Homec4science
Diffusion lammps (efficient_neuronet) Browse

lammps/src/USER-OMP/pair_lj_cut_tip4p_cut_omp.cpp
4f65214374a5efficient_neuronet

Last Change
Raw File

pair_lj_cut_tip4p_cut_omp.cpp
View Options

/* ----------------------------------------------------------------------
LAMMPS - Large-scale Atomic/Molecular Massively Parallel Simulator
http://lammps.sandia.gov, Sandia National Laboratories
Steve Plimpton, sjplimp@sandia.gov
This software is distributed under the GNU General Public License.
See the README file in the top-level LAMMPS directory.
------------------------------------------------------------------------- */
/* ----------------------------------------------------------------------
Contributing author: Axel Kohlmeyer (Temple U)
------------------------------------------------------------------------- */
#include <math.h>
#include "pair_lj_cut_tip4p_cut_omp.h"
#include "atom.h"
#include "domain.h"
#include "comm.h"
#include "force.h"
#include "neighbor.h"
#include "error.h"
#include "memory.h"
#include "neigh_list.h"
#include "suffix.h"
using namespace LAMMPS_NS;
#define EWALD_F 1.12837917
#define EWALD_P 0.3275911
#define A1 0.254829592
#define A2 -0.284496736
#define A3 1.421413741
#define A4 -1.453152027
#define A5 1.061405429
/* ---------------------------------------------------------------------- */
PairLJCutTIP4PCutOMP::PairLJCutTIP4PCutOMP(LAMMPS *lmp) :
PairLJCutTIP4PCut(lmp), ThrOMP(lmp, THR_PAIR)
{
suffix_flag |= Suffix::OMP;
respa_enable = 0;
newsite_thr = NULL;
hneigh_thr = NULL;
// TIP4P cannot compute virial as F dot r
// due to finding bonded H atoms which are not near O atom
no_virial_fdotr_compute = 1;
}
/* ---------------------------------------------------------------------- */
PairLJCutTIP4PCutOMP::~PairLJCutTIP4PCutOMP()
{
memory->destroy(hneigh_thr);
memory->destroy(newsite_thr);
}
/* ---------------------------------------------------------------------- */
void PairLJCutTIP4PCutOMP::compute(int eflag, int vflag)
{
if (eflag || vflag) ev_setup(eflag,vflag);
else evflag = vflag_fdotr = 0;
const int nlocal = atom->nlocal;
const int nall = nlocal + atom->nghost;
// reallocate hneigh_thr & newsite_thr if necessary
// initialize hneigh_thr[0] to -1 on steps when reneighboring occurred
// initialize hneigh_thr[2] to 0 every step
if (atom->nmax > nmax) {
nmax = atom->nmax;
memory->destroy(hneigh_thr);
memory->create(hneigh_thr,nmax,"pair:hneigh_thr");
memory->destroy(newsite_thr);
memory->create(newsite_thr,nmax,"pair:newsite_thr");
}
int i;
// tag entire list as completely invalid after a neighbor
// list update, since that can change the order of atoms.
if (neighbor->ago == 0)
for (i = 0; i < nall; i++) hneigh_thr[i].a = -1;
// indicate that the coordinates for the M point need to
// be updated. this needs to be done in every step.
for (i = 0; i < nall; i++) hneigh_thr[i].t = 0;
const int nthreads = comm->nthreads;
const int inum = list->inum;
#if defined(_OPENMP)
#pragma omp parallel default(none) shared(eflag,vflag)
#endif
{
int ifrom, ito, tid;
loop_setup_thr(ifrom, ito, tid, inum, nthreads);
ThrData *thr = fix->get_thr(tid);
thr->timer(Timer::START);
ev_setup_thr(eflag, vflag, nall, eatom, vatom, thr);
if (evflag) {
if (eflag) {
if (vflag) eval<1,1,1>(ifrom, ito, thr);
else eval<1,1,0>(ifrom, ito, thr);
} else {
if (vflag) eval<1,0,1>(ifrom, ito, thr);
else eval<1,0,0>(ifrom, ito, thr);
}
} else eval<0,0,0>(ifrom, ito, thr);
thr->timer(Timer::PAIR);
reduce_thr(this, eflag, vflag, thr);
} // end of omp parallel region
}
/* ---------------------------------------------------------------------- */
template <int EVFLAG, int EFLAG, int VFLAG>
void PairLJCutTIP4PCutOMP::eval(int iifrom, int iito, ThrData * const thr)
{
double qtmp,xtmp,ytmp,ztmp,delx,dely,delz,evdwl,ecoul;
double r,rsq,r2inv,r6inv,forcecoul,forcelj,cforce;
double factor_coul,factor_lj;
double v[6],xH1[3],xH2[3];
double fdx,fdy,fdz,fOx,fOy,fOz,fHx,fHy,fHz;
dbl3_t x1,x2;
int *ilist,*jlist,*numneigh,**firstneigh;
int i,j,ii,jj,jnum,itype,jtype,key;
int n,vlist[6];
int iH1,iH2,jH1,jH2;
evdwl = ecoul = 0.0;
const dbl3_t * _noalias const x = (dbl3_t *) atom->x[0];
dbl3_t * _noalias const f = (dbl3_t *) thr->get_f()[0];
const double * _noalias const q = atom->q;
const int * _noalias const type = atom->type;
const int nlocal = atom->nlocal;
const double * _noalias const special_coul = force->special_coul;
const double * _noalias const special_lj = force->special_lj;
const double qqrd2e = force->qqrd2e;
const double cut_coulsqplus = (cut_coul+2.0*qdist) * (cut_coul+2.0*qdist);
double fxtmp,fytmp,fztmp;
ilist = list->ilist;
numneigh = list->numneigh;
firstneigh = list->firstneigh;
// loop over neighbors of my atoms
for (ii = iifrom; ii < iito; ++ii) {
i = ilist[ii];
qtmp = q[i];
xtmp = x[i].x;
ytmp = x[i].y;
ztmp = x[i].z;
itype = type[i];
// if atom I = water O, set x1 = offset charge site
// else x1 = x of atom I
// NOTE: to make this part thread safe, we need to
// make sure that the hneigh_thr[][] entries only get
// updated, when all data is in place. worst case,
// some calculation is repeated, but since the results
// will be the same, there is no race condition.
if (itype == typeO) {
if (hneigh_thr[i].a < 0) {
iH1 = atom->map(atom->tag[i] + 1);
iH2 = atom->map(atom->tag[i] + 2);
if (iH1 == -1 || iH2 == -1)
error->one(FLERR,"TIP4P hydrogen is missing");
if (atom->type[iH1] != typeH || atom->type[iH2] != typeH)
error->one(FLERR,"TIP4P hydrogen has incorrect atom type");
compute_newsite_thr(x[i],x[iH1],x[iH2],newsite_thr[i]);
hneigh_thr[i].t = 1;
hneigh_thr[i].b = iH2;
hneigh_thr[i].a = iH1;
} else {
iH1 = hneigh_thr[i].a;
iH2 = hneigh_thr[i].b;
if (hneigh_thr[i].t == 0) {
compute_newsite_thr(x[i],x[iH1],x[iH2],newsite_thr[i]);
hneigh_thr[i].t = 1;
}
}
x1 = newsite_thr[i];
} else x1 = x[i];
jlist = firstneigh[i];
jnum = numneigh[i];
fxtmp=fytmp=fztmp=0.0;
for (jj = 0; jj < jnum; jj++) {
j = jlist[jj];
factor_lj = special_lj[sbmask(j)];
factor_coul = special_coul[sbmask(j)];
j &= NEIGHMASK;
delx = xtmp - x[j].x;
dely = ytmp - x[j].y;
delz = ztmp - x[j].z;
rsq = delx*delx + dely*dely + delz*delz;
jtype = type[j];
// LJ interaction based on true rsq
if (rsq < cut_ljsq[itype][jtype]) {
r2inv = 1.0/rsq;
r6inv = r2inv*r2inv*r2inv;
forcelj = r6inv * (lj1[itype][jtype]*r6inv - lj2[itype][jtype]);
forcelj *= factor_lj * r2inv;
fxtmp += delx*forcelj;
fytmp += dely*forcelj;
fztmp += delz*forcelj;
f[j].x -= delx*forcelj;
f[j].y -= dely*forcelj;
f[j].z -= delz*forcelj;
if (EFLAG) {
evdwl = r6inv*(lj3[itype][jtype]*r6inv-lj4[itype][jtype]) -
offset[itype][jtype];
evdwl *= factor_lj;
} else evdwl = 0.0;
if (EVFLAG) ev_tally_thr(this,i,j,nlocal, /* newton_pair = */ 1,
evdwl,0.0,forcelj,delx,dely,delz,thr);
}
// adjust rsq and delxyz for off-site O charge(s) if necessary
// but only if they are within reach
// NOTE: to make this part thread safe, we need to
// make sure that the hneigh_thr[][] entries only get
// updated, when all data is in place. worst case,
// some calculation is repeated, but since the results
// will be the same, there is no race condition.
if (rsq < cut_coulsqplus) {
if (itype == typeO || jtype == typeO) {
// if atom J = water O, set x2 = offset charge site
// else x2 = x of atom J
if (jtype == typeO) {
if (hneigh_thr[j].a < 0) {
jH1 = atom->map(atom->tag[j] + 1);
jH2 = atom->map(atom->tag[j] + 2);
if (jH1 == -1 || jH2 == -1)
error->one(FLERR,"TIP4P hydrogen is missing");
if (atom->type[jH1] != typeH || atom->type[jH2] != typeH)
error->one(FLERR,"TIP4P hydrogen has incorrect atom type");
compute_newsite_thr(x[j],x[jH1],x[jH2],newsite_thr[j]);
hneigh_thr[j].t = 1;
hneigh_thr[j].b = jH2;
hneigh_thr[j].a = jH1;
} else {
jH1 = hneigh_thr[j].a;
jH2 = hneigh_thr[j].b;
if (hneigh_thr[j].t == 0) {
compute_newsite_thr(x[j],x[jH1],x[jH2],newsite_thr[j]);
hneigh_thr[j].t = 1;
}
}
x2 = newsite_thr[j];
} else x2 = x[j];
delx = x1.x - x2.x;
dely = x1.y - x2.y;
delz = x1.z - x2.z;
rsq = delx*delx + dely*dely + delz*delz;
}
// Coulombic interaction based on modified rsq
if (rsq < cut_coulsq) {
r = sqrt(rsq);
r2inv = 1.0 / rsq;
forcecoul = qqrd2e * qtmp * q[j] / r;
cforce = factor_coul * forcecoul * r2inv;
// if i,j are not O atoms, force is applied directly
// if i or j are O atoms, force is on fictitious atom & partitioned
// force partitioning due to Feenstra, J Comp Chem, 20, 786 (1999)
// f_f = fictitious force, fO = f_f (1 - 2 alpha), fH = alpha f_f
// preserves total force and torque on water molecule
// virial = sum(r x F) where each water's atoms are near xi and xj
// vlist stores 2,4,6 atoms whose forces contribute to virial
if (EVFLAG) {
n = 0;
key = 0;
}
if (itype != typeO) {
fxtmp += delx * cforce;
fytmp += dely * cforce;
fztmp += delz * cforce;
if (VFLAG) {
v[0] = x[i].x * delx * cforce;
v[1] = x[i].y * dely * cforce;
v[2] = x[i].z * delz * cforce;
v[3] = x[i].x * dely * cforce;
v[4] = x[i].x * delz * cforce;
v[5] = x[i].y * delz * cforce;
}
if (EVFLAG) vlist[n++] = i;
} else {
if (EVFLAG) key++;
fdx = delx*cforce;
fdy = dely*cforce;
fdz = delz*cforce;
fOx = fdx*(1 - alpha);
fOy = fdy*(1 - alpha);
fOz = fdz*(1 - alpha);
fHx = 0.5*alpha * fdx;
fHy = 0.5*alpha * fdy;
fHz = 0.5*alpha * fdz;
fxtmp += fOx;
fytmp += fOy;
fztmp += fOz;
f[iH1].x += fHx;
f[iH1].y += fHy;
f[iH1].z += fHz;
f[iH2].x += fHx;
f[iH2].y += fHy;
f[iH2].z += fHz;
if (VFLAG) {
domain->closest_image(&x[i].x,&x[iH1].x,xH1);
domain->closest_image(&x[i].x,&x[iH2].x,xH2);
v[0] = x[i].x*fOx + xH1[0]*fHx + xH2[0]*fHx;
v[1] = x[i].y*fOy + xH1[1]*fHy + xH2[1]*fHy;
v[2] = x[i].z*fOz + xH1[2]*fHz + xH2[2]*fHz;
v[3] = x[i].x*fOy + xH1[0]*fHy + xH2[0]*fHy;
v[4] = x[i].x*fOz + xH1[0]*fHz + xH2[0]*fHz;
v[5] = x[i].y*fOz + xH1[1]*fHz + xH2[1]*fHz;
}
if (EVFLAG) {
vlist[n++] = i;
vlist[n++] = iH1;
vlist[n++] = iH2;
}
}
if (jtype != typeO) {
f[j].x -= delx * cforce;
f[j].y -= dely * cforce;
f[j].z -= delz * cforce;
if (VFLAG) {
v[0] -= x[j].x * delx * cforce;
v[1] -= x[j].y * dely * cforce;
v[2] -= x[j].z * delz * cforce;
v[3] -= x[j].x * dely * cforce;
v[4] -= x[j].x * delz * cforce;
v[5] -= x[j].y * delz * cforce;
}
if (EVFLAG) vlist[n++] = j;
} else {
if (EVFLAG) key += 2;
fdx = -delx*cforce;
fdy = -dely*cforce;
fdz = -delz*cforce;
fOx = fdx*(1 - alpha);
fOy = fdy*(1 - alpha);
fOz = fdz*(1 - alpha);
fHx = 0.5*alpha * fdx;
fHy = 0.5*alpha * fdy;
fHz = 0.5*alpha * fdz;
f[j].x += fOx;
f[j].y += fOy;
f[j].z += fOz;
f[jH1].x += fHx;
f[jH1].y += fHy;
f[jH1].z += fHz;
f[jH2].x += fHx;
f[jH2].y += fHy;
f[jH2].z += fHz;
if (VFLAG) {
domain->closest_image(&x[j].x,&x[jH1].x,xH1);
domain->closest_image(&x[j].x,&x[jH2].x,xH2);
v[0] += x[j].x*fOx + xH1[0]*fHx + xH2[0]*fHx;
v[1] += x[j].y*fOy + xH1[1]*fHy + xH2[1]*fHy;
v[2] += x[j].z*fOz + xH1[2]*fHz + xH2[2]*fHz;
v[3] += x[j].x*fOy + xH1[0]*fHy + xH2[0]*fHy;
v[4] += x[j].x*fOz + xH1[0]*fHz + xH2[0]*fHz;
v[5] += x[j].y*fOz + xH1[1]*fHz + xH2[1]*fHz;
}
if (EVFLAG) {
vlist[n++] = j;
vlist[n++] = jH1;
vlist[n++] = jH2;
}
}
if (EFLAG) {
ecoul = qqrd2e * qtmp * q[j] / r;
ecoul *= factor_coul;
} else ecoul = 0.0;
if (EVFLAG) ev_tally_list_thr(this,key,vlist,v,ecoul,alpha,thr);
}
}
}
f[i].x += fxtmp;
f[i].y += fytmp;
f[i].z += fztmp;
}
}
/* ----------------------------------------------------------------------
compute position xM of fictitious charge site for O atom and 2 H atoms
return it as xM
------------------------------------------------------------------------- */
void PairLJCutTIP4PCutOMP::compute_newsite_thr(const dbl3_t &xO,
const dbl3_t &xH1,
const dbl3_t &xH2,
dbl3_t &xM) const
{
double delx1 = xH1.x - xO.x;
double dely1 = xH1.y - xO.y;
double delz1 = xH1.z - xO.z;
domain->minimum_image(delx1,dely1,delz1);
double delx2 = xH2.x - xO.x;
double dely2 = xH2.y - xO.y;
double delz2 = xH2.z - xO.z;
domain->minimum_image(delx2,dely2,delz2);
const double prefac = alpha * 0.5;
xM.x = xO.x + prefac * (delx1 + delx2);
xM.y = xO.y + prefac * (dely1 + dely2);
xM.z = xO.z + prefac * (delz1 + delz2);
}
/* ---------------------------------------------------------------------- */
double PairLJCutTIP4PCutOMP::memory_usage()
{
double bytes = memory_usage_thr();
bytes += PairLJCutTIP4PCut::memory_usage();
return bytes;
}
c4science · Help