Merge release-5-0 into master
[alexxy/gromacs.git] / src / gromacs / gmxlib / nonbonded / nb_kernel_c / nb_kernel_ElecRFCut_VdwLJSh_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_ElecRFCut_VdwLJSh_GeomP1P1_VF_c
51  * Electrostatics interaction: ReactionField
52  * VdW interaction:            LennardJones
53  * Geometry:                   Particle-Particle
54  * Calculate force/pot:        PotentialAndForce
55  */
56 void
57 nb_kernel_ElecRFCut_VdwLJSh_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
84     x                = xx[0];
85     f                = ff[0];
86
87     nri              = nlist->nri;
88     iinr             = nlist->iinr;
89     jindex           = nlist->jindex;
90     jjnr             = nlist->jjnr;
91     shiftidx         = nlist->shift;
92     gid              = nlist->gid;
93     shiftvec         = fr->shift_vec[0];
94     fshift           = fr->fshift[0];
95     facel            = fr->epsfac;
96     charge           = mdatoms->chargeA;
97     krf              = fr->ic->k_rf;
98     krf2             = krf*2.0;
99     crf              = fr->ic->c_rf;
100     nvdwtype         = fr->ntype;
101     vdwparam         = fr->nbfp;
102     vdwtype          = mdatoms->typeA;
103
104     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
105     rcutoff          = fr->rcoulomb;
106     rcutoff2         = rcutoff*rcutoff;
107
108     sh_vdw_invrcut6  = fr->ic->sh_invrc6;
109     rvdw             = fr->rvdw;
110
111     outeriter        = 0;
112     inneriter        = 0;
113
114     /* Start outer loop over neighborlists */
115     for(iidx=0; iidx<nri; iidx++)
116     {
117         /* Load shift vector for this list */
118         i_shift_offset   = DIM*shiftidx[iidx];
119         shX              = shiftvec[i_shift_offset+XX];
120         shY              = shiftvec[i_shift_offset+YY];
121         shZ              = shiftvec[i_shift_offset+ZZ];
122
123         /* Load limits for loop over neighbors */
124         j_index_start    = jindex[iidx];
125         j_index_end      = jindex[iidx+1];
126
127         /* Get outer coordinate index */
128         inr              = iinr[iidx];
129         i_coord_offset   = DIM*inr;
130
131         /* Load i particle coords and add shift vector */
132         ix0              = shX + x[i_coord_offset+DIM*0+XX];
133         iy0              = shY + x[i_coord_offset+DIM*0+YY];
134         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
135
136         fix0             = 0.0;
137         fiy0             = 0.0;
138         fiz0             = 0.0;
139
140         /* Load parameters for i particles */
141         iq0              = facel*charge[inr+0];
142         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
143
144         /* Reset potential sums */
145         velecsum         = 0.0;
146         vvdwsum          = 0.0;
147
148         /* Start inner kernel loop */
149         for(jidx=j_index_start; jidx<j_index_end; jidx++)
150         {
151             /* Get j neighbor index, and coordinate index */
152             jnr              = jjnr[jidx];
153             j_coord_offset   = DIM*jnr;
154
155             /* load j atom coordinates */
156             jx0              = x[j_coord_offset+DIM*0+XX];
157             jy0              = x[j_coord_offset+DIM*0+YY];
158             jz0              = x[j_coord_offset+DIM*0+ZZ];
159
160             /* Calculate displacement vector */
161             dx00             = ix0 - jx0;
162             dy00             = iy0 - jy0;
163             dz00             = iz0 - jz0;
164
165             /* Calculate squared distance and things based on it */
166             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
167
168             rinv00           = gmx_invsqrt(rsq00);
169
170             rinvsq00         = rinv00*rinv00;
171
172             /* Load parameters for j particles */
173             jq0              = charge[jnr+0];
174             vdwjidx0         = 2*vdwtype[jnr+0];
175
176             /**************************
177              * CALCULATE INTERACTIONS *
178              **************************/
179
180             if (rsq00<rcutoff2)
181             {
182
183             qq00             = iq0*jq0;
184             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
185             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
186
187             /* REACTION-FIELD ELECTROSTATICS */
188             velec            = qq00*(rinv00+krf*rsq00-crf);
189             felec            = qq00*(rinv00*rinvsq00-krf2);
190
191             /* LENNARD-JONES DISPERSION/REPULSION */
192
193             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
194             vvdw6            = c6_00*rinvsix;
195             vvdw12           = c12_00*rinvsix*rinvsix;
196             vvdw             = (vvdw12 - c12_00*sh_vdw_invrcut6*sh_vdw_invrcut6)*(1.0/12.0) - (vvdw6 - c6_00*sh_vdw_invrcut6)*(1.0/6.0);
197             fvdw             = (vvdw12-vvdw6)*rinvsq00;
198
199             /* Update potential sums from outer loop */
200             velecsum        += velec;
201             vvdwsum         += vvdw;
202
203             fscal            = felec+fvdw;
204
205             /* Calculate temporary vectorial force */
206             tx               = fscal*dx00;
207             ty               = fscal*dy00;
208             tz               = fscal*dz00;
209
210             /* Update vectorial force */
211             fix0            += tx;
212             fiy0            += ty;
213             fiz0            += tz;
214             f[j_coord_offset+DIM*0+XX] -= tx;
215             f[j_coord_offset+DIM*0+YY] -= ty;
216             f[j_coord_offset+DIM*0+ZZ] -= tz;
217
218             }
219
220             /* Inner loop uses 49 flops */
221         }
222         /* End of innermost loop */
223
224         tx = ty = tz = 0;
225         f[i_coord_offset+DIM*0+XX] += fix0;
226         f[i_coord_offset+DIM*0+YY] += fiy0;
227         f[i_coord_offset+DIM*0+ZZ] += fiz0;
228         tx                         += fix0;
229         ty                         += fiy0;
230         tz                         += fiz0;
231         fshift[i_shift_offset+XX]  += tx;
232         fshift[i_shift_offset+YY]  += ty;
233         fshift[i_shift_offset+ZZ]  += tz;
234
235         ggid                        = gid[iidx];
236         /* Update potential energies */
237         kernel_data->energygrp_elec[ggid] += velecsum;
238         kernel_data->energygrp_vdw[ggid] += vvdwsum;
239
240         /* Increment number of inner iterations */
241         inneriter                  += j_index_end - j_index_start;
242
243         /* Outer loop uses 15 flops */
244     }
245
246     /* Increment number of outer iterations */
247     outeriter        += nri;
248
249     /* Update outer/inner flops */
250
251     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*49);
252 }
253 /*
254  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwLJSh_GeomP1P1_F_c
255  * Electrostatics interaction: ReactionField
256  * VdW interaction:            LennardJones
257  * Geometry:                   Particle-Particle
258  * Calculate force/pot:        Force
259  */
260 void
261 nb_kernel_ElecRFCut_VdwLJSh_GeomP1P1_F_c
262                     (t_nblist                    * gmx_restrict       nlist,
263                      rvec                        * gmx_restrict          xx,
264                      rvec                        * gmx_restrict          ff,
265                      t_forcerec                  * gmx_restrict          fr,
266                      t_mdatoms                   * gmx_restrict     mdatoms,
267                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
268                      t_nrnb                      * gmx_restrict        nrnb)
269 {
270     int              i_shift_offset,i_coord_offset,j_coord_offset;
271     int              j_index_start,j_index_end;
272     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
273     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
274     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
275     real             *shiftvec,*fshift,*x,*f;
276     int              vdwioffset0;
277     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
278     int              vdwjidx0;
279     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
280     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
281     real             velec,felec,velecsum,facel,crf,krf,krf2;
282     real             *charge;
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
288     x                = xx[0];
289     f                = ff[0];
290
291     nri              = nlist->nri;
292     iinr             = nlist->iinr;
293     jindex           = nlist->jindex;
294     jjnr             = nlist->jjnr;
295     shiftidx         = nlist->shift;
296     gid              = nlist->gid;
297     shiftvec         = fr->shift_vec[0];
298     fshift           = fr->fshift[0];
299     facel            = fr->epsfac;
300     charge           = mdatoms->chargeA;
301     krf              = fr->ic->k_rf;
302     krf2             = krf*2.0;
303     crf              = fr->ic->c_rf;
304     nvdwtype         = fr->ntype;
305     vdwparam         = fr->nbfp;
306     vdwtype          = mdatoms->typeA;
307
308     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
309     rcutoff          = fr->rcoulomb;
310     rcutoff2         = rcutoff*rcutoff;
311
312     sh_vdw_invrcut6  = fr->ic->sh_invrc6;
313     rvdw             = fr->rvdw;
314
315     outeriter        = 0;
316     inneriter        = 0;
317
318     /* Start outer loop over neighborlists */
319     for(iidx=0; iidx<nri; iidx++)
320     {
321         /* Load shift vector for this list */
322         i_shift_offset   = DIM*shiftidx[iidx];
323         shX              = shiftvec[i_shift_offset+XX];
324         shY              = shiftvec[i_shift_offset+YY];
325         shZ              = shiftvec[i_shift_offset+ZZ];
326
327         /* Load limits for loop over neighbors */
328         j_index_start    = jindex[iidx];
329         j_index_end      = jindex[iidx+1];
330
331         /* Get outer coordinate index */
332         inr              = iinr[iidx];
333         i_coord_offset   = DIM*inr;
334
335         /* Load i particle coords and add shift vector */
336         ix0              = shX + x[i_coord_offset+DIM*0+XX];
337         iy0              = shY + x[i_coord_offset+DIM*0+YY];
338         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
339
340         fix0             = 0.0;
341         fiy0             = 0.0;
342         fiz0             = 0.0;
343
344         /* Load parameters for i particles */
345         iq0              = facel*charge[inr+0];
346         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
347
348         /* Start inner kernel loop */
349         for(jidx=j_index_start; jidx<j_index_end; jidx++)
350         {
351             /* Get j neighbor index, and coordinate index */
352             jnr              = jjnr[jidx];
353             j_coord_offset   = DIM*jnr;
354
355             /* load j atom coordinates */
356             jx0              = x[j_coord_offset+DIM*0+XX];
357             jy0              = x[j_coord_offset+DIM*0+YY];
358             jz0              = x[j_coord_offset+DIM*0+ZZ];
359
360             /* Calculate displacement vector */
361             dx00             = ix0 - jx0;
362             dy00             = iy0 - jy0;
363             dz00             = iz0 - jz0;
364
365             /* Calculate squared distance and things based on it */
366             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
367
368             rinv00           = gmx_invsqrt(rsq00);
369
370             rinvsq00         = rinv00*rinv00;
371
372             /* Load parameters for j particles */
373             jq0              = charge[jnr+0];
374             vdwjidx0         = 2*vdwtype[jnr+0];
375
376             /**************************
377              * CALCULATE INTERACTIONS *
378              **************************/
379
380             if (rsq00<rcutoff2)
381             {
382
383             qq00             = iq0*jq0;
384             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
385             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
386
387             /* REACTION-FIELD ELECTROSTATICS */
388             felec            = qq00*(rinv00*rinvsq00-krf2);
389
390             /* LENNARD-JONES DISPERSION/REPULSION */
391
392             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
393             fvdw             = (c12_00*rinvsix-c6_00)*rinvsix*rinvsq00;
394
395             fscal            = felec+fvdw;
396
397             /* Calculate temporary vectorial force */
398             tx               = fscal*dx00;
399             ty               = fscal*dy00;
400             tz               = fscal*dz00;
401
402             /* Update vectorial force */
403             fix0            += tx;
404             fiy0            += ty;
405             fiz0            += tz;
406             f[j_coord_offset+DIM*0+XX] -= tx;
407             f[j_coord_offset+DIM*0+YY] -= ty;
408             f[j_coord_offset+DIM*0+ZZ] -= tz;
409
410             }
411
412             /* Inner loop uses 34 flops */
413         }
414         /* End of innermost loop */
415
416         tx = ty = tz = 0;
417         f[i_coord_offset+DIM*0+XX] += fix0;
418         f[i_coord_offset+DIM*0+YY] += fiy0;
419         f[i_coord_offset+DIM*0+ZZ] += fiz0;
420         tx                         += fix0;
421         ty                         += fiy0;
422         tz                         += fiz0;
423         fshift[i_shift_offset+XX]  += tx;
424         fshift[i_shift_offset+YY]  += ty;
425         fshift[i_shift_offset+ZZ]  += tz;
426
427         /* Increment number of inner iterations */
428         inneriter                  += j_index_end - j_index_start;
429
430         /* Outer loop uses 13 flops */
431     }
432
433     /* Increment number of outer iterations */
434     outeriter        += nri;
435
436     /* Update outer/inner flops */
437
438     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*34);
439 }