src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecNone_VdwLJEw_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_ElecNone_VdwLJEw_GeomP1P1_VF_c
  49  * Electrostatics interaction: None
  50  * VdW interaction:            LJEwald
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecNone_VdwLJEw_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     int              nvdwtype;
  76     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  77     int              *vdwtype;
  78     real             *vdwparam;
  79     real             c6grid_00;
  80     real             ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,sh_lj_ewald;
  81     real             *vdwgridparam;
  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     nvdwtype         = fr->ntype;
  95     vdwparam         = fr->nbfp;
  96     vdwtype          = mdatoms->typeA;
  97     vdwgridparam     = fr->ljpme_c6grid;
  98     ewclj            = fr->ewaldcoeff_lj;
  99     sh_lj_ewald      = fr->ic->sh_lj_ewald;
 100     ewclj2           = ewclj*ewclj;
 101     ewclj6           = ewclj2*ewclj2*ewclj2;
 102
 103     outeriter        = 0;
 104     inneriter        = 0;
 105
 106     /* Start outer loop over neighborlists */
 107     for(iidx=0; iidx<nri; iidx++)
 108     {
 109         /* Load shift vector for this list */
 110         i_shift_offset   = DIM*shiftidx[iidx];
 111         shX              = shiftvec[i_shift_offset+XX];
 112         shY              = shiftvec[i_shift_offset+YY];
 113         shZ              = shiftvec[i_shift_offset+ZZ];
 114
 115         /* Load limits for loop over neighbors */
 116         j_index_start    = jindex[iidx];
 117         j_index_end      = jindex[iidx+1];
 118
 119         /* Get outer coordinate index */
 120         inr              = iinr[iidx];
 121         i_coord_offset   = DIM*inr;
 122
 123         /* Load i particle coords and add shift vector */
 124         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 125         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 126         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 127
 128         fix0             = 0.0;
 129         fiy0             = 0.0;
 130         fiz0             = 0.0;
 131
 132         /* Load parameters for i particles */
 133         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 134
 135         /* Reset potential sums */
 136         vvdwsum          = 0.0;
 137
 138         /* Start inner kernel loop */
 139         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 140         {
 141             /* Get j neighbor index, and coordinate index */
 142             jnr              = jjnr[jidx];
 143             j_coord_offset   = DIM*jnr;
 144
 145             /* load j atom coordinates */
 146             jx0              = x[j_coord_offset+DIM*0+XX];
 147             jy0              = x[j_coord_offset+DIM*0+YY];
 148             jz0              = x[j_coord_offset+DIM*0+ZZ];
 149
 150             /* Calculate displacement vector */
 151             dx00             = ix0 - jx0;
 152             dy00             = iy0 - jy0;
 153             dz00             = iz0 - jz0;
 154
 155             /* Calculate squared distance and things based on it */
 156             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 157
 158             rinv00           = gmx_invsqrt(rsq00);
 159
 160             rinvsq00         = rinv00*rinv00;
 161
 162             /* Load parameters for j particles */
 163             vdwjidx0         = 2*vdwtype[jnr+0];
 164
 165             /**************************
 166              * CALCULATE INTERACTIONS *
 167              **************************/
 168
 169             r00              = rsq00*rinv00;
 170
 171             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 172             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 173             c6grid_00        = vdwgridparam[vdwioffset0+vdwjidx0];
 174
 175             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 176             ewcljrsq         = ewclj2*rsq00;
 177             exponent         = exp(-ewcljrsq);
 178             poly             = exponent*(1.0 + ewcljrsq + ewcljrsq*ewcljrsq*0.5);
 179             vvdw6            = (c6_00-c6grid_00*(1.0-poly))*rinvsix;
 180             vvdw12           = c12_00*rinvsix*rinvsix;
 181             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 182             fvdw             = (vvdw12 - vvdw6 - c6grid_00*(1.0/6.0)*exponent*ewclj6)*rinvsq00;
 183
 184             /* Update potential sums from outer loop */
 185             vvdwsum         += vvdw;
 186
 187             fscal            = fvdw;
 188
 189             /* Calculate temporary vectorial force */
 190             tx               = fscal*dx00;
 191             ty               = fscal*dy00;
 192             tz               = fscal*dz00;
 193
 194             /* Update vectorial force */
 195             fix0            += tx;
 196             fiy0            += ty;
 197             fiz0            += tz;
 198             f[j_coord_offset+DIM*0+XX] -= tx;
 199             f[j_coord_offset+DIM*0+YY] -= ty;
 200             f[j_coord_offset+DIM*0+ZZ] -= tz;
 201
 202             /* Inner loop uses 49 flops */
 203         }
 204         /* End of innermost loop */
 205
 206         tx = ty = tz = 0;
 207         f[i_coord_offset+DIM*0+XX] += fix0;
 208         f[i_coord_offset+DIM*0+YY] += fiy0;
 209         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 210         tx                         += fix0;
 211         ty                         += fiy0;
 212         tz                         += fiz0;
 213         fshift[i_shift_offset+XX]  += tx;
 214         fshift[i_shift_offset+YY]  += ty;
 215         fshift[i_shift_offset+ZZ]  += tz;
 216
 217         ggid                        = gid[iidx];
 218         /* Update potential energies */
 219         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 220
 221         /* Increment number of inner iterations */
 222         inneriter                  += j_index_end - j_index_start;
 223
 224         /* Outer loop uses 13 flops */
 225     }
 226
 227     /* Increment number of outer iterations */
 228     outeriter        += nri;
 229
 230     /* Update outer/inner flops */
 231
 232     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_VF,outeriter*13 + inneriter*49);
 233 }
 234 /*
 235  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwLJEw_GeomP1P1_F_c
 236  * Electrostatics interaction: None
 237  * VdW interaction:            LJEwald
 238  * Geometry:                   Particle-Particle
 239  * Calculate force/pot:        Force
 240  */
 241 void
 242 nb_kernel_ElecNone_VdwLJEw_GeomP1P1_F_c
 243                     (t_nblist                    * gmx_restrict       nlist,
 244                      rvec                        * gmx_restrict          xx,
 245                      rvec                        * gmx_restrict          ff,
 246                      t_forcerec                  * gmx_restrict          fr,
 247                      t_mdatoms                   * gmx_restrict     mdatoms,
 248                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 249                      t_nrnb                      * gmx_restrict        nrnb)
 250 {
 251     int              i_shift_offset,i_coord_offset,j_coord_offset;
 252     int              j_index_start,j_index_end;
 253     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 254     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 255     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 256     real             *shiftvec,*fshift,*x,*f;
 257     int              vdwioffset0;
 258     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 259     int              vdwjidx0;
 260     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 261     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 262     int              nvdwtype;
 263     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 264     int              *vdwtype;
 265     real             *vdwparam;
 266     real             c6grid_00;
 267     real             ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,sh_lj_ewald;
 268     real             *vdwgridparam;
 269
 270     x                = xx[0];
 271     f                = ff[0];
 272
 273     nri              = nlist->nri;
 274     iinr             = nlist->iinr;
 275     jindex           = nlist->jindex;
 276     jjnr             = nlist->jjnr;
 277     shiftidx         = nlist->shift;
 278     gid              = nlist->gid;
 279     shiftvec         = fr->shift_vec[0];
 280     fshift           = fr->fshift[0];
 281     nvdwtype         = fr->ntype;
 282     vdwparam         = fr->nbfp;
 283     vdwtype          = mdatoms->typeA;
 284     vdwgridparam     = fr->ljpme_c6grid;
 285     ewclj            = fr->ewaldcoeff_lj;
 286     sh_lj_ewald      = fr->ic->sh_lj_ewald;
 287     ewclj2           = ewclj*ewclj;
 288     ewclj6           = ewclj2*ewclj2*ewclj2;
 289
 290     outeriter        = 0;
 291     inneriter        = 0;
 292
 293     /* Start outer loop over neighborlists */
 294     for(iidx=0; iidx<nri; iidx++)
 295     {
 296         /* Load shift vector for this list */
 297         i_shift_offset   = DIM*shiftidx[iidx];
 298         shX              = shiftvec[i_shift_offset+XX];
 299         shY              = shiftvec[i_shift_offset+YY];
 300         shZ              = shiftvec[i_shift_offset+ZZ];
 301
 302         /* Load limits for loop over neighbors */
 303         j_index_start    = jindex[iidx];
 304         j_index_end      = jindex[iidx+1];
 305
 306         /* Get outer coordinate index */
 307         inr              = iinr[iidx];
 308         i_coord_offset   = DIM*inr;
 309
 310         /* Load i particle coords and add shift vector */
 311         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 312         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 313         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 314
 315         fix0             = 0.0;
 316         fiy0             = 0.0;
 317         fiz0             = 0.0;
 318
 319         /* Load parameters for i particles */
 320         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 321
 322         /* Start inner kernel loop */
 323         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 324         {
 325             /* Get j neighbor index, and coordinate index */
 326             jnr              = jjnr[jidx];
 327             j_coord_offset   = DIM*jnr;
 328
 329             /* load j atom coordinates */
 330             jx0              = x[j_coord_offset+DIM*0+XX];
 331             jy0              = x[j_coord_offset+DIM*0+YY];
 332             jz0              = x[j_coord_offset+DIM*0+ZZ];
 333
 334             /* Calculate displacement vector */
 335             dx00             = ix0 - jx0;
 336             dy00             = iy0 - jy0;
 337             dz00             = iz0 - jz0;
 338
 339             /* Calculate squared distance and things based on it */
 340             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 341
 342             rinv00           = gmx_invsqrt(rsq00);
 343
 344             rinvsq00         = rinv00*rinv00;
 345
 346             /* Load parameters for j particles */
 347             vdwjidx0         = 2*vdwtype[jnr+0];
 348
 349             /**************************
 350              * CALCULATE INTERACTIONS *
 351              **************************/
 352
 353             r00              = rsq00*rinv00;
 354
 355             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 356             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 357             c6grid_00        = vdwgridparam[vdwioffset0+vdwjidx0];
 358
 359             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 360             ewcljrsq         = ewclj2*rsq00;
 361             exponent         = exp(-ewcljrsq);
 362             poly             = exponent*(1.0 + ewcljrsq + ewcljrsq*ewcljrsq*0.5);
 363             fvdw             = (((c12_00*rinvsix - c6_00 + c6grid_00*(1.0-poly))*rinvsix) - c6grid_00*(1.0/6.0)*exponent*ewclj6)*rinvsq00;
 364
 365             fscal            = fvdw;
 366
 367             /* Calculate temporary vectorial force */
 368             tx               = fscal*dx00;
 369             ty               = fscal*dy00;
 370             tz               = fscal*dz00;
 371
 372             /* Update vectorial force */
 373             fix0            += tx;
 374             fiy0            += ty;
 375             fiz0            += tz;
 376             f[j_coord_offset+DIM*0+XX] -= tx;
 377             f[j_coord_offset+DIM*0+YY] -= ty;
 378             f[j_coord_offset+DIM*0+ZZ] -= tz;
 379
 380             /* Inner loop uses 44 flops */
 381         }
 382         /* End of innermost loop */
 383
 384         tx = ty = tz = 0;
 385         f[i_coord_offset+DIM*0+XX] += fix0;
 386         f[i_coord_offset+DIM*0+YY] += fiy0;
 387         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 388         tx                         += fix0;
 389         ty                         += fiy0;
 390         tz                         += fiz0;
 391         fshift[i_shift_offset+XX]  += tx;
 392         fshift[i_shift_offset+YY]  += ty;
 393         fshift[i_shift_offset+ZZ]  += tz;
 394
 395         /* Increment number of inner iterations */
 396         inneriter                  += j_index_end - j_index_start;
 397
 398         /* Outer loop uses 12 flops */
 399     }
 400
 401     /* Increment number of outer iterations */
 402     outeriter        += nri;
 403
 404     /* Update outer/inner flops */
 405
 406     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_F,outeriter*12 + inneriter*44);
 407 }