2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2012,2013, 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.
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.
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.
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.
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.
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.
36 * Note: this file was generated by the GROMACS c kernel generator.
44 #include "../nb_kernel.h"
45 #include "types/simple.h"
50 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_VF_c
51 * Electrostatics interaction: ReactionField
52 * VdW interaction: Buckingham
53 * Geometry: Water3-Particle
54 * Calculate force/pot: PotentialAndForce
57 nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_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)
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;
73 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
75 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
77 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
79 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
80 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
81 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
82 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
83 real velec,felec,velecsum,facel,crf,krf,krf2;
86 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
89 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
96 jindex = nlist->jindex;
98 shiftidx = nlist->shift;
100 shiftvec = fr->shift_vec[0];
101 fshift = fr->fshift[0];
103 charge = mdatoms->chargeA;
107 nvdwtype = fr->ntype;
109 vdwtype = mdatoms->typeA;
111 /* Setup water-specific parameters */
112 inr = nlist->iinr[0];
113 iq0 = facel*charge[inr+0];
114 iq1 = facel*charge[inr+1];
115 iq2 = facel*charge[inr+2];
116 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
118 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
119 rcutoff = fr->rcoulomb;
120 rcutoff2 = rcutoff*rcutoff;
122 rswitch = fr->rvdw_switch;
123 /* Setup switch parameters */
125 swV3 = -10.0/(d*d*d);
126 swV4 = 15.0/(d*d*d*d);
127 swV5 = -6.0/(d*d*d*d*d);
128 swF2 = -30.0/(d*d*d);
129 swF3 = 60.0/(d*d*d*d);
130 swF4 = -30.0/(d*d*d*d*d);
135 /* Start outer loop over neighborlists */
136 for(iidx=0; iidx<nri; iidx++)
138 /* Load shift vector for this list */
139 i_shift_offset = DIM*shiftidx[iidx];
140 shX = shiftvec[i_shift_offset+XX];
141 shY = shiftvec[i_shift_offset+YY];
142 shZ = shiftvec[i_shift_offset+ZZ];
144 /* Load limits for loop over neighbors */
145 j_index_start = jindex[iidx];
146 j_index_end = jindex[iidx+1];
148 /* Get outer coordinate index */
150 i_coord_offset = DIM*inr;
152 /* Load i particle coords and add shift vector */
153 ix0 = shX + x[i_coord_offset+DIM*0+XX];
154 iy0 = shY + x[i_coord_offset+DIM*0+YY];
155 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
156 ix1 = shX + x[i_coord_offset+DIM*1+XX];
157 iy1 = shY + x[i_coord_offset+DIM*1+YY];
158 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
159 ix2 = shX + x[i_coord_offset+DIM*2+XX];
160 iy2 = shY + x[i_coord_offset+DIM*2+YY];
161 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
173 /* Reset potential sums */
177 /* Start inner kernel loop */
178 for(jidx=j_index_start; jidx<j_index_end; jidx++)
180 /* Get j neighbor index, and coordinate index */
182 j_coord_offset = DIM*jnr;
184 /* load j atom coordinates */
185 jx0 = x[j_coord_offset+DIM*0+XX];
186 jy0 = x[j_coord_offset+DIM*0+YY];
187 jz0 = x[j_coord_offset+DIM*0+ZZ];
189 /* Calculate displacement vector */
200 /* Calculate squared distance and things based on it */
201 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
202 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
203 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
205 rinv00 = gmx_invsqrt(rsq00);
206 rinv10 = gmx_invsqrt(rsq10);
207 rinv20 = gmx_invsqrt(rsq20);
209 rinvsq00 = rinv00*rinv00;
210 rinvsq10 = rinv10*rinv10;
211 rinvsq20 = rinv20*rinv20;
213 /* Load parameters for j particles */
215 vdwjidx0 = 3*vdwtype[jnr+0];
217 /**************************
218 * CALCULATE INTERACTIONS *
219 **************************/
227 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
228 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
229 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
231 /* REACTION-FIELD ELECTROSTATICS */
232 velec = qq00*(rinv00+krf*rsq00-crf);
233 felec = qq00*(rinv00*rinvsq00-krf2);
235 /* BUCKINGHAM DISPERSION/REPULSION */
236 rinvsix = rinvsq00*rinvsq00*rinvsq00;
237 vvdw6 = c6_00*rinvsix;
239 vvdwexp = cexp1_00*exp(-br);
240 vvdw = vvdwexp - vvdw6*(1.0/6.0);
241 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
244 d = (d>0.0) ? d : 0.0;
246 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
248 dsw = d2*(swF2+d*(swF3+d*swF4));
250 /* Evaluate switch function */
251 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
252 fvdw = fvdw*sw - rinv00*vvdw*dsw;
255 /* Update potential sums from outer loop */
261 /* Calculate temporary vectorial force */
266 /* Update vectorial force */
270 f[j_coord_offset+DIM*0+XX] -= tx;
271 f[j_coord_offset+DIM*0+YY] -= ty;
272 f[j_coord_offset+DIM*0+ZZ] -= tz;
276 /**************************
277 * CALCULATE INTERACTIONS *
278 **************************/
285 /* REACTION-FIELD ELECTROSTATICS */
286 velec = qq10*(rinv10+krf*rsq10-crf);
287 felec = qq10*(rinv10*rinvsq10-krf2);
289 /* Update potential sums from outer loop */
294 /* Calculate temporary vectorial force */
299 /* Update vectorial force */
303 f[j_coord_offset+DIM*0+XX] -= tx;
304 f[j_coord_offset+DIM*0+YY] -= ty;
305 f[j_coord_offset+DIM*0+ZZ] -= tz;
309 /**************************
310 * CALCULATE INTERACTIONS *
311 **************************/
318 /* REACTION-FIELD ELECTROSTATICS */
319 velec = qq20*(rinv20+krf*rsq20-crf);
320 felec = qq20*(rinv20*rinvsq20-krf2);
322 /* Update potential sums from outer loop */
327 /* Calculate temporary vectorial force */
332 /* Update vectorial force */
336 f[j_coord_offset+DIM*0+XX] -= tx;
337 f[j_coord_offset+DIM*0+YY] -= ty;
338 f[j_coord_offset+DIM*0+ZZ] -= tz;
342 /* Inner loop uses 153 flops */
344 /* End of innermost loop */
347 f[i_coord_offset+DIM*0+XX] += fix0;
348 f[i_coord_offset+DIM*0+YY] += fiy0;
349 f[i_coord_offset+DIM*0+ZZ] += fiz0;
353 f[i_coord_offset+DIM*1+XX] += fix1;
354 f[i_coord_offset+DIM*1+YY] += fiy1;
355 f[i_coord_offset+DIM*1+ZZ] += fiz1;
359 f[i_coord_offset+DIM*2+XX] += fix2;
360 f[i_coord_offset+DIM*2+YY] += fiy2;
361 f[i_coord_offset+DIM*2+ZZ] += fiz2;
365 fshift[i_shift_offset+XX] += tx;
366 fshift[i_shift_offset+YY] += ty;
367 fshift[i_shift_offset+ZZ] += tz;
370 /* Update potential energies */
371 kernel_data->energygrp_elec[ggid] += velecsum;
372 kernel_data->energygrp_vdw[ggid] += vvdwsum;
374 /* Increment number of inner iterations */
375 inneriter += j_index_end - j_index_start;
377 /* Outer loop uses 32 flops */
380 /* Increment number of outer iterations */
383 /* Update outer/inner flops */
385 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_VF,outeriter*32 + inneriter*153);
388 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_F_c
389 * Electrostatics interaction: ReactionField
390 * VdW interaction: Buckingham
391 * Geometry: Water3-Particle
392 * Calculate force/pot: Force
395 nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_F_c
396 (t_nblist * gmx_restrict nlist,
397 rvec * gmx_restrict xx,
398 rvec * gmx_restrict ff,
399 t_forcerec * gmx_restrict fr,
400 t_mdatoms * gmx_restrict mdatoms,
401 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
402 t_nrnb * gmx_restrict nrnb)
404 int i_shift_offset,i_coord_offset,j_coord_offset;
405 int j_index_start,j_index_end;
406 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
407 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
408 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
409 real *shiftvec,*fshift,*x,*f;
411 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
413 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
415 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
417 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
418 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
419 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
420 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
421 real velec,felec,velecsum,facel,crf,krf,krf2;
424 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
427 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
434 jindex = nlist->jindex;
436 shiftidx = nlist->shift;
438 shiftvec = fr->shift_vec[0];
439 fshift = fr->fshift[0];
441 charge = mdatoms->chargeA;
445 nvdwtype = fr->ntype;
447 vdwtype = mdatoms->typeA;
449 /* Setup water-specific parameters */
450 inr = nlist->iinr[0];
451 iq0 = facel*charge[inr+0];
452 iq1 = facel*charge[inr+1];
453 iq2 = facel*charge[inr+2];
454 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
456 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
457 rcutoff = fr->rcoulomb;
458 rcutoff2 = rcutoff*rcutoff;
460 rswitch = fr->rvdw_switch;
461 /* Setup switch parameters */
463 swV3 = -10.0/(d*d*d);
464 swV4 = 15.0/(d*d*d*d);
465 swV5 = -6.0/(d*d*d*d*d);
466 swF2 = -30.0/(d*d*d);
467 swF3 = 60.0/(d*d*d*d);
468 swF4 = -30.0/(d*d*d*d*d);
473 /* Start outer loop over neighborlists */
474 for(iidx=0; iidx<nri; iidx++)
476 /* Load shift vector for this list */
477 i_shift_offset = DIM*shiftidx[iidx];
478 shX = shiftvec[i_shift_offset+XX];
479 shY = shiftvec[i_shift_offset+YY];
480 shZ = shiftvec[i_shift_offset+ZZ];
482 /* Load limits for loop over neighbors */
483 j_index_start = jindex[iidx];
484 j_index_end = jindex[iidx+1];
486 /* Get outer coordinate index */
488 i_coord_offset = DIM*inr;
490 /* Load i particle coords and add shift vector */
491 ix0 = shX + x[i_coord_offset+DIM*0+XX];
492 iy0 = shY + x[i_coord_offset+DIM*0+YY];
493 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
494 ix1 = shX + x[i_coord_offset+DIM*1+XX];
495 iy1 = shY + x[i_coord_offset+DIM*1+YY];
496 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
497 ix2 = shX + x[i_coord_offset+DIM*2+XX];
498 iy2 = shY + x[i_coord_offset+DIM*2+YY];
499 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
511 /* Start inner kernel loop */
512 for(jidx=j_index_start; jidx<j_index_end; jidx++)
514 /* Get j neighbor index, and coordinate index */
516 j_coord_offset = DIM*jnr;
518 /* load j atom coordinates */
519 jx0 = x[j_coord_offset+DIM*0+XX];
520 jy0 = x[j_coord_offset+DIM*0+YY];
521 jz0 = x[j_coord_offset+DIM*0+ZZ];
523 /* Calculate displacement vector */
534 /* Calculate squared distance and things based on it */
535 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
536 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
537 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
539 rinv00 = gmx_invsqrt(rsq00);
540 rinv10 = gmx_invsqrt(rsq10);
541 rinv20 = gmx_invsqrt(rsq20);
543 rinvsq00 = rinv00*rinv00;
544 rinvsq10 = rinv10*rinv10;
545 rinvsq20 = rinv20*rinv20;
547 /* Load parameters for j particles */
549 vdwjidx0 = 3*vdwtype[jnr+0];
551 /**************************
552 * CALCULATE INTERACTIONS *
553 **************************/
561 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
562 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
563 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
565 /* REACTION-FIELD ELECTROSTATICS */
566 felec = qq00*(rinv00*rinvsq00-krf2);
568 /* BUCKINGHAM DISPERSION/REPULSION */
569 rinvsix = rinvsq00*rinvsq00*rinvsq00;
570 vvdw6 = c6_00*rinvsix;
572 vvdwexp = cexp1_00*exp(-br);
573 vvdw = vvdwexp - vvdw6*(1.0/6.0);
574 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
577 d = (d>0.0) ? d : 0.0;
579 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
581 dsw = d2*(swF2+d*(swF3+d*swF4));
583 /* Evaluate switch function */
584 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
585 fvdw = fvdw*sw - rinv00*vvdw*dsw;
589 /* Calculate temporary vectorial force */
594 /* Update vectorial force */
598 f[j_coord_offset+DIM*0+XX] -= tx;
599 f[j_coord_offset+DIM*0+YY] -= ty;
600 f[j_coord_offset+DIM*0+ZZ] -= tz;
604 /**************************
605 * CALCULATE INTERACTIONS *
606 **************************/
613 /* REACTION-FIELD ELECTROSTATICS */
614 felec = qq10*(rinv10*rinvsq10-krf2);
618 /* Calculate temporary vectorial force */
623 /* Update vectorial force */
627 f[j_coord_offset+DIM*0+XX] -= tx;
628 f[j_coord_offset+DIM*0+YY] -= ty;
629 f[j_coord_offset+DIM*0+ZZ] -= tz;
633 /**************************
634 * CALCULATE INTERACTIONS *
635 **************************/
642 /* REACTION-FIELD ELECTROSTATICS */
643 felec = qq20*(rinv20*rinvsq20-krf2);
647 /* Calculate temporary vectorial force */
652 /* Update vectorial force */
656 f[j_coord_offset+DIM*0+XX] -= tx;
657 f[j_coord_offset+DIM*0+YY] -= ty;
658 f[j_coord_offset+DIM*0+ZZ] -= tz;
662 /* Inner loop uses 136 flops */
664 /* End of innermost loop */
667 f[i_coord_offset+DIM*0+XX] += fix0;
668 f[i_coord_offset+DIM*0+YY] += fiy0;
669 f[i_coord_offset+DIM*0+ZZ] += fiz0;
673 f[i_coord_offset+DIM*1+XX] += fix1;
674 f[i_coord_offset+DIM*1+YY] += fiy1;
675 f[i_coord_offset+DIM*1+ZZ] += fiz1;
679 f[i_coord_offset+DIM*2+XX] += fix2;
680 f[i_coord_offset+DIM*2+YY] += fiy2;
681 f[i_coord_offset+DIM*2+ZZ] += fiz2;
685 fshift[i_shift_offset+XX] += tx;
686 fshift[i_shift_offset+YY] += ty;
687 fshift[i_shift_offset+ZZ] += tz;
689 /* Increment number of inner iterations */
690 inneriter += j_index_end - j_index_start;
692 /* Outer loop uses 30 flops */
695 /* Increment number of outer iterations */
698 /* Update outer/inner flops */
700 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_F,outeriter*30 + inneriter*136);