File: | gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecEw_VdwBham_GeomW4W4_c.c |
Location: | line 125, column 5 |
Description: | Value stored to 'sh_ewald' is never read |
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 | #ifdef HAVE_CONFIG_H1 |
39 | #include <config.h> |
40 | #endif |
41 | |
42 | #include <math.h> |
43 | |
44 | #include "../nb_kernel.h" |
45 | #include "types/simple.h" |
46 | #include "gromacs/math/vec.h" |
47 | #include "nrnb.h" |
48 | |
49 | /* |
50 | * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwBham_GeomW4W4_VF_c |
51 | * Electrostatics interaction: Ewald |
52 | * VdW interaction: Buckingham |
53 | * Geometry: Water4-Water4 |
54 | * Calculate force/pot: PotentialAndForce |
55 | */ |
56 | void |
57 | nb_kernel_ElecEw_VdwBham_GeomW4W4_VF_c |
58 | (t_nblist * gmx_restrict__restrict nlist, |
59 | rvec * gmx_restrict__restrict xx, |
60 | rvec * gmx_restrict__restrict ff, |
61 | t_forcerec * gmx_restrict__restrict fr, |
62 | t_mdatoms * gmx_restrict__restrict mdatoms, |
63 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
64 | t_nrnb * gmx_restrict__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 vdwioffset1; |
75 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
76 | int vdwioffset2; |
77 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
78 | int vdwioffset3; |
79 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
80 | int vdwjidx0; |
81 | real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0; |
82 | int vdwjidx1; |
83 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
84 | int vdwjidx2; |
85 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
86 | int vdwjidx3; |
87 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
88 | real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00; |
89 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
90 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
91 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
92 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
93 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
94 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
95 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
96 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
97 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
98 | real velec,felec,velecsum,facel,crf,krf,krf2; |
99 | real *charge; |
100 | int nvdwtype; |
101 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
102 | int *vdwtype; |
103 | real *vdwparam; |
104 | int ewitab; |
105 | real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace; |
106 | real *ewtab; |
107 | |
108 | x = xx[0]; |
109 | f = ff[0]; |
110 | |
111 | nri = nlist->nri; |
112 | iinr = nlist->iinr; |
113 | jindex = nlist->jindex; |
114 | jjnr = nlist->jjnr; |
115 | shiftidx = nlist->shift; |
116 | gid = nlist->gid; |
117 | shiftvec = fr->shift_vec[0]; |
118 | fshift = fr->fshift[0]; |
119 | facel = fr->epsfac; |
120 | charge = mdatoms->chargeA; |
121 | nvdwtype = fr->ntype; |
122 | vdwparam = fr->nbfp; |
123 | vdwtype = mdatoms->typeA; |
124 | |
125 | sh_ewald = fr->ic->sh_ewald; |
Value stored to 'sh_ewald' is never read | |
126 | ewtab = fr->ic->tabq_coul_FDV0; |
127 | ewtabscale = fr->ic->tabq_scale; |
128 | ewtabhalfspace = 0.5/ewtabscale; |
129 | |
130 | /* Setup water-specific parameters */ |
131 | inr = nlist->iinr[0]; |
132 | iq1 = facel*charge[inr+1]; |
133 | iq2 = facel*charge[inr+2]; |
134 | iq3 = facel*charge[inr+3]; |
135 | vdwioffset0 = 3*nvdwtype*vdwtype[inr+0]; |
136 | |
137 | jq1 = charge[inr+1]; |
138 | jq2 = charge[inr+2]; |
139 | jq3 = charge[inr+3]; |
140 | vdwjidx0 = 3*vdwtype[inr+0]; |
141 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
142 | cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
143 | cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2]; |
144 | qq11 = iq1*jq1; |
145 | qq12 = iq1*jq2; |
146 | qq13 = iq1*jq3; |
147 | qq21 = iq2*jq1; |
148 | qq22 = iq2*jq2; |
149 | qq23 = iq2*jq3; |
150 | qq31 = iq3*jq1; |
151 | qq32 = iq3*jq2; |
152 | qq33 = iq3*jq3; |
153 | |
154 | outeriter = 0; |
155 | inneriter = 0; |
156 | |
157 | /* Start outer loop over neighborlists */ |
158 | for(iidx=0; iidx<nri; iidx++) |
159 | { |
160 | /* Load shift vector for this list */ |
161 | i_shift_offset = DIM3*shiftidx[iidx]; |
162 | shX = shiftvec[i_shift_offset+XX0]; |
163 | shY = shiftvec[i_shift_offset+YY1]; |
164 | shZ = shiftvec[i_shift_offset+ZZ2]; |
165 | |
166 | /* Load limits for loop over neighbors */ |
167 | j_index_start = jindex[iidx]; |
168 | j_index_end = jindex[iidx+1]; |
169 | |
170 | /* Get outer coordinate index */ |
171 | inr = iinr[iidx]; |
172 | i_coord_offset = DIM3*inr; |
173 | |
174 | /* Load i particle coords and add shift vector */ |
175 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
176 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
177 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
178 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
179 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
180 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
181 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
182 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
183 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
184 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
185 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
186 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
187 | |
188 | fix0 = 0.0; |
189 | fiy0 = 0.0; |
190 | fiz0 = 0.0; |
191 | fix1 = 0.0; |
192 | fiy1 = 0.0; |
193 | fiz1 = 0.0; |
194 | fix2 = 0.0; |
195 | fiy2 = 0.0; |
196 | fiz2 = 0.0; |
197 | fix3 = 0.0; |
198 | fiy3 = 0.0; |
199 | fiz3 = 0.0; |
200 | |
201 | /* Reset potential sums */ |
202 | velecsum = 0.0; |
203 | vvdwsum = 0.0; |
204 | |
205 | /* Start inner kernel loop */ |
206 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
207 | { |
208 | /* Get j neighbor index, and coordinate index */ |
209 | jnr = jjnr[jidx]; |
210 | j_coord_offset = DIM3*jnr; |
211 | |
212 | /* load j atom coordinates */ |
213 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
214 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
215 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
216 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
217 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
218 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
219 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
220 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
221 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
222 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
223 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
224 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
225 | |
226 | /* Calculate displacement vector */ |
227 | dx00 = ix0 - jx0; |
228 | dy00 = iy0 - jy0; |
229 | dz00 = iz0 - jz0; |
230 | dx11 = ix1 - jx1; |
231 | dy11 = iy1 - jy1; |
232 | dz11 = iz1 - jz1; |
233 | dx12 = ix1 - jx2; |
234 | dy12 = iy1 - jy2; |
235 | dz12 = iz1 - jz2; |
236 | dx13 = ix1 - jx3; |
237 | dy13 = iy1 - jy3; |
238 | dz13 = iz1 - jz3; |
239 | dx21 = ix2 - jx1; |
240 | dy21 = iy2 - jy1; |
241 | dz21 = iz2 - jz1; |
242 | dx22 = ix2 - jx2; |
243 | dy22 = iy2 - jy2; |
244 | dz22 = iz2 - jz2; |
245 | dx23 = ix2 - jx3; |
246 | dy23 = iy2 - jy3; |
247 | dz23 = iz2 - jz3; |
248 | dx31 = ix3 - jx1; |
249 | dy31 = iy3 - jy1; |
250 | dz31 = iz3 - jz1; |
251 | dx32 = ix3 - jx2; |
252 | dy32 = iy3 - jy2; |
253 | dz32 = iz3 - jz2; |
254 | dx33 = ix3 - jx3; |
255 | dy33 = iy3 - jy3; |
256 | dz33 = iz3 - jz3; |
257 | |
258 | /* Calculate squared distance and things based on it */ |
259 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
260 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
261 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
262 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
263 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
264 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
265 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
266 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
267 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
268 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
269 | |
270 | rinv00 = gmx_invsqrt(rsq00)gmx_software_invsqrt(rsq00); |
271 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
272 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
273 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
274 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
275 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
276 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
277 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
278 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
279 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
280 | |
281 | rinvsq00 = rinv00*rinv00; |
282 | rinvsq11 = rinv11*rinv11; |
283 | rinvsq12 = rinv12*rinv12; |
284 | rinvsq13 = rinv13*rinv13; |
285 | rinvsq21 = rinv21*rinv21; |
286 | rinvsq22 = rinv22*rinv22; |
287 | rinvsq23 = rinv23*rinv23; |
288 | rinvsq31 = rinv31*rinv31; |
289 | rinvsq32 = rinv32*rinv32; |
290 | rinvsq33 = rinv33*rinv33; |
291 | |
292 | /************************** |
293 | * CALCULATE INTERACTIONS * |
294 | **************************/ |
295 | |
296 | r00 = rsq00*rinv00; |
297 | |
298 | /* BUCKINGHAM DISPERSION/REPULSION */ |
299 | rinvsix = rinvsq00*rinvsq00*rinvsq00; |
300 | vvdw6 = c6_00*rinvsix; |
301 | br = cexp2_00*r00; |
302 | vvdwexp = cexp1_00*exp(-br); |
303 | vvdw = vvdwexp - vvdw6*(1.0/6.0); |
304 | fvdw = (br*vvdwexp-vvdw6)*rinvsq00; |
305 | |
306 | /* Update potential sums from outer loop */ |
307 | vvdwsum += vvdw; |
308 | |
309 | fscal = fvdw; |
310 | |
311 | /* Calculate temporary vectorial force */ |
312 | tx = fscal*dx00; |
313 | ty = fscal*dy00; |
314 | tz = fscal*dz00; |
315 | |
316 | /* Update vectorial force */ |
317 | fix0 += tx; |
318 | fiy0 += ty; |
319 | fiz0 += tz; |
320 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
321 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
322 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
323 | |
324 | /************************** |
325 | * CALCULATE INTERACTIONS * |
326 | **************************/ |
327 | |
328 | r11 = rsq11*rinv11; |
329 | |
330 | /* EWALD ELECTROSTATICS */ |
331 | |
332 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
333 | ewrt = r11*ewtabscale; |
334 | ewitab = ewrt; |
335 | eweps = ewrt-ewitab; |
336 | ewitab = 4*ewitab; |
337 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
338 | velec = qq11*(rinv11-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
339 | felec = qq11*rinv11*(rinvsq11-felec); |
340 | |
341 | /* Update potential sums from outer loop */ |
342 | velecsum += velec; |
343 | |
344 | fscal = felec; |
345 | |
346 | /* Calculate temporary vectorial force */ |
347 | tx = fscal*dx11; |
348 | ty = fscal*dy11; |
349 | tz = fscal*dz11; |
350 | |
351 | /* Update vectorial force */ |
352 | fix1 += tx; |
353 | fiy1 += ty; |
354 | fiz1 += tz; |
355 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
356 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
357 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
358 | |
359 | /************************** |
360 | * CALCULATE INTERACTIONS * |
361 | **************************/ |
362 | |
363 | r12 = rsq12*rinv12; |
364 | |
365 | /* EWALD ELECTROSTATICS */ |
366 | |
367 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
368 | ewrt = r12*ewtabscale; |
369 | ewitab = ewrt; |
370 | eweps = ewrt-ewitab; |
371 | ewitab = 4*ewitab; |
372 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
373 | velec = qq12*(rinv12-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
374 | felec = qq12*rinv12*(rinvsq12-felec); |
375 | |
376 | /* Update potential sums from outer loop */ |
377 | velecsum += velec; |
378 | |
379 | fscal = felec; |
380 | |
381 | /* Calculate temporary vectorial force */ |
382 | tx = fscal*dx12; |
383 | ty = fscal*dy12; |
384 | tz = fscal*dz12; |
385 | |
386 | /* Update vectorial force */ |
387 | fix1 += tx; |
388 | fiy1 += ty; |
389 | fiz1 += tz; |
390 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
391 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
392 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
393 | |
394 | /************************** |
395 | * CALCULATE INTERACTIONS * |
396 | **************************/ |
397 | |
398 | r13 = rsq13*rinv13; |
399 | |
400 | /* EWALD ELECTROSTATICS */ |
401 | |
402 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
403 | ewrt = r13*ewtabscale; |
404 | ewitab = ewrt; |
405 | eweps = ewrt-ewitab; |
406 | ewitab = 4*ewitab; |
407 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
408 | velec = qq13*(rinv13-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
409 | felec = qq13*rinv13*(rinvsq13-felec); |
410 | |
411 | /* Update potential sums from outer loop */ |
412 | velecsum += velec; |
413 | |
414 | fscal = felec; |
415 | |
416 | /* Calculate temporary vectorial force */ |
417 | tx = fscal*dx13; |
418 | ty = fscal*dy13; |
419 | tz = fscal*dz13; |
420 | |
421 | /* Update vectorial force */ |
422 | fix1 += tx; |
423 | fiy1 += ty; |
424 | fiz1 += tz; |
425 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
426 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
427 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
428 | |
429 | /************************** |
430 | * CALCULATE INTERACTIONS * |
431 | **************************/ |
432 | |
433 | r21 = rsq21*rinv21; |
434 | |
435 | /* EWALD ELECTROSTATICS */ |
436 | |
437 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
438 | ewrt = r21*ewtabscale; |
439 | ewitab = ewrt; |
440 | eweps = ewrt-ewitab; |
441 | ewitab = 4*ewitab; |
442 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
443 | velec = qq21*(rinv21-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
444 | felec = qq21*rinv21*(rinvsq21-felec); |
445 | |
446 | /* Update potential sums from outer loop */ |
447 | velecsum += velec; |
448 | |
449 | fscal = felec; |
450 | |
451 | /* Calculate temporary vectorial force */ |
452 | tx = fscal*dx21; |
453 | ty = fscal*dy21; |
454 | tz = fscal*dz21; |
455 | |
456 | /* Update vectorial force */ |
457 | fix2 += tx; |
458 | fiy2 += ty; |
459 | fiz2 += tz; |
460 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
461 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
462 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
463 | |
464 | /************************** |
465 | * CALCULATE INTERACTIONS * |
466 | **************************/ |
467 | |
468 | r22 = rsq22*rinv22; |
469 | |
470 | /* EWALD ELECTROSTATICS */ |
471 | |
472 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
473 | ewrt = r22*ewtabscale; |
474 | ewitab = ewrt; |
475 | eweps = ewrt-ewitab; |
476 | ewitab = 4*ewitab; |
477 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
478 | velec = qq22*(rinv22-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
479 | felec = qq22*rinv22*(rinvsq22-felec); |
480 | |
481 | /* Update potential sums from outer loop */ |
482 | velecsum += velec; |
483 | |
484 | fscal = felec; |
485 | |
486 | /* Calculate temporary vectorial force */ |
487 | tx = fscal*dx22; |
488 | ty = fscal*dy22; |
489 | tz = fscal*dz22; |
490 | |
491 | /* Update vectorial force */ |
492 | fix2 += tx; |
493 | fiy2 += ty; |
494 | fiz2 += tz; |
495 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
496 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
497 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
498 | |
499 | /************************** |
500 | * CALCULATE INTERACTIONS * |
501 | **************************/ |
502 | |
503 | r23 = rsq23*rinv23; |
504 | |
505 | /* EWALD ELECTROSTATICS */ |
506 | |
507 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
508 | ewrt = r23*ewtabscale; |
509 | ewitab = ewrt; |
510 | eweps = ewrt-ewitab; |
511 | ewitab = 4*ewitab; |
512 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
513 | velec = qq23*(rinv23-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
514 | felec = qq23*rinv23*(rinvsq23-felec); |
515 | |
516 | /* Update potential sums from outer loop */ |
517 | velecsum += velec; |
518 | |
519 | fscal = felec; |
520 | |
521 | /* Calculate temporary vectorial force */ |
522 | tx = fscal*dx23; |
523 | ty = fscal*dy23; |
524 | tz = fscal*dz23; |
525 | |
526 | /* Update vectorial force */ |
527 | fix2 += tx; |
528 | fiy2 += ty; |
529 | fiz2 += tz; |
530 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
531 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
532 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
533 | |
534 | /************************** |
535 | * CALCULATE INTERACTIONS * |
536 | **************************/ |
537 | |
538 | r31 = rsq31*rinv31; |
539 | |
540 | /* EWALD ELECTROSTATICS */ |
541 | |
542 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
543 | ewrt = r31*ewtabscale; |
544 | ewitab = ewrt; |
545 | eweps = ewrt-ewitab; |
546 | ewitab = 4*ewitab; |
547 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
548 | velec = qq31*(rinv31-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
549 | felec = qq31*rinv31*(rinvsq31-felec); |
550 | |
551 | /* Update potential sums from outer loop */ |
552 | velecsum += velec; |
553 | |
554 | fscal = felec; |
555 | |
556 | /* Calculate temporary vectorial force */ |
557 | tx = fscal*dx31; |
558 | ty = fscal*dy31; |
559 | tz = fscal*dz31; |
560 | |
561 | /* Update vectorial force */ |
562 | fix3 += tx; |
563 | fiy3 += ty; |
564 | fiz3 += tz; |
565 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
566 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
567 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
568 | |
569 | /************************** |
570 | * CALCULATE INTERACTIONS * |
571 | **************************/ |
572 | |
573 | r32 = rsq32*rinv32; |
574 | |
575 | /* EWALD ELECTROSTATICS */ |
576 | |
577 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
578 | ewrt = r32*ewtabscale; |
579 | ewitab = ewrt; |
580 | eweps = ewrt-ewitab; |
581 | ewitab = 4*ewitab; |
582 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
583 | velec = qq32*(rinv32-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
584 | felec = qq32*rinv32*(rinvsq32-felec); |
585 | |
586 | /* Update potential sums from outer loop */ |
587 | velecsum += velec; |
588 | |
589 | fscal = felec; |
590 | |
591 | /* Calculate temporary vectorial force */ |
592 | tx = fscal*dx32; |
593 | ty = fscal*dy32; |
594 | tz = fscal*dz32; |
595 | |
596 | /* Update vectorial force */ |
597 | fix3 += tx; |
598 | fiy3 += ty; |
599 | fiz3 += tz; |
600 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
601 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
602 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
603 | |
604 | /************************** |
605 | * CALCULATE INTERACTIONS * |
606 | **************************/ |
607 | |
608 | r33 = rsq33*rinv33; |
609 | |
610 | /* EWALD ELECTROSTATICS */ |
611 | |
612 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
613 | ewrt = r33*ewtabscale; |
614 | ewitab = ewrt; |
615 | eweps = ewrt-ewitab; |
616 | ewitab = 4*ewitab; |
617 | felec = ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
618 | velec = qq33*(rinv33-(ewtab[ewitab+2]-ewtabhalfspace*eweps*(ewtab[ewitab]+felec))); |
619 | felec = qq33*rinv33*(rinvsq33-felec); |
620 | |
621 | /* Update potential sums from outer loop */ |
622 | velecsum += velec; |
623 | |
624 | fscal = felec; |
625 | |
626 | /* Calculate temporary vectorial force */ |
627 | tx = fscal*dx33; |
628 | ty = fscal*dy33; |
629 | tz = fscal*dz33; |
630 | |
631 | /* Update vectorial force */ |
632 | fix3 += tx; |
633 | fiy3 += ty; |
634 | fiz3 += tz; |
635 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
636 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
637 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
638 | |
639 | /* Inner loop uses 421 flops */ |
640 | } |
641 | /* End of innermost loop */ |
642 | |
643 | tx = ty = tz = 0; |
644 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
645 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
646 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
647 | tx += fix0; |
648 | ty += fiy0; |
649 | tz += fiz0; |
650 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
651 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
652 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
653 | tx += fix1; |
654 | ty += fiy1; |
655 | tz += fiz1; |
656 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
657 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
658 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
659 | tx += fix2; |
660 | ty += fiy2; |
661 | tz += fiz2; |
662 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
663 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
664 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
665 | tx += fix3; |
666 | ty += fiy3; |
667 | tz += fiz3; |
668 | fshift[i_shift_offset+XX0] += tx; |
669 | fshift[i_shift_offset+YY1] += ty; |
670 | fshift[i_shift_offset+ZZ2] += tz; |
671 | |
672 | ggid = gid[iidx]; |
673 | /* Update potential energies */ |
674 | kernel_data->energygrp_elec[ggid] += velecsum; |
675 | kernel_data->energygrp_vdw[ggid] += vvdwsum; |
676 | |
677 | /* Increment number of inner iterations */ |
678 | inneriter += j_index_end - j_index_start; |
679 | |
680 | /* Outer loop uses 41 flops */ |
681 | } |
682 | |
683 | /* Increment number of outer iterations */ |
684 | outeriter += nri; |
685 | |
686 | /* Update outer/inner flops */ |
687 | |
688 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*41 + inneriter*421)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W4W4_VF] += outeriter*41 + inneriter*421; |
689 | } |
690 | /* |
691 | * Gromacs nonbonded kernel: nb_kernel_ElecEw_VdwBham_GeomW4W4_F_c |
692 | * Electrostatics interaction: Ewald |
693 | * VdW interaction: Buckingham |
694 | * Geometry: Water4-Water4 |
695 | * Calculate force/pot: Force |
696 | */ |
697 | void |
698 | nb_kernel_ElecEw_VdwBham_GeomW4W4_F_c |
699 | (t_nblist * gmx_restrict__restrict nlist, |
700 | rvec * gmx_restrict__restrict xx, |
701 | rvec * gmx_restrict__restrict ff, |
702 | t_forcerec * gmx_restrict__restrict fr, |
703 | t_mdatoms * gmx_restrict__restrict mdatoms, |
704 | nb_kernel_data_t gmx_unused__attribute__ ((unused)) * gmx_restrict__restrict kernel_data, |
705 | t_nrnb * gmx_restrict__restrict nrnb) |
706 | { |
707 | int i_shift_offset,i_coord_offset,j_coord_offset; |
708 | int j_index_start,j_index_end; |
709 | int nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter; |
710 | real shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2; |
711 | int *iinr,*jindex,*jjnr,*shiftidx,*gid; |
712 | real *shiftvec,*fshift,*x,*f; |
713 | int vdwioffset0; |
714 | real ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0; |
715 | int vdwioffset1; |
716 | real ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1; |
717 | int vdwioffset2; |
718 | real ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2; |
719 | int vdwioffset3; |
720 | real ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3; |
721 | int vdwjidx0; |
722 | real jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0; |
723 | int vdwjidx1; |
724 | real jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1; |
725 | int vdwjidx2; |
726 | real jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2; |
727 | int vdwjidx3; |
728 | real jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3; |
729 | real dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00; |
730 | real dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11,cexp1_11,cexp2_11; |
731 | real dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12,cexp1_12,cexp2_12; |
732 | real dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13,cexp1_13,cexp2_13; |
733 | real dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21,cexp1_21,cexp2_21; |
734 | real dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22,cexp1_22,cexp2_22; |
735 | real dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23,cexp1_23,cexp2_23; |
736 | real dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31,cexp1_31,cexp2_31; |
737 | real dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32,cexp1_32,cexp2_32; |
738 | real dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33,cexp1_33,cexp2_33; |
739 | real velec,felec,velecsum,facel,crf,krf,krf2; |
740 | real *charge; |
741 | int nvdwtype; |
742 | real rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6; |
743 | int *vdwtype; |
744 | real *vdwparam; |
745 | int ewitab; |
746 | real ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace; |
747 | real *ewtab; |
748 | |
749 | x = xx[0]; |
750 | f = ff[0]; |
751 | |
752 | nri = nlist->nri; |
753 | iinr = nlist->iinr; |
754 | jindex = nlist->jindex; |
755 | jjnr = nlist->jjnr; |
756 | shiftidx = nlist->shift; |
757 | gid = nlist->gid; |
758 | shiftvec = fr->shift_vec[0]; |
759 | fshift = fr->fshift[0]; |
760 | facel = fr->epsfac; |
761 | charge = mdatoms->chargeA; |
762 | nvdwtype = fr->ntype; |
763 | vdwparam = fr->nbfp; |
764 | vdwtype = mdatoms->typeA; |
765 | |
766 | sh_ewald = fr->ic->sh_ewald; |
767 | ewtab = fr->ic->tabq_coul_F; |
768 | ewtabscale = fr->ic->tabq_scale; |
769 | ewtabhalfspace = 0.5/ewtabscale; |
770 | |
771 | /* Setup water-specific parameters */ |
772 | inr = nlist->iinr[0]; |
773 | iq1 = facel*charge[inr+1]; |
774 | iq2 = facel*charge[inr+2]; |
775 | iq3 = facel*charge[inr+3]; |
776 | vdwioffset0 = 3*nvdwtype*vdwtype[inr+0]; |
777 | |
778 | jq1 = charge[inr+1]; |
779 | jq2 = charge[inr+2]; |
780 | jq3 = charge[inr+3]; |
781 | vdwjidx0 = 3*vdwtype[inr+0]; |
782 | c6_00 = vdwparam[vdwioffset0+vdwjidx0]; |
783 | cexp1_00 = vdwparam[vdwioffset0+vdwjidx0+1]; |
784 | cexp2_00 = vdwparam[vdwioffset0+vdwjidx0+2]; |
785 | qq11 = iq1*jq1; |
786 | qq12 = iq1*jq2; |
787 | qq13 = iq1*jq3; |
788 | qq21 = iq2*jq1; |
789 | qq22 = iq2*jq2; |
790 | qq23 = iq2*jq3; |
791 | qq31 = iq3*jq1; |
792 | qq32 = iq3*jq2; |
793 | qq33 = iq3*jq3; |
794 | |
795 | outeriter = 0; |
796 | inneriter = 0; |
797 | |
798 | /* Start outer loop over neighborlists */ |
799 | for(iidx=0; iidx<nri; iidx++) |
800 | { |
801 | /* Load shift vector for this list */ |
802 | i_shift_offset = DIM3*shiftidx[iidx]; |
803 | shX = shiftvec[i_shift_offset+XX0]; |
804 | shY = shiftvec[i_shift_offset+YY1]; |
805 | shZ = shiftvec[i_shift_offset+ZZ2]; |
806 | |
807 | /* Load limits for loop over neighbors */ |
808 | j_index_start = jindex[iidx]; |
809 | j_index_end = jindex[iidx+1]; |
810 | |
811 | /* Get outer coordinate index */ |
812 | inr = iinr[iidx]; |
813 | i_coord_offset = DIM3*inr; |
814 | |
815 | /* Load i particle coords and add shift vector */ |
816 | ix0 = shX + x[i_coord_offset+DIM3*0+XX0]; |
817 | iy0 = shY + x[i_coord_offset+DIM3*0+YY1]; |
818 | iz0 = shZ + x[i_coord_offset+DIM3*0+ZZ2]; |
819 | ix1 = shX + x[i_coord_offset+DIM3*1+XX0]; |
820 | iy1 = shY + x[i_coord_offset+DIM3*1+YY1]; |
821 | iz1 = shZ + x[i_coord_offset+DIM3*1+ZZ2]; |
822 | ix2 = shX + x[i_coord_offset+DIM3*2+XX0]; |
823 | iy2 = shY + x[i_coord_offset+DIM3*2+YY1]; |
824 | iz2 = shZ + x[i_coord_offset+DIM3*2+ZZ2]; |
825 | ix3 = shX + x[i_coord_offset+DIM3*3+XX0]; |
826 | iy3 = shY + x[i_coord_offset+DIM3*3+YY1]; |
827 | iz3 = shZ + x[i_coord_offset+DIM3*3+ZZ2]; |
828 | |
829 | fix0 = 0.0; |
830 | fiy0 = 0.0; |
831 | fiz0 = 0.0; |
832 | fix1 = 0.0; |
833 | fiy1 = 0.0; |
834 | fiz1 = 0.0; |
835 | fix2 = 0.0; |
836 | fiy2 = 0.0; |
837 | fiz2 = 0.0; |
838 | fix3 = 0.0; |
839 | fiy3 = 0.0; |
840 | fiz3 = 0.0; |
841 | |
842 | /* Start inner kernel loop */ |
843 | for(jidx=j_index_start; jidx<j_index_end; jidx++) |
844 | { |
845 | /* Get j neighbor index, and coordinate index */ |
846 | jnr = jjnr[jidx]; |
847 | j_coord_offset = DIM3*jnr; |
848 | |
849 | /* load j atom coordinates */ |
850 | jx0 = x[j_coord_offset+DIM3*0+XX0]; |
851 | jy0 = x[j_coord_offset+DIM3*0+YY1]; |
852 | jz0 = x[j_coord_offset+DIM3*0+ZZ2]; |
853 | jx1 = x[j_coord_offset+DIM3*1+XX0]; |
854 | jy1 = x[j_coord_offset+DIM3*1+YY1]; |
855 | jz1 = x[j_coord_offset+DIM3*1+ZZ2]; |
856 | jx2 = x[j_coord_offset+DIM3*2+XX0]; |
857 | jy2 = x[j_coord_offset+DIM3*2+YY1]; |
858 | jz2 = x[j_coord_offset+DIM3*2+ZZ2]; |
859 | jx3 = x[j_coord_offset+DIM3*3+XX0]; |
860 | jy3 = x[j_coord_offset+DIM3*3+YY1]; |
861 | jz3 = x[j_coord_offset+DIM3*3+ZZ2]; |
862 | |
863 | /* Calculate displacement vector */ |
864 | dx00 = ix0 - jx0; |
865 | dy00 = iy0 - jy0; |
866 | dz00 = iz0 - jz0; |
867 | dx11 = ix1 - jx1; |
868 | dy11 = iy1 - jy1; |
869 | dz11 = iz1 - jz1; |
870 | dx12 = ix1 - jx2; |
871 | dy12 = iy1 - jy2; |
872 | dz12 = iz1 - jz2; |
873 | dx13 = ix1 - jx3; |
874 | dy13 = iy1 - jy3; |
875 | dz13 = iz1 - jz3; |
876 | dx21 = ix2 - jx1; |
877 | dy21 = iy2 - jy1; |
878 | dz21 = iz2 - jz1; |
879 | dx22 = ix2 - jx2; |
880 | dy22 = iy2 - jy2; |
881 | dz22 = iz2 - jz2; |
882 | dx23 = ix2 - jx3; |
883 | dy23 = iy2 - jy3; |
884 | dz23 = iz2 - jz3; |
885 | dx31 = ix3 - jx1; |
886 | dy31 = iy3 - jy1; |
887 | dz31 = iz3 - jz1; |
888 | dx32 = ix3 - jx2; |
889 | dy32 = iy3 - jy2; |
890 | dz32 = iz3 - jz2; |
891 | dx33 = ix3 - jx3; |
892 | dy33 = iy3 - jy3; |
893 | dz33 = iz3 - jz3; |
894 | |
895 | /* Calculate squared distance and things based on it */ |
896 | rsq00 = dx00*dx00+dy00*dy00+dz00*dz00; |
897 | rsq11 = dx11*dx11+dy11*dy11+dz11*dz11; |
898 | rsq12 = dx12*dx12+dy12*dy12+dz12*dz12; |
899 | rsq13 = dx13*dx13+dy13*dy13+dz13*dz13; |
900 | rsq21 = dx21*dx21+dy21*dy21+dz21*dz21; |
901 | rsq22 = dx22*dx22+dy22*dy22+dz22*dz22; |
902 | rsq23 = dx23*dx23+dy23*dy23+dz23*dz23; |
903 | rsq31 = dx31*dx31+dy31*dy31+dz31*dz31; |
904 | rsq32 = dx32*dx32+dy32*dy32+dz32*dz32; |
905 | rsq33 = dx33*dx33+dy33*dy33+dz33*dz33; |
906 | |
907 | rinv00 = gmx_invsqrt(rsq00)gmx_software_invsqrt(rsq00); |
908 | rinv11 = gmx_invsqrt(rsq11)gmx_software_invsqrt(rsq11); |
909 | rinv12 = gmx_invsqrt(rsq12)gmx_software_invsqrt(rsq12); |
910 | rinv13 = gmx_invsqrt(rsq13)gmx_software_invsqrt(rsq13); |
911 | rinv21 = gmx_invsqrt(rsq21)gmx_software_invsqrt(rsq21); |
912 | rinv22 = gmx_invsqrt(rsq22)gmx_software_invsqrt(rsq22); |
913 | rinv23 = gmx_invsqrt(rsq23)gmx_software_invsqrt(rsq23); |
914 | rinv31 = gmx_invsqrt(rsq31)gmx_software_invsqrt(rsq31); |
915 | rinv32 = gmx_invsqrt(rsq32)gmx_software_invsqrt(rsq32); |
916 | rinv33 = gmx_invsqrt(rsq33)gmx_software_invsqrt(rsq33); |
917 | |
918 | rinvsq00 = rinv00*rinv00; |
919 | rinvsq11 = rinv11*rinv11; |
920 | rinvsq12 = rinv12*rinv12; |
921 | rinvsq13 = rinv13*rinv13; |
922 | rinvsq21 = rinv21*rinv21; |
923 | rinvsq22 = rinv22*rinv22; |
924 | rinvsq23 = rinv23*rinv23; |
925 | rinvsq31 = rinv31*rinv31; |
926 | rinvsq32 = rinv32*rinv32; |
927 | rinvsq33 = rinv33*rinv33; |
928 | |
929 | /************************** |
930 | * CALCULATE INTERACTIONS * |
931 | **************************/ |
932 | |
933 | r00 = rsq00*rinv00; |
934 | |
935 | /* BUCKINGHAM DISPERSION/REPULSION */ |
936 | rinvsix = rinvsq00*rinvsq00*rinvsq00; |
937 | vvdw6 = c6_00*rinvsix; |
938 | br = cexp2_00*r00; |
939 | vvdwexp = cexp1_00*exp(-br); |
940 | fvdw = (br*vvdwexp-vvdw6)*rinvsq00; |
941 | |
942 | fscal = fvdw; |
943 | |
944 | /* Calculate temporary vectorial force */ |
945 | tx = fscal*dx00; |
946 | ty = fscal*dy00; |
947 | tz = fscal*dz00; |
948 | |
949 | /* Update vectorial force */ |
950 | fix0 += tx; |
951 | fiy0 += ty; |
952 | fiz0 += tz; |
953 | f[j_coord_offset+DIM3*0+XX0] -= tx; |
954 | f[j_coord_offset+DIM3*0+YY1] -= ty; |
955 | f[j_coord_offset+DIM3*0+ZZ2] -= tz; |
956 | |
957 | /************************** |
958 | * CALCULATE INTERACTIONS * |
959 | **************************/ |
960 | |
961 | r11 = rsq11*rinv11; |
962 | |
963 | /* EWALD ELECTROSTATICS */ |
964 | |
965 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
966 | ewrt = r11*ewtabscale; |
967 | ewitab = ewrt; |
968 | eweps = ewrt-ewitab; |
969 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
970 | felec = qq11*rinv11*(rinvsq11-felec); |
971 | |
972 | fscal = felec; |
973 | |
974 | /* Calculate temporary vectorial force */ |
975 | tx = fscal*dx11; |
976 | ty = fscal*dy11; |
977 | tz = fscal*dz11; |
978 | |
979 | /* Update vectorial force */ |
980 | fix1 += tx; |
981 | fiy1 += ty; |
982 | fiz1 += tz; |
983 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
984 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
985 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
986 | |
987 | /************************** |
988 | * CALCULATE INTERACTIONS * |
989 | **************************/ |
990 | |
991 | r12 = rsq12*rinv12; |
992 | |
993 | /* EWALD ELECTROSTATICS */ |
994 | |
995 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
996 | ewrt = r12*ewtabscale; |
997 | ewitab = ewrt; |
998 | eweps = ewrt-ewitab; |
999 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1000 | felec = qq12*rinv12*(rinvsq12-felec); |
1001 | |
1002 | fscal = felec; |
1003 | |
1004 | /* Calculate temporary vectorial force */ |
1005 | tx = fscal*dx12; |
1006 | ty = fscal*dy12; |
1007 | tz = fscal*dz12; |
1008 | |
1009 | /* Update vectorial force */ |
1010 | fix1 += tx; |
1011 | fiy1 += ty; |
1012 | fiz1 += tz; |
1013 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1014 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1015 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1016 | |
1017 | /************************** |
1018 | * CALCULATE INTERACTIONS * |
1019 | **************************/ |
1020 | |
1021 | r13 = rsq13*rinv13; |
1022 | |
1023 | /* EWALD ELECTROSTATICS */ |
1024 | |
1025 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1026 | ewrt = r13*ewtabscale; |
1027 | ewitab = ewrt; |
1028 | eweps = ewrt-ewitab; |
1029 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1030 | felec = qq13*rinv13*(rinvsq13-felec); |
1031 | |
1032 | fscal = felec; |
1033 | |
1034 | /* Calculate temporary vectorial force */ |
1035 | tx = fscal*dx13; |
1036 | ty = fscal*dy13; |
1037 | tz = fscal*dz13; |
1038 | |
1039 | /* Update vectorial force */ |
1040 | fix1 += tx; |
1041 | fiy1 += ty; |
1042 | fiz1 += tz; |
1043 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1044 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1045 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1046 | |
1047 | /************************** |
1048 | * CALCULATE INTERACTIONS * |
1049 | **************************/ |
1050 | |
1051 | r21 = rsq21*rinv21; |
1052 | |
1053 | /* EWALD ELECTROSTATICS */ |
1054 | |
1055 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1056 | ewrt = r21*ewtabscale; |
1057 | ewitab = ewrt; |
1058 | eweps = ewrt-ewitab; |
1059 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1060 | felec = qq21*rinv21*(rinvsq21-felec); |
1061 | |
1062 | fscal = felec; |
1063 | |
1064 | /* Calculate temporary vectorial force */ |
1065 | tx = fscal*dx21; |
1066 | ty = fscal*dy21; |
1067 | tz = fscal*dz21; |
1068 | |
1069 | /* Update vectorial force */ |
1070 | fix2 += tx; |
1071 | fiy2 += ty; |
1072 | fiz2 += tz; |
1073 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1074 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1075 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1076 | |
1077 | /************************** |
1078 | * CALCULATE INTERACTIONS * |
1079 | **************************/ |
1080 | |
1081 | r22 = rsq22*rinv22; |
1082 | |
1083 | /* EWALD ELECTROSTATICS */ |
1084 | |
1085 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1086 | ewrt = r22*ewtabscale; |
1087 | ewitab = ewrt; |
1088 | eweps = ewrt-ewitab; |
1089 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1090 | felec = qq22*rinv22*(rinvsq22-felec); |
1091 | |
1092 | fscal = felec; |
1093 | |
1094 | /* Calculate temporary vectorial force */ |
1095 | tx = fscal*dx22; |
1096 | ty = fscal*dy22; |
1097 | tz = fscal*dz22; |
1098 | |
1099 | /* Update vectorial force */ |
1100 | fix2 += tx; |
1101 | fiy2 += ty; |
1102 | fiz2 += tz; |
1103 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1104 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1105 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1106 | |
1107 | /************************** |
1108 | * CALCULATE INTERACTIONS * |
1109 | **************************/ |
1110 | |
1111 | r23 = rsq23*rinv23; |
1112 | |
1113 | /* EWALD ELECTROSTATICS */ |
1114 | |
1115 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1116 | ewrt = r23*ewtabscale; |
1117 | ewitab = ewrt; |
1118 | eweps = ewrt-ewitab; |
1119 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1120 | felec = qq23*rinv23*(rinvsq23-felec); |
1121 | |
1122 | fscal = felec; |
1123 | |
1124 | /* Calculate temporary vectorial force */ |
1125 | tx = fscal*dx23; |
1126 | ty = fscal*dy23; |
1127 | tz = fscal*dz23; |
1128 | |
1129 | /* Update vectorial force */ |
1130 | fix2 += tx; |
1131 | fiy2 += ty; |
1132 | fiz2 += tz; |
1133 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1134 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1135 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1136 | |
1137 | /************************** |
1138 | * CALCULATE INTERACTIONS * |
1139 | **************************/ |
1140 | |
1141 | r31 = rsq31*rinv31; |
1142 | |
1143 | /* EWALD ELECTROSTATICS */ |
1144 | |
1145 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1146 | ewrt = r31*ewtabscale; |
1147 | ewitab = ewrt; |
1148 | eweps = ewrt-ewitab; |
1149 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1150 | felec = qq31*rinv31*(rinvsq31-felec); |
1151 | |
1152 | fscal = felec; |
1153 | |
1154 | /* Calculate temporary vectorial force */ |
1155 | tx = fscal*dx31; |
1156 | ty = fscal*dy31; |
1157 | tz = fscal*dz31; |
1158 | |
1159 | /* Update vectorial force */ |
1160 | fix3 += tx; |
1161 | fiy3 += ty; |
1162 | fiz3 += tz; |
1163 | f[j_coord_offset+DIM3*1+XX0] -= tx; |
1164 | f[j_coord_offset+DIM3*1+YY1] -= ty; |
1165 | f[j_coord_offset+DIM3*1+ZZ2] -= tz; |
1166 | |
1167 | /************************** |
1168 | * CALCULATE INTERACTIONS * |
1169 | **************************/ |
1170 | |
1171 | r32 = rsq32*rinv32; |
1172 | |
1173 | /* EWALD ELECTROSTATICS */ |
1174 | |
1175 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1176 | ewrt = r32*ewtabscale; |
1177 | ewitab = ewrt; |
1178 | eweps = ewrt-ewitab; |
1179 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1180 | felec = qq32*rinv32*(rinvsq32-felec); |
1181 | |
1182 | fscal = felec; |
1183 | |
1184 | /* Calculate temporary vectorial force */ |
1185 | tx = fscal*dx32; |
1186 | ty = fscal*dy32; |
1187 | tz = fscal*dz32; |
1188 | |
1189 | /* Update vectorial force */ |
1190 | fix3 += tx; |
1191 | fiy3 += ty; |
1192 | fiz3 += tz; |
1193 | f[j_coord_offset+DIM3*2+XX0] -= tx; |
1194 | f[j_coord_offset+DIM3*2+YY1] -= ty; |
1195 | f[j_coord_offset+DIM3*2+ZZ2] -= tz; |
1196 | |
1197 | /************************** |
1198 | * CALCULATE INTERACTIONS * |
1199 | **************************/ |
1200 | |
1201 | r33 = rsq33*rinv33; |
1202 | |
1203 | /* EWALD ELECTROSTATICS */ |
1204 | |
1205 | /* Calculate Ewald table index by multiplying r with scale and truncate to integer */ |
1206 | ewrt = r33*ewtabscale; |
1207 | ewitab = ewrt; |
1208 | eweps = ewrt-ewitab; |
1209 | felec = (1.0-eweps)*ewtab[ewitab]+eweps*ewtab[ewitab+1]; |
1210 | felec = qq33*rinv33*(rinvsq33-felec); |
1211 | |
1212 | fscal = felec; |
1213 | |
1214 | /* Calculate temporary vectorial force */ |
1215 | tx = fscal*dx33; |
1216 | ty = fscal*dy33; |
1217 | tz = fscal*dz33; |
1218 | |
1219 | /* Update vectorial force */ |
1220 | fix3 += tx; |
1221 | fiy3 += ty; |
1222 | fiz3 += tz; |
1223 | f[j_coord_offset+DIM3*3+XX0] -= tx; |
1224 | f[j_coord_offset+DIM3*3+YY1] -= ty; |
1225 | f[j_coord_offset+DIM3*3+ZZ2] -= tz; |
1226 | |
1227 | /* Inner loop uses 355 flops */ |
1228 | } |
1229 | /* End of innermost loop */ |
1230 | |
1231 | tx = ty = tz = 0; |
1232 | f[i_coord_offset+DIM3*0+XX0] += fix0; |
1233 | f[i_coord_offset+DIM3*0+YY1] += fiy0; |
1234 | f[i_coord_offset+DIM3*0+ZZ2] += fiz0; |
1235 | tx += fix0; |
1236 | ty += fiy0; |
1237 | tz += fiz0; |
1238 | f[i_coord_offset+DIM3*1+XX0] += fix1; |
1239 | f[i_coord_offset+DIM3*1+YY1] += fiy1; |
1240 | f[i_coord_offset+DIM3*1+ZZ2] += fiz1; |
1241 | tx += fix1; |
1242 | ty += fiy1; |
1243 | tz += fiz1; |
1244 | f[i_coord_offset+DIM3*2+XX0] += fix2; |
1245 | f[i_coord_offset+DIM3*2+YY1] += fiy2; |
1246 | f[i_coord_offset+DIM3*2+ZZ2] += fiz2; |
1247 | tx += fix2; |
1248 | ty += fiy2; |
1249 | tz += fiz2; |
1250 | f[i_coord_offset+DIM3*3+XX0] += fix3; |
1251 | f[i_coord_offset+DIM3*3+YY1] += fiy3; |
1252 | f[i_coord_offset+DIM3*3+ZZ2] += fiz3; |
1253 | tx += fix3; |
1254 | ty += fiy3; |
1255 | tz += fiz3; |
1256 | fshift[i_shift_offset+XX0] += tx; |
1257 | fshift[i_shift_offset+YY1] += ty; |
1258 | fshift[i_shift_offset+ZZ2] += tz; |
1259 | |
1260 | /* Increment number of inner iterations */ |
1261 | inneriter += j_index_end - j_index_start; |
1262 | |
1263 | /* Outer loop uses 39 flops */ |
1264 | } |
1265 | |
1266 | /* Increment number of outer iterations */ |
1267 | outeriter += nri; |
1268 | |
1269 | /* Update outer/inner flops */ |
1270 | |
1271 | inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*39 + inneriter*355)(nrnb)->n[eNR_NBKERNEL_ELEC_VDW_W4W4_F] += outeriter*39 + inneriter *355; |
1272 | } |