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.
44 #include "../nb_kernel.h"
45 #include "gromacs/legacyheaders/types/simple.h"
46 #include "gromacs/math/vec.h"
47 #include "gromacs/legacyheaders/nrnb.h"
50 * Gromacs nonbonded kernel: nb_kernel_ElecEwSw_VdwNone_GeomW4P1_VF_c
51 * Electrostatics interaction: Ewald
52 * VdW interaction: None
53 * Geometry: Water4-Particle
54 * Calculate force/pot: PotentialAndForce
57 nb_kernel_ElecEwSw_VdwNone_GeomW4P1_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 ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
75 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
77 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
79 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
80 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
81 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
82 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
83 real velec,felec,velecsum,facel,crf,krf,krf2;
86 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
88 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
95 jindex = nlist->jindex;
97 shiftidx = nlist->shift;
99 shiftvec = fr->shift_vec[0];
100 fshift = fr->fshift[0];
102 charge = mdatoms->chargeA;
104 sh_ewald = fr->ic->sh_ewald;
105 ewtab = fr->ic->tabq_coul_FDV0;
106 ewtabscale = fr->ic->tabq_scale;
107 ewtabhalfspace = 0.5/ewtabscale;
109 /* Setup water-specific parameters */
110 inr = nlist->iinr[0];
111 iq1 = facel*charge[inr+1];
112 iq2 = facel*charge[inr+2];
113 iq3 = facel*charge[inr+3];
115 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
116 rcutoff = fr->rcoulomb;
117 rcutoff2 = rcutoff*rcutoff;
119 rswitch = fr->rcoulomb_switch;
120 /* Setup switch parameters */
122 swV3 = -10.0/(d*d*d);
123 swV4 = 15.0/(d*d*d*d);
124 swV5 = -6.0/(d*d*d*d*d);
125 swF2 = -30.0/(d*d*d);
126 swF3 = 60.0/(d*d*d*d);
127 swF4 = -30.0/(d*d*d*d*d);
132 /* Start outer loop over neighborlists */
133 for(iidx=0; iidx<nri; iidx++)
135 /* Load shift vector for this list */
136 i_shift_offset = DIM*shiftidx[iidx];
137 shX = shiftvec[i_shift_offset+XX];
138 shY = shiftvec[i_shift_offset+YY];
139 shZ = shiftvec[i_shift_offset+ZZ];
141 /* Load limits for loop over neighbors */
142 j_index_start = jindex[iidx];
143 j_index_end = jindex[iidx+1];
145 /* Get outer coordinate index */
147 i_coord_offset = DIM*inr;
149 /* Load i particle coords and add shift vector */
150 ix1 = shX + x[i_coord_offset+DIM*1+XX];
151 iy1 = shY + x[i_coord_offset+DIM*1+YY];
152 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
153 ix2 = shX + x[i_coord_offset+DIM*2+XX];
154 iy2 = shY + x[i_coord_offset+DIM*2+YY];
155 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
156 ix3 = shX + x[i_coord_offset+DIM*3+XX];
157 iy3 = shY + x[i_coord_offset+DIM*3+YY];
158 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
170 /* Reset potential sums */
173 /* Start inner kernel loop */
174 for(jidx=j_index_start; jidx<j_index_end; jidx++)
176 /* Get j neighbor index, and coordinate index */
178 j_coord_offset = DIM*jnr;
180 /* load j atom coordinates */
181 jx0 = x[j_coord_offset+DIM*0+XX];
182 jy0 = x[j_coord_offset+DIM*0+YY];
183 jz0 = x[j_coord_offset+DIM*0+ZZ];
185 /* Calculate displacement vector */
196 /* Calculate squared distance and things based on it */
197 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
198 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
199 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
201 rinv10 = gmx_invsqrt(rsq10);
202 rinv20 = gmx_invsqrt(rsq20);
203 rinv30 = gmx_invsqrt(rsq30);
205 rinvsq10 = rinv10*rinv10;
206 rinvsq20 = rinv20*rinv20;
207 rinvsq30 = rinv30*rinv30;
209 /* Load parameters for j particles */
212 /**************************
213 * CALCULATE INTERACTIONS *
214 **************************/
223 /* EWALD ELECTROSTATICS */
225 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
226 ewrt = r10*ewtabscale;
230 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
231 velec = qq10*(rinv10-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
232 felec = qq10*rinv10*(rinvsq10-felec);
235 d = (d>0.0) ? d : 0.0;
237 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
239 dsw = d2*(swF2+d*(swF3+d*swF4));
241 /* Evaluate switch function */
242 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
243 felec = felec*sw - rinv10*velec*dsw;
246 /* Update potential sums from outer loop */
251 /* Calculate temporary vectorial force */
256 /* Update vectorial force */
260 f[j_coord_offset+DIM*0+XX] -= tx;
261 f[j_coord_offset+DIM*0+YY] -= ty;
262 f[j_coord_offset+DIM*0+ZZ] -= tz;
266 /**************************
267 * CALCULATE INTERACTIONS *
268 **************************/
277 /* EWALD ELECTROSTATICS */
279 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
280 ewrt = r20*ewtabscale;
284 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
285 velec = qq20*(rinv20-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
286 felec = qq20*rinv20*(rinvsq20-felec);
289 d = (d>0.0) ? d : 0.0;
291 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
293 dsw = d2*(swF2+d*(swF3+d*swF4));
295 /* Evaluate switch function */
296 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
297 felec = felec*sw - rinv20*velec*dsw;
300 /* Update potential sums from outer loop */
305 /* Calculate temporary vectorial force */
310 /* Update vectorial force */
314 f[j_coord_offset+DIM*0+XX] -= tx;
315 f[j_coord_offset+DIM*0+YY] -= ty;
316 f[j_coord_offset+DIM*0+ZZ] -= tz;
320 /**************************
321 * CALCULATE INTERACTIONS *
322 **************************/
331 /* EWALD ELECTROSTATICS */
333 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
334 ewrt = r30*ewtabscale;
338 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
339 velec = qq30*(rinv30-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
340 felec = qq30*rinv30*(rinvsq30-felec);
343 d = (d>0.0) ? d : 0.0;
345 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
347 dsw = d2*(swF2+d*(swF3+d*swF4));
349 /* Evaluate switch function */
350 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
351 felec = felec*sw - rinv30*velec*dsw;
354 /* Update potential sums from outer loop */
359 /* Calculate temporary vectorial force */
364 /* Update vectorial force */
368 f[j_coord_offset+DIM*0+XX] -= tx;
369 f[j_coord_offset+DIM*0+YY] -= ty;
370 f[j_coord_offset+DIM*0+ZZ] -= tz;
374 /* Inner loop uses 177 flops */
376 /* End of innermost loop */
379 f[i_coord_offset+DIM*1+XX] += fix1;
380 f[i_coord_offset+DIM*1+YY] += fiy1;
381 f[i_coord_offset+DIM*1+ZZ] += fiz1;
385 f[i_coord_offset+DIM*2+XX] += fix2;
386 f[i_coord_offset+DIM*2+YY] += fiy2;
387 f[i_coord_offset+DIM*2+ZZ] += fiz2;
391 f[i_coord_offset+DIM*3+XX] += fix3;
392 f[i_coord_offset+DIM*3+YY] += fiy3;
393 f[i_coord_offset+DIM*3+ZZ] += fiz3;
397 fshift[i_shift_offset+XX] += tx;
398 fshift[i_shift_offset+YY] += ty;
399 fshift[i_shift_offset+ZZ] += tz;
402 /* Update potential energies */
403 kernel_data->energygrp_elec[ggid] += velecsum;
405 /* Increment number of inner iterations */
406 inneriter += j_index_end - j_index_start;
408 /* Outer loop uses 31 flops */
411 /* Increment number of outer iterations */
414 /* Update outer/inner flops */
416 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_VF,outeriter*31 + inneriter*177);
419 * Gromacs nonbonded kernel: nb_kernel_ElecEwSw_VdwNone_GeomW4P1_F_c
420 * Electrostatics interaction: Ewald
421 * VdW interaction: None
422 * Geometry: Water4-Particle
423 * Calculate force/pot: Force
426 nb_kernel_ElecEwSw_VdwNone_GeomW4P1_F_c
427 (t_nblist * gmx_restrict nlist,
428 rvec * gmx_restrict xx,
429 rvec * gmx_restrict ff,
430 t_forcerec * gmx_restrict fr,
431 t_mdatoms * gmx_restrict mdatoms,
432 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
433 t_nrnb * gmx_restrict nrnb)
435 int i_shift_offset,i_coord_offset,j_coord_offset;
436 int j_index_start,j_index_end;
437 int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
438 real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
439 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
440 real *shiftvec,*fshift,*x,*f;
442 real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
444 real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
446 real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
448 real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
449 real dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10,cexp1_10,cexp2_10;
450 real dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20,cexp1_20,cexp2_20;
451 real dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30,cexp1_30,cexp2_30;
452 real velec,felec,velecsum,facel,crf,krf,krf2;
455 real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace;
457 real rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
464 jindex = nlist->jindex;
466 shiftidx = nlist->shift;
468 shiftvec = fr->shift_vec[0];
469 fshift = fr->fshift[0];
471 charge = mdatoms->chargeA;
473 sh_ewald = fr->ic->sh_ewald;
474 ewtab = fr->ic->tabq_coul_FDV0;
475 ewtabscale = fr->ic->tabq_scale;
476 ewtabhalfspace = 0.5/ewtabscale;
478 /* Setup water-specific parameters */
479 inr = nlist->iinr[0];
480 iq1 = facel*charge[inr+1];
481 iq2 = facel*charge[inr+2];
482 iq3 = facel*charge[inr+3];
484 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
485 rcutoff = fr->rcoulomb;
486 rcutoff2 = rcutoff*rcutoff;
488 rswitch = fr->rcoulomb_switch;
489 /* Setup switch parameters */
491 swV3 = -10.0/(d*d*d);
492 swV4 = 15.0/(d*d*d*d);
493 swV5 = -6.0/(d*d*d*d*d);
494 swF2 = -30.0/(d*d*d);
495 swF3 = 60.0/(d*d*d*d);
496 swF4 = -30.0/(d*d*d*d*d);
501 /* Start outer loop over neighborlists */
502 for(iidx=0; iidx<nri; iidx++)
504 /* Load shift vector for this list */
505 i_shift_offset = DIM*shiftidx[iidx];
506 shX = shiftvec[i_shift_offset+XX];
507 shY = shiftvec[i_shift_offset+YY];
508 shZ = shiftvec[i_shift_offset+ZZ];
510 /* Load limits for loop over neighbors */
511 j_index_start = jindex[iidx];
512 j_index_end = jindex[iidx+1];
514 /* Get outer coordinate index */
516 i_coord_offset = DIM*inr;
518 /* Load i particle coords and add shift vector */
519 ix1 = shX + x[i_coord_offset+DIM*1+XX];
520 iy1 = shY + x[i_coord_offset+DIM*1+YY];
521 iz1 = shZ + x[i_coord_offset+DIM*1+ZZ];
522 ix2 = shX + x[i_coord_offset+DIM*2+XX];
523 iy2 = shY + x[i_coord_offset+DIM*2+YY];
524 iz2 = shZ + x[i_coord_offset+DIM*2+ZZ];
525 ix3 = shX + x[i_coord_offset+DIM*3+XX];
526 iy3 = shY + x[i_coord_offset+DIM*3+YY];
527 iz3 = shZ + x[i_coord_offset+DIM*3+ZZ];
539 /* Start inner kernel loop */
540 for(jidx=j_index_start; jidx<j_index_end; jidx++)
542 /* Get j neighbor index, and coordinate index */
544 j_coord_offset = DIM*jnr;
546 /* load j atom coordinates */
547 jx0 = x[j_coord_offset+DIM*0+XX];
548 jy0 = x[j_coord_offset+DIM*0+YY];
549 jz0 = x[j_coord_offset+DIM*0+ZZ];
551 /* Calculate displacement vector */
562 /* Calculate squared distance and things based on it */
563 rsq10 = dx10*dx10+dy10*dy10+dz10*dz10;
564 rsq20 = dx20*dx20+dy20*dy20+dz20*dz20;
565 rsq30 = dx30*dx30+dy30*dy30+dz30*dz30;
567 rinv10 = gmx_invsqrt(rsq10);
568 rinv20 = gmx_invsqrt(rsq20);
569 rinv30 = gmx_invsqrt(rsq30);
571 rinvsq10 = rinv10*rinv10;
572 rinvsq20 = rinv20*rinv20;
573 rinvsq30 = rinv30*rinv30;
575 /* Load parameters for j particles */
578 /**************************
579 * CALCULATE INTERACTIONS *
580 **************************/
589 /* EWALD ELECTROSTATICS */
591 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
592 ewrt = r10*ewtabscale;
596 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
597 velec = qq10*(rinv10-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
598 felec = qq10*rinv10*(rinvsq10-felec);
601 d = (d>0.0) ? d : 0.0;
603 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
605 dsw = d2*(swF2+d*(swF3+d*swF4));
607 /* Evaluate switch function */
608 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
609 felec = felec*sw - rinv10*velec*dsw;
613 /* Calculate temporary vectorial force */
618 /* Update vectorial force */
622 f[j_coord_offset+DIM*0+XX] -= tx;
623 f[j_coord_offset+DIM*0+YY] -= ty;
624 f[j_coord_offset+DIM*0+ZZ] -= tz;
628 /**************************
629 * CALCULATE INTERACTIONS *
630 **************************/
639 /* EWALD ELECTROSTATICS */
641 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
642 ewrt = r20*ewtabscale;
646 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
647 velec = qq20*(rinv20-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
648 felec = qq20*rinv20*(rinvsq20-felec);
651 d = (d>0.0) ? d : 0.0;
653 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
655 dsw = d2*(swF2+d*(swF3+d*swF4));
657 /* Evaluate switch function */
658 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
659 felec = felec*sw - rinv20*velec*dsw;
663 /* Calculate temporary vectorial force */
668 /* Update vectorial force */
672 f[j_coord_offset+DIM*0+XX] -= tx;
673 f[j_coord_offset+DIM*0+YY] -= ty;
674 f[j_coord_offset+DIM*0+ZZ] -= tz;
678 /**************************
679 * CALCULATE INTERACTIONS *
680 **************************/
689 /* EWALD ELECTROSTATICS */
691 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
692 ewrt = r30*ewtabscale;
696 felec = ewtab[ewitab]+eweps*ewtab[ewitab+1];
697 velec = qq30*(rinv30-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec)));
698 felec = qq30*rinv30*(rinvsq30-felec);
701 d = (d>0.0) ? d : 0.0;
703 sw = 1.0+d2*d*(swV3+d*(swV4+d*swV5));
705 dsw = d2*(swF2+d*(swF3+d*swF4));
707 /* Evaluate switch function */
708 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
709 felec = felec*sw - rinv30*velec*dsw;
713 /* Calculate temporary vectorial force */
718 /* Update vectorial force */
722 f[j_coord_offset+DIM*0+XX] -= tx;
723 f[j_coord_offset+DIM*0+YY] -= ty;
724 f[j_coord_offset+DIM*0+ZZ] -= tz;
728 /* Inner loop uses 171 flops */
730 /* End of innermost loop */
733 f[i_coord_offset+DIM*1+XX] += fix1;
734 f[i_coord_offset+DIM*1+YY] += fiy1;
735 f[i_coord_offset+DIM*1+ZZ] += fiz1;
739 f[i_coord_offset+DIM*2+XX] += fix2;
740 f[i_coord_offset+DIM*2+YY] += fiy2;
741 f[i_coord_offset+DIM*2+ZZ] += fiz2;
745 f[i_coord_offset+DIM*3+XX] += fix3;
746 f[i_coord_offset+DIM*3+YY] += fiy3;
747 f[i_coord_offset+DIM*3+ZZ] += fiz3;
751 fshift[i_shift_offset+XX] += tx;
752 fshift[i_shift_offset+YY] += ty;
753 fshift[i_shift_offset+ZZ] += tz;
755 /* Increment number of inner iterations */
756 inneriter += j_index_end - j_index_start;
758 /* Outer loop uses 30 flops */
761 /* Increment number of outer iterations */
764 /* Update outer/inner flops */
766 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_W4_F,outeriter*30 + inneriter*171);