src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCoul_VdwLJ_GeomW4P1_c.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS c kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "nrnb.h"
  46
  47 /*
  48  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwLJ_GeomW4P1_VF_c
  49  * Electrostatics interaction: Coulomb
  50  * VdW interaction:            LennardJones
  51  * Geometry:                   Water4-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecCoul_VdwLJ_GeomW4P1_VF_c
  56                     (t_nblist                    * gmx_restrict       nlist,
  57                      rvec                        * gmx_restrict          xx,
  58                      rvec                        * gmx_restrict          ff,
  59                      t_forcerec                  * gmx_restrict          fr,
  60                      t_mdatoms                   * gmx_restrict     mdatoms,
  61                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  62                      t_nrnb                      * gmx_restrict        nrnb)
  63 {
  64     int              i_shift_offset,i_coord_offset,j_coord_offset;
  65     int              j_index_start,j_index_end;
  66     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
  67     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
  68     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  69     real             *shiftvec,*fshift,*x,*f;
  70     int              vdwioffset0;
  71     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  72     int              vdwioffset1;
  73     real             ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  74     int              vdwioffset2;
  75     real             ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  76     int              vdwioffset3;
  77     real             ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  78     int              vdwjidx0;
  79     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  80     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
  81     real             dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
  82     real             dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
  83     real             dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
  84     real             velec,felec,velecsum,facel,crf,krf,krf2;
  85     real             *charge;
  86     int              nvdwtype;
  87     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  88     int              *vdwtype;
  89     real             *vdwparam;
  90
  91     x                = xx[0];
  92     f                = ff[0];
  93
  94     nri              = nlist->nri;
  95     iinr             = nlist->iinr;
  96     jindex           = nlist->jindex;
  97     jjnr             = nlist->jjnr;
  98     shiftidx         = nlist->shift;
  99     gid              = nlist->gid;
 100     shiftvec         = fr->shift_vec[0];
 101     fshift           = fr->fshift[0];
 102     facel            = fr->epsfac;
 103     charge           = mdatoms->chargeA;
 104     nvdwtype         = fr->ntype;
 105     vdwparam         = fr->nbfp;
 106     vdwtype          = mdatoms->typeA;
 107
 108     /* Setup water-specific parameters */
 109     inr              = nlist->iinr[0];
 110     iq1              = facel*charge[inr+1];
 111     iq2              = facel*charge[inr+2];
 112     iq3              = facel*charge[inr+3];
 113     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 114
 115     outeriter        = 0;
 116     inneriter        = 0;
 117
 118     /* Start outer loop over neighborlists */
 119     for(iidx=0; iidx<nri; iidx++)
 120     {
 121         /* Load shift vector for this list */
 122         i_shift_offset   = DIM*shiftidx[iidx];
 123         shX              = shiftvec[i_shift_offset+XX];
 124         shY              = shiftvec[i_shift_offset+YY];
 125         shZ              = shiftvec[i_shift_offset+ZZ];
 126
 127         /* Load limits for loop over neighbors */
 128         j_index_start    = jindex[iidx];
 129         j_index_end      = jindex[iidx+1];
 130
 131         /* Get outer coordinate index */
 132         inr              = iinr[iidx];
 133         i_coord_offset   = DIM*inr;
 134
 135         /* Load i particle coords and add shift vector */
 136         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 137         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 138         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 139         ix1              = shX + x[i_coord_offset+DIM*1+XX];
 140         iy1              = shY + x[i_coord_offset+DIM*1+YY];
 141         iz1              = shZ + x[i_coord_offset+DIM*1+ZZ];
 142         ix2              = shX + x[i_coord_offset+DIM*2+XX];
 143         iy2              = shY + x[i_coord_offset+DIM*2+YY];
 144         iz2              = shZ + x[i_coord_offset+DIM*2+ZZ];
 145         ix3              = shX + x[i_coord_offset+DIM*3+XX];
 146         iy3              = shY + x[i_coord_offset+DIM*3+YY];
 147         iz3              = shZ + x[i_coord_offset+DIM*3+ZZ];
 148
 149         fix0             = 0.0;
 150         fiy0             = 0.0;
 151         fiz0             = 0.0;
 152         fix1             = 0.0;
 153         fiy1             = 0.0;
 154         fiz1             = 0.0;
 155         fix2             = 0.0;
 156         fiy2             = 0.0;
 157         fiz2             = 0.0;
 158         fix3             = 0.0;
 159         fiy3             = 0.0;
 160         fiz3             = 0.0;
 161
 162         /* Reset potential sums */
 163         velecsum         = 0.0;
 164         vvdwsum          = 0.0;
 165
 166         /* Start inner kernel loop */
 167         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 168         {
 169             /* Get j neighbor index, and coordinate index */
 170             jnr              = jjnr[jidx];
 171             j_coord_offset   = DIM*jnr;
 172
 173             /* load j atom coordinates */
 174             jx0              = x[j_coord_offset+DIM*0+XX];
 175             jy0              = x[j_coord_offset+DIM*0+YY];
 176             jz0              = x[j_coord_offset+DIM*0+ZZ];
 177
 178             /* Calculate displacement vector */
 179             dx00             = ix0 - jx0;
 180             dy00             = iy0 - jy0;
 181             dz00             = iz0 - jz0;
 182             dx10             = ix1 - jx0;
 183             dy10             = iy1 - jy0;
 184             dz10             = iz1 - jz0;
 185             dx20             = ix2 - jx0;
 186             dy20             = iy2 - jy0;
 187             dz20             = iz2 - jz0;
 188             dx30             = ix3 - jx0;
 189             dy30             = iy3 - jy0;
 190             dz30             = iz3 - jz0;
 191
 192             /* Calculate squared distance and things based on it */
 193             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 194             rsq10            = dx10*dx10+dy10*dy10+dz10*dz10;
 195             rsq20            = dx20*dx20+dy20*dy20+dz20*dz20;
 196             rsq30            = dx30*dx30+dy30*dy30+dz30*dz30;
 197
 198             rinv10           = gmx_invsqrt(rsq10);
 199             rinv20           = gmx_invsqrt(rsq20);
 200             rinv30           = gmx_invsqrt(rsq30);
 201
 202             rinvsq00         = 1.0/rsq00;
 203             rinvsq10         = rinv10*rinv10;
 204             rinvsq20         = rinv20*rinv20;
 205             rinvsq30         = rinv30*rinv30;
 206
 207             /* Load parameters for j particles */
 208             jq0              = charge[jnr+0];
 209             vdwjidx0         = 2*vdwtype[jnr+0];
 210
 211             /**************************
 212              * CALCULATE INTERACTIONS *
 213              **************************/
 214
 215             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 216             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 217
 218             /* LENNARD-JONES DISPERSION/REPULSION */
 219
 220             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 221             vvdw6            = c6_00*rinvsix;
 222             vvdw12           = c12_00*rinvsix*rinvsix;
 223             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 224             fvdw             = (vvdw12-vvdw6)*rinvsq00;
 225
 226             /* Update potential sums from outer loop */
 227             vvdwsum         += vvdw;
 228
 229             fscal            = fvdw;
 230
 231             /* Calculate temporary vectorial force */
 232             tx               = fscal*dx00;
 233             ty               = fscal*dy00;
 234             tz               = fscal*dz00;
 235
 236             /* Update vectorial force */
 237             fix0            += tx;
 238             fiy0            += ty;
 239             fiz0            += tz;
 240             f[j_coord_offset+DIM*0+XX] -= tx;
 241             f[j_coord_offset+DIM*0+YY] -= ty;
 242             f[j_coord_offset+DIM*0+ZZ] -= tz;
 243
 244             /**************************
 245              * CALCULATE INTERACTIONS *
 246              **************************/
 247
 248             qq10             = iq1*jq0;
 249
 250             /* COULOMB ELECTROSTATICS */
 251             velec            = qq10*rinv10;
 252             felec            = velec*rinvsq10;
 253
 254             /* Update potential sums from outer loop */
 255             velecsum        += velec;
 256
 257             fscal            = felec;
 258
 259             /* Calculate temporary vectorial force */
 260             tx               = fscal*dx10;
 261             ty               = fscal*dy10;
 262             tz               = fscal*dz10;
 263
 264             /* Update vectorial force */
 265             fix1            += tx;
 266             fiy1            += ty;
 267             fiz1            += tz;
 268             f[j_coord_offset+DIM*0+XX] -= tx;
 269             f[j_coord_offset+DIM*0+YY] -= ty;
 270             f[j_coord_offset+DIM*0+ZZ] -= tz;
 271
 272             /**************************
 273              * CALCULATE INTERACTIONS *
 274              **************************/
 275
 276             qq20             = iq2*jq0;
 277
 278             /* COULOMB ELECTROSTATICS */
 279             velec            = qq20*rinv20;
 280             felec            = velec*rinvsq20;
 281
 282             /* Update potential sums from outer loop */
 283             velecsum        += velec;
 284
 285             fscal            = felec;
 286
 287             /* Calculate temporary vectorial force */
 288             tx               = fscal*dx20;
 289             ty               = fscal*dy20;
 290             tz               = fscal*dz20;
 291
 292             /* Update vectorial force */
 293             fix2            += tx;
 294             fiy2            += ty;
 295             fiz2            += tz;
 296             f[j_coord_offset+DIM*0+XX] -= tx;
 297             f[j_coord_offset+DIM*0+YY] -= ty;
 298             f[j_coord_offset+DIM*0+ZZ] -= tz;
 299
 300             /**************************
 301              * CALCULATE INTERACTIONS *
 302              **************************/
 303
 304             qq30             = iq3*jq0;
 305
 306             /* COULOMB ELECTROSTATICS */
 307             velec            = qq30*rinv30;
 308             felec            = velec*rinvsq30;
 309
 310             /* Update potential sums from outer loop */
 311             velecsum        += velec;
 312
 313             fscal            = felec;
 314
 315             /* Calculate temporary vectorial force */
 316             tx               = fscal*dx30;
 317             ty               = fscal*dy30;
 318             tz               = fscal*dz30;
 319
 320             /* Update vectorial force */
 321             fix3            += tx;
 322             fiy3            += ty;
 323             fiz3            += tz;
 324             f[j_coord_offset+DIM*0+XX] -= tx;
 325             f[j_coord_offset+DIM*0+YY] -= ty;
 326             f[j_coord_offset+DIM*0+ZZ] -= tz;
 327
 328             /* Inner loop uses 116 flops */
 329         }
 330         /* End of innermost loop */
 331
 332         tx = ty = tz = 0;
 333         f[i_coord_offset+DIM*0+XX] += fix0;
 334         f[i_coord_offset+DIM*0+YY] += fiy0;
 335         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 336         tx                         += fix0;
 337         ty                         += fiy0;
 338         tz                         += fiz0;
 339         f[i_coord_offset+DIM*1+XX] += fix1;
 340         f[i_coord_offset+DIM*1+YY] += fiy1;
 341         f[i_coord_offset+DIM*1+ZZ] += fiz1;
 342         tx                         += fix1;
 343         ty                         += fiy1;
 344         tz                         += fiz1;
 345         f[i_coord_offset+DIM*2+XX] += fix2;
 346         f[i_coord_offset+DIM*2+YY] += fiy2;
 347         f[i_coord_offset+DIM*2+ZZ] += fiz2;
 348         tx                         += fix2;
 349         ty                         += fiy2;
 350         tz                         += fiz2;
 351         f[i_coord_offset+DIM*3+XX] += fix3;
 352         f[i_coord_offset+DIM*3+YY] += fiy3;
 353         f[i_coord_offset+DIM*3+ZZ] += fiz3;
 354         tx                         += fix3;
 355         ty                         += fiy3;
 356         tz                         += fiz3;
 357         fshift[i_shift_offset+XX]  += tx;
 358         fshift[i_shift_offset+YY]  += ty;
 359         fshift[i_shift_offset+ZZ]  += tz;
 360
 361         ggid                        = gid[iidx];
 362         /* Update potential energies */
 363         kernel_data->energygrp_elec[ggid] += velecsum;
 364         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 365
 366         /* Increment number of inner iterations */
 367         inneriter                  += j_index_end - j_index_start;
 368
 369         /* Outer loop uses 41 flops */
 370     }
 371
 372     /* Increment number of outer iterations */
 373     outeriter        += nri;
 374
 375     /* Update outer/inner flops */
 376
 377     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*41 + inneriter*116);
 378 }
 379 /*
 380  * Gromacs nonbonded kernel:   nb_kernel_ElecCoul_VdwLJ_GeomW4P1_F_c
 381  * Electrostatics interaction: Coulomb
 382  * VdW interaction:            LennardJones
 383  * Geometry:                   Water4-Particle
 384  * Calculate force/pot:        Force
 385  */
 386 void
 387 nb_kernel_ElecCoul_VdwLJ_GeomW4P1_F_c
 388                     (t_nblist                    * gmx_restrict       nlist,
 389                      rvec                        * gmx_restrict          xx,
 390                      rvec                        * gmx_restrict          ff,
 391                      t_forcerec                  * gmx_restrict          fr,
 392                      t_mdatoms                   * gmx_restrict     mdatoms,
 393                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 394                      t_nrnb                      * gmx_restrict        nrnb)
 395 {
 396     int              i_shift_offset,i_coord_offset,j_coord_offset;
 397     int              j_index_start,j_index_end;
 398     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 399     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 400     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 401     real             *shiftvec,*fshift,*x,*f;
 402     int              vdwioffset0;
 403     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 404     int              vdwioffset1;
 405     real             ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 406     int              vdwioffset2;
 407     real             ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 408     int              vdwioffset3;
 409     real             ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 410     int              vdwjidx0;
 411     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 412     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 413     real             dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
 414     real             dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
 415     real             dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
 416     real             velec,felec,velecsum,facel,crf,krf,krf2;
 417     real             *charge;
 418     int              nvdwtype;
 419     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 420     int              *vdwtype;
 421     real             *vdwparam;
 422
 423     x                = xx[0];
 424     f                = ff[0];
 425
 426     nri              = nlist->nri;
 427     iinr             = nlist->iinr;
 428     jindex           = nlist->jindex;
 429     jjnr             = nlist->jjnr;
 430     shiftidx         = nlist->shift;
 431     gid              = nlist->gid;
 432     shiftvec         = fr->shift_vec[0];
 433     fshift           = fr->fshift[0];
 434     facel            = fr->epsfac;
 435     charge           = mdatoms->chargeA;
 436     nvdwtype         = fr->ntype;
 437     vdwparam         = fr->nbfp;
 438     vdwtype          = mdatoms->typeA;
 439
 440     /* Setup water-specific parameters */
 441     inr              = nlist->iinr[0];
 442     iq1              = facel*charge[inr+1];
 443     iq2              = facel*charge[inr+2];
 444     iq3              = facel*charge[inr+3];
 445     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 446
 447     outeriter        = 0;
 448     inneriter        = 0;
 449
 450     /* Start outer loop over neighborlists */
 451     for(iidx=0; iidx<nri; iidx++)
 452     {
 453         /* Load shift vector for this list */
 454         i_shift_offset   = DIM*shiftidx[iidx];
 455         shX              = shiftvec[i_shift_offset+XX];
 456         shY              = shiftvec[i_shift_offset+YY];
 457         shZ              = shiftvec[i_shift_offset+ZZ];
 458
 459         /* Load limits for loop over neighbors */
 460         j_index_start    = jindex[iidx];
 461         j_index_end      = jindex[iidx+1];
 462
 463         /* Get outer coordinate index */
 464         inr              = iinr[iidx];
 465         i_coord_offset   = DIM*inr;
 466
 467         /* Load i particle coords and add shift vector */
 468         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 469         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 470         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 471         ix1              = shX + x[i_coord_offset+DIM*1+XX];
 472         iy1              = shY + x[i_coord_offset+DIM*1+YY];
 473         iz1              = shZ + x[i_coord_offset+DIM*1+ZZ];
 474         ix2              = shX + x[i_coord_offset+DIM*2+XX];
 475         iy2              = shY + x[i_coord_offset+DIM*2+YY];
 476         iz2              = shZ + x[i_coord_offset+DIM*2+ZZ];
 477         ix3              = shX + x[i_coord_offset+DIM*3+XX];
 478         iy3              = shY + x[i_coord_offset+DIM*3+YY];
 479         iz3              = shZ + x[i_coord_offset+DIM*3+ZZ];
 480
 481         fix0             = 0.0;
 482         fiy0             = 0.0;
 483         fiz0             = 0.0;
 484         fix1             = 0.0;
 485         fiy1             = 0.0;
 486         fiz1             = 0.0;
 487         fix2             = 0.0;
 488         fiy2             = 0.0;
 489         fiz2             = 0.0;
 490         fix3             = 0.0;
 491         fiy3             = 0.0;
 492         fiz3             = 0.0;
 493
 494         /* Start inner kernel loop */
 495         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 496         {
 497             /* Get j neighbor index, and coordinate index */
 498             jnr              = jjnr[jidx];
 499             j_coord_offset   = DIM*jnr;
 500
 501             /* load j atom coordinates */
 502             jx0              = x[j_coord_offset+DIM*0+XX];
 503             jy0              = x[j_coord_offset+DIM*0+YY];
 504             jz0              = x[j_coord_offset+DIM*0+ZZ];
 505
 506             /* Calculate displacement vector */
 507             dx00             = ix0 - jx0;
 508             dy00             = iy0 - jy0;
 509             dz00             = iz0 - jz0;
 510             dx10             = ix1 - jx0;
 511             dy10             = iy1 - jy0;
 512             dz10             = iz1 - jz0;
 513             dx20             = ix2 - jx0;
 514             dy20             = iy2 - jy0;
 515             dz20             = iz2 - jz0;
 516             dx30             = ix3 - jx0;
 517             dy30             = iy3 - jy0;
 518             dz30             = iz3 - jz0;
 519
 520             /* Calculate squared distance and things based on it */
 521             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 522             rsq10            = dx10*dx10+dy10*dy10+dz10*dz10;
 523             rsq20            = dx20*dx20+dy20*dy20+dz20*dz20;
 524             rsq30            = dx30*dx30+dy30*dy30+dz30*dz30;
 525
 526             rinv10           = gmx_invsqrt(rsq10);
 527             rinv20           = gmx_invsqrt(rsq20);
 528             rinv30           = gmx_invsqrt(rsq30);
 529
 530             rinvsq00         = 1.0/rsq00;
 531             rinvsq10         = rinv10*rinv10;
 532             rinvsq20         = rinv20*rinv20;
 533             rinvsq30         = rinv30*rinv30;
 534
 535             /* Load parameters for j particles */
 536             jq0              = charge[jnr+0];
 537             vdwjidx0         = 2*vdwtype[jnr+0];
 538
 539             /**************************
 540              * CALCULATE INTERACTIONS *
 541              **************************/
 542
 543             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 544             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 545
 546             /* LENNARD-JONES DISPERSION/REPULSION */
 547
 548             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 549             fvdw             = (c12_00*rinvsix-c6_00)*rinvsix*rinvsq00;
 550
 551             fscal            = fvdw;
 552
 553             /* Calculate temporary vectorial force */
 554             tx               = fscal*dx00;
 555             ty               = fscal*dy00;
 556             tz               = fscal*dz00;
 557
 558             /* Update vectorial force */
 559             fix0            += tx;
 560             fiy0            += ty;
 561             fiz0            += tz;
 562             f[j_coord_offset+DIM*0+XX] -= tx;
 563             f[j_coord_offset+DIM*0+YY] -= ty;
 564             f[j_coord_offset+DIM*0+ZZ] -= tz;
 565
 566             /**************************
 567              * CALCULATE INTERACTIONS *
 568              **************************/
 569
 570             qq10             = iq1*jq0;
 571
 572             /* COULOMB ELECTROSTATICS */
 573             velec            = qq10*rinv10;
 574             felec            = velec*rinvsq10;
 575
 576             fscal            = felec;
 577
 578             /* Calculate temporary vectorial force */
 579             tx               = fscal*dx10;
 580             ty               = fscal*dy10;
 581             tz               = fscal*dz10;
 582
 583             /* Update vectorial force */
 584             fix1            += tx;
 585             fiy1            += ty;
 586             fiz1            += tz;
 587             f[j_coord_offset+DIM*0+XX] -= tx;
 588             f[j_coord_offset+DIM*0+YY] -= ty;
 589             f[j_coord_offset+DIM*0+ZZ] -= tz;
 590
 591             /**************************
 592              * CALCULATE INTERACTIONS *
 593              **************************/
 594
 595             qq20             = iq2*jq0;
 596
 597             /* COULOMB ELECTROSTATICS */
 598             velec            = qq20*rinv20;
 599             felec            = velec*rinvsq20;
 600
 601             fscal            = felec;
 602
 603             /* Calculate temporary vectorial force */
 604             tx               = fscal*dx20;
 605             ty               = fscal*dy20;
 606             tz               = fscal*dz20;
 607
 608             /* Update vectorial force */
 609             fix2            += tx;
 610             fiy2            += ty;
 611             fiz2            += tz;
 612             f[j_coord_offset+DIM*0+XX] -= tx;
 613             f[j_coord_offset+DIM*0+YY] -= ty;
 614             f[j_coord_offset+DIM*0+ZZ] -= tz;
 615
 616             /**************************
 617              * CALCULATE INTERACTIONS *
 618              **************************/
 619
 620             qq30             = iq3*jq0;
 621
 622             /* COULOMB ELECTROSTATICS */
 623             velec            = qq30*rinv30;
 624             felec            = velec*rinvsq30;
 625
 626             fscal            = felec;
 627
 628             /* Calculate temporary vectorial force */
 629             tx               = fscal*dx30;
 630             ty               = fscal*dy30;
 631             tz               = fscal*dz30;
 632
 633             /* Update vectorial force */
 634             fix3            += tx;
 635             fiy3            += ty;
 636             fiz3            += tz;
 637             f[j_coord_offset+DIM*0+XX] -= tx;
 638             f[j_coord_offset+DIM*0+YY] -= ty;
 639             f[j_coord_offset+DIM*0+ZZ] -= tz;
 640
 641             /* Inner loop uses 108 flops */
 642         }
 643         /* End of innermost loop */
 644
 645         tx = ty = tz = 0;
 646         f[i_coord_offset+DIM*0+XX] += fix0;
 647         f[i_coord_offset+DIM*0+YY] += fiy0;
 648         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 649         tx                         += fix0;
 650         ty                         += fiy0;
 651         tz                         += fiz0;
 652         f[i_coord_offset+DIM*1+XX] += fix1;
 653         f[i_coord_offset+DIM*1+YY] += fiy1;
 654         f[i_coord_offset+DIM*1+ZZ] += fiz1;
 655         tx                         += fix1;
 656         ty                         += fiy1;
 657         tz                         += fiz1;
 658         f[i_coord_offset+DIM*2+XX] += fix2;
 659         f[i_coord_offset+DIM*2+YY] += fiy2;
 660         f[i_coord_offset+DIM*2+ZZ] += fiz2;
 661         tx                         += fix2;
 662         ty                         += fiy2;
 663         tz                         += fiz2;
 664         f[i_coord_offset+DIM*3+XX] += fix3;
 665         f[i_coord_offset+DIM*3+YY] += fiy3;
 666         f[i_coord_offset+DIM*3+ZZ] += fiz3;
 667         tx                         += fix3;
 668         ty                         += fiy3;
 669         tz                         += fiz3;
 670         fshift[i_shift_offset+XX]  += tx;
 671         fshift[i_shift_offset+YY]  += ty;
 672         fshift[i_shift_offset+ZZ]  += tz;
 673
 674         /* Increment number of inner iterations */
 675         inneriter                  += j_index_end - j_index_start;
 676
 677         /* Outer loop uses 39 flops */
 678     }
 679
 680     /* Increment number of outer iterations */
 681     outeriter        += nri;
 682
 683     /* Update outer/inner flops */
 684
 685     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*39 + inneriter*108);
 686 }