4d13337accc426782e0af89953d4830b89a12d62
[alexxy/gromacs.git] / src / gromacs / gmxlib / nonbonded / nb_kernel_c / nb_kernel_ElecRFCut_VdwBhamSh_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 "gromacs/legacyheaders/types/simple.h"
44 #include "gromacs/math/vec.h"
45 #include "gromacs/legacyheaders/nrnb.h"
46
47 /*
48  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwBhamSh_GeomP1P1_VF_c
49  * Electrostatics interaction: ReactionField
50  * VdW interaction:            Buckingham
51  * Geometry:                   Particle-Particle
52  * Calculate force/pot:        PotentialAndForce
53  */
54 void
55 nb_kernel_ElecRFCut_VdwBhamSh_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     real             velec,felec,velecsum,facel,crf,krf,krf2;
76     real             *charge;
77     int              nvdwtype;
78     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
79     int              *vdwtype;
80     real             *vdwparam;
81
82     x                = xx[0];
83     f                = ff[0];
84
85     nri              = nlist->nri;
86     iinr             = nlist->iinr;
87     jindex           = nlist->jindex;
88     jjnr             = nlist->jjnr;
89     shiftidx         = nlist->shift;
90     gid              = nlist->gid;
91     shiftvec         = fr->shift_vec[0];
92     fshift           = fr->fshift[0];
93     facel            = fr->epsfac;
94     charge           = mdatoms->chargeA;
95     krf              = fr->ic->k_rf;
96     krf2             = krf*2.0;
97     crf              = fr->ic->c_rf;
98     nvdwtype         = fr->ntype;
99     vdwparam         = fr->nbfp;
100     vdwtype          = mdatoms->typeA;
101
102     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
103     rcutoff          = fr->rcoulomb;
104     rcutoff2         = rcutoff*rcutoff;
105
106     sh_vdw_invrcut6  = fr->ic->sh_invrc6;
107     rvdw             = fr->rvdw;
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         iq0              = facel*charge[inr+0];
140         vdwioffset0      = 3*nvdwtype*vdwtype[inr+0];
141
142         /* Reset potential sums */
143         velecsum         = 0.0;
144         vvdwsum          = 0.0;
145
146         /* Start inner kernel loop */
147         for(jidx=j_index_start; jidx<j_index_end; jidx++)
148         {
149             /* Get j neighbor index, and coordinate index */
150             jnr              = jjnr[jidx];
151             j_coord_offset   = DIM*jnr;
152
153             /* load j atom coordinates */
154             jx0              = x[j_coord_offset+DIM*0+XX];
155             jy0              = x[j_coord_offset+DIM*0+YY];
156             jz0              = x[j_coord_offset+DIM*0+ZZ];
157
158             /* Calculate displacement vector */
159             dx00             = ix0 - jx0;
160             dy00             = iy0 - jy0;
161             dz00             = iz0 - jz0;
162
163             /* Calculate squared distance and things based on it */
164             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
165
166             rinv00           = gmx_invsqrt(rsq00);
167
168             rinvsq00         = rinv00*rinv00;
169
170             /* Load parameters for j particles */
171             jq0              = charge[jnr+0];
172             vdwjidx0         = 3*vdwtype[jnr+0];
173
174             /**************************
175              * CALCULATE INTERACTIONS *
176              **************************/
177
178             if (rsq00<rcutoff2)
179             {
180
181             r00              = rsq00*rinv00;
182
183             qq00             = iq0*jq0;
184             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
185             cexp1_00         = vdwparam[vdwioffset0+vdwjidx0+1];
186             cexp2_00         = vdwparam[vdwioffset0+vdwjidx0+2];
187
188             /* REACTION-FIELD ELECTROSTATICS */
189             velec            = qq00*(rinv00+krf*rsq00-crf);
190             felec            = qq00*(rinv00*rinvsq00-krf2);
191
192             /* BUCKINGHAM DISPERSION/REPULSION */
193             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
194             vvdw6            = c6_00*rinvsix;
195             br               = cexp2_00*r00;
196             vvdwexp          = cexp1_00*exp(-br);
197             vvdw             = (vvdwexp-cexp1_00*exp(-cexp2_00*rvdw)) - (vvdw6 - c6_00*sh_vdw_invrcut6)*(1.0/6.0);
198             fvdw             = (br*vvdwexp-vvdw6)*rinvsq00;
199
200             /* Update potential sums from outer loop */
201             velecsum        += velec;
202             vvdwsum         += vvdw;
203
204             fscal            = felec+fvdw;
205
206             /* Calculate temporary vectorial force */
207             tx               = fscal*dx00;
208             ty               = fscal*dy00;
209             tz               = fscal*dz00;
210
211             /* Update vectorial force */
212             fix0            += tx;
213             fiy0            += ty;
214             fiz0            += tz;
215             f[j_coord_offset+DIM*0+XX] -= tx;
216             f[j_coord_offset+DIM*0+YY] -= ty;
217             f[j_coord_offset+DIM*0+ZZ] -= tz;
218
219             }
220
221             /* Inner loop uses 102 flops */
222         }
223         /* End of innermost loop */
224
225         tx = ty = tz = 0;
226         f[i_coord_offset+DIM*0+XX] += fix0;
227         f[i_coord_offset+DIM*0+YY] += fiy0;
228         f[i_coord_offset+DIM*0+ZZ] += fiz0;
229         tx                         += fix0;
230         ty                         += fiy0;
231         tz                         += fiz0;
232         fshift[i_shift_offset+XX]  += tx;
233         fshift[i_shift_offset+YY]  += ty;
234         fshift[i_shift_offset+ZZ]  += tz;
235
236         ggid                        = gid[iidx];
237         /* Update potential energies */
238         kernel_data->energygrp_elec[ggid] += velecsum;
239         kernel_data->energygrp_vdw[ggid] += vvdwsum;
240
241         /* Increment number of inner iterations */
242         inneriter                  += j_index_end - j_index_start;
243
244         /* Outer loop uses 15 flops */
245     }
246
247     /* Increment number of outer iterations */
248     outeriter        += nri;
249
250     /* Update outer/inner flops */
251
252     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_VF,outeriter*15 + inneriter*102);
253 }
254 /*
255  * Gromacs nonbonded kernel:   nb_kernel_ElecRFCut_VdwBhamSh_GeomP1P1_F_c
256  * Electrostatics interaction: ReactionField
257  * VdW interaction:            Buckingham
258  * Geometry:                   Particle-Particle
259  * Calculate force/pot:        Force
260  */
261 void
262 nb_kernel_ElecRFCut_VdwBhamSh_GeomP1P1_F_c
263                     (t_nblist                    * gmx_restrict       nlist,
264                      rvec                        * gmx_restrict          xx,
265                      rvec                        * gmx_restrict          ff,
266                      t_forcerec                  * gmx_restrict          fr,
267                      t_mdatoms                   * gmx_restrict     mdatoms,
268                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
269                      t_nrnb                      * gmx_restrict        nrnb)
270 {
271     int              i_shift_offset,i_coord_offset,j_coord_offset;
272     int              j_index_start,j_index_end;
273     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
274     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
275     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
276     real             *shiftvec,*fshift,*x,*f;
277     int              vdwioffset0;
278     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
279     int              vdwjidx0;
280     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
281     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
282     real             velec,felec,velecsum,facel,crf,krf,krf2;
283     real             *charge;
284     int              nvdwtype;
285     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
286     int              *vdwtype;
287     real             *vdwparam;
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     facel            = fr->epsfac;
301     charge           = mdatoms->chargeA;
302     krf              = fr->ic->k_rf;
303     krf2             = krf*2.0;
304     crf              = fr->ic->c_rf;
305     nvdwtype         = fr->ntype;
306     vdwparam         = fr->nbfp;
307     vdwtype          = mdatoms->typeA;
308
309     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
310     rcutoff          = fr->rcoulomb;
311     rcutoff2         = rcutoff*rcutoff;
312
313     sh_vdw_invrcut6  = fr->ic->sh_invrc6;
314     rvdw             = fr->rvdw;
315
316     outeriter        = 0;
317     inneriter        = 0;
318
319     /* Start outer loop over neighborlists */
320     for(iidx=0; iidx<nri; iidx++)
321     {
322         /* Load shift vector for this list */
323         i_shift_offset   = DIM*shiftidx[iidx];
324         shX              = shiftvec[i_shift_offset+XX];
325         shY              = shiftvec[i_shift_offset+YY];
326         shZ              = shiftvec[i_shift_offset+ZZ];
327
328         /* Load limits for loop over neighbors */
329         j_index_start    = jindex[iidx];
330         j_index_end      = jindex[iidx+1];
331
332         /* Get outer coordinate index */
333         inr              = iinr[iidx];
334         i_coord_offset   = DIM*inr;
335
336         /* Load i particle coords and add shift vector */
337         ix0              = shX + x[i_coord_offset+DIM*0+XX];
338         iy0              = shY + x[i_coord_offset+DIM*0+YY];
339         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
340
341         fix0             = 0.0;
342         fiy0             = 0.0;
343         fiz0             = 0.0;
344
345         /* Load parameters for i particles */
346         iq0              = facel*charge[inr+0];
347         vdwioffset0      = 3*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             jq0              = charge[jnr+0];
375             vdwjidx0         = 3*vdwtype[jnr+0];
376
377             /**************************
378              * CALCULATE INTERACTIONS *
379              **************************/
380
381             if (rsq00<rcutoff2)
382             {
383
384             r00              = rsq00*rinv00;
385
386             qq00             = iq0*jq0;
387             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
388             cexp1_00         = vdwparam[vdwioffset0+vdwjidx0+1];
389             cexp2_00         = vdwparam[vdwioffset0+vdwjidx0+2];
390
391             /* REACTION-FIELD ELECTROSTATICS */
392             felec            = qq00*(rinv00*rinvsq00-krf2);
393
394             /* BUCKINGHAM DISPERSION/REPULSION */
395             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
396             vvdw6            = c6_00*rinvsix;
397             br               = cexp2_00*r00;
398             vvdwexp          = cexp1_00*exp(-br);
399             fvdw             = (br*vvdwexp-vvdw6)*rinvsq00;
400
401             fscal            = felec+fvdw;
402
403             /* Calculate temporary vectorial force */
404             tx               = fscal*dx00;
405             ty               = fscal*dy00;
406             tz               = fscal*dz00;
407
408             /* Update vectorial force */
409             fix0            += tx;
410             fiy0            += ty;
411             fiz0            += tz;
412             f[j_coord_offset+DIM*0+XX] -= tx;
413             f[j_coord_offset+DIM*0+YY] -= ty;
414             f[j_coord_offset+DIM*0+ZZ] -= tz;
415
416             }
417
418             /* Inner loop uses 63 flops */
419         }
420         /* End of innermost loop */
421
422         tx = ty = tz = 0;
423         f[i_coord_offset+DIM*0+XX] += fix0;
424         f[i_coord_offset+DIM*0+YY] += fiy0;
425         f[i_coord_offset+DIM*0+ZZ] += fiz0;
426         tx                         += fix0;
427         ty                         += fiy0;
428         tz                         += fiz0;
429         fshift[i_shift_offset+XX]  += tx;
430         fshift[i_shift_offset+YY]  += ty;
431         fshift[i_shift_offset+ZZ]  += tz;
432
433         /* Increment number of inner iterations */
434         inneriter                  += j_index_end - j_index_start;
435
436         /* Outer loop uses 13 flops */
437     }
438
439     /* Increment number of outer iterations */
440     outeriter        += nri;
441
442     /* Update outer/inner flops */
443
444     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_F,outeriter*13 + inneriter*63);
445 }