src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecEwSh_VdwNone_GeomP1P1_c.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013, 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 #ifdef HAVE_CONFIG_H
  39 #include <config.h>
  40 #endif
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "types/simple.h"
  46 #include "vec.h"
  47 #include "nrnb.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwNone_GeomP1P1_VF_c
  51  * Electrostatics interaction: Ewald
  52  * VdW interaction:            None
  53  * Geometry:                   Particle-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecEwSh_VdwNone_GeomP1P1_VF_c
  58                     (t_nblist                    * gmx_restrict       nlist,
  59                      rvec                        * gmx_restrict          xx,
  60                      rvec                        * gmx_restrict          ff,
  61                      t_forcerec                  * gmx_restrict          fr,
  62                      t_mdatoms                   * gmx_restrict     mdatoms,
  63                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  64                      t_nrnb                      * gmx_restrict        nrnb)
  65 {
  66     int              i_shift_offset,i_coord_offset,j_coord_offset;
  67     int              j_index_start,j_index_end;
  68     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
  69     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
  70     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  71     real             *shiftvec,*fshift,*x,*f;
  72     int              vdwioffset0;
  73     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  74     int              vdwjidx0;
  75     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  76     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
  77     real             velec,felec,velecsum,facel,crf,krf,krf2;
  78     real             *charge;
  79     int              ewitab;
  80     real             ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
  81     real             *ewtab;
  82
  83     x                = xx[0];
  84     f                = ff[0];
  85
  86     nri              = nlist->nri;
  87     iinr             = nlist->iinr;
  88     jindex           = nlist->jindex;
  89     jjnr             = nlist->jjnr;
  90     shiftidx         = nlist->shift;
  91     gid              = nlist->gid;
  92     shiftvec         = fr->shift_vec[0];
  93     fshift           = fr->fshift[0];
  94     facel            = fr->epsfac;
  95     charge           = mdatoms->chargeA;
  96
  97     sh_ewald         = fr->ic->sh_ewald;
  98     ewtab            = fr->ic->tabq_coul_FDV0;
  99     ewtabscale       = fr->ic->tabq_scale;
 100     ewtabhalfspace   = 0.5/ewtabscale;
 101
 102     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 103     rcutoff          = fr->rcoulomb;
 104     rcutoff2         = rcutoff*rcutoff;
 105
 106     outeriter        = 0;
 107     inneriter        = 0;
 108
 109     /* Start outer loop over neighborlists */
 110     for(iidx=0; iidx<nri; iidx++)
 111     {
 112         /* Load shift vector for this list */
 113         i_shift_offset   = DIM*shiftidx[iidx];
 114         shX              = shiftvec[i_shift_offset+XX];
 115         shY              = shiftvec[i_shift_offset+YY];
 116         shZ              = shiftvec[i_shift_offset+ZZ];
 117
 118         /* Load limits for loop over neighbors */
 119         j_index_start    = jindex[iidx];
 120         j_index_end      = jindex[iidx+1];
 121
 122         /* Get outer coordinate index */
 123         inr              = iinr[iidx];
 124         i_coord_offset   = DIM*inr;
 125
 126         /* Load i particle coords and add shift vector */
 127         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 128         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 129         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 130
 131         fix0             = 0.0;
 132         fiy0             = 0.0;
 133         fiz0             = 0.0;
 134
 135         /* Load parameters for i particles */
 136         iq0              = facel*charge[inr+0];
 137
 138         /* Reset potential sums */
 139         velecsum         = 0.0;
 140
 141         /* Start inner kernel loop */
 142         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 143         {
 144             /* Get j neighbor index, and coordinate index */
 145             jnr              = jjnr[jidx];
 146             j_coord_offset   = DIM*jnr;
 147
 148             /* load j atom coordinates */
 149             jx0              = x[j_coord_offset+DIM*0+XX];
 150             jy0              = x[j_coord_offset+DIM*0+YY];
 151             jz0              = x[j_coord_offset+DIM*0+ZZ];
 152
 153             /* Calculate displacement vector */
 154             dx00             = ix0 - jx0;
 155             dy00             = iy0 - jy0;
 156             dz00             = iz0 - jz0;
 157
 158             /* Calculate squared distance and things based on it */
 159             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 160
 161             rinv00           = gmx_invsqrt(rsq00);
 162
 163             rinvsq00         = rinv00*rinv00;
 164
 165             /* Load parameters for j particles */
 166             jq0              = charge[jnr+0];
 167
 168             /**************************
 169              * CALCULATE INTERACTIONS *
 170              **************************/
 171
 172             if (rsq00<rcutoff2)
 173             {
 174
 175             r00              = rsq00*rinv00;
 176
 177             qq00             = iq0*jq0;
 178
 179             /* EWALD ELECTROSTATICS */
 180
 181             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 182             ewrt             = r00*ewtabscale;
 183             ewitab           = ewrt;
 184             eweps            = ewrt-ewitab;
 185             ewitab           = 4*ewitab;
 186             felec            = ewtab[ewitab]+eweps*ewtab[ewitab+1];
 187             velec            = qq00*((rinv00-sh_ewald)-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
 188             felec            = qq00*rinv00*(rinvsq00-felec);
 189
 190             /* Update potential sums from outer loop */
 191             velecsum        += velec;
 192
 193             fscal            = felec;
 194
 195             /* Calculate temporary vectorial force */
 196             tx               = fscal*dx00;
 197             ty               = fscal*dy00;
 198             tz               = fscal*dz00;
 199
 200             /* Update vectorial force */
 201             fix0            += tx;
 202             fiy0            += ty;
 203             fiz0            += tz;
 204             f[j_coord_offset+DIM*0+XX] -= tx;
 205             f[j_coord_offset+DIM*0+YY] -= ty;
 206             f[j_coord_offset+DIM*0+ZZ] -= tz;
 207
 208             }
 209
 210             /* Inner loop uses 42 flops */
 211         }
 212         /* End of innermost loop */
 213
 214         tx = ty = tz = 0;
 215         f[i_coord_offset+DIM*0+XX] += fix0;
 216         f[i_coord_offset+DIM*0+YY] += fiy0;
 217         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 218         tx                         += fix0;
 219         ty                         += fiy0;
 220         tz                         += fiz0;
 221         fshift[i_shift_offset+XX]  += tx;
 222         fshift[i_shift_offset+YY]  += ty;
 223         fshift[i_shift_offset+ZZ]  += tz;
 224
 225         ggid                        = gid[iidx];
 226         /* Update potential energies */
 227         kernel_data->energygrp_elec[ggid] += velecsum;
 228
 229         /* Increment number of inner iterations */
 230         inneriter                  += j_index_end - j_index_start;
 231
 232         /* Outer loop uses 14 flops */
 233     }
 234
 235     /* Increment number of outer iterations */
 236     outeriter        += nri;
 237
 238     /* Update outer/inner flops */
 239
 240     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VF,outeriter*14 + inneriter*42);
 241 }
 242 /*
 243  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwNone_GeomP1P1_F_c
 244  * Electrostatics interaction: Ewald
 245  * VdW interaction:            None
 246  * Geometry:                   Particle-Particle
 247  * Calculate force/pot:        Force
 248  */
 249 void
 250 nb_kernel_ElecEwSh_VdwNone_GeomP1P1_F_c
 251                     (t_nblist                    * gmx_restrict       nlist,
 252                      rvec                        * gmx_restrict          xx,
 253                      rvec                        * gmx_restrict          ff,
 254                      t_forcerec                  * gmx_restrict          fr,
 255                      t_mdatoms                   * gmx_restrict     mdatoms,
 256                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 257                      t_nrnb                      * gmx_restrict        nrnb)
 258 {
 259     int              i_shift_offset,i_coord_offset,j_coord_offset;
 260     int              j_index_start,j_index_end;
 261     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 262     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 263     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 264     real             *shiftvec,*fshift,*x,*f;
 265     int              vdwioffset0;
 266     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 267     int              vdwjidx0;
 268     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 269     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 270     real             velec,felec,velecsum,facel,crf,krf,krf2;
 271     real             *charge;
 272     int              ewitab;
 273     real             ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
 274     real             *ewtab;
 275
 276     x                = xx[0];
 277     f                = ff[0];
 278
 279     nri              = nlist->nri;
 280     iinr             = nlist->iinr;
 281     jindex           = nlist->jindex;
 282     jjnr             = nlist->jjnr;
 283     shiftidx         = nlist->shift;
 284     gid              = nlist->gid;
 285     shiftvec         = fr->shift_vec[0];
 286     fshift           = fr->fshift[0];
 287     facel            = fr->epsfac;
 288     charge           = mdatoms->chargeA;
 289
 290     sh_ewald         = fr->ic->sh_ewald;
 291     ewtab            = fr->ic->tabq_coul_F;
 292     ewtabscale       = fr->ic->tabq_scale;
 293     ewtabhalfspace   = 0.5/ewtabscale;
 294
 295     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 296     rcutoff          = fr->rcoulomb;
 297     rcutoff2         = rcutoff*rcutoff;
 298
 299     outeriter        = 0;
 300     inneriter        = 0;
 301
 302     /* Start outer loop over neighborlists */
 303     for(iidx=0; iidx<nri; iidx++)
 304     {
 305         /* Load shift vector for this list */
 306         i_shift_offset   = DIM*shiftidx[iidx];
 307         shX              = shiftvec[i_shift_offset+XX];
 308         shY              = shiftvec[i_shift_offset+YY];
 309         shZ              = shiftvec[i_shift_offset+ZZ];
 310
 311         /* Load limits for loop over neighbors */
 312         j_index_start    = jindex[iidx];
 313         j_index_end      = jindex[iidx+1];
 314
 315         /* Get outer coordinate index */
 316         inr              = iinr[iidx];
 317         i_coord_offset   = DIM*inr;
 318
 319         /* Load i particle coords and add shift vector */
 320         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 321         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 322         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 323
 324         fix0             = 0.0;
 325         fiy0             = 0.0;
 326         fiz0             = 0.0;
 327
 328         /* Load parameters for i particles */
 329         iq0              = facel*charge[inr+0];
 330
 331         /* Start inner kernel loop */
 332         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 333         {
 334             /* Get j neighbor index, and coordinate index */
 335             jnr              = jjnr[jidx];
 336             j_coord_offset   = DIM*jnr;
 337
 338             /* load j atom coordinates */
 339             jx0              = x[j_coord_offset+DIM*0+XX];
 340             jy0              = x[j_coord_offset+DIM*0+YY];
 341             jz0              = x[j_coord_offset+DIM*0+ZZ];
 342
 343             /* Calculate displacement vector */
 344             dx00             = ix0 - jx0;
 345             dy00             = iy0 - jy0;
 346             dz00             = iz0 - jz0;
 347
 348             /* Calculate squared distance and things based on it */
 349             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 350
 351             rinv00           = gmx_invsqrt(rsq00);
 352
 353             rinvsq00         = rinv00*rinv00;
 354
 355             /* Load parameters for j particles */
 356             jq0              = charge[jnr+0];
 357
 358             /**************************
 359              * CALCULATE INTERACTIONS *
 360              **************************/
 361
 362             if (rsq00<rcutoff2)
 363             {
 364
 365             r00              = rsq00*rinv00;
 366
 367             qq00             = iq0*jq0;
 368
 369             /* EWALD ELECTROSTATICS */
 370
 371             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 372             ewrt             = r00*ewtabscale;
 373             ewitab           = ewrt;
 374             eweps            = ewrt-ewitab;
 375             felec            = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
 376             felec            = qq00*rinv00*(rinvsq00-felec);
 377
 378             fscal            = felec;
 379
 380             /* Calculate temporary vectorial force */
 381             tx               = fscal*dx00;
 382             ty               = fscal*dy00;
 383             tz               = fscal*dz00;
 384
 385             /* Update vectorial force */
 386             fix0            += tx;
 387             fiy0            += ty;
 388             fiz0            += tz;
 389             f[j_coord_offset+DIM*0+XX] -= tx;
 390             f[j_coord_offset+DIM*0+YY] -= ty;
 391             f[j_coord_offset+DIM*0+ZZ] -= tz;
 392
 393             }
 394
 395             /* Inner loop uses 34 flops */
 396         }
 397         /* End of innermost loop */
 398
 399         tx = ty = tz = 0;
 400         f[i_coord_offset+DIM*0+XX] += fix0;
 401         f[i_coord_offset+DIM*0+YY] += fiy0;
 402         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 403         tx                         += fix0;
 404         ty                         += fiy0;
 405         tz                         += fiz0;
 406         fshift[i_shift_offset+XX]  += tx;
 407         fshift[i_shift_offset+YY]  += ty;
 408         fshift[i_shift_offset+ZZ]  += tz;
 409
 410         /* Increment number of inner iterations */
 411         inneriter                  += j_index_end - j_index_start;
 412
 413         /* Outer loop uses 13 flops */
 414     }
 415
 416     /* Increment number of outer iterations */
 417     outeriter        += nri;
 418
 419     /* Update outer/inner flops */
 420
 421     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_F,outeriter*13 + inneriter*34);
 422 }