src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecNone_VdwLJSw_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_VdwLJSw_GeomP1P1_VF_c
  49  * Electrostatics interaction: None
  50  * VdW interaction:            LennardJones
  51  * Geometry:                   Particle-Particle
  52  * Calculate force/pot:        PotentialAndForce
  53  */
  54 void
  55 nb_kernel_ElecNone_VdwLJSw_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             rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
  80
  81     x                = xx[0];
  82     f                = ff[0];
  83
  84     nri              = nlist->nri;
  85     iinr             = nlist->iinr;
  86     jindex           = nlist->jindex;
  87     jjnr             = nlist->jjnr;
  88     shiftidx         = nlist->shift;
  89     gid              = nlist->gid;
  90     shiftvec         = fr->shift_vec[0];
  91     fshift           = fr->fshift[0];
  92     nvdwtype         = fr->ntype;
  93     vdwparam         = fr->nbfp;
  94     vdwtype          = mdatoms->typeA;
  95
  96     rcutoff          = fr->rvdw;
  97     rcutoff2         = rcutoff*rcutoff;
  98
  99     rswitch          = fr->rvdw_switch;
 100     /* Setup switch parameters */
 101     d                = rcutoff-rswitch;
 102     swV3             = -10.0/(d*d*d);
 103     swV4             =  15.0/(d*d*d*d);
 104     swV5             =  -6.0/(d*d*d*d*d);
 105     swF2             = -30.0/(d*d*d);
 106     swF3             =  60.0/(d*d*d*d);
 107     swF4             = -30.0/(d*d*d*d*d);
 108
 109     outeriter        = 0;
 110     inneriter        = 0;
 111
 112     /* Start outer loop over neighborlists */
 113     for(iidx=0; iidx<nri; iidx++)
 114     {
 115         /* Load shift vector for this list */
 116         i_shift_offset   = DIM*shiftidx[iidx];
 117         shX              = shiftvec[i_shift_offset+XX];
 118         shY              = shiftvec[i_shift_offset+YY];
 119         shZ              = shiftvec[i_shift_offset+ZZ];
 120
 121         /* Load limits for loop over neighbors */
 122         j_index_start    = jindex[iidx];
 123         j_index_end      = jindex[iidx+1];
 124
 125         /* Get outer coordinate index */
 126         inr              = iinr[iidx];
 127         i_coord_offset   = DIM*inr;
 128
 129         /* Load i particle coords and add shift vector */
 130         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 131         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 132         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 133
 134         fix0             = 0.0;
 135         fiy0             = 0.0;
 136         fiz0             = 0.0;
 137
 138         /* Load parameters for i particles */
 139         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 140
 141         /* Reset potential sums */
 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             vdwjidx0         = 2*vdwtype[jnr+0];
 170
 171             /**************************
 172              * CALCULATE INTERACTIONS *
 173              **************************/
 174
 175             if (rsq00<rcutoff2)
 176             {
 177
 178             r00              = rsq00*rinv00;
 179
 180             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 181             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 182
 183             /* LENNARD-JONES DISPERSION/REPULSION */
 184
 185             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 186             vvdw6            = c6_00*rinvsix;
 187             vvdw12           = c12_00*rinvsix*rinvsix;
 188             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 189             fvdw             = (vvdw12-vvdw6)*rinvsq00;
 190
 191             d                = r00-rswitch;
 192             d                = (d>0.0) ? d : 0.0;
 193             d2               = d*d;
 194             sw               = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
 195
 196             dsw              = d2*(swF2+d*(swF3+d*swF4));
 197
 198             /* Evaluate switch function */
 199             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 200             fvdw             = fvdw*sw - rinv00*vvdw*dsw;
 201             vvdw            *= sw;
 202
 203             /* Update potential sums from outer loop */
 204             vvdwsum         += vvdw;
 205
 206             fscal            = 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             }
 222
 223             /* Inner loop uses 53 flops */
 224         }
 225         /* End of innermost loop */
 226
 227         tx = ty = tz = 0;
 228         f[i_coord_offset+DIM*0+XX] += fix0;
 229         f[i_coord_offset+DIM*0+YY] += fiy0;
 230         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 231         tx                         += fix0;
 232         ty                         += fiy0;
 233         tz                         += fiz0;
 234         fshift[i_shift_offset+XX]  += tx;
 235         fshift[i_shift_offset+YY]  += ty;
 236         fshift[i_shift_offset+ZZ]  += tz;
 237
 238         ggid                        = gid[iidx];
 239         /* Update potential energies */
 240         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 241
 242         /* Increment number of inner iterations */
 243         inneriter                  += j_index_end - j_index_start;
 244
 245         /* Outer loop uses 13 flops */
 246     }
 247
 248     /* Increment number of outer iterations */
 249     outeriter        += nri;
 250
 251     /* Update outer/inner flops */
 252
 253     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_VF,outeriter*13 + inneriter*53);
 254 }
 255 /*
 256  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwLJSw_GeomP1P1_F_c
 257  * Electrostatics interaction: None
 258  * VdW interaction:            LennardJones
 259  * Geometry:                   Particle-Particle
 260  * Calculate force/pot:        Force
 261  */
 262 void
 263 nb_kernel_ElecNone_VdwLJSw_GeomP1P1_F_c
 264                     (t_nblist                    * gmx_restrict       nlist,
 265                      rvec                        * gmx_restrict          xx,
 266                      rvec                        * gmx_restrict          ff,
 267                      t_forcerec                  * gmx_restrict          fr,
 268                      t_mdatoms                   * gmx_restrict     mdatoms,
 269                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 270                      t_nrnb                      * gmx_restrict        nrnb)
 271 {
 272     int              i_shift_offset,i_coord_offset,j_coord_offset;
 273     int              j_index_start,j_index_end;
 274     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 275     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 276     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 277     real             *shiftvec,*fshift,*x,*f;
 278     int              vdwioffset0;
 279     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 280     int              vdwjidx0;
 281     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 282     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 283     int              nvdwtype;
 284     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 285     int              *vdwtype;
 286     real             *vdwparam;
 287     real             rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
 288
 289     x                = xx[0];
 290     f                = ff[0];
 291
 292     nri              = nlist->nri;
 293     iinr             = nlist->iinr;
 294     jindex           = nlist->jindex;
 295     jjnr             = nlist->jjnr;
 296     shiftidx         = nlist->shift;
 297     gid              = nlist->gid;
 298     shiftvec         = fr->shift_vec[0];
 299     fshift           = fr->fshift[0];
 300     nvdwtype         = fr->ntype;
 301     vdwparam         = fr->nbfp;
 302     vdwtype          = mdatoms->typeA;
 303
 304     rcutoff          = fr->rvdw;
 305     rcutoff2         = rcutoff*rcutoff;
 306
 307     rswitch          = fr->rvdw_switch;
 308     /* Setup switch parameters */
 309     d                = rcutoff-rswitch;
 310     swV3             = -10.0/(d*d*d);
 311     swV4             =  15.0/(d*d*d*d);
 312     swV5             =  -6.0/(d*d*d*d*d);
 313     swF2             = -30.0/(d*d*d);
 314     swF3             =  60.0/(d*d*d*d);
 315     swF4             = -30.0/(d*d*d*d*d);
 316
 317     outeriter        = 0;
 318     inneriter        = 0;
 319
 320     /* Start outer loop over neighborlists */
 321     for(iidx=0; iidx<nri; iidx++)
 322     {
 323         /* Load shift vector for this list */
 324         i_shift_offset   = DIM*shiftidx[iidx];
 325         shX              = shiftvec[i_shift_offset+XX];
 326         shY              = shiftvec[i_shift_offset+YY];
 327         shZ              = shiftvec[i_shift_offset+ZZ];
 328
 329         /* Load limits for loop over neighbors */
 330         j_index_start    = jindex[iidx];
 331         j_index_end      = jindex[iidx+1];
 332
 333         /* Get outer coordinate index */
 334         inr              = iinr[iidx];
 335         i_coord_offset   = DIM*inr;
 336
 337         /* Load i particle coords and add shift vector */
 338         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 339         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 340         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 341
 342         fix0             = 0.0;
 343         fiy0             = 0.0;
 344         fiz0             = 0.0;
 345
 346         /* Load parameters for i particles */
 347         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 348
 349         /* Start inner kernel loop */
 350         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 351         {
 352             /* Get j neighbor index, and coordinate index */
 353             jnr              = jjnr[jidx];
 354             j_coord_offset   = DIM*jnr;
 355
 356             /* load j atom coordinates */
 357             jx0              = x[j_coord_offset+DIM*0+XX];
 358             jy0              = x[j_coord_offset+DIM*0+YY];
 359             jz0              = x[j_coord_offset+DIM*0+ZZ];
 360
 361             /* Calculate displacement vector */
 362             dx00             = ix0 - jx0;
 363             dy00             = iy0 - jy0;
 364             dz00             = iz0 - jz0;
 365
 366             /* Calculate squared distance and things based on it */
 367             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 368
 369             rinv00           = gmx_invsqrt(rsq00);
 370
 371             rinvsq00         = rinv00*rinv00;
 372
 373             /* Load parameters for j particles */
 374             vdwjidx0         = 2*vdwtype[jnr+0];
 375
 376             /**************************
 377              * CALCULATE INTERACTIONS *
 378              **************************/
 379
 380             if (rsq00<rcutoff2)
 381             {
 382
 383             r00              = rsq00*rinv00;
 384
 385             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 386             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 387
 388             /* LENNARD-JONES DISPERSION/REPULSION */
 389
 390             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 391             vvdw6            = c6_00*rinvsix;
 392             vvdw12           = c12_00*rinvsix*rinvsix;
 393             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 394             fvdw             = (vvdw12-vvdw6)*rinvsq00;
 395
 396             d                = r00-rswitch;
 397             d                = (d>0.0) ? d : 0.0;
 398             d2               = d*d;
 399             sw               = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
 400
 401             dsw              = d2*(swF2+d*(swF3+d*swF4));
 402
 403             /* Evaluate switch function */
 404             /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
 405             fvdw             = fvdw*sw - rinv00*vvdw*dsw;
 406
 407             fscal            = fvdw;
 408
 409             /* Calculate temporary vectorial force */
 410             tx               = fscal*dx00;
 411             ty               = fscal*dy00;
 412             tz               = fscal*dz00;
 413
 414             /* Update vectorial force */
 415             fix0            += tx;
 416             fiy0            += ty;
 417             fiz0            += tz;
 418             f[j_coord_offset+DIM*0+XX] -= tx;
 419             f[j_coord_offset+DIM*0+YY] -= ty;
 420             f[j_coord_offset+DIM*0+ZZ] -= tz;
 421
 422             }
 423
 424             /* Inner loop uses 51 flops */
 425         }
 426         /* End of innermost loop */
 427
 428         tx = ty = tz = 0;
 429         f[i_coord_offset+DIM*0+XX] += fix0;
 430         f[i_coord_offset+DIM*0+YY] += fiy0;
 431         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 432         tx                         += fix0;
 433         ty                         += fiy0;
 434         tz                         += fiz0;
 435         fshift[i_shift_offset+XX]  += tx;
 436         fshift[i_shift_offset+YY]  += ty;
 437         fshift[i_shift_offset+ZZ]  += tz;
 438
 439         /* Increment number of inner iterations */
 440         inneriter                  += j_index_end - j_index_start;
 441
 442         /* Outer loop uses 12 flops */
 443     }
 444
 445     /* Increment number of outer iterations */
 446     outeriter        += nri;
 447
 448     /* Update outer/inner flops */
 449
 450     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_F,outeriter*12 + inneriter*51);
 451 }