src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecRFCut_VdwCSTab_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_ElecRFCut_VdwCSTab_GeomP1P1_VF_c
  49  * Electrostatics interaction: ReactionField
  50  * VdW interaction:            CubicSplineTable
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecRFCut_VdwCSTab_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              vfitab;
  82     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
  83     real             *vftab;
  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     krf              = fr->ic->k_rf;
  99     krf2             = krf*2.0;
 100     crf              = fr->ic->c_rf;
 101     nvdwtype         = fr->ntype;
 102     vdwparam         = fr->nbfp;
 103     vdwtype          = mdatoms->typeA;
 104
 105     vftab            = kernel_data->table_vdw->data;
 106     vftabscale       = kernel_data->table_vdw->scale;
 107
 108     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 109     rcutoff          = fr->rcoulomb;
 110     rcutoff2         = rcutoff*rcutoff;
 111
 112     outeriter        = 0;
 113     inneriter        = 0;
 114
 115     /* Start outer loop over neighborlists */
 116     for(iidx=0; iidx<nri; iidx++)
 117     {
 118         /* Load shift vector for this list */
 119         i_shift_offset   = DIM*shiftidx[iidx];
 120         shX              = shiftvec[i_shift_offset+XX];
 121         shY              = shiftvec[i_shift_offset+YY];
 122         shZ              = shiftvec[i_shift_offset+ZZ];
 123
 124         /* Load limits for loop over neighbors */
 125         j_index_start    = jindex[iidx];
 126         j_index_end      = jindex[iidx+1];
 127
 128         /* Get outer coordinate index */
 129         inr              = iinr[iidx];
 130         i_coord_offset   = DIM*inr;
 131
 132         /* Load i particle coords and add shift vector */
 133         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 134         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 135         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 136
 137         fix0             = 0.0;
 138         fiy0             = 0.0;
 139         fiz0             = 0.0;
 140
 141         /* Load parameters for i particles */
 142         iq0              = facel*charge[inr+0];
 143         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 144
 145         /* Reset potential sums */
 146         velecsum         = 0.0;
 147         vvdwsum          = 0.0;
 148
 149         /* Start inner kernel loop */
 150         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 151         {
 152             /* Get j neighbor index, and coordinate index */
 153             jnr              = jjnr[jidx];
 154             j_coord_offset   = DIM*jnr;
 155
 156             /* load j atom coordinates */
 157             jx0              = x[j_coord_offset+DIM*0+XX];
 158             jy0              = x[j_coord_offset+DIM*0+YY];
 159             jz0              = x[j_coord_offset+DIM*0+ZZ];
 160
 161             /* Calculate displacement vector */
 162             dx00             = ix0 - jx0;
 163             dy00             = iy0 - jy0;
 164             dz00             = iz0 - jz0;
 165
 166             /* Calculate squared distance and things based on it */
 167             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 168
 169             rinv00           = gmx_invsqrt(rsq00);
 170
 171             rinvsq00         = rinv00*rinv00;
 172
 173             /* Load parameters for j particles */
 174             jq0              = charge[jnr+0];
 175             vdwjidx0         = 2*vdwtype[jnr+0];
 176
 177             /**************************
 178              * CALCULATE INTERACTIONS *
 179              **************************/
 180
 181             if (rsq00<rcutoff2)
 182             {
 183
 184             r00              = rsq00*rinv00;
 185
 186             qq00             = iq0*jq0;
 187             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 188             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 189
 190             /* Calculate table index by multiplying r with table scale and truncate to integer */
 191             rt               = r00*vftabscale;
 192             vfitab           = rt;
 193             vfeps            = rt-vfitab;
 194             vfitab           = 2*4*vfitab;
 195
 196             /* REACTION-FIELD ELECTROSTATICS */
 197             velec            = qq00*(rinv00+krf*rsq00-crf);
 198             felec            = qq00*(rinv00*rinvsq00-krf2);
 199
 200             /* CUBIC SPLINE TABLE DISPERSION */
 201             vfitab          += 0;
 202             Y                = vftab[vfitab];
 203             F                = vftab[vfitab+1];
 204             Geps             = vfeps*vftab[vfitab+2];
 205             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 206             Fp               = F+Geps+Heps2;
 207             VV               = Y+vfeps*Fp;
 208             vvdw6            = c6_00*VV;
 209             FF               = Fp+Geps+2.0*Heps2;
 210             fvdw6            = c6_00*FF;
 211
 212             /* CUBIC SPLINE TABLE REPULSION */
 213             Y                = vftab[vfitab+4];
 214             F                = vftab[vfitab+5];
 215             Geps             = vfeps*vftab[vfitab+6];
 216             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 217             Fp               = F+Geps+Heps2;
 218             VV               = Y+vfeps*Fp;
 219             vvdw12           = c12_00*VV;
 220             FF               = Fp+Geps+2.0*Heps2;
 221             fvdw12           = c12_00*FF;
 222             vvdw             = vvdw12+vvdw6;
 223             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 224
 225             /* Update potential sums from outer loop */
 226             velecsum        += velec;
 227             vvdwsum         += vvdw;
 228
 229             fscal            = felec+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
 246             /* Inner loop uses 66 flops */
 247         }
 248         /* End of innermost loop */
 249
 250         tx = ty = tz = 0;
 251         f[i_coord_offset+DIM*0+XX] += fix0;
 252         f[i_coord_offset+DIM*0+YY] += fiy0;
 253         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 254         tx                         += fix0;
 255         ty                         += fiy0;
 256         tz                         += fiz0;
 257         fshift[i_shift_offset+XX]  += tx;
 258         fshift[i_shift_offset+YY]  += ty;
 259         fshift[i_shift_offset+ZZ]  += tz;
 260
 261         ggid                        = gid[iidx];
 262         /* Update potential energies */
 263         kernel_data->energygrp_elec[ggid] += velecsum;
 264         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 265
 266         /* Increment number of inner iterations */
 267         inneriter                  += j_index_end - j_index_start;
 268
 269         /* Outer loop uses 15 flops */
 270     }
 271
 272     /* Increment number of outer iterations */
 273     outeriter        += nri;
 274
 275     /* Update outer/inner flops */
 276
 277     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*66);
 278 }
 279 /*
 280  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_F_c
 281  * Electrostatics interaction: ReactionField
 282  * VdW interaction:            CubicSplineTable
 283  * Geometry:                   Particle-Particle
 284  * Calculate force/pot:        Force
 285  */
 286 void
 287 nb_kernel_ElecRFCut_VdwCSTab_GeomP1P1_F_c
 288                     (t_nblist                    * gmx_restrict       nlist,
 289                      rvec                        * gmx_restrict          xx,
 290                      rvec                        * gmx_restrict          ff,
 291                      t_forcerec                  * gmx_restrict          fr,
 292                      t_mdatoms                   * gmx_restrict     mdatoms,
 293                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 294                      t_nrnb                      * gmx_restrict        nrnb)
 295 {
 296     int              i_shift_offset,i_coord_offset,j_coord_offset;
 297     int              j_index_start,j_index_end;
 298     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 299     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 300     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 301     real             *shiftvec,*fshift,*x,*f;
 302     int              vdwioffset0;
 303     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 304     int              vdwjidx0;
 305     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 306     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 307     real             velec,felec,velecsum,facel,crf,krf,krf2;
 308     real             *charge;
 309     int              nvdwtype;
 310     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 311     int              *vdwtype;
 312     real             *vdwparam;
 313     int              vfitab;
 314     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 315     real             *vftab;
 316
 317     x                = xx[0];
 318     f                = ff[0];
 319
 320     nri              = nlist->nri;
 321     iinr             = nlist->iinr;
 322     jindex           = nlist->jindex;
 323     jjnr             = nlist->jjnr;
 324     shiftidx         = nlist->shift;
 325     gid              = nlist->gid;
 326     shiftvec         = fr->shift_vec[0];
 327     fshift           = fr->fshift[0];
 328     facel            = fr->epsfac;
 329     charge           = mdatoms->chargeA;
 330     krf              = fr->ic->k_rf;
 331     krf2             = krf*2.0;
 332     crf              = fr->ic->c_rf;
 333     nvdwtype         = fr->ntype;
 334     vdwparam         = fr->nbfp;
 335     vdwtype          = mdatoms->typeA;
 336
 337     vftab            = kernel_data->table_vdw->data;
 338     vftabscale       = kernel_data->table_vdw->scale;
 339
 340     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 341     rcutoff          = fr->rcoulomb;
 342     rcutoff2         = rcutoff*rcutoff;
 343
 344     outeriter        = 0;
 345     inneriter        = 0;
 346
 347     /* Start outer loop over neighborlists */
 348     for(iidx=0; iidx<nri; iidx++)
 349     {
 350         /* Load shift vector for this list */
 351         i_shift_offset   = DIM*shiftidx[iidx];
 352         shX              = shiftvec[i_shift_offset+XX];
 353         shY              = shiftvec[i_shift_offset+YY];
 354         shZ              = shiftvec[i_shift_offset+ZZ];
 355
 356         /* Load limits for loop over neighbors */
 357         j_index_start    = jindex[iidx];
 358         j_index_end      = jindex[iidx+1];
 359
 360         /* Get outer coordinate index */
 361         inr              = iinr[iidx];
 362         i_coord_offset   = DIM*inr;
 363
 364         /* Load i particle coords and add shift vector */
 365         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 366         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 367         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 368
 369         fix0             = 0.0;
 370         fiy0             = 0.0;
 371         fiz0             = 0.0;
 372
 373         /* Load parameters for i particles */
 374         iq0              = facel*charge[inr+0];
 375         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 376
 377         /* Start inner kernel loop */
 378         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 379         {
 380             /* Get j neighbor index, and coordinate index */
 381             jnr              = jjnr[jidx];
 382             j_coord_offset   = DIM*jnr;
 383
 384             /* load j atom coordinates */
 385             jx0              = x[j_coord_offset+DIM*0+XX];
 386             jy0              = x[j_coord_offset+DIM*0+YY];
 387             jz0              = x[j_coord_offset+DIM*0+ZZ];
 388
 389             /* Calculate displacement vector */
 390             dx00             = ix0 - jx0;
 391             dy00             = iy0 - jy0;
 392             dz00             = iz0 - jz0;
 393
 394             /* Calculate squared distance and things based on it */
 395             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 396
 397             rinv00           = gmx_invsqrt(rsq00);
 398
 399             rinvsq00         = rinv00*rinv00;
 400
 401             /* Load parameters for j particles */
 402             jq0              = charge[jnr+0];
 403             vdwjidx0         = 2*vdwtype[jnr+0];
 404
 405             /**************************
 406              * CALCULATE INTERACTIONS *
 407              **************************/
 408
 409             if (rsq00<rcutoff2)
 410             {
 411
 412             r00              = rsq00*rinv00;
 413
 414             qq00             = iq0*jq0;
 415             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 416             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 417
 418             /* Calculate table index by multiplying r with table scale and truncate to integer */
 419             rt               = r00*vftabscale;
 420             vfitab           = rt;
 421             vfeps            = rt-vfitab;
 422             vfitab           = 2*4*vfitab;
 423
 424             /* REACTION-FIELD ELECTROSTATICS */
 425             felec            = qq00*(rinv00*rinvsq00-krf2);
 426
 427             /* CUBIC SPLINE TABLE DISPERSION */
 428             vfitab          += 0;
 429             F                = vftab[vfitab+1];
 430             Geps             = vfeps*vftab[vfitab+2];
 431             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 432             Fp               = F+Geps+Heps2;
 433             FF               = Fp+Geps+2.0*Heps2;
 434             fvdw6            = c6_00*FF;
 435
 436             /* CUBIC SPLINE TABLE REPULSION */
 437             F                = vftab[vfitab+5];
 438             Geps             = vfeps*vftab[vfitab+6];
 439             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 440             Fp               = F+Geps+Heps2;
 441             FF               = Fp+Geps+2.0*Heps2;
 442             fvdw12           = c12_00*FF;
 443             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 444
 445             fscal            = felec+fvdw;
 446
 447             /* Calculate temporary vectorial force */
 448             tx               = fscal*dx00;
 449             ty               = fscal*dy00;
 450             tz               = fscal*dz00;
 451
 452             /* Update vectorial force */
 453             fix0            += tx;
 454             fiy0            += ty;
 455             fiz0            += tz;
 456             f[j_coord_offset+DIM*0+XX] -= tx;
 457             f[j_coord_offset+DIM*0+YY] -= ty;
 458             f[j_coord_offset+DIM*0+ZZ] -= tz;
 459
 460             }
 461
 462             /* Inner loop uses 53 flops */
 463         }
 464         /* End of innermost loop */
 465
 466         tx = ty = tz = 0;
 467         f[i_coord_offset+DIM*0+XX] += fix0;
 468         f[i_coord_offset+DIM*0+YY] += fiy0;
 469         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 470         tx                         += fix0;
 471         ty                         += fiy0;
 472         tz                         += fiz0;
 473         fshift[i_shift_offset+XX]  += tx;
 474         fshift[i_shift_offset+YY]  += ty;
 475         fshift[i_shift_offset+ZZ]  += tz;
 476
 477         /* Increment number of inner iterations */
 478         inneriter                  += j_index_end - j_index_start;
 479
 480         /* Outer loop uses 13 flops */
 481     }
 482
 483     /* Increment number of outer iterations */
 484     outeriter        += nri;
 485
 486     /* Update outer/inner flops */
 487
 488     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*53);
 489 }