2 * This file is part of the GROMACS molecular simulation package.
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.
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.
42 #include "../nb_kernel.h"
43 #include "types/simple.h"
44 #include "gromacs/math/vec.h"
48 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_VF_c
49 * Electrostatics interaction: ReactionField
50 * VdW interaction: Buckingham
51 * Geometry: Water3-Particle
52 * Calculate force/pot: PotentialAndForce
55 nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_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)
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;
71 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
73 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
75 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
77 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
78 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
79 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
80 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
81 real velec,felec,velecsum,facel,crf,krf,krf2;
84 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
87 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
94 jindex = nlist->jindex;
96 shiftidx = nlist->shift;
98 shiftvec = fr->shift_vec[0];
99 fshift = fr->fshift[0];
101 charge = mdatoms->chargeA;
105 nvdwtype = fr->ntype;
107 vdwtype = mdatoms->typeA;
109 /* Setup water-specific parameters */
110 inr = nlist->iinr[0];
111 iq0 = facel*charge[inr+0];
112 iq1 = facel*charge[inr+1];
113 iq2 = facel*charge[inr+2];
114 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
116 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
117 rcutoff = fr->rcoulomb;
118 rcutoff2 = rcutoff*rcutoff;
120 rswitch = fr->rvdw_switch;
121 /* Setup switch parameters */
123 swV3 = -10.0/(d*d*d);
124 swV4 = 15.0/(d*d*d*d);
125 swV5 = -6.0/(d*d*d*d*d);
126 swF2 = -30.0/(d*d*d);
127 swF3 = 60.0/(d*d*d*d);
128 swF4 = -30.0/(d*d*d*d*d);
133 /* Start outer loop over neighborlists */
134 for(iidx=0; iidx<nri; iidx++)
136 /* Load shift vector for this list */
137 i_shift_offset = DIM*shiftidx[iidx];
138 shX = shiftvec[i_shift_offset+XX];
139 shY = shiftvec[i_shift_offset+YY];
140 shZ = shiftvec[i_shift_offset+ZZ];
142 /* Load limits for loop over neighbors */
143 j_index_start = jindex[iidx];
144 j_index_end = jindex[iidx+1];
146 /* Get outer coordinate index */
148 i_coord_offset = DIM*inr;
150 /* Load i particle coords and add shift vector */
151 ix0 = shX + x[i_coord_offset+DIM*0+XX];
152 iy0 = shY + x[i_coord_offset+DIM*0+YY];
153 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
154 ix1 = shX + x[i_coord_offset+DIM*1+XX];
155 iy1 = shY + x[i_coord_offset+DIM*1+YY];
156 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
157 ix2 = shX + x[i_coord_offset+DIM*2+XX];
158 iy2 = shY + x[i_coord_offset+DIM*2+YY];
159 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
171 /* Reset potential sums */
175 /* Start inner kernel loop */
176 for(jidx=j_index_start; jidx<j_index_end; jidx++)
178 /* Get j neighbor index, and coordinate index */
180 j_coord_offset = DIM*jnr;
182 /* load j atom coordinates */
183 jx0 = x[j_coord_offset+DIM*0+XX];
184 jy0 = x[j_coord_offset+DIM*0+YY];
185 jz0 = x[j_coord_offset+DIM*0+ZZ];
187 /* Calculate displacement vector */
198 /* Calculate squared distance and things based on it */
199 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
200 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
201 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
203 rinv00 = gmx_invsqrt(rsq00);
204 rinv10 = gmx_invsqrt(rsq10);
205 rinv20 = gmx_invsqrt(rsq20);
207 rinvsq00 = rinv00*rinv00;
208 rinvsq10 = rinv10*rinv10;
209 rinvsq20 = rinv20*rinv20;
211 /* Load parameters for j particles */
213 vdwjidx0 = 3*vdwtype[jnr+0];
215 /**************************
216 * CALCULATE INTERACTIONS *
217 **************************/
225 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
226 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
227 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
229 /* REACTION-FIELD ELECTROSTATICS */
230 velec = qq00*(rinv00+krf*rsq00-crf);
231 felec = qq00*(rinv00*rinvsq00-krf2);
233 /* BUCKINGHAM DISPERSION/REPULSION */
234 rinvsix = rinvsq00*rinvsq00*rinvsq00;
235 vvdw6 = c6_00*rinvsix;
237 vvdwexp = cexp1_00*exp(-br);
238 vvdw = vvdwexp - vvdw6*(1.0/6.0);
239 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
242 d = (d>0.0) ? d : 0.0;
244 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
246 dsw = d2*(swF2+d*(swF3+d*swF4));
248 /* Evaluate switch function */
249 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
250 fvdw = fvdw*sw - rinv00*vvdw*dsw;
253 /* Update potential sums from outer loop */
259 /* Calculate temporary vectorial force */
264 /* Update vectorial force */
268 f[j_coord_offset+DIM*0+XX] -= tx;
269 f[j_coord_offset+DIM*0+YY] -= ty;
270 f[j_coord_offset+DIM*0+ZZ] -= tz;
274 /**************************
275 * CALCULATE INTERACTIONS *
276 **************************/
283 /* REACTION-FIELD ELECTROSTATICS */
284 velec = qq10*(rinv10+krf*rsq10-crf);
285 felec = qq10*(rinv10*rinvsq10-krf2);
287 /* Update potential sums from outer loop */
292 /* Calculate temporary vectorial force */
297 /* Update vectorial force */
301 f[j_coord_offset+DIM*0+XX] -= tx;
302 f[j_coord_offset+DIM*0+YY] -= ty;
303 f[j_coord_offset+DIM*0+ZZ] -= tz;
307 /**************************
308 * CALCULATE INTERACTIONS *
309 **************************/
316 /* REACTION-FIELD ELECTROSTATICS */
317 velec = qq20*(rinv20+krf*rsq20-crf);
318 felec = qq20*(rinv20*rinvsq20-krf2);
320 /* Update potential sums from outer loop */
325 /* Calculate temporary vectorial force */
330 /* Update vectorial force */
334 f[j_coord_offset+DIM*0+XX] -= tx;
335 f[j_coord_offset+DIM*0+YY] -= ty;
336 f[j_coord_offset+DIM*0+ZZ] -= tz;
340 /* Inner loop uses 153 flops */
342 /* End of innermost loop */
345 f[i_coord_offset+DIM*0+XX] += fix0;
346 f[i_coord_offset+DIM*0+YY] += fiy0;
347 f[i_coord_offset+DIM*0+ZZ] += fiz0;
351 f[i_coord_offset+DIM*1+XX] += fix1;
352 f[i_coord_offset+DIM*1+YY] += fiy1;
353 f[i_coord_offset+DIM*1+ZZ] += fiz1;
357 f[i_coord_offset+DIM*2+XX] += fix2;
358 f[i_coord_offset+DIM*2+YY] += fiy2;
359 f[i_coord_offset+DIM*2+ZZ] += fiz2;
363 fshift[i_shift_offset+XX] += tx;
364 fshift[i_shift_offset+YY] += ty;
365 fshift[i_shift_offset+ZZ] += tz;
368 /* Update potential energies */
369 kernel_data->energygrp_elec[ggid] += velecsum;
370 kernel_data->energygrp_vdw[ggid] += vvdwsum;
372 /* Increment number of inner iterations */
373 inneriter += j_index_end - j_index_start;
375 /* Outer loop uses 32 flops */
378 /* Increment number of outer iterations */
381 /* Update outer/inner flops */
383 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_VF,outeriter*32 + inneriter*153);
386 * Gromacs nonbonded kernel: nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_F_c
387 * Electrostatics interaction: ReactionField
388 * VdW interaction: Buckingham
389 * Geometry: Water3-Particle
390 * Calculate force/pot: Force
393 nb_kernel_ElecRFCut_VdwBhamSw_GeomW3P1_F_c
394 (t_nblist * gmx_restrict nlist,
395 rvec * gmx_restrict xx,
396 rvec * gmx_restrict ff,
397 t_forcerec * gmx_restrict fr,
398 t_mdatoms * gmx_restrict mdatoms,
399 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
400 t_nrnb * gmx_restrict nrnb)
402 int i_shift_offset,i_coord_offset,j_coord_offset;
403 int j_index_start,j_index_end;
404 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
405 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
406 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
407 real *shiftvec,*fshift,*x,*f;
409 real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
411 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
413 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
415 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
416 real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
417 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
418 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
419 real velec,felec,velecsum,facel,crf,krf,krf2;
422 real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
425 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
432 jindex = nlist->jindex;
434 shiftidx = nlist->shift;
436 shiftvec = fr->shift_vec[0];
437 fshift = fr->fshift[0];
439 charge = mdatoms->chargeA;
443 nvdwtype = fr->ntype;
445 vdwtype = mdatoms->typeA;
447 /* Setup water-specific parameters */
448 inr = nlist->iinr[0];
449 iq0 = facel*charge[inr+0];
450 iq1 = facel*charge[inr+1];
451 iq2 = facel*charge[inr+2];
452 vdwioffset0 = 3*nvdwtype*vdwtype[inr+0];
454 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
455 rcutoff = fr->rcoulomb;
456 rcutoff2 = rcutoff*rcutoff;
458 rswitch = fr->rvdw_switch;
459 /* Setup switch parameters */
461 swV3 = -10.0/(d*d*d);
462 swV4 = 15.0/(d*d*d*d);
463 swV5 = -6.0/(d*d*d*d*d);
464 swF2 = -30.0/(d*d*d);
465 swF3 = 60.0/(d*d*d*d);
466 swF4 = -30.0/(d*d*d*d*d);
471 /* Start outer loop over neighborlists */
472 for(iidx=0; iidx<nri; iidx++)
474 /* Load shift vector for this list */
475 i_shift_offset = DIM*shiftidx[iidx];
476 shX = shiftvec[i_shift_offset+XX];
477 shY = shiftvec[i_shift_offset+YY];
478 shZ = shiftvec[i_shift_offset+ZZ];
480 /* Load limits for loop over neighbors */
481 j_index_start = jindex[iidx];
482 j_index_end = jindex[iidx+1];
484 /* Get outer coordinate index */
486 i_coord_offset = DIM*inr;
488 /* Load i particle coords and add shift vector */
489 ix0 = shX + x[i_coord_offset+DIM*0+XX];
490 iy0 = shY + x[i_coord_offset+DIM*0+YY];
491 iz0 = shZ + x[i_coord_offset+DIM*0+ZZ];
492 ix1 = shX + x[i_coord_offset+DIM*1+XX];
493 iy1 = shY + x[i_coord_offset+DIM*1+YY];
494 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
495 ix2 = shX + x[i_coord_offset+DIM*2+XX];
496 iy2 = shY + x[i_coord_offset+DIM*2+YY];
497 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
509 /* Start inner kernel loop */
510 for(jidx=j_index_start; jidx<j_index_end; jidx++)
512 /* Get j neighbor index, and coordinate index */
514 j_coord_offset = DIM*jnr;
516 /* load j atom coordinates */
517 jx0 = x[j_coord_offset+DIM*0+XX];
518 jy0 = x[j_coord_offset+DIM*0+YY];
519 jz0 = x[j_coord_offset+DIM*0+ZZ];
521 /* Calculate displacement vector */
532 /* Calculate squared distance and things based on it */
533 rsq00 = dx00*dx00+dy00*dy00+dz00*dz00;
534 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
535 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
537 rinv00 = gmx_invsqrt(rsq00);
538 rinv10 = gmx_invsqrt(rsq10);
539 rinv20 = gmx_invsqrt(rsq20);
541 rinvsq00 = rinv00*rinv00;
542 rinvsq10 = rinv10*rinv10;
543 rinvsq20 = rinv20*rinv20;
545 /* Load parameters for j particles */
547 vdwjidx0 = 3*vdwtype[jnr+0];
549 /**************************
550 * CALCULATE INTERACTIONS *
551 **************************/
559 c6_00 = vdwparam[vdwioffset0+vdwjidx0];
560 cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1];
561 cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2];
563 /* REACTION-FIELD ELECTROSTATICS */
564 felec = qq00*(rinv00*rinvsq00-krf2);
566 /* BUCKINGHAM DISPERSION/REPULSION */
567 rinvsix = rinvsq00*rinvsq00*rinvsq00;
568 vvdw6 = c6_00*rinvsix;
570 vvdwexp = cexp1_00*exp(-br);
571 vvdw = vvdwexp - vvdw6*(1.0/6.0);
572 fvdw = (br*vvdwexp-vvdw6)*rinvsq00;
575 d = (d>0.0) ? d : 0.0;
577 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
579 dsw = d2*(swF2+d*(swF3+d*swF4));
581 /* Evaluate switch function */
582 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
583 fvdw = fvdw*sw - rinv00*vvdw*dsw;
587 /* Calculate temporary vectorial force */
592 /* Update vectorial force */
596 f[j_coord_offset+DIM*0+XX] -= tx;
597 f[j_coord_offset+DIM*0+YY] -= ty;
598 f[j_coord_offset+DIM*0+ZZ] -= tz;
602 /**************************
603 * CALCULATE INTERACTIONS *
604 **************************/
611 /* REACTION-FIELD ELECTROSTATICS */
612 felec = qq10*(rinv10*rinvsq10-krf2);
616 /* Calculate temporary vectorial force */
621 /* Update vectorial force */
625 f[j_coord_offset+DIM*0+XX] -= tx;
626 f[j_coord_offset+DIM*0+YY] -= ty;
627 f[j_coord_offset+DIM*0+ZZ] -= tz;
631 /**************************
632 * CALCULATE INTERACTIONS *
633 **************************/
640 /* REACTION-FIELD ELECTROSTATICS */
641 felec = qq20*(rinv20*rinvsq20-krf2);
645 /* Calculate temporary vectorial force */
650 /* Update vectorial force */
654 f[j_coord_offset+DIM*0+XX] -= tx;
655 f[j_coord_offset+DIM*0+YY] -= ty;
656 f[j_coord_offset+DIM*0+ZZ] -= tz;
660 /* Inner loop uses 136 flops */
662 /* End of innermost loop */
665 f[i_coord_offset+DIM*0+XX] += fix0;
666 f[i_coord_offset+DIM*0+YY] += fiy0;
667 f[i_coord_offset+DIM*0+ZZ] += fiz0;
671 f[i_coord_offset+DIM*1+XX] += fix1;
672 f[i_coord_offset+DIM*1+YY] += fiy1;
673 f[i_coord_offset+DIM*1+ZZ] += fiz1;
677 f[i_coord_offset+DIM*2+XX] += fix2;
678 f[i_coord_offset+DIM*2+YY] += fiy2;
679 f[i_coord_offset+DIM*2+ZZ] += fiz2;
683 fshift[i_shift_offset+XX] += tx;
684 fshift[i_shift_offset+YY] += ty;
685 fshift[i_shift_offset+ZZ] += tz;
687 /* Increment number of inner iterations */
688 inneriter += j_index_end - j_index_start;
690 /* Outer loop uses 30 flops */
693 /* Increment number of outer iterations */
696 /* Update outer/inner flops */
698 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W3_F,outeriter*30 + inneriter*136);