src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecCSTab_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 "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/nrnb.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwCSTab_GeomP1P1_VF_c
  51  * Electrostatics interaction: CubicSplineTable
  52  * VdW interaction:            CubicSplineTable
  53  * Geometry:                   Particle-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecCSTab_VdwCSTab_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              nvdwtype;
  80     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  81     int              *vdwtype;
  82     real             *vdwparam;
  83     int              vfitab;
  84     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
  85     real             *vftab;
  86
  87     x                = xx[0];
  88     f                = ff[0];
  89
  90     nri              = nlist->nri;
  91     iinr             = nlist->iinr;
  92     jindex           = nlist->jindex;
  93     jjnr             = nlist->jjnr;
  94     shiftidx         = nlist->shift;
  95     gid              = nlist->gid;
  96     shiftvec         = fr->shift_vec[0];
  97     fshift           = fr->fshift[0];
  98     facel            = fr->epsfac;
  99     charge           = mdatoms->chargeA;
 100     nvdwtype         = fr->ntype;
 101     vdwparam         = fr->nbfp;
 102     vdwtype          = mdatoms->typeA;
 103
 104     vftab            = kernel_data->table_elec_vdw->data;
 105     vftabscale       = kernel_data->table_elec_vdw->scale;
 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      = 2*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             /* Load parameters for j particles */
 167             jq0              = charge[jnr+0];
 168             vdwjidx0         = 2*vdwtype[jnr+0];
 169
 170             /**************************
 171              * CALCULATE INTERACTIONS *
 172              **************************/
 173
 174             r00              = rsq00*rinv00;
 175
 176             qq00             = iq0*jq0;
 177             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 178             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 179
 180             /* Calculate table index by multiplying r with table scale and truncate to integer */
 181             rt               = r00*vftabscale;
 182             vfitab           = rt;
 183             vfeps            = rt-vfitab;
 184             vfitab           = 3*4*vfitab;
 185
 186             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 187             Y                = vftab[vfitab];
 188             F                = vftab[vfitab+1];
 189             Geps             = vfeps*vftab[vfitab+2];
 190             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 191             Fp               = F+Geps+Heps2;
 192             VV               = Y+vfeps*Fp;
 193             velec            = qq00*VV;
 194             FF               = Fp+Geps+2.0*Heps2;
 195             felec            = -qq00*FF*vftabscale*rinv00;
 196
 197             /* CUBIC SPLINE TABLE DISPERSION */
 198             vfitab          += 4;
 199             Y                = vftab[vfitab];
 200             F                = vftab[vfitab+1];
 201             Geps             = vfeps*vftab[vfitab+2];
 202             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 203             Fp               = F+Geps+Heps2;
 204             VV               = Y+vfeps*Fp;
 205             vvdw6            = c6_00*VV;
 206             FF               = Fp+Geps+2.0*Heps2;
 207             fvdw6            = c6_00*FF;
 208
 209             /* CUBIC SPLINE TABLE REPULSION */
 210             Y                = vftab[vfitab+4];
 211             F                = vftab[vfitab+5];
 212             Geps             = vfeps*vftab[vfitab+6];
 213             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 214             Fp               = F+Geps+Heps2;
 215             VV               = Y+vfeps*Fp;
 216             vvdw12           = c12_00*VV;
 217             FF               = Fp+Geps+2.0*Heps2;
 218             fvdw12           = c12_00*FF;
 219             vvdw             = vvdw12+vvdw6;
 220             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 221
 222             /* Update potential sums from outer loop */
 223             velecsum        += velec;
 224             vvdwsum         += vvdw;
 225
 226             fscal            = felec+fvdw;
 227
 228             /* Calculate temporary vectorial force */
 229             tx               = fscal*dx00;
 230             ty               = fscal*dy00;
 231             tz               = fscal*dz00;
 232
 233             /* Update vectorial force */
 234             fix0            += tx;
 235             fiy0            += ty;
 236             fiz0            += tz;
 237             f[j_coord_offset+DIM*0+XX] -= tx;
 238             f[j_coord_offset+DIM*0+YY] -= ty;
 239             f[j_coord_offset+DIM*0+ZZ] -= tz;
 240
 241             /* Inner loop uses 73 flops */
 242         }
 243         /* End of innermost loop */
 244
 245         tx = ty = tz = 0;
 246         f[i_coord_offset+DIM*0+XX] += fix0;
 247         f[i_coord_offset+DIM*0+YY] += fiy0;
 248         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 249         tx                         += fix0;
 250         ty                         += fiy0;
 251         tz                         += fiz0;
 252         fshift[i_shift_offset+XX]  += tx;
 253         fshift[i_shift_offset+YY]  += ty;
 254         fshift[i_shift_offset+ZZ]  += tz;
 255
 256         ggid                        = gid[iidx];
 257         /* Update potential energies */
 258         kernel_data->energygrp_elec[ggid] += velecsum;
 259         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 260
 261         /* Increment number of inner iterations */
 262         inneriter                  += j_index_end - j_index_start;
 263
 264         /* Outer loop uses 15 flops */
 265     }
 266
 267     /* Increment number of outer iterations */
 268     outeriter        += nri;
 269
 270     /* Update outer/inner flops */
 271
 272     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*73);
 273 }
 274 /*
 275  * Gromacs nonbonded kernel:   nb_kernel_ElecCSTab_VdwCSTab_GeomP1P1_F_c
 276  * Electrostatics interaction: CubicSplineTable
 277  * VdW interaction:            CubicSplineTable
 278  * Geometry:                   Particle-Particle
 279  * Calculate force/pot:        Force
 280  */
 281 void
 282 nb_kernel_ElecCSTab_VdwCSTab_GeomP1P1_F_c
 283                     (t_nblist                    * gmx_restrict       nlist,
 284                      rvec                        * gmx_restrict          xx,
 285                      rvec                        * gmx_restrict          ff,
 286                      t_forcerec                  * gmx_restrict          fr,
 287                      t_mdatoms                   * gmx_restrict     mdatoms,
 288                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 289                      t_nrnb                      * gmx_restrict        nrnb)
 290 {
 291     int              i_shift_offset,i_coord_offset,j_coord_offset;
 292     int              j_index_start,j_index_end;
 293     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 294     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 295     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 296     real             *shiftvec,*fshift,*x,*f;
 297     int              vdwioffset0;
 298     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 299     int              vdwjidx0;
 300     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 301     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 302     real             velec,felec,velecsum,facel,crf,krf,krf2;
 303     real             *charge;
 304     int              nvdwtype;
 305     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 306     int              *vdwtype;
 307     real             *vdwparam;
 308     int              vfitab;
 309     real             rt,vfeps,vftabscale,Y,F,Geps,Heps2,Fp,VV,FF;
 310     real             *vftab;
 311
 312     x                = xx[0];
 313     f                = ff[0];
 314
 315     nri              = nlist->nri;
 316     iinr             = nlist->iinr;
 317     jindex           = nlist->jindex;
 318     jjnr             = nlist->jjnr;
 319     shiftidx         = nlist->shift;
 320     gid              = nlist->gid;
 321     shiftvec         = fr->shift_vec[0];
 322     fshift           = fr->fshift[0];
 323     facel            = fr->epsfac;
 324     charge           = mdatoms->chargeA;
 325     nvdwtype         = fr->ntype;
 326     vdwparam         = fr->nbfp;
 327     vdwtype          = mdatoms->typeA;
 328
 329     vftab            = kernel_data->table_elec_vdw->data;
 330     vftabscale       = kernel_data->table_elec_vdw->scale;
 331
 332     outeriter        = 0;
 333     inneriter        = 0;
 334
 335     /* Start outer loop over neighborlists */
 336     for(iidx=0; iidx<nri; iidx++)
 337     {
 338         /* Load shift vector for this list */
 339         i_shift_offset   = DIM*shiftidx[iidx];
 340         shX              = shiftvec[i_shift_offset+XX];
 341         shY              = shiftvec[i_shift_offset+YY];
 342         shZ              = shiftvec[i_shift_offset+ZZ];
 343
 344         /* Load limits for loop over neighbors */
 345         j_index_start    = jindex[iidx];
 346         j_index_end      = jindex[iidx+1];
 347
 348         /* Get outer coordinate index */
 349         inr              = iinr[iidx];
 350         i_coord_offset   = DIM*inr;
 351
 352         /* Load i particle coords and add shift vector */
 353         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 354         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 355         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 356
 357         fix0             = 0.0;
 358         fiy0             = 0.0;
 359         fiz0             = 0.0;
 360
 361         /* Load parameters for i particles */
 362         iq0              = facel*charge[inr+0];
 363         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 364
 365         /* Start inner kernel loop */
 366         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 367         {
 368             /* Get j neighbor index, and coordinate index */
 369             jnr              = jjnr[jidx];
 370             j_coord_offset   = DIM*jnr;
 371
 372             /* load j atom coordinates */
 373             jx0              = x[j_coord_offset+DIM*0+XX];
 374             jy0              = x[j_coord_offset+DIM*0+YY];
 375             jz0              = x[j_coord_offset+DIM*0+ZZ];
 376
 377             /* Calculate displacement vector */
 378             dx00             = ix0 - jx0;
 379             dy00             = iy0 - jy0;
 380             dz00             = iz0 - jz0;
 381
 382             /* Calculate squared distance and things based on it */
 383             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 384
 385             rinv00           = gmx_invsqrt(rsq00);
 386
 387             /* Load parameters for j particles */
 388             jq0              = charge[jnr+0];
 389             vdwjidx0         = 2*vdwtype[jnr+0];
 390
 391             /**************************
 392              * CALCULATE INTERACTIONS *
 393              **************************/
 394
 395             r00              = rsq00*rinv00;
 396
 397             qq00             = iq0*jq0;
 398             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 399             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 400
 401             /* Calculate table index by multiplying r with table scale and truncate to integer */
 402             rt               = r00*vftabscale;
 403             vfitab           = rt;
 404             vfeps            = rt-vfitab;
 405             vfitab           = 3*4*vfitab;
 406
 407             /* CUBIC SPLINE TABLE ELECTROSTATICS */
 408             F                = vftab[vfitab+1];
 409             Geps             = vfeps*vftab[vfitab+2];
 410             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 411             Fp               = F+Geps+Heps2;
 412             FF               = Fp+Geps+2.0*Heps2;
 413             felec            = -qq00*FF*vftabscale*rinv00;
 414
 415             /* CUBIC SPLINE TABLE DISPERSION */
 416             vfitab          += 4;
 417             F                = vftab[vfitab+1];
 418             Geps             = vfeps*vftab[vfitab+2];
 419             Heps2            = vfeps*vfeps*vftab[vfitab+3];
 420             Fp               = F+Geps+Heps2;
 421             FF               = Fp+Geps+2.0*Heps2;
 422             fvdw6            = c6_00*FF;
 423
 424             /* CUBIC SPLINE TABLE REPULSION */
 425             F                = vftab[vfitab+5];
 426             Geps             = vfeps*vftab[vfitab+6];
 427             Heps2            = vfeps*vfeps*vftab[vfitab+7];
 428             Fp               = F+Geps+Heps2;
 429             FF               = Fp+Geps+2.0*Heps2;
 430             fvdw12           = c12_00*FF;
 431             fvdw             = -(fvdw6+fvdw12)*vftabscale*rinv00;
 432
 433             fscal            = felec+fvdw;
 434
 435             /* Calculate temporary vectorial force */
 436             tx               = fscal*dx00;
 437             ty               = fscal*dy00;
 438             tz               = fscal*dz00;
 439
 440             /* Update vectorial force */
 441             fix0            += tx;
 442             fiy0            += ty;
 443             fiz0            += tz;
 444             f[j_coord_offset+DIM*0+XX] -= tx;
 445             f[j_coord_offset+DIM*0+YY] -= ty;
 446             f[j_coord_offset+DIM*0+ZZ] -= tz;
 447
 448             /* Inner loop uses 61 flops */
 449         }
 450         /* End of innermost loop */
 451
 452         tx = ty = tz = 0;
 453         f[i_coord_offset+DIM*0+XX] += fix0;
 454         f[i_coord_offset+DIM*0+YY] += fiy0;
 455         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 456         tx                         += fix0;
 457         ty                         += fiy0;
 458         tz                         += fiz0;
 459         fshift[i_shift_offset+XX]  += tx;
 460         fshift[i_shift_offset+YY]  += ty;
 461         fshift[i_shift_offset+ZZ]  += tz;
 462
 463         /* Increment number of inner iterations */
 464         inneriter                  += j_index_end - j_index_start;
 465
 466         /* Outer loop uses 13 flops */
 467     }
 468
 469     /* Increment number of outer iterations */
 470     outeriter        += nri;
 471
 472     /* Update outer/inner flops */
 473
 474     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*61);
 475 }