chunk.html - dipole vector and total dipole for each chunk
|
||||
"displace/atom"_compute_displace_atom.html - displacement of each atom
|
||||
"dpd"_compute_dpd.html -
|
||||
"dpd/atom"_compute_dpd_atom.html -
|
||||
"edpd/temp/atom"_compute_edpd_temp_atom.html -
|
||||
"entropy/atom"_compute_entropy_atom.html -
|
||||
"edpd/temp/atom"_compute_edpd_temp_atom.html - per-atom temperature for each eDPD particle in a group
|
||||
"entropy/atom"_compute_entropy_atom.html - pair entropy fingerprint of each atom
|
||||
"erotate/asphere"_compute_erotate_asphere.html - rotational energy of aspherical particles
|
||||
"erotate/rigid"_compute_erotate_rigid.html - rotational energy of rigid bodies
|
||||
"erotate/sphere"_compute_erotate_sphere.html - rotational energy of spherical particles
|
||||
"erotate/sphere/atom"_compute_erotate_sphere.html - rotational energy for each spherical particle
|
||||
"erotate/sphere/atom"_compute_erotate_sphere_atom.html -
|
||||
"erotate/sphere/atom"_compute_erotate_sphere_atom.html - rotational energy for each spherical particle
|
||||
"event/displace"_compute_event_displace.html - detect event on atom displacement
|
||||
"fep"_compute_fep.html -
|
||||
"force/tally"_compute_tally.html -
|
||||
|
|
@ -218,17 +216,17 @@ compute"_Commands_compute.html doc page are followed by one or more of
|
|||
"heat/flux"_compute_heat_flux.html - heat flux through a group of atoms
|
||||
"heat/flux/tally"_compute_tally.html -
|
||||
"hexorder/atom"_compute_hexorder_atom.html - bond orientational order parameter q6
|
||||
"improper"_compute_improper.html -
|
||||
"improper"_compute_improper.html - energy of each improper sub-style
|
||||
"improper/local"_compute_improper_local.html - angle of each improper
|
||||
"inertia/chunk"_compute_inertia_chunk.html - inertia tensor for each chunk
|
||||
"ke"_compute_ke.html - translational kinetic energy
|
||||
"ke/atom"_compute_ke_atom.html - kinetic energy for each atom
|
||||
"ke/atom/eff"_compute_ke_atom_eff.html -
|
||||
"ke/eff"_compute_ke_eff.html -
|
||||
"ke/atom/eff"_compute_ke_atom_eff.html - per-atom translational and radial kinetic energy in the electron force field model
|
||||
"ke/eff"_compute_ke_eff.html - kinetic energy of a group of nuclei and electrons in the electron force field model
|
||||
"ke/rigid"_compute_ke_rigid.html - translational kinetic energy of rigid bodies
|
||||
"meso/e/atom"_compute_meso_e_atom.html -
|
||||
"meso/rho/atom"_compute_meso_rho_atom.html -
|
||||
"meso/t/atom"_compute_meso_t_atom.html -
|
||||
"meso/e/atom"_compute_meso_e_atom.html - per-atom internal energy of Smooth-Particle Hydrodynamics atoms
|
||||
"meso/rho/atom"_compute_meso_rho_atom.html - per-atom mesoscopic density of Smooth-Particle Hydrodynamics atoms
|
||||
"meso/t/atom"_compute_meso_t_atom.html - per-atom internal temperature of Smooth-Particle Hydrodynamics atoms
|
||||
"msd"_compute_msd.html - mean-squared displacement of group of atoms
|
||||
"msd/chunk"_compute_msd_chunk.html - mean-squared displacement for each chunk
|
||||
"msd/nongauss"_compute_msd_nongauss.html - MSD and non-Gaussian parameter of group of atoms
|
||||
|
|
@ -242,73 +240,72 @@ compute"_Commands_compute.html doc page are followed by one or more of
|
|||
"pe/tally"_compute_tally.html -
|
||||
"plasticity/atom"_compute_plasticity_atom.html - Peridynamic plasticity for each atom
|
||||
"pressure"_compute_pressure.html - total pressure and pressure tensor
|
||||
"pressure/cylinder"_compute_pressure_cylinder.html -
|
||||
"pressure/uef"_compute_pressure_uef.html -
|
||||
"pressure/cylinder"_compute_pressure_cylinder.html - pressure tensor in cylindrical coordinates
|
||||
"pressure/uef"_compute_pressure_uef.html - pressure tensor in the reference frame of an applied flow field
|
||||
"property/atom"_compute_property_atom.html - convert atom attributes to per-atom vectors/arrays
|
||||
"property/chunk"_compute_property_chunk.html - extract various per-chunk attributes
|
||||
"property/local"_compute_property_local.html - convert local attributes to localvectors/arrays
|
||||
"ptm/atom"_compute_ptm_atom.html -
|
||||
"ptm/atom"_compute_ptm_atom.html - determines the local lattice structure based on the Polyhedral Template Matching method
|
||||
"rdf"_compute_rdf.html - radial distribution function g(r) histogram of group of atoms
|
||||
"reduce"_compute_reduce.html - combine per-atom quantities into a single global value
|
||||
"reduce/chunk"_compute_reduce_chunk.html - reduce per-atom quantities within each chunk
|
||||
"reduce/region"_compute_reduce.html - same as compute reduce, within a region
|
||||
"rigid/local"_compute_rigid_local.html - extract rigid body attributes
|
||||
"saed"_compute_saed.html -
|
||||
"saed"_compute_saed.html - electron diffraction intensity on a mesh of reciprocal lattice nodes
|
||||
"slice"_compute_slice.html - extract values from global vector or array
|
||||
"smd/contact/radius"_compute_smd_contact_radius.html -
|
||||
"smd/damage"_compute_smd_damage.html -
|
||||
"smd/damage"_compute_smd_damage.html - damage status of SPH particles in Smooth Mach Dynamics
|
||||
"smd/hourglass/error"_compute_smd_hourglass_error.html -
|
||||
"smd/internal/energy"_compute_smd_internal_energy.html -
|
||||
"smd/plastic/strain"_compute_smd_plastic_strain.html -
|
||||
"smd/plastic/strain/rate"_compute_smd_plastic_strain_rate.html -
|
||||
"smd/rho"_compute_smd_rho.html -
|
||||
"smd/tlsph/defgrad"_compute_smd_tlsph_defgrad.html -
|
||||
"smd/tlsph/dt"_compute_smd_tlsph_dt.html -
|
||||
"smd/internal/energy"_compute_smd_internal_energy.html - per-particle enthalpy in Smooth Mach Dynamics
|
||||
"smd/plastic/strain"_compute_smd_plastic_strain.html - equivalent plastic strain per particle in Smooth Mach Dynamics
|
||||
"smd/plastic/strain/rate"_compute_smd_plastic_strain_rate.html - time rate of the equivalent plastic strain in Smooth Mach Dynamics
|
||||
"smd/rho"_compute_smd_rho.html - per-particle mass density in Smooth Mach Dynamics
|
||||
"smd/tlsph/defgrad"_compute_smd_tlsph_defgrad.html - deformation gradient in Smooth Mach Dynamics
|
||||
"smd/tlsph/dt"_compute_smd_tlsph_dt.html - CFL-stable time increment per particle in Smooth Mach Dynamics
|
||||
"smd/tlsph/num/neighs"_compute_smd_tlsph_num_neighs.html -
|
||||
"smd/tlsph/shape"_compute_smd_tlsph_shape.html -
|
||||
"smd/tlsph/strain"_compute_smd_tlsph_strain.html -
|
||||
"smd/tlsph/strain/rate"_compute_smd_tlsph_strain_rate.html -
|
||||
"smd/tlsph/stress"_compute_smd_tlsph_stress.html -
|
||||
"smd/triangle/vertices"_compute_smd_triangle_vertices.html -
|
||||
"smd/tlsph/stress"_compute_smd_tlsph_stress.html - per-particle Cauchy stress tensor for SPH particles
|
||||
"smd/triangle/vertices"_compute_smd_triangle_vertices.html -
|
||||
"smd/ulsph/num/neighs"_compute_smd_ulsph_num_neighs.html -
|
||||
"smd/ulsph/strain"_compute_smd_ulsph_strain.html -
|
||||
"smd/ulsph/strain/rate"_compute_smd_ulsph_strain_rate.html -
|
||||
"smd/ulsph/stress"_compute_smd_ulsph_stress.html -
|
||||
"smd/vol"_compute_smd_vol.html -
|
||||
"smd/ulsph/stress"_compute_smd_ulsph_stress.html - per-particle Cauchy stress tensor and von Mises equivalent stress in Smooth Mach Dynamics
|
||||
"smd/vol"_compute_smd_vol.html - per-particle volumes and their sum in Smooth Mach Dynamics
|
||||
"sna/atom"_compute_sna_atom.html - calculate bispectrum coefficients for each atom
|
||||
"snad/atom"_compute_sna_atom.html - derivative of bispectrum coefficients for each atom
|
||||
"snav/atom"_compute_sna_atom.html - virial contribution from bispectrum coefficients for each atom
|
||||
"spin"_compute_spin.html -
|
||||
"spin"_compute_spin.html - magnetic quantities for a system of atoms having spins
|
||||
"stress/atom"_compute_stress_atom.html - stress tensor for each atom
|
||||
"stress/mop"_compute_stress_mop.html -
|
||||
"stress/mop/profile"_compute_stress_mop.html -
|
||||
"stress/mop"_compute_stress_mop.html - normal components of the local stress tensor using the method of planes
|
||||
"stress/mop/profile"_compute_stress_mop.html - profile of the normal components of the local stress tensor using the method of planes
|
||||
"stress/tally"_compute_tally.html -
|
||||
"tdpd/cc/atom"_compute_tdpd_cc_atom.html -
|
||||
"tdpd/cc/atom"_compute_tdpd_cc_atom.html - per-atom chemical concentration of a specified species for each tDPD particle
|
||||
"temp"_compute_temp.html - temperature of group of atoms
|
||||
"temp/asphere"_compute_temp_asphere.html - temperature of aspherical particles
|
||||
"temp/body"_compute_temp_body.html - temperature of body particles
|
||||
"temp/chunk"_compute_temp_chunk.html - temperature of each chunk
|
||||
"temp/com"_compute_temp_com.html - temperature after subtracting center-of-mass velocity
|
||||
"temp/cs"_compute_temp_cs.html -
|
||||
"temp/cs"_compute_temp_cs.html - temperature based on the center-of-mass velocity of atom pairs that are bonded to each other
|
||||
"temp/deform"_compute_temp_deform.html - temperature excluding box deformation velocity
|
||||
"temp/deform/eff"_compute_temp_deform_eff.html -
|
||||
"temp/drude"_compute_temp_drude.html -
|
||||
"temp/eff"_compute_temp_eff.html -
|
||||
"temp/deform/eff"_compute_temp_deform_eff.html - temperature excluding box deformation velocity in the electron force field model
|
||||
"temp/drude"_compute_temp_drude.html - temperature of Core-Drude pairs
|
||||
"temp/eff"_compute_temp_eff.html - temperature of a group of nuclei and electrons in the electron force field model
|
||||
"temp/partial"_compute_temp_partial.html - temperature excluding one or more dimensions of velocity
|
||||
"temp/profile"_compute_temp_profile.html - temperature excluding a binned velocity profile
|
||||
"temp/ramp"_compute_temp_ramp.html - temperature excluding ramped velocity component
|
||||
"temp/region"_compute_temp_region.html - temperature of a region of atoms
|
||||
"temp/region/eff"_compute_temp_region_eff.html -
|
||||
"temp/rotate"_compute_temp_rotate.html -
|
||||
"temp/region/eff"_compute_temp_region_eff.html - temperature of a region of nuclei and electrons in the electron force field model
|
||||
"temp/rotate"_compute_temp_rotate.html - temperature of a group of atoms after subtracting out their center-of-mass and angular velocities
|
||||
"temp/sphere"_compute_temp_sphere.html - temperature of spherical particles
|
||||
"temp/uef"_compute_temp_uef.html -
|
||||
"temp/uef"_compute_temp_uef.html - kinetic energy tensor in the reference frame of an applied flow field
|
||||
"ti"_compute_ti.html - thermodynamic integration free energy values
|
||||
"torque/chunk"_compute_torque_chunk.html - torque applied on each chunk
|
||||
"vacf"_compute_vacf.html - velocity auto-correlation function of group of atoms
|
||||
"vcm/chunk"_compute_vcm_chunk.html - velocity of center-of-mass for each chunk
|
||||
"voronoi/atom"_compute_voronoi_atom.html - Voronoi volume and neighbors for each atom
|
||||
"xrd"_compute_xrd.html - :ul
|
||||
"xrd"_compute_xrd.html - x-ray diffraction intensity on a mesh of reciprocal lattice nodes :ul
|
||||
|
||||
[Restrictions:] none
|
||||
|
||||
|
|
|
|||
|
|
@ -306,12 +306,6 @@ void PPPMKokkos<DeviceType>::init()
|
|||
|
||||
if (me == 0) {
|
||||
|
||||
#ifdef FFT_SINGLE
|
||||
const char fft_prec[] = "single";
|
||||
#else
|
||||
const char fft_prec[] = "double";
|
||||
#endif
|
||||
|
||||
if (screen) {
|
||||
fprintf(screen," G vector (1/distance) = %g\n",g_ewald);
|
||||
fprintf(screen," grid = %d %d %d\n",nx_pppm,ny_pppm,nz_pppm);
|
||||
|
|
@ -320,7 +314,7 @@ void PPPMKokkos<DeviceType>::init()
|
|||
estimated_accuracy);
|
||||
fprintf(screen," estimated relative force accuracy = %g\n",
|
||||
estimated_accuracy/two_charge_force);
|
||||
fprintf(screen," using %s precision FFTs\n",fft_prec);
|
||||
fprintf(screen," using " LMP_FFT_PREC " precision " LMP_FFT_LIB "\n");
|
||||
fprintf(screen," 3d grid and FFT values/proc = %d %d\n",
|
||||
ngrid_max,nfft_both_max);
|
||||
}
|
||||
|
|
@ -332,7 +326,7 @@ void PPPMKokkos<DeviceType>::init()
|
|||
estimated_accuracy);
|
||||
fprintf(logfile," estimated relative force accuracy = %g\n",
|
||||
estimated_accuracy/two_charge_force);
|
||||
fprintf(logfile," using %s precision FFTs\n",fft_prec);
|
||||
fprintf(logfile," using " LMP_FFT_PREC " precision " LMP_FFT_LIB "\n");
|
||||
fprintf(logfile," 3d grid and FFT values/proc = %d %d\n",
|
||||
ngrid_max,nfft_both_max);
|
||||
}
|
||||
|
|
|
|||
|
|
@ -350,12 +350,6 @@ void PPPM::init()
|
|||
|
||||
if (me == 0) {
|
||||
|
||||
#ifdef FFT_SINGLE
|
||||
const char fft_prec[] = "single";
|
||||
#else
|
||||
const char fft_prec[] = "double";
|
||||
#endif
|
||||
|
||||
if (screen) {
|
||||
fprintf(screen," G vector (1/distance) = %g\n",g_ewald);
|
||||
fprintf(screen," grid = %d %d %d\n",nx_pppm,ny_pppm,nz_pppm);
|
||||
|
|
@ -364,7 +358,7 @@ void PPPM::init()
|
|||
estimated_accuracy);
|
||||
fprintf(screen," estimated relative force accuracy = %g\n",
|
||||
estimated_accuracy/two_charge_force);
|
||||
fprintf(screen," using %s precision FFTs\n",fft_prec);
|
||||
fprintf(screen," using " LMP_FFT_PREC " precision " LMP_FFT_LIB "\n");
|
||||
fprintf(screen," 3d grid and FFT values/proc = %d %d\n",
|
||||
ngrid_max,nfft_both_max);
|
||||
}
|
||||
|
|
@ -376,7 +370,7 @@ void PPPM::init()
|
|||
estimated_accuracy);
|
||||
fprintf(logfile," estimated relative force accuracy = %g\n",
|
||||
estimated_accuracy/two_charge_force);
|
||||
fprintf(logfile," using %s precision FFTs\n",fft_prec);
|
||||
fprintf(logfile," using " LMP_FFT_PREC " precision " LMP_FFT_LIB "\n");
|
||||
fprintf(logfile," 3d grid and FFT values/proc = %d %d\n",
|
||||
ngrid_max,nfft_both_max);
|
||||
}
|
||||
|
|
|
|||
|
|
@ -23,11 +23,22 @@ KSpaceStyle(pppm,PPPM)
|
|||
#include "lmptype.h"
|
||||
#include <mpi.h>
|
||||
|
||||
#if defined(FFT_FFTW3)
|
||||
#define LMP_FFT_LIB "FFTW3"
|
||||
#elif defined(FFT_MKL)
|
||||
#define LMP_FFT_LIB "MKL FFT"
|
||||
#else
|
||||
#define LMP_FFT_LIB "KISS FFT"
|
||||
#endif
|
||||
|
||||
#ifdef FFT_SINGLE
|
||||
typedef float FFT_SCALAR;
|
||||
#define LMP_FFT_PREC "single"
|
||||
#define MPI_FFT_SCALAR MPI_FLOAT
|
||||
#else
|
||||
|
||||
typedef double FFT_SCALAR;
|
||||
#define LMP_FFT_PREC "double"
|
||||
#define MPI_FFT_SCALAR MPI_DOUBLE
|
||||
#endif
|
||||
|
||||
|
|
|
|||
|
|
@ -474,12 +474,6 @@ void PPPMDisp::init()
|
|||
MPI_Allreduce(&nfft_both,&nfft_both_max,1,MPI_INT,MPI_MAX,world);
|
||||
|
||||
if (me == 0) {
|
||||
#ifdef FFT_SINGLE
|
||||
const char fft_prec[] = "single";
|
||||
#else
|
||||
const char fft_prec[] = "double";
|
||||
#endif
|
||||
|
||||
if (screen) {
|
||||
fprintf(screen," Coulomb G vector (1/distance)= %g\n",g_ewald);
|
||||
fprintf(screen," Coulomb grid = %d %d %d\n",nx_pppm,ny_pppm,nz_pppm);
|
||||
|
|
@ -488,7 +482,7 @@ void PPPMDisp::init()
|
|||
acc);
|
||||
fprintf(screen," Coulomb estimated relative force accuracy = %g\n",
|
||||
acc/two_charge_force);
|
||||
fprintf(screen," using %s precision FFTs\n",fft_prec);
|
||||
fprintf(screen," using " LMP_FFT_PREC " precision " LMP_FFT_LIB "\n");
|
||||
fprintf(screen," 3d grid and FFT values/proc = %d %d\n",
|
||||
ngrid_max, nfft_both_max);
|
||||
}
|
||||
|
|
@ -501,7 +495,7 @@ void PPPMDisp::init()
|
|||
acc);
|
||||
fprintf(logfile," Coulomb estimated relative force accuracy = %g\n",
|
||||
acc/two_charge_force);
|
||||
fprintf(logfile," using %s precision FFTs\n",fft_prec);
|
||||
fprintf(logfile," using " LMP_FFT_PREC " precision " LMP_FFT_LIB "\n");
|
||||
fprintf(logfile," 3d grid and FFT values/proc = %d %d\n",
|
||||
ngrid_max, nfft_both_max);
|
||||
}
|
||||
|
|
|
|||
|
|
@ -23,11 +23,21 @@ KSpaceStyle(pppm/disp,PPPMDisp)
|
|||
#include "lmptype.h"
|
||||
#include <mpi.h>
|
||||
|
||||
#if defined(FFT_FFTW3)
|
||||
#define LMP_FFT_LIB "FFTW3"
|
||||
#elif defined(FFT_MKL)
|
||||
#define LMP_FFT_LIB "MKL FFT"
|
||||
#else
|
||||
#define LMP_FFT_LIB "KISS FFT"
|
||||
#endif
|
||||
|
||||
#ifdef FFT_SINGLE
|
||||
typedef float FFT_SCALAR;
|
||||
#define LMP_FFT_PREC "single"
|
||||
#define MPI_FFT_SCALAR MPI_FLOAT
|
||||
#else
|
||||
typedef double FFT_SCALAR;
|
||||
#define LMP_FFT_PREC "double"
|
||||
#define MPI_FFT_SCALAR MPI_DOUBLE
|
||||
#endif
|
||||
|
||||
|
|
|
|||
|
|
@ -55,7 +55,7 @@ AtomVecPeri::AtomVecPeri(LAMMPS *lmp) : AtomVec(lmp)
|
|||
comm_f_only = 1;
|
||||
size_forward = 4;
|
||||
size_reverse = 3;
|
||||
size_border = 11;
|
||||
size_border = 12;
|
||||
size_velocity = 3;
|
||||
size_data_atom = 7;
|
||||
size_data_vel = 4;
|
||||
|
|
@ -357,6 +357,7 @@ int AtomVecPeri::pack_border(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = vfrac[j];
|
||||
buf[m++] = rmass[j];
|
||||
buf[m++] = s0[j];
|
||||
buf[m++] = x0[j][0];
|
||||
buf[m++] = x0[j][1];
|
||||
|
|
@ -381,6 +382,7 @@ int AtomVecPeri::pack_border(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = vfrac[j];
|
||||
buf[m++] = rmass[j];
|
||||
buf[m++] = s0[j];
|
||||
buf[m++] = x0[j][0];
|
||||
buf[m++] = x0[j][1];
|
||||
|
|
@ -414,6 +416,7 @@ int AtomVecPeri::pack_border_vel(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = vfrac[j];
|
||||
buf[m++] = rmass[j];
|
||||
buf[m++] = s0[j];
|
||||
buf[m++] = x0[j][0];
|
||||
buf[m++] = x0[j][1];
|
||||
|
|
@ -442,6 +445,7 @@ int AtomVecPeri::pack_border_vel(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = vfrac[j];
|
||||
buf[m++] = rmass[j];
|
||||
buf[m++] = s0[j];
|
||||
buf[m++] = x0[j][0];
|
||||
buf[m++] = x0[j][1];
|
||||
|
|
@ -463,6 +467,7 @@ int AtomVecPeri::pack_border_vel(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = vfrac[j];
|
||||
buf[m++] = rmass[j];
|
||||
buf[m++] = s0[j];
|
||||
buf[m++] = x0[j][0];
|
||||
buf[m++] = x0[j][1];
|
||||
|
|
@ -497,6 +502,7 @@ int AtomVecPeri::pack_border_hybrid(int n, int *list, double *buf)
|
|||
for (i = 0; i < n; i++) {
|
||||
j = list[i];
|
||||
buf[m++] = vfrac[j];
|
||||
buf[m++] = rmass[j];
|
||||
buf[m++] = s0[j];
|
||||
buf[m++] = x0[j][0];
|
||||
buf[m++] = x0[j][1];
|
||||
|
|
@ -522,6 +528,7 @@ void AtomVecPeri::unpack_border(int n, int first, double *buf)
|
|||
type[i] = (int) ubuf(buf[m++]).i;
|
||||
mask[i] = (int) ubuf(buf[m++]).i;
|
||||
vfrac[i] = buf[m++];
|
||||
rmass[i] = buf[m++];
|
||||
s0[i] = buf[m++];
|
||||
x0[i][0] = buf[m++];
|
||||
x0[i][1] = buf[m++];
|
||||
|
|
@ -551,6 +558,7 @@ void AtomVecPeri::unpack_border_vel(int n, int first, double *buf)
|
|||
type[i] = (int) ubuf(buf[m++]).i;
|
||||
mask[i] = (int) ubuf(buf[m++]).i;
|
||||
vfrac[i] = buf[m++];
|
||||
rmass[i] = buf[m++];
|
||||
s0[i] = buf[m++];
|
||||
x0[i][0] = buf[m++];
|
||||
x0[i][1] = buf[m++];
|
||||
|
|
@ -576,6 +584,7 @@ int AtomVecPeri::unpack_border_hybrid(int n, int first, double *buf)
|
|||
last = first + n;
|
||||
for (i = first; i < last; i++) {
|
||||
vfrac[i] = buf[m++];
|
||||
rmass[i] = buf[m++];
|
||||
s0[i] = buf[m++];
|
||||
x0[i][0] = buf[m++];
|
||||
x0[i][1] = buf[m++];
|
||||
|
|
|
|||
|
|
@ -41,7 +41,7 @@ AtomVecEllipsoid::AtomVecEllipsoid(LAMMPS *lmp) : AtomVec(lmp)
|
|||
comm_x_only = comm_f_only = 0;
|
||||
size_forward = 7;
|
||||
size_reverse = 6;
|
||||
size_border = 14;
|
||||
size_border = 15;
|
||||
size_velocity = 6;
|
||||
size_data_atom = 7;
|
||||
size_data_vel = 7;
|
||||
|
|
@ -544,6 +544,7 @@ int AtomVecEllipsoid::pack_border(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(tag[j]).d;
|
||||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = rmass[j];
|
||||
if (ellipsoid[j] < 0) buf[m++] = ubuf(0).d;
|
||||
else {
|
||||
buf[m++] = ubuf(1).d;
|
||||
|
|
@ -576,6 +577,7 @@ int AtomVecEllipsoid::pack_border(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(tag[j]).d;
|
||||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = rmass[j];
|
||||
if (ellipsoid[j] < 0) buf[m++] = ubuf(0).d;
|
||||
else {
|
||||
buf[m++] = ubuf(1).d;
|
||||
|
|
@ -618,6 +620,7 @@ int AtomVecEllipsoid::pack_border_vel(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(tag[j]).d;
|
||||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = rmass[j];
|
||||
if (ellipsoid[j] < 0) buf[m++] = ubuf(0).d;
|
||||
else {
|
||||
buf[m++] = ubuf(1).d;
|
||||
|
|
@ -657,6 +660,7 @@ int AtomVecEllipsoid::pack_border_vel(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(tag[j]).d;
|
||||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = rmass[j];
|
||||
if (ellipsoid[j] < 0) buf[m++] = ubuf(0).d;
|
||||
else {
|
||||
buf[m++] = ubuf(1).d;
|
||||
|
|
@ -689,6 +693,7 @@ int AtomVecEllipsoid::pack_border_vel(int n, int *list, double *buf,
|
|||
buf[m++] = ubuf(tag[j]).d;
|
||||
buf[m++] = ubuf(type[j]).d;
|
||||
buf[m++] = ubuf(mask[j]).d;
|
||||
buf[m++] = rmass[j];
|
||||
if (ellipsoid[j] < 0) buf[m++] = ubuf(0).d;
|
||||
else {
|
||||
buf[m++] = ubuf(1).d;
|
||||
|
|
@ -735,6 +740,7 @@ int AtomVecEllipsoid::pack_border_hybrid(int n, int *list, double *buf)
|
|||
m = 0;
|
||||
for (i = 0; i < n; i++) {
|
||||
j = list[i];
|
||||
buf[m++] = rmass[j];
|
||||
if (ellipsoid[j] < 0) buf[m++] = ubuf(0).d;
|
||||
else {
|
||||
buf[m++] = ubuf(1).d;
|
||||
|
|
@ -769,6 +775,7 @@ void AtomVecEllipsoid::unpack_border(int n, int first, double *buf)
|
|||
tag[i] = (tagint) ubuf(buf[m++]).i;
|
||||
type[i] = (int) ubuf(buf[m++]).i;
|
||||
mask[i] = (int) ubuf(buf[m++]).i;
|
||||
rmass[i] = buf[m++];
|
||||
ellipsoid[i] = (int) ubuf(buf[m++]).i;
|
||||
if (ellipsoid[i] == 0) ellipsoid[i] = -1;
|
||||
else {
|
||||
|
|
@ -812,6 +819,7 @@ void AtomVecEllipsoid::unpack_border_vel(int n, int first, double *buf)
|
|||
tag[i] = (tagint) ubuf(buf[m++]).i;
|
||||
type[i] = (int) ubuf(buf[m++]).i;
|
||||
mask[i] = (int) ubuf(buf[m++]).i;
|
||||
rmass[i] = buf[m++];
|
||||
ellipsoid[i] = (int) ubuf(buf[m++]).i;
|
||||
if (ellipsoid[i] == 0) ellipsoid[i] = -1;
|
||||
else {
|
||||
|
|
@ -854,6 +862,7 @@ int AtomVecEllipsoid::unpack_border_hybrid(int n, int first, double *buf)
|
|||
m = 0;
|
||||
last = first + n;
|
||||
for (i = first; i < last; i++) {
|
||||
rmass[i] = buf[m++];
|
||||
ellipsoid[i] = (int) ubuf(buf[m++]).i;
|
||||
if (ellipsoid[i] == 0) ellipsoid[i] = -1;
|
||||
else {
|
||||
|
|
|
|||
|
|
@ -13,6 +13,7 @@
|
|||
|
||||
#include <mpi.h>
|
||||
#include <cstring>
|
||||
#include <cstdlib>
|
||||
#include "utils.h"
|
||||
#include "lammps.h"
|
||||
#include "error.h"
|
||||
|
|
|
|||
Loading…
Reference in New Issue