src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecEw_VdwBham_GeomP1P1_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_ElecEw_VdwBham_GeomP1P1_VF_c
  49  * Electrostatics interaction: Ewald
  50  * VdW interaction:            Buckingham
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecEw_VdwBham_GeomP1P1_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              vdwjidx0;
  73     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  74     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
  75     real             velec,felec,velecsum,facel,crf,krf,krf2;
  76     real             *charge;
  77     int              nvdwtype;
  78     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  79     int              *vdwtype;
  80     real             *vdwparam;
  81     int              ewitab;
  82     real             ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
  83     real             *ewtab;
  84
  85     x                = xx[0];
  86     f                = ff[0];
  87
  88     nri              = nlist->nri;
  89     iinr             = nlist->iinr;
  90     jindex           = nlist->jindex;
  91     jjnr             = nlist->jjnr;
  92     shiftidx         = nlist->shift;
  93     gid              = nlist->gid;
  94     shiftvec         = fr->shift_vec[0];
  95     fshift           = fr->fshift[0];
  96     facel            = fr->epsfac;
  97     charge           = mdatoms->chargeA;
  98     nvdwtype         = fr->ntype;
  99     vdwparam         = fr->nbfp;
 100     vdwtype          = mdatoms->typeA;
 101
 102     sh_ewald         = fr->ic->sh_ewald;
 103     ewtab            = fr->ic->tabq_coul_FDV0;
 104     ewtabscale       = fr->ic->tabq_scale;
 105     ewtabhalfspace   = 0.5/ewtabscale;
 106
 107     outeriter        = 0;
 108     inneriter        = 0;
 109
 110     /* Start outer loop over neighborlists */
 111     for(iidx=0; iidx<nri; iidx++)
 112     {
 113         /* Load shift vector for this list */
 114         i_shift_offset   = DIM*shiftidx[iidx];
 115         shX              = shiftvec[i_shift_offset+XX];
 116         shY              = shiftvec[i_shift_offset+YY];
 117         shZ              = shiftvec[i_shift_offset+ZZ];
 118
 119         /* Load limits for loop over neighbors */
 120         j_index_start    = jindex[iidx];
 121         j_index_end      = jindex[iidx+1];
 122
 123         /* Get outer coordinate index */
 124         inr              = iinr[iidx];
 125         i_coord_offset   = DIM*inr;
 126
 127         /* Load i particle coords and add shift vector */
 128         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 129         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 130         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 131
 132         fix0             = 0.0;
 133         fiy0             = 0.0;
 134         fiz0             = 0.0;
 135
 136         /* Load parameters for i particles */
 137         iq0              = facel*charge[inr+0];
 138         vdwioffset0      = 3*nvdwtype*vdwtype[inr+0];
 139
 140         /* Reset potential sums */
 141         velecsum         = 0.0;
 142         vvdwsum          = 0.0;
 143
 144         /* Start inner kernel loop */
 145         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 146         {
 147             /* Get j neighbor index, and coordinate index */
 148             jnr              = jjnr[jidx];
 149             j_coord_offset   = DIM*jnr;
 150
 151             /* load j atom coordinates */
 152             jx0              = x[j_coord_offset+DIM*0+XX];
 153             jy0              = x[j_coord_offset+DIM*0+YY];
 154             jz0              = x[j_coord_offset+DIM*0+ZZ];
 155
 156             /* Calculate displacement vector */
 157             dx00             = ix0 - jx0;
 158             dy00             = iy0 - jy0;
 159             dz00             = iz0 - jz0;
 160
 161             /* Calculate squared distance and things based on it */
 162             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 163
 164             rinv00           = gmx_invsqrt(rsq00);
 165
 166             rinvsq00         = rinv00*rinv00;
 167
 168             /* Load parameters for j particles */
 169             jq0              = charge[jnr+0];
 170             vdwjidx0         = 3*vdwtype[jnr+0];
 171
 172             /**************************
 173              * CALCULATE INTERACTIONS *
 174              **************************/
 175
 176             r00              = rsq00*rinv00;
 177
 178             qq00             = iq0*jq0;
 179             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 180             cexp1_00         = vdwparam[vdwioffset0+vdwjidx0+1];
 181             cexp2_00         = vdwparam[vdwioffset0+vdwjidx0+2];
 182
 183             /* EWALD ELECTROSTATICS */
 184
 185             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 186             ewrt             = r00*ewtabscale;
 187             ewitab           = ewrt;
 188             eweps            = ewrt-ewitab;
 189             ewitab           = 4*ewitab;
 190             felec            = ewtab[ewitab]+eweps*ewtab[ewitab+1];
 191             velec            = qq00*(rinv00-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
 192             felec            = qq00*rinv00*(rinvsq00-felec);
 193
 194             /* BUCKINGHAM DISPERSION/REPULSION */
 195             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 196             vvdw6            = c6_00*rinvsix;
 197             br               = cexp2_00*r00;
 198             vvdwexp          = cexp1_00*exp(-br);
 199             vvdw             = vvdwexp - vvdw6*(1.0/6.0);
 200             fvdw             = (br*vvdwexp-vvdw6)*rinvsq00;
 201
 202             /* Update potential sums from outer loop */
 203             velecsum        += velec;
 204             vvdwsum         += vvdw;
 205
 206             fscal            = felec+fvdw;
 207
 208             /* Calculate temporary vectorial force */
 209             tx               = fscal*dx00;
 210             ty               = fscal*dy00;
 211             tz               = fscal*dz00;
 212
 213             /* Update vectorial force */
 214             fix0            += tx;
 215             fiy0            += ty;
 216             fiz0            += tz;
 217             f[j_coord_offset+DIM*0+XX] -= tx;
 218             f[j_coord_offset+DIM*0+YY] -= ty;
 219             f[j_coord_offset+DIM*0+ZZ] -= tz;
 220
 221             /* Inner loop uses 79 flops */
 222         }
 223         /* End of innermost loop */
 224
 225         tx = ty = tz = 0;
 226         f[i_coord_offset+DIM*0+XX] += fix0;
 227         f[i_coord_offset+DIM*0+YY] += fiy0;
 228         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 229         tx                         += fix0;
 230         ty                         += fiy0;
 231         tz                         += fiz0;
 232         fshift[i_shift_offset+XX]  += tx;
 233         fshift[i_shift_offset+YY]  += ty;
 234         fshift[i_shift_offset+ZZ]  += tz;
 235
 236         ggid                        = gid[iidx];
 237         /* Update potential energies */
 238         kernel_data->energygrp_elec[ggid] += velecsum;
 239         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 240
 241         /* Increment number of inner iterations */
 242         inneriter                  += j_index_end - j_index_start;
 243
 244         /* Outer loop uses 15 flops */
 245     }
 246
 247     /* Increment number of outer iterations */
 248     outeriter        += nri;
 249
 250     /* Update outer/inner flops */
 251
 252     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*79);
 253 }
 254 /*
 255  * Gromacs nonbonded kernel:   nb_kernel_ElecEw_VdwBham_GeomP1P1_F_c
 256  * Electrostatics interaction: Ewald
 257  * VdW interaction:            Buckingham
 258  * Geometry:                   Particle-Particle
 259  * Calculate force/pot:        Force
 260  */
 261 void
 262 nb_kernel_ElecEw_VdwBham_GeomP1P1_F_c
 263                     (t_nblist                    * gmx_restrict       nlist,
 264                      rvec                        * gmx_restrict          xx,
 265                      rvec                        * gmx_restrict          ff,
 266                      t_forcerec                  * gmx_restrict          fr,
 267                      t_mdatoms                   * gmx_restrict     mdatoms,
 268                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 269                      t_nrnb                      * gmx_restrict        nrnb)
 270 {
 271     int              i_shift_offset,i_coord_offset,j_coord_offset;
 272     int              j_index_start,j_index_end;
 273     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 274     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 275     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 276     real             *shiftvec,*fshift,*x,*f;
 277     int              vdwioffset0;
 278     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 279     int              vdwjidx0;
 280     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 281     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 282     real             velec,felec,velecsum,facel,crf,krf,krf2;
 283     real             *charge;
 284     int              nvdwtype;
 285     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 286     int              *vdwtype;
 287     real             *vdwparam;
 288     int              ewitab;
 289     real             ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
 290     real             *ewtab;
 291
 292     x                = xx[0];
 293     f                = ff[0];
 294
 295     nri              = nlist->nri;
 296     iinr             = nlist->iinr;
 297     jindex           = nlist->jindex;
 298     jjnr             = nlist->jjnr;
 299     shiftidx         = nlist->shift;
 300     gid              = nlist->gid;
 301     shiftvec         = fr->shift_vec[0];
 302     fshift           = fr->fshift[0];
 303     facel            = fr->epsfac;
 304     charge           = mdatoms->chargeA;
 305     nvdwtype         = fr->ntype;
 306     vdwparam         = fr->nbfp;
 307     vdwtype          = mdatoms->typeA;
 308
 309     sh_ewald         = fr->ic->sh_ewald;
 310     ewtab            = fr->ic->tabq_coul_F;
 311     ewtabscale       = fr->ic->tabq_scale;
 312     ewtabhalfspace   = 0.5/ewtabscale;
 313
 314     outeriter        = 0;
 315     inneriter        = 0;
 316
 317     /* Start outer loop over neighborlists */
 318     for(iidx=0; iidx<nri; iidx++)
 319     {
 320         /* Load shift vector for this list */
 321         i_shift_offset   = DIM*shiftidx[iidx];
 322         shX              = shiftvec[i_shift_offset+XX];
 323         shY              = shiftvec[i_shift_offset+YY];
 324         shZ              = shiftvec[i_shift_offset+ZZ];
 325
 326         /* Load limits for loop over neighbors */
 327         j_index_start    = jindex[iidx];
 328         j_index_end      = jindex[iidx+1];
 329
 330         /* Get outer coordinate index */
 331         inr              = iinr[iidx];
 332         i_coord_offset   = DIM*inr;
 333
 334         /* Load i particle coords and add shift vector */
 335         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 336         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 337         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 338
 339         fix0             = 0.0;
 340         fiy0             = 0.0;
 341         fiz0             = 0.0;
 342
 343         /* Load parameters for i particles */
 344         iq0              = facel*charge[inr+0];
 345         vdwioffset0      = 3*nvdwtype*vdwtype[inr+0];
 346
 347         /* Start inner kernel loop */
 348         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 349         {
 350             /* Get j neighbor index, and coordinate index */
 351             jnr              = jjnr[jidx];
 352             j_coord_offset   = DIM*jnr;
 353
 354             /* load j atom coordinates */
 355             jx0              = x[j_coord_offset+DIM*0+XX];
 356             jy0              = x[j_coord_offset+DIM*0+YY];
 357             jz0              = x[j_coord_offset+DIM*0+ZZ];
 358
 359             /* Calculate displacement vector */
 360             dx00             = ix0 - jx0;
 361             dy00             = iy0 - jy0;
 362             dz00             = iz0 - jz0;
 363
 364             /* Calculate squared distance and things based on it */
 365             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 366
 367             rinv00           = gmx_invsqrt(rsq00);
 368
 369             rinvsq00         = rinv00*rinv00;
 370
 371             /* Load parameters for j particles */
 372             jq0              = charge[jnr+0];
 373             vdwjidx0         = 3*vdwtype[jnr+0];
 374
 375             /**************************
 376              * CALCULATE INTERACTIONS *
 377              **************************/
 378
 379             r00              = rsq00*rinv00;
 380
 381             qq00             = iq0*jq0;
 382             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 383             cexp1_00         = vdwparam[vdwioffset0+vdwjidx0+1];
 384             cexp2_00         = vdwparam[vdwioffset0+vdwjidx0+2];
 385
 386             /* EWALD ELECTROSTATICS */
 387
 388             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 389             ewrt             = r00*ewtabscale;
 390             ewitab           = ewrt;
 391             eweps            = ewrt-ewitab;
 392             felec            = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1];
 393             felec            = qq00*rinv00*(rinvsq00-felec);
 394
 395             /* BUCKINGHAM DISPERSION/REPULSION */
 396             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 397             vvdw6            = c6_00*rinvsix;
 398             br               = cexp2_00*r00;
 399             vvdwexp          = cexp1_00*exp(-br);
 400             fvdw             = (br*vvdwexp-vvdw6)*rinvsq00;
 401
 402             fscal            = felec+fvdw;
 403
 404             /* Calculate temporary vectorial force */
 405             tx               = fscal*dx00;
 406             ty               = fscal*dy00;
 407             tz               = fscal*dz00;
 408
 409             /* Update vectorial force */
 410             fix0            += tx;
 411             fiy0            += ty;
 412             fiz0            += tz;
 413             f[j_coord_offset+DIM*0+XX] -= tx;
 414             f[j_coord_offset+DIM*0+YY] -= ty;
 415             f[j_coord_offset+DIM*0+ZZ] -= tz;
 416
 417             /* Inner loop uses 69 flops */
 418         }
 419         /* End of innermost loop */
 420
 421         tx = ty = tz = 0;
 422         f[i_coord_offset+DIM*0+XX] += fix0;
 423         f[i_coord_offset+DIM*0+YY] += fiy0;
 424         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 425         tx                         += fix0;
 426         ty                         += fiy0;
 427         tz                         += fiz0;
 428         fshift[i_shift_offset+XX]  += tx;
 429         fshift[i_shift_offset+YY]  += ty;
 430         fshift[i_shift_offset+ZZ]  += tz;
 431
 432         /* Increment number of inner iterations */
 433         inneriter                  += j_index_end - j_index_start;
 434
 435         /* Outer loop uses 13 flops */
 436     }
 437
 438     /* Increment number of outer iterations */
 439     outeriter        += nri;
 440
 441     /* Update outer/inner flops */
 442
 443     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*69);
 444 }