src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecRF_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 "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 /*
  48  * Gromacs nonbonded kernel:   nb_kernel_ElecRF_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_ElecRF_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     outeriter        = 0;
 109     inneriter        = 0;
 110
 111     /* Start outer loop over neighborlists */
 112     for(iidx=0; iidx<nri; iidx++)
 113     {
 114         /* Load shift vector for this list */
 115         i_shift_offset   = DIM*shiftidx[iidx];
 116         shX              = shiftvec[i_shift_offset+XX];
 117         shY              = shiftvec[i_shift_offset+YY];
 118         shZ              = shiftvec[i_shift_offset+ZZ];
 119
 120         /* Load limits for loop over neighbors */
 121         j_index_start    = jindex[iidx];
 122         j_index_end      = jindex[iidx+1];
 123
 124         /* Get outer coordinate index */
 125         inr              = iinr[iidx];
 126         i_coord_offset   = DIM*inr;
 127
 128         /* Load i particle coords and add shift vector */
 129         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 130         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 131         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 132
 133         fix0             = 0.0;
 134         fiy0             = 0.0;
 135         fiz0             = 0.0;
 136
 137         /* Load parameters for i particles */
 138         iq0              = facel*charge[inr+0];
 139         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 140
 141         /* Reset potential sums */
 142         velecsum         = 0.0;
 143         vvdwsum          = 0.0;
 144
 145         /* Start inner kernel loop */
 146         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 147         {
 148             /* Get j neighbor index, and coordinate index */
 149             jnr              = jjnr[jidx];
 150             j_coord_offset   = DIM*jnr;
 151
 152             /* load j atom coordinates */
 153             jx0              = x[j_coord_offset+DIM*0+XX];
 154             jy0              = x[j_coord_offset+DIM*0+YY];
 155             jz0              = x[j_coord_offset+DIM*0+ZZ];
 156
 157             /* Calculate displacement vector */
 158             dx00             = ix0 - jx0;
 159             dy00             = iy0 - jy0;
 160             dz00             = iz0 - jz0;
 161
 162             /* Calculate squared distance and things based on it */
 163             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 164
 165             rinv00           = gmx_invsqrt(rsq00);
 166
 167             rinvsq00         = rinv00*rinv00;
 168
 169             /* Load parameters for j particles */
 170             jq0              = charge[jnr+0];
 171             vdwjidx0         = 2*vdwtype[jnr+0];
 172
 173             /**************************
 174              * CALCULATE INTERACTIONS *
 175              **************************/
 176
 177             r00              = rsq00*rinv00;
 178
 179             qq00             = iq0*jq0;
 180             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 181             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 182
 183             /* Calculate table index by multiplying r with table scale and truncate to integer */
 184             rt               = r00*vftabscale;
 185             vfitab           = rt;
 186             vfeps            = rt-vfitab;
 187             vfitab           = 2*4*vfitab;
 188
 189             /* REACTION-FIELD ELECTROSTATICS */
 190             velec            = qq00*(rinv00+krf*rsq00-crf);
 191             felec            = qq00*(rinv00*rinvsq00-krf2);
 192
 193             /* CUBIC SPLINE TABLE DISPERSION */
 194             vfitab          += 0;
 195             Y                = vftab[vfitab];
 196             F                = vftab[vfitab+1];
 197             Geps             = vfeps*vftab[vfitab+2];
 198             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 199             Fp               = F+Geps+Heps2;
 200             VV               = Y+vfeps*Fp;
 201             vvdw6            = c6_00*VV;
 202             FF               = Fp+Geps+2.0*Heps2;
 203             fvdw6            = c6_00*FF;
 204
 205             /* CUBIC SPLINE TABLE REPULSION */
 206             Y                = vftab[vfitab+4];
 207             F                = vftab[vfitab+5];
 208             Geps             = vfeps*vftab[vfitab+6];
 209             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 210             Fp               = F+Geps+Heps2;
 211             VV               = Y+vfeps*Fp;
 212             vvdw12           = c12_00*VV;
 213             FF               = Fp+Geps+2.0*Heps2;
 214             fvdw12           = c12_00*FF;
 215             vvdw             = vvdw12+vvdw6;
 216             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 217
 218             /* Update potential sums from outer loop */
 219             velecsum        += velec;
 220             vvdwsum         += vvdw;
 221
 222             fscal            = felec+fvdw;
 223
 224             /* Calculate temporary vectorial force */
 225             tx               = fscal*dx00;
 226             ty               = fscal*dy00;
 227             tz               = fscal*dz00;
 228
 229             /* Update vectorial force */
 230             fix0            += tx;
 231             fiy0            += ty;
 232             fiz0            += tz;
 233             f[j_coord_offset+DIM*0+XX] -= tx;
 234             f[j_coord_offset+DIM*0+YY] -= ty;
 235             f[j_coord_offset+DIM*0+ZZ] -= tz;
 236
 237             /* Inner loop uses 66 flops */
 238         }
 239         /* End of innermost loop */
 240
 241         tx = ty = tz = 0;
 242         f[i_coord_offset+DIM*0+XX] += fix0;
 243         f[i_coord_offset+DIM*0+YY] += fiy0;
 244         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 245         tx                         += fix0;
 246         ty                         += fiy0;
 247         tz                         += fiz0;
 248         fshift[i_shift_offset+XX]  += tx;
 249         fshift[i_shift_offset+YY]  += ty;
 250         fshift[i_shift_offset+ZZ]  += tz;
 251
 252         ggid                        = gid[iidx];
 253         /* Update potential energies */
 254         kernel_data->energygrp_elec[ggid] += velecsum;
 255         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 256
 257         /* Increment number of inner iterations */
 258         inneriter                  += j_index_end - j_index_start;
 259
 260         /* Outer loop uses 15 flops */
 261     }
 262
 263     /* Increment number of outer iterations */
 264     outeriter        += nri;
 265
 266     /* Update outer/inner flops */
 267
 268     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*66);
 269 }
 270 /*
 271  * Gromacs nonbonded kernel:   nb_kernel_ElecRF_VdwCSTab_GeomP1P1_F_c
 272  * Electrostatics interaction: ReactionField
 273  * VdW interaction:            CubicSplineTable
 274  * Geometry:                   Particle-Particle
 275  * Calculate force/pot:        Force
 276  */
 277 void
 278 nb_kernel_ElecRF_VdwCSTab_GeomP1P1_F_c
 279                     (t_nblist                    * gmx_restrict       nlist,
 280                      rvec                        * gmx_restrict          xx,
 281                      rvec                        * gmx_restrict          ff,
 282                      t_forcerec                  * gmx_restrict          fr,
 283                      t_mdatoms                   * gmx_restrict     mdatoms,
 284                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 285                      t_nrnb                      * gmx_restrict        nrnb)
 286 {
 287     int              i_shift_offset,i_coord_offset,j_coord_offset;
 288     int              j_index_start,j_index_end;
 289     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 290     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 291     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 292     real             *shiftvec,*fshift,*x,*f;
 293     int              vdwioffset0;
 294     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 295     int              vdwjidx0;
 296     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 297     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 298     real             velec,felec,velecsum,facel,crf,krf,krf2;
 299     real             *charge;
 300     int              nvdwtype;
 301     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 302     int              *vdwtype;
 303     real             *vdwparam;
 304     int              vfitab;
 305     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 306     real             *vftab;
 307
 308     x                = xx[0];
 309     f                = ff[0];
 310
 311     nri              = nlist->nri;
 312     iinr             = nlist->iinr;
 313     jindex           = nlist->jindex;
 314     jjnr             = nlist->jjnr;
 315     shiftidx         = nlist->shift;
 316     gid              = nlist->gid;
 317     shiftvec         = fr->shift_vec[0];
 318     fshift           = fr->fshift[0];
 319     facel            = fr->epsfac;
 320     charge           = mdatoms->chargeA;
 321     krf              = fr->ic->k_rf;
 322     krf2             = krf*2.0;
 323     crf              = fr->ic->c_rf;
 324     nvdwtype         = fr->ntype;
 325     vdwparam         = fr->nbfp;
 326     vdwtype          = mdatoms->typeA;
 327
 328     vftab            = kernel_data->table_vdw->data;
 329     vftabscale       = kernel_data->table_vdw->scale;
 330
 331     outeriter        = 0;
 332     inneriter        = 0;
 333
 334     /* Start outer loop over neighborlists */
 335     for(iidx=0; iidx<nri; iidx++)
 336     {
 337         /* Load shift vector for this list */
 338         i_shift_offset   = DIM*shiftidx[iidx];
 339         shX              = shiftvec[i_shift_offset+XX];
 340         shY              = shiftvec[i_shift_offset+YY];
 341         shZ              = shiftvec[i_shift_offset+ZZ];
 342
 343         /* Load limits for loop over neighbors */
 344         j_index_start    = jindex[iidx];
 345         j_index_end      = jindex[iidx+1];
 346
 347         /* Get outer coordinate index */
 348         inr              = iinr[iidx];
 349         i_coord_offset   = DIM*inr;
 350
 351         /* Load i particle coords and add shift vector */
 352         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 353         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 354         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 355
 356         fix0             = 0.0;
 357         fiy0             = 0.0;
 358         fiz0             = 0.0;
 359
 360         /* Load parameters for i particles */
 361         iq0              = facel*charge[inr+0];
 362         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 363
 364         /* Start inner kernel loop */
 365         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 366         {
 367             /* Get j neighbor index, and coordinate index */
 368             jnr              = jjnr[jidx];
 369             j_coord_offset   = DIM*jnr;
 370
 371             /* load j atom coordinates */
 372             jx0              = x[j_coord_offset+DIM*0+XX];
 373             jy0              = x[j_coord_offset+DIM*0+YY];
 374             jz0              = x[j_coord_offset+DIM*0+ZZ];
 375
 376             /* Calculate displacement vector */
 377             dx00             = ix0 - jx0;
 378             dy00             = iy0 - jy0;
 379             dz00             = iz0 - jz0;
 380
 381             /* Calculate squared distance and things based on it */
 382             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 383
 384             rinv00           = gmx_invsqrt(rsq00);
 385
 386             rinvsq00         = rinv00*rinv00;
 387
 388             /* Load parameters for j particles */
 389             jq0              = charge[jnr+0];
 390             vdwjidx0         = 2*vdwtype[jnr+0];
 391
 392             /**************************
 393              * CALCULATE INTERACTIONS *
 394              **************************/
 395
 396             r00              = rsq00*rinv00;
 397
 398             qq00             = iq0*jq0;
 399             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 400             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 401
 402             /* Calculate table index by multiplying r with table scale and truncate to integer */
 403             rt               = r00*vftabscale;
 404             vfitab           = rt;
 405             vfeps            = rt-vfitab;
 406             vfitab           = 2*4*vfitab;
 407
 408             /* REACTION-FIELD ELECTROSTATICS */
 409             felec            = qq00*(rinv00*rinvsq00-krf2);
 410
 411             /* CUBIC SPLINE TABLE DISPERSION */
 412             vfitab          += 0;
 413             F                = vftab[vfitab+1];
 414             Geps             = vfeps*vftab[vfitab+2];
 415             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 416             Fp               = F+Geps+Heps2;
 417             FF               = Fp+Geps+2.0*Heps2;
 418             fvdw6            = c6_00*FF;
 419
 420             /* CUBIC SPLINE TABLE REPULSION */
 421             F                = vftab[vfitab+5];
 422             Geps             = vfeps*vftab[vfitab+6];
 423             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 424             Fp               = F+Geps+Heps2;
 425             FF               = Fp+Geps+2.0*Heps2;
 426             fvdw12           = c12_00*FF;
 427             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 428
 429             fscal            = felec+fvdw;
 430
 431             /* Calculate temporary vectorial force */
 432             tx               = fscal*dx00;
 433             ty               = fscal*dy00;
 434             tz               = fscal*dz00;
 435
 436             /* Update vectorial force */
 437             fix0            += tx;
 438             fiy0            += ty;
 439             fiz0            += tz;
 440             f[j_coord_offset+DIM*0+XX] -= tx;
 441             f[j_coord_offset+DIM*0+YY] -= ty;
 442             f[j_coord_offset+DIM*0+ZZ] -= tz;
 443
 444             /* Inner loop uses 53 flops */
 445         }
 446         /* End of innermost loop */
 447
 448         tx = ty = tz = 0;
 449         f[i_coord_offset+DIM*0+XX] += fix0;
 450         f[i_coord_offset+DIM*0+YY] += fiy0;
 451         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 452         tx                         += fix0;
 453         ty                         += fiy0;
 454         tz                         += fiz0;
 455         fshift[i_shift_offset+XX]  += tx;
 456         fshift[i_shift_offset+YY]  += ty;
 457         fshift[i_shift_offset+ZZ]  += tz;
 458
 459         /* Increment number of inner iterations */
 460         inneriter                  += j_index_end - j_index_start;
 461
 462         /* Outer loop uses 13 flops */
 463     }
 464
 465     /* Increment number of outer iterations */
 466     outeriter        += nri;
 467
 468     /* Update outer/inner flops */
 469
 470     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*53);
 471 }