2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2012,2013, 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 sse2_double kernel generator.
44 #include "../nb_kernel.h"
45 #include "types/simple.h"
49 #include "gromacs/simd/math_x86_sse2_double.h"
50 #include "kernelutil_x86_sse2_double.h"
53 * Gromacs nonbonded kernel: nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_VF_sse2_double
54 * Electrostatics interaction: Ewald
55 * VdW interaction: LennardJones
56 * Geometry: Water4-Water4
57 * Calculate force/pot: PotentialAndForce
60 nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_VF_sse2_double
61 (t_nblist * gmx_restrict nlist,
62 rvec * gmx_restrict xx,
63 rvec * gmx_restrict ff,
64 t_forcerec * gmx_restrict fr,
65 t_mdatoms * gmx_restrict mdatoms,
66 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
67 t_nrnb * gmx_restrict nrnb)
69 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
70 * just 0 for non-waters.
71 * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
72 * jnr indices corresponding to data put in the four positions in the SIMD register.
74 int i_shift_offset,i_coord_offset,outeriter,inneriter;
75 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
77 int j_coord_offsetA,j_coord_offsetB;
78 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
80 real *shiftvec,*fshift,*x,*f;
81 __m128d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
83 __m128d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
85 __m128d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
87 __m128d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
89 __m128d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
90 int vdwjidx0A,vdwjidx0B;
91 __m128d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
92 int vdwjidx1A,vdwjidx1B;
93 __m128d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
94 int vdwjidx2A,vdwjidx2B;
95 __m128d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
96 int vdwjidx3A,vdwjidx3B;
97 __m128d jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
98 __m128d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
99 __m128d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
100 __m128d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
101 __m128d dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
102 __m128d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
103 __m128d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
104 __m128d dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
105 __m128d dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
106 __m128d dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
107 __m128d dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
108 __m128d velec,felec,velecsum,facel,crf,krf,krf2;
111 __m128d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
114 __m128d one_sixth = _mm_set1_pd(1.0/6.0);
115 __m128d one_twelfth = _mm_set1_pd(1.0/12.0);
117 __m128d ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
119 __m128d rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
120 real rswitch_scalar,d_scalar;
121 __m128d dummy_mask,cutoff_mask;
122 __m128d signbit = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
123 __m128d one = _mm_set1_pd(1.0);
124 __m128d two = _mm_set1_pd(2.0);
130 jindex = nlist->jindex;
132 shiftidx = nlist->shift;
134 shiftvec = fr->shift_vec[0];
135 fshift = fr->fshift[0];
136 facel = _mm_set1_pd(fr->epsfac);
137 charge = mdatoms->chargeA;
138 nvdwtype = fr->ntype;
140 vdwtype = mdatoms->typeA;
142 sh_ewald = _mm_set1_pd(fr->ic->sh_ewald);
143 ewtab = fr->ic->tabq_coul_FDV0;
144 ewtabscale = _mm_set1_pd(fr->ic->tabq_scale);
145 ewtabhalfspace = _mm_set1_pd(0.5/fr->ic->tabq_scale);
147 /* Setup water-specific parameters */
148 inr = nlist->iinr[0];
149 iq1 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
150 iq2 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
151 iq3 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
152 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
154 jq1 = _mm_set1_pd(charge[inr+1]);
155 jq2 = _mm_set1_pd(charge[inr+2]);
156 jq3 = _mm_set1_pd(charge[inr+3]);
157 vdwjidx0A = 2*vdwtype[inr+0];
158 c6_00 = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A]);
159 c12_00 = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A+1]);
160 qq11 = _mm_mul_pd(iq1,jq1);
161 qq12 = _mm_mul_pd(iq1,jq2);
162 qq13 = _mm_mul_pd(iq1,jq3);
163 qq21 = _mm_mul_pd(iq2,jq1);
164 qq22 = _mm_mul_pd(iq2,jq2);
165 qq23 = _mm_mul_pd(iq2,jq3);
166 qq31 = _mm_mul_pd(iq3,jq1);
167 qq32 = _mm_mul_pd(iq3,jq2);
168 qq33 = _mm_mul_pd(iq3,jq3);
170 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
171 rcutoff_scalar = fr->rcoulomb;
172 rcutoff = _mm_set1_pd(rcutoff_scalar);
173 rcutoff2 = _mm_mul_pd(rcutoff,rcutoff);
175 rswitch_scalar = fr->rcoulomb_switch;
176 rswitch = _mm_set1_pd(rswitch_scalar);
177 /* Setup switch parameters */
178 d_scalar = rcutoff_scalar-rswitch_scalar;
179 d = _mm_set1_pd(d_scalar);
180 swV3 = _mm_set1_pd(-10.0/(d_scalar*d_scalar*d_scalar));
181 swV4 = _mm_set1_pd( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
182 swV5 = _mm_set1_pd( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
183 swF2 = _mm_set1_pd(-30.0/(d_scalar*d_scalar*d_scalar));
184 swF3 = _mm_set1_pd( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
185 swF4 = _mm_set1_pd(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
187 /* Avoid stupid compiler warnings */
195 /* Start outer loop over neighborlists */
196 for(iidx=0; iidx<nri; iidx++)
198 /* Load shift vector for this list */
199 i_shift_offset = DIM*shiftidx[iidx];
201 /* Load limits for loop over neighbors */
202 j_index_start = jindex[iidx];
203 j_index_end = jindex[iidx+1];
205 /* Get outer coordinate index */
207 i_coord_offset = DIM*inr;
209 /* Load i particle coords and add shift vector */
210 gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
211 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
213 fix0 = _mm_setzero_pd();
214 fiy0 = _mm_setzero_pd();
215 fiz0 = _mm_setzero_pd();
216 fix1 = _mm_setzero_pd();
217 fiy1 = _mm_setzero_pd();
218 fiz1 = _mm_setzero_pd();
219 fix2 = _mm_setzero_pd();
220 fiy2 = _mm_setzero_pd();
221 fiz2 = _mm_setzero_pd();
222 fix3 = _mm_setzero_pd();
223 fiy3 = _mm_setzero_pd();
224 fiz3 = _mm_setzero_pd();
226 /* Reset potential sums */
227 velecsum = _mm_setzero_pd();
228 vvdwsum = _mm_setzero_pd();
230 /* Start inner kernel loop */
231 for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
234 /* Get j neighbor index, and coordinate index */
237 j_coord_offsetA = DIM*jnrA;
238 j_coord_offsetB = DIM*jnrB;
240 /* load j atom coordinates */
241 gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
242 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
243 &jy2,&jz2,&jx3,&jy3,&jz3);
245 /* Calculate displacement vector */
246 dx00 = _mm_sub_pd(ix0,jx0);
247 dy00 = _mm_sub_pd(iy0,jy0);
248 dz00 = _mm_sub_pd(iz0,jz0);
249 dx11 = _mm_sub_pd(ix1,jx1);
250 dy11 = _mm_sub_pd(iy1,jy1);
251 dz11 = _mm_sub_pd(iz1,jz1);
252 dx12 = _mm_sub_pd(ix1,jx2);
253 dy12 = _mm_sub_pd(iy1,jy2);
254 dz12 = _mm_sub_pd(iz1,jz2);
255 dx13 = _mm_sub_pd(ix1,jx3);
256 dy13 = _mm_sub_pd(iy1,jy3);
257 dz13 = _mm_sub_pd(iz1,jz3);
258 dx21 = _mm_sub_pd(ix2,jx1);
259 dy21 = _mm_sub_pd(iy2,jy1);
260 dz21 = _mm_sub_pd(iz2,jz1);
261 dx22 = _mm_sub_pd(ix2,jx2);
262 dy22 = _mm_sub_pd(iy2,jy2);
263 dz22 = _mm_sub_pd(iz2,jz2);
264 dx23 = _mm_sub_pd(ix2,jx3);
265 dy23 = _mm_sub_pd(iy2,jy3);
266 dz23 = _mm_sub_pd(iz2,jz3);
267 dx31 = _mm_sub_pd(ix3,jx1);
268 dy31 = _mm_sub_pd(iy3,jy1);
269 dz31 = _mm_sub_pd(iz3,jz1);
270 dx32 = _mm_sub_pd(ix3,jx2);
271 dy32 = _mm_sub_pd(iy3,jy2);
272 dz32 = _mm_sub_pd(iz3,jz2);
273 dx33 = _mm_sub_pd(ix3,jx3);
274 dy33 = _mm_sub_pd(iy3,jy3);
275 dz33 = _mm_sub_pd(iz3,jz3);
277 /* Calculate squared distance and things based on it */
278 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
279 rsq11 = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
280 rsq12 = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
281 rsq13 = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
282 rsq21 = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
283 rsq22 = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
284 rsq23 = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
285 rsq31 = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
286 rsq32 = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
287 rsq33 = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
289 rinv00 = gmx_mm_invsqrt_pd(rsq00);
290 rinv11 = gmx_mm_invsqrt_pd(rsq11);
291 rinv12 = gmx_mm_invsqrt_pd(rsq12);
292 rinv13 = gmx_mm_invsqrt_pd(rsq13);
293 rinv21 = gmx_mm_invsqrt_pd(rsq21);
294 rinv22 = gmx_mm_invsqrt_pd(rsq22);
295 rinv23 = gmx_mm_invsqrt_pd(rsq23);
296 rinv31 = gmx_mm_invsqrt_pd(rsq31);
297 rinv32 = gmx_mm_invsqrt_pd(rsq32);
298 rinv33 = gmx_mm_invsqrt_pd(rsq33);
300 rinvsq00 = _mm_mul_pd(rinv00,rinv00);
301 rinvsq11 = _mm_mul_pd(rinv11,rinv11);
302 rinvsq12 = _mm_mul_pd(rinv12,rinv12);
303 rinvsq13 = _mm_mul_pd(rinv13,rinv13);
304 rinvsq21 = _mm_mul_pd(rinv21,rinv21);
305 rinvsq22 = _mm_mul_pd(rinv22,rinv22);
306 rinvsq23 = _mm_mul_pd(rinv23,rinv23);
307 rinvsq31 = _mm_mul_pd(rinv31,rinv31);
308 rinvsq32 = _mm_mul_pd(rinv32,rinv32);
309 rinvsq33 = _mm_mul_pd(rinv33,rinv33);
311 fjx0 = _mm_setzero_pd();
312 fjy0 = _mm_setzero_pd();
313 fjz0 = _mm_setzero_pd();
314 fjx1 = _mm_setzero_pd();
315 fjy1 = _mm_setzero_pd();
316 fjz1 = _mm_setzero_pd();
317 fjx2 = _mm_setzero_pd();
318 fjy2 = _mm_setzero_pd();
319 fjz2 = _mm_setzero_pd();
320 fjx3 = _mm_setzero_pd();
321 fjy3 = _mm_setzero_pd();
322 fjz3 = _mm_setzero_pd();
324 /**************************
325 * CALCULATE INTERACTIONS *
326 **************************/
328 if (gmx_mm_any_lt(rsq00,rcutoff2))
331 r00 = _mm_mul_pd(rsq00,rinv00);
333 /* LENNARD-JONES DISPERSION/REPULSION */
335 rinvsix = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
336 vvdw6 = _mm_mul_pd(c6_00,rinvsix);
337 vvdw12 = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
338 vvdw = _mm_sub_pd( _mm_mul_pd(vvdw12,one_twelfth) , _mm_mul_pd(vvdw6,one_sixth) );
339 fvdw = _mm_mul_pd(_mm_sub_pd(vvdw12,vvdw6),rinvsq00);
341 d = _mm_sub_pd(r00,rswitch);
342 d = _mm_max_pd(d,_mm_setzero_pd());
343 d2 = _mm_mul_pd(d,d);
344 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
346 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
348 /* Evaluate switch function */
349 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
350 fvdw = _mm_sub_pd( _mm_mul_pd(fvdw,sw) , _mm_mul_pd(rinv00,_mm_mul_pd(vvdw,dsw)) );
351 vvdw = _mm_mul_pd(vvdw,sw);
352 cutoff_mask = _mm_cmplt_pd(rsq00,rcutoff2);
354 /* Update potential sum for this i atom from the interaction with this j atom. */
355 vvdw = _mm_and_pd(vvdw,cutoff_mask);
356 vvdwsum = _mm_add_pd(vvdwsum,vvdw);
360 fscal = _mm_and_pd(fscal,cutoff_mask);
362 /* Calculate temporary vectorial force */
363 tx = _mm_mul_pd(fscal,dx00);
364 ty = _mm_mul_pd(fscal,dy00);
365 tz = _mm_mul_pd(fscal,dz00);
367 /* Update vectorial force */
368 fix0 = _mm_add_pd(fix0,tx);
369 fiy0 = _mm_add_pd(fiy0,ty);
370 fiz0 = _mm_add_pd(fiz0,tz);
372 fjx0 = _mm_add_pd(fjx0,tx);
373 fjy0 = _mm_add_pd(fjy0,ty);
374 fjz0 = _mm_add_pd(fjz0,tz);
378 /**************************
379 * CALCULATE INTERACTIONS *
380 **************************/
382 if (gmx_mm_any_lt(rsq11,rcutoff2))
385 r11 = _mm_mul_pd(rsq11,rinv11);
387 /* EWALD ELECTROSTATICS */
389 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
390 ewrt = _mm_mul_pd(r11,ewtabscale);
391 ewitab = _mm_cvttpd_epi32(ewrt);
392 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
393 ewitab = _mm_slli_epi32(ewitab,2);
394 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
395 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
396 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
397 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
398 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
399 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
400 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
401 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
402 velec = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
403 felec = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
405 d = _mm_sub_pd(r11,rswitch);
406 d = _mm_max_pd(d,_mm_setzero_pd());
407 d2 = _mm_mul_pd(d,d);
408 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
410 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
412 /* Evaluate switch function */
413 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
414 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv11,_mm_mul_pd(velec,dsw)) );
415 velec = _mm_mul_pd(velec,sw);
416 cutoff_mask = _mm_cmplt_pd(rsq11,rcutoff2);
418 /* Update potential sum for this i atom from the interaction with this j atom. */
419 velec = _mm_and_pd(velec,cutoff_mask);
420 velecsum = _mm_add_pd(velecsum,velec);
424 fscal = _mm_and_pd(fscal,cutoff_mask);
426 /* Calculate temporary vectorial force */
427 tx = _mm_mul_pd(fscal,dx11);
428 ty = _mm_mul_pd(fscal,dy11);
429 tz = _mm_mul_pd(fscal,dz11);
431 /* Update vectorial force */
432 fix1 = _mm_add_pd(fix1,tx);
433 fiy1 = _mm_add_pd(fiy1,ty);
434 fiz1 = _mm_add_pd(fiz1,tz);
436 fjx1 = _mm_add_pd(fjx1,tx);
437 fjy1 = _mm_add_pd(fjy1,ty);
438 fjz1 = _mm_add_pd(fjz1,tz);
442 /**************************
443 * CALCULATE INTERACTIONS *
444 **************************/
446 if (gmx_mm_any_lt(rsq12,rcutoff2))
449 r12 = _mm_mul_pd(rsq12,rinv12);
451 /* EWALD ELECTROSTATICS */
453 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
454 ewrt = _mm_mul_pd(r12,ewtabscale);
455 ewitab = _mm_cvttpd_epi32(ewrt);
456 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
457 ewitab = _mm_slli_epi32(ewitab,2);
458 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
459 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
460 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
461 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
462 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
463 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
464 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
465 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
466 velec = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
467 felec = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
469 d = _mm_sub_pd(r12,rswitch);
470 d = _mm_max_pd(d,_mm_setzero_pd());
471 d2 = _mm_mul_pd(d,d);
472 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
474 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
476 /* Evaluate switch function */
477 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
478 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv12,_mm_mul_pd(velec,dsw)) );
479 velec = _mm_mul_pd(velec,sw);
480 cutoff_mask = _mm_cmplt_pd(rsq12,rcutoff2);
482 /* Update potential sum for this i atom from the interaction with this j atom. */
483 velec = _mm_and_pd(velec,cutoff_mask);
484 velecsum = _mm_add_pd(velecsum,velec);
488 fscal = _mm_and_pd(fscal,cutoff_mask);
490 /* Calculate temporary vectorial force */
491 tx = _mm_mul_pd(fscal,dx12);
492 ty = _mm_mul_pd(fscal,dy12);
493 tz = _mm_mul_pd(fscal,dz12);
495 /* Update vectorial force */
496 fix1 = _mm_add_pd(fix1,tx);
497 fiy1 = _mm_add_pd(fiy1,ty);
498 fiz1 = _mm_add_pd(fiz1,tz);
500 fjx2 = _mm_add_pd(fjx2,tx);
501 fjy2 = _mm_add_pd(fjy2,ty);
502 fjz2 = _mm_add_pd(fjz2,tz);
506 /**************************
507 * CALCULATE INTERACTIONS *
508 **************************/
510 if (gmx_mm_any_lt(rsq13,rcutoff2))
513 r13 = _mm_mul_pd(rsq13,rinv13);
515 /* EWALD ELECTROSTATICS */
517 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
518 ewrt = _mm_mul_pd(r13,ewtabscale);
519 ewitab = _mm_cvttpd_epi32(ewrt);
520 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
521 ewitab = _mm_slli_epi32(ewitab,2);
522 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
523 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
524 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
525 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
526 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
527 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
528 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
529 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
530 velec = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
531 felec = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
533 d = _mm_sub_pd(r13,rswitch);
534 d = _mm_max_pd(d,_mm_setzero_pd());
535 d2 = _mm_mul_pd(d,d);
536 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
538 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
540 /* Evaluate switch function */
541 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
542 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv13,_mm_mul_pd(velec,dsw)) );
543 velec = _mm_mul_pd(velec,sw);
544 cutoff_mask = _mm_cmplt_pd(rsq13,rcutoff2);
546 /* Update potential sum for this i atom from the interaction with this j atom. */
547 velec = _mm_and_pd(velec,cutoff_mask);
548 velecsum = _mm_add_pd(velecsum,velec);
552 fscal = _mm_and_pd(fscal,cutoff_mask);
554 /* Calculate temporary vectorial force */
555 tx = _mm_mul_pd(fscal,dx13);
556 ty = _mm_mul_pd(fscal,dy13);
557 tz = _mm_mul_pd(fscal,dz13);
559 /* Update vectorial force */
560 fix1 = _mm_add_pd(fix1,tx);
561 fiy1 = _mm_add_pd(fiy1,ty);
562 fiz1 = _mm_add_pd(fiz1,tz);
564 fjx3 = _mm_add_pd(fjx3,tx);
565 fjy3 = _mm_add_pd(fjy3,ty);
566 fjz3 = _mm_add_pd(fjz3,tz);
570 /**************************
571 * CALCULATE INTERACTIONS *
572 **************************/
574 if (gmx_mm_any_lt(rsq21,rcutoff2))
577 r21 = _mm_mul_pd(rsq21,rinv21);
579 /* EWALD ELECTROSTATICS */
581 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
582 ewrt = _mm_mul_pd(r21,ewtabscale);
583 ewitab = _mm_cvttpd_epi32(ewrt);
584 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
585 ewitab = _mm_slli_epi32(ewitab,2);
586 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
587 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
588 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
589 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
590 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
591 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
592 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
593 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
594 velec = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
595 felec = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
597 d = _mm_sub_pd(r21,rswitch);
598 d = _mm_max_pd(d,_mm_setzero_pd());
599 d2 = _mm_mul_pd(d,d);
600 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
602 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
604 /* Evaluate switch function */
605 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
606 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv21,_mm_mul_pd(velec,dsw)) );
607 velec = _mm_mul_pd(velec,sw);
608 cutoff_mask = _mm_cmplt_pd(rsq21,rcutoff2);
610 /* Update potential sum for this i atom from the interaction with this j atom. */
611 velec = _mm_and_pd(velec,cutoff_mask);
612 velecsum = _mm_add_pd(velecsum,velec);
616 fscal = _mm_and_pd(fscal,cutoff_mask);
618 /* Calculate temporary vectorial force */
619 tx = _mm_mul_pd(fscal,dx21);
620 ty = _mm_mul_pd(fscal,dy21);
621 tz = _mm_mul_pd(fscal,dz21);
623 /* Update vectorial force */
624 fix2 = _mm_add_pd(fix2,tx);
625 fiy2 = _mm_add_pd(fiy2,ty);
626 fiz2 = _mm_add_pd(fiz2,tz);
628 fjx1 = _mm_add_pd(fjx1,tx);
629 fjy1 = _mm_add_pd(fjy1,ty);
630 fjz1 = _mm_add_pd(fjz1,tz);
634 /**************************
635 * CALCULATE INTERACTIONS *
636 **************************/
638 if (gmx_mm_any_lt(rsq22,rcutoff2))
641 r22 = _mm_mul_pd(rsq22,rinv22);
643 /* EWALD ELECTROSTATICS */
645 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
646 ewrt = _mm_mul_pd(r22,ewtabscale);
647 ewitab = _mm_cvttpd_epi32(ewrt);
648 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
649 ewitab = _mm_slli_epi32(ewitab,2);
650 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
651 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
652 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
653 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
654 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
655 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
656 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
657 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
658 velec = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
659 felec = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
661 d = _mm_sub_pd(r22,rswitch);
662 d = _mm_max_pd(d,_mm_setzero_pd());
663 d2 = _mm_mul_pd(d,d);
664 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
666 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
668 /* Evaluate switch function */
669 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
670 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv22,_mm_mul_pd(velec,dsw)) );
671 velec = _mm_mul_pd(velec,sw);
672 cutoff_mask = _mm_cmplt_pd(rsq22,rcutoff2);
674 /* Update potential sum for this i atom from the interaction with this j atom. */
675 velec = _mm_and_pd(velec,cutoff_mask);
676 velecsum = _mm_add_pd(velecsum,velec);
680 fscal = _mm_and_pd(fscal,cutoff_mask);
682 /* Calculate temporary vectorial force */
683 tx = _mm_mul_pd(fscal,dx22);
684 ty = _mm_mul_pd(fscal,dy22);
685 tz = _mm_mul_pd(fscal,dz22);
687 /* Update vectorial force */
688 fix2 = _mm_add_pd(fix2,tx);
689 fiy2 = _mm_add_pd(fiy2,ty);
690 fiz2 = _mm_add_pd(fiz2,tz);
692 fjx2 = _mm_add_pd(fjx2,tx);
693 fjy2 = _mm_add_pd(fjy2,ty);
694 fjz2 = _mm_add_pd(fjz2,tz);
698 /**************************
699 * CALCULATE INTERACTIONS *
700 **************************/
702 if (gmx_mm_any_lt(rsq23,rcutoff2))
705 r23 = _mm_mul_pd(rsq23,rinv23);
707 /* EWALD ELECTROSTATICS */
709 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
710 ewrt = _mm_mul_pd(r23,ewtabscale);
711 ewitab = _mm_cvttpd_epi32(ewrt);
712 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
713 ewitab = _mm_slli_epi32(ewitab,2);
714 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
715 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
716 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
717 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
718 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
719 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
720 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
721 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
722 velec = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
723 felec = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
725 d = _mm_sub_pd(r23,rswitch);
726 d = _mm_max_pd(d,_mm_setzero_pd());
727 d2 = _mm_mul_pd(d,d);
728 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
730 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
732 /* Evaluate switch function */
733 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
734 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv23,_mm_mul_pd(velec,dsw)) );
735 velec = _mm_mul_pd(velec,sw);
736 cutoff_mask = _mm_cmplt_pd(rsq23,rcutoff2);
738 /* Update potential sum for this i atom from the interaction with this j atom. */
739 velec = _mm_and_pd(velec,cutoff_mask);
740 velecsum = _mm_add_pd(velecsum,velec);
744 fscal = _mm_and_pd(fscal,cutoff_mask);
746 /* Calculate temporary vectorial force */
747 tx = _mm_mul_pd(fscal,dx23);
748 ty = _mm_mul_pd(fscal,dy23);
749 tz = _mm_mul_pd(fscal,dz23);
751 /* Update vectorial force */
752 fix2 = _mm_add_pd(fix2,tx);
753 fiy2 = _mm_add_pd(fiy2,ty);
754 fiz2 = _mm_add_pd(fiz2,tz);
756 fjx3 = _mm_add_pd(fjx3,tx);
757 fjy3 = _mm_add_pd(fjy3,ty);
758 fjz3 = _mm_add_pd(fjz3,tz);
762 /**************************
763 * CALCULATE INTERACTIONS *
764 **************************/
766 if (gmx_mm_any_lt(rsq31,rcutoff2))
769 r31 = _mm_mul_pd(rsq31,rinv31);
771 /* EWALD ELECTROSTATICS */
773 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
774 ewrt = _mm_mul_pd(r31,ewtabscale);
775 ewitab = _mm_cvttpd_epi32(ewrt);
776 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
777 ewitab = _mm_slli_epi32(ewitab,2);
778 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
779 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
780 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
781 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
782 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
783 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
784 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
785 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
786 velec = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
787 felec = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
789 d = _mm_sub_pd(r31,rswitch);
790 d = _mm_max_pd(d,_mm_setzero_pd());
791 d2 = _mm_mul_pd(d,d);
792 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
794 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
796 /* Evaluate switch function */
797 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
798 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv31,_mm_mul_pd(velec,dsw)) );
799 velec = _mm_mul_pd(velec,sw);
800 cutoff_mask = _mm_cmplt_pd(rsq31,rcutoff2);
802 /* Update potential sum for this i atom from the interaction with this j atom. */
803 velec = _mm_and_pd(velec,cutoff_mask);
804 velecsum = _mm_add_pd(velecsum,velec);
808 fscal = _mm_and_pd(fscal,cutoff_mask);
810 /* Calculate temporary vectorial force */
811 tx = _mm_mul_pd(fscal,dx31);
812 ty = _mm_mul_pd(fscal,dy31);
813 tz = _mm_mul_pd(fscal,dz31);
815 /* Update vectorial force */
816 fix3 = _mm_add_pd(fix3,tx);
817 fiy3 = _mm_add_pd(fiy3,ty);
818 fiz3 = _mm_add_pd(fiz3,tz);
820 fjx1 = _mm_add_pd(fjx1,tx);
821 fjy1 = _mm_add_pd(fjy1,ty);
822 fjz1 = _mm_add_pd(fjz1,tz);
826 /**************************
827 * CALCULATE INTERACTIONS *
828 **************************/
830 if (gmx_mm_any_lt(rsq32,rcutoff2))
833 r32 = _mm_mul_pd(rsq32,rinv32);
835 /* EWALD ELECTROSTATICS */
837 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
838 ewrt = _mm_mul_pd(r32,ewtabscale);
839 ewitab = _mm_cvttpd_epi32(ewrt);
840 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
841 ewitab = _mm_slli_epi32(ewitab,2);
842 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
843 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
844 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
845 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
846 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
847 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
848 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
849 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
850 velec = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
851 felec = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
853 d = _mm_sub_pd(r32,rswitch);
854 d = _mm_max_pd(d,_mm_setzero_pd());
855 d2 = _mm_mul_pd(d,d);
856 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
858 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
860 /* Evaluate switch function */
861 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
862 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv32,_mm_mul_pd(velec,dsw)) );
863 velec = _mm_mul_pd(velec,sw);
864 cutoff_mask = _mm_cmplt_pd(rsq32,rcutoff2);
866 /* Update potential sum for this i atom from the interaction with this j atom. */
867 velec = _mm_and_pd(velec,cutoff_mask);
868 velecsum = _mm_add_pd(velecsum,velec);
872 fscal = _mm_and_pd(fscal,cutoff_mask);
874 /* Calculate temporary vectorial force */
875 tx = _mm_mul_pd(fscal,dx32);
876 ty = _mm_mul_pd(fscal,dy32);
877 tz = _mm_mul_pd(fscal,dz32);
879 /* Update vectorial force */
880 fix3 = _mm_add_pd(fix3,tx);
881 fiy3 = _mm_add_pd(fiy3,ty);
882 fiz3 = _mm_add_pd(fiz3,tz);
884 fjx2 = _mm_add_pd(fjx2,tx);
885 fjy2 = _mm_add_pd(fjy2,ty);
886 fjz2 = _mm_add_pd(fjz2,tz);
890 /**************************
891 * CALCULATE INTERACTIONS *
892 **************************/
894 if (gmx_mm_any_lt(rsq33,rcutoff2))
897 r33 = _mm_mul_pd(rsq33,rinv33);
899 /* EWALD ELECTROSTATICS */
901 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
902 ewrt = _mm_mul_pd(r33,ewtabscale);
903 ewitab = _mm_cvttpd_epi32(ewrt);
904 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
905 ewitab = _mm_slli_epi32(ewitab,2);
906 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
907 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
908 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
909 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
910 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
911 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
912 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
913 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
914 velec = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
915 felec = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
917 d = _mm_sub_pd(r33,rswitch);
918 d = _mm_max_pd(d,_mm_setzero_pd());
919 d2 = _mm_mul_pd(d,d);
920 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
922 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
924 /* Evaluate switch function */
925 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
926 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv33,_mm_mul_pd(velec,dsw)) );
927 velec = _mm_mul_pd(velec,sw);
928 cutoff_mask = _mm_cmplt_pd(rsq33,rcutoff2);
930 /* Update potential sum for this i atom from the interaction with this j atom. */
931 velec = _mm_and_pd(velec,cutoff_mask);
932 velecsum = _mm_add_pd(velecsum,velec);
936 fscal = _mm_and_pd(fscal,cutoff_mask);
938 /* Calculate temporary vectorial force */
939 tx = _mm_mul_pd(fscal,dx33);
940 ty = _mm_mul_pd(fscal,dy33);
941 tz = _mm_mul_pd(fscal,dz33);
943 /* Update vectorial force */
944 fix3 = _mm_add_pd(fix3,tx);
945 fiy3 = _mm_add_pd(fiy3,ty);
946 fiz3 = _mm_add_pd(fiz3,tz);
948 fjx3 = _mm_add_pd(fjx3,tx);
949 fjy3 = _mm_add_pd(fjy3,ty);
950 fjz3 = _mm_add_pd(fjz3,tz);
954 gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
956 /* Inner loop uses 647 flops */
963 j_coord_offsetA = DIM*jnrA;
965 /* load j atom coordinates */
966 gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
967 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
968 &jy2,&jz2,&jx3,&jy3,&jz3);
970 /* Calculate displacement vector */
971 dx00 = _mm_sub_pd(ix0,jx0);
972 dy00 = _mm_sub_pd(iy0,jy0);
973 dz00 = _mm_sub_pd(iz0,jz0);
974 dx11 = _mm_sub_pd(ix1,jx1);
975 dy11 = _mm_sub_pd(iy1,jy1);
976 dz11 = _mm_sub_pd(iz1,jz1);
977 dx12 = _mm_sub_pd(ix1,jx2);
978 dy12 = _mm_sub_pd(iy1,jy2);
979 dz12 = _mm_sub_pd(iz1,jz2);
980 dx13 = _mm_sub_pd(ix1,jx3);
981 dy13 = _mm_sub_pd(iy1,jy3);
982 dz13 = _mm_sub_pd(iz1,jz3);
983 dx21 = _mm_sub_pd(ix2,jx1);
984 dy21 = _mm_sub_pd(iy2,jy1);
985 dz21 = _mm_sub_pd(iz2,jz1);
986 dx22 = _mm_sub_pd(ix2,jx2);
987 dy22 = _mm_sub_pd(iy2,jy2);
988 dz22 = _mm_sub_pd(iz2,jz2);
989 dx23 = _mm_sub_pd(ix2,jx3);
990 dy23 = _mm_sub_pd(iy2,jy3);
991 dz23 = _mm_sub_pd(iz2,jz3);
992 dx31 = _mm_sub_pd(ix3,jx1);
993 dy31 = _mm_sub_pd(iy3,jy1);
994 dz31 = _mm_sub_pd(iz3,jz1);
995 dx32 = _mm_sub_pd(ix3,jx2);
996 dy32 = _mm_sub_pd(iy3,jy2);
997 dz32 = _mm_sub_pd(iz3,jz2);
998 dx33 = _mm_sub_pd(ix3,jx3);
999 dy33 = _mm_sub_pd(iy3,jy3);
1000 dz33 = _mm_sub_pd(iz3,jz3);
1002 /* Calculate squared distance and things based on it */
1003 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1004 rsq11 = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1005 rsq12 = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1006 rsq13 = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
1007 rsq21 = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1008 rsq22 = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1009 rsq23 = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
1010 rsq31 = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
1011 rsq32 = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
1012 rsq33 = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
1014 rinv00 = gmx_mm_invsqrt_pd(rsq00);
1015 rinv11 = gmx_mm_invsqrt_pd(rsq11);
1016 rinv12 = gmx_mm_invsqrt_pd(rsq12);
1017 rinv13 = gmx_mm_invsqrt_pd(rsq13);
1018 rinv21 = gmx_mm_invsqrt_pd(rsq21);
1019 rinv22 = gmx_mm_invsqrt_pd(rsq22);
1020 rinv23 = gmx_mm_invsqrt_pd(rsq23);
1021 rinv31 = gmx_mm_invsqrt_pd(rsq31);
1022 rinv32 = gmx_mm_invsqrt_pd(rsq32);
1023 rinv33 = gmx_mm_invsqrt_pd(rsq33);
1025 rinvsq00 = _mm_mul_pd(rinv00,rinv00);
1026 rinvsq11 = _mm_mul_pd(rinv11,rinv11);
1027 rinvsq12 = _mm_mul_pd(rinv12,rinv12);
1028 rinvsq13 = _mm_mul_pd(rinv13,rinv13);
1029 rinvsq21 = _mm_mul_pd(rinv21,rinv21);
1030 rinvsq22 = _mm_mul_pd(rinv22,rinv22);
1031 rinvsq23 = _mm_mul_pd(rinv23,rinv23);
1032 rinvsq31 = _mm_mul_pd(rinv31,rinv31);
1033 rinvsq32 = _mm_mul_pd(rinv32,rinv32);
1034 rinvsq33 = _mm_mul_pd(rinv33,rinv33);
1036 fjx0 = _mm_setzero_pd();
1037 fjy0 = _mm_setzero_pd();
1038 fjz0 = _mm_setzero_pd();
1039 fjx1 = _mm_setzero_pd();
1040 fjy1 = _mm_setzero_pd();
1041 fjz1 = _mm_setzero_pd();
1042 fjx2 = _mm_setzero_pd();
1043 fjy2 = _mm_setzero_pd();
1044 fjz2 = _mm_setzero_pd();
1045 fjx3 = _mm_setzero_pd();
1046 fjy3 = _mm_setzero_pd();
1047 fjz3 = _mm_setzero_pd();
1049 /**************************
1050 * CALCULATE INTERACTIONS *
1051 **************************/
1053 if (gmx_mm_any_lt(rsq00,rcutoff2))
1056 r00 = _mm_mul_pd(rsq00,rinv00);
1058 /* LENNARD-JONES DISPERSION/REPULSION */
1060 rinvsix = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1061 vvdw6 = _mm_mul_pd(c6_00,rinvsix);
1062 vvdw12 = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
1063 vvdw = _mm_sub_pd( _mm_mul_pd(vvdw12,one_twelfth) , _mm_mul_pd(vvdw6,one_sixth) );
1064 fvdw = _mm_mul_pd(_mm_sub_pd(vvdw12,vvdw6),rinvsq00);
1066 d = _mm_sub_pd(r00,rswitch);
1067 d = _mm_max_pd(d,_mm_setzero_pd());
1068 d2 = _mm_mul_pd(d,d);
1069 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1071 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1073 /* Evaluate switch function */
1074 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1075 fvdw = _mm_sub_pd( _mm_mul_pd(fvdw,sw) , _mm_mul_pd(rinv00,_mm_mul_pd(vvdw,dsw)) );
1076 vvdw = _mm_mul_pd(vvdw,sw);
1077 cutoff_mask = _mm_cmplt_pd(rsq00,rcutoff2);
1079 /* Update potential sum for this i atom from the interaction with this j atom. */
1080 vvdw = _mm_and_pd(vvdw,cutoff_mask);
1081 vvdw = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
1082 vvdwsum = _mm_add_pd(vvdwsum,vvdw);
1086 fscal = _mm_and_pd(fscal,cutoff_mask);
1088 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1090 /* Calculate temporary vectorial force */
1091 tx = _mm_mul_pd(fscal,dx00);
1092 ty = _mm_mul_pd(fscal,dy00);
1093 tz = _mm_mul_pd(fscal,dz00);
1095 /* Update vectorial force */
1096 fix0 = _mm_add_pd(fix0,tx);
1097 fiy0 = _mm_add_pd(fiy0,ty);
1098 fiz0 = _mm_add_pd(fiz0,tz);
1100 fjx0 = _mm_add_pd(fjx0,tx);
1101 fjy0 = _mm_add_pd(fjy0,ty);
1102 fjz0 = _mm_add_pd(fjz0,tz);
1106 /**************************
1107 * CALCULATE INTERACTIONS *
1108 **************************/
1110 if (gmx_mm_any_lt(rsq11,rcutoff2))
1113 r11 = _mm_mul_pd(rsq11,rinv11);
1115 /* EWALD ELECTROSTATICS */
1117 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1118 ewrt = _mm_mul_pd(r11,ewtabscale);
1119 ewitab = _mm_cvttpd_epi32(ewrt);
1120 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1121 ewitab = _mm_slli_epi32(ewitab,2);
1122 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1123 ewtabD = _mm_setzero_pd();
1124 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1125 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1126 ewtabFn = _mm_setzero_pd();
1127 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1128 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1129 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1130 velec = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
1131 felec = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
1133 d = _mm_sub_pd(r11,rswitch);
1134 d = _mm_max_pd(d,_mm_setzero_pd());
1135 d2 = _mm_mul_pd(d,d);
1136 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1138 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1140 /* Evaluate switch function */
1141 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1142 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv11,_mm_mul_pd(velec,dsw)) );
1143 velec = _mm_mul_pd(velec,sw);
1144 cutoff_mask = _mm_cmplt_pd(rsq11,rcutoff2);
1146 /* Update potential sum for this i atom from the interaction with this j atom. */
1147 velec = _mm_and_pd(velec,cutoff_mask);
1148 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1149 velecsum = _mm_add_pd(velecsum,velec);
1153 fscal = _mm_and_pd(fscal,cutoff_mask);
1155 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1157 /* Calculate temporary vectorial force */
1158 tx = _mm_mul_pd(fscal,dx11);
1159 ty = _mm_mul_pd(fscal,dy11);
1160 tz = _mm_mul_pd(fscal,dz11);
1162 /* Update vectorial force */
1163 fix1 = _mm_add_pd(fix1,tx);
1164 fiy1 = _mm_add_pd(fiy1,ty);
1165 fiz1 = _mm_add_pd(fiz1,tz);
1167 fjx1 = _mm_add_pd(fjx1,tx);
1168 fjy1 = _mm_add_pd(fjy1,ty);
1169 fjz1 = _mm_add_pd(fjz1,tz);
1173 /**************************
1174 * CALCULATE INTERACTIONS *
1175 **************************/
1177 if (gmx_mm_any_lt(rsq12,rcutoff2))
1180 r12 = _mm_mul_pd(rsq12,rinv12);
1182 /* EWALD ELECTROSTATICS */
1184 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1185 ewrt = _mm_mul_pd(r12,ewtabscale);
1186 ewitab = _mm_cvttpd_epi32(ewrt);
1187 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1188 ewitab = _mm_slli_epi32(ewitab,2);
1189 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1190 ewtabD = _mm_setzero_pd();
1191 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1192 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1193 ewtabFn = _mm_setzero_pd();
1194 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1195 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1196 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1197 velec = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
1198 felec = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
1200 d = _mm_sub_pd(r12,rswitch);
1201 d = _mm_max_pd(d,_mm_setzero_pd());
1202 d2 = _mm_mul_pd(d,d);
1203 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1205 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1207 /* Evaluate switch function */
1208 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1209 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv12,_mm_mul_pd(velec,dsw)) );
1210 velec = _mm_mul_pd(velec,sw);
1211 cutoff_mask = _mm_cmplt_pd(rsq12,rcutoff2);
1213 /* Update potential sum for this i atom from the interaction with this j atom. */
1214 velec = _mm_and_pd(velec,cutoff_mask);
1215 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1216 velecsum = _mm_add_pd(velecsum,velec);
1220 fscal = _mm_and_pd(fscal,cutoff_mask);
1222 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1224 /* Calculate temporary vectorial force */
1225 tx = _mm_mul_pd(fscal,dx12);
1226 ty = _mm_mul_pd(fscal,dy12);
1227 tz = _mm_mul_pd(fscal,dz12);
1229 /* Update vectorial force */
1230 fix1 = _mm_add_pd(fix1,tx);
1231 fiy1 = _mm_add_pd(fiy1,ty);
1232 fiz1 = _mm_add_pd(fiz1,tz);
1234 fjx2 = _mm_add_pd(fjx2,tx);
1235 fjy2 = _mm_add_pd(fjy2,ty);
1236 fjz2 = _mm_add_pd(fjz2,tz);
1240 /**************************
1241 * CALCULATE INTERACTIONS *
1242 **************************/
1244 if (gmx_mm_any_lt(rsq13,rcutoff2))
1247 r13 = _mm_mul_pd(rsq13,rinv13);
1249 /* EWALD ELECTROSTATICS */
1251 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1252 ewrt = _mm_mul_pd(r13,ewtabscale);
1253 ewitab = _mm_cvttpd_epi32(ewrt);
1254 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1255 ewitab = _mm_slli_epi32(ewitab,2);
1256 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1257 ewtabD = _mm_setzero_pd();
1258 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1259 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1260 ewtabFn = _mm_setzero_pd();
1261 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1262 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1263 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1264 velec = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
1265 felec = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
1267 d = _mm_sub_pd(r13,rswitch);
1268 d = _mm_max_pd(d,_mm_setzero_pd());
1269 d2 = _mm_mul_pd(d,d);
1270 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1272 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1274 /* Evaluate switch function */
1275 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1276 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv13,_mm_mul_pd(velec,dsw)) );
1277 velec = _mm_mul_pd(velec,sw);
1278 cutoff_mask = _mm_cmplt_pd(rsq13,rcutoff2);
1280 /* Update potential sum for this i atom from the interaction with this j atom. */
1281 velec = _mm_and_pd(velec,cutoff_mask);
1282 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1283 velecsum = _mm_add_pd(velecsum,velec);
1287 fscal = _mm_and_pd(fscal,cutoff_mask);
1289 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1291 /* Calculate temporary vectorial force */
1292 tx = _mm_mul_pd(fscal,dx13);
1293 ty = _mm_mul_pd(fscal,dy13);
1294 tz = _mm_mul_pd(fscal,dz13);
1296 /* Update vectorial force */
1297 fix1 = _mm_add_pd(fix1,tx);
1298 fiy1 = _mm_add_pd(fiy1,ty);
1299 fiz1 = _mm_add_pd(fiz1,tz);
1301 fjx3 = _mm_add_pd(fjx3,tx);
1302 fjy3 = _mm_add_pd(fjy3,ty);
1303 fjz3 = _mm_add_pd(fjz3,tz);
1307 /**************************
1308 * CALCULATE INTERACTIONS *
1309 **************************/
1311 if (gmx_mm_any_lt(rsq21,rcutoff2))
1314 r21 = _mm_mul_pd(rsq21,rinv21);
1316 /* EWALD ELECTROSTATICS */
1318 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1319 ewrt = _mm_mul_pd(r21,ewtabscale);
1320 ewitab = _mm_cvttpd_epi32(ewrt);
1321 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1322 ewitab = _mm_slli_epi32(ewitab,2);
1323 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1324 ewtabD = _mm_setzero_pd();
1325 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1326 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1327 ewtabFn = _mm_setzero_pd();
1328 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1329 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1330 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1331 velec = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
1332 felec = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
1334 d = _mm_sub_pd(r21,rswitch);
1335 d = _mm_max_pd(d,_mm_setzero_pd());
1336 d2 = _mm_mul_pd(d,d);
1337 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1339 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1341 /* Evaluate switch function */
1342 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1343 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv21,_mm_mul_pd(velec,dsw)) );
1344 velec = _mm_mul_pd(velec,sw);
1345 cutoff_mask = _mm_cmplt_pd(rsq21,rcutoff2);
1347 /* Update potential sum for this i atom from the interaction with this j atom. */
1348 velec = _mm_and_pd(velec,cutoff_mask);
1349 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1350 velecsum = _mm_add_pd(velecsum,velec);
1354 fscal = _mm_and_pd(fscal,cutoff_mask);
1356 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1358 /* Calculate temporary vectorial force */
1359 tx = _mm_mul_pd(fscal,dx21);
1360 ty = _mm_mul_pd(fscal,dy21);
1361 tz = _mm_mul_pd(fscal,dz21);
1363 /* Update vectorial force */
1364 fix2 = _mm_add_pd(fix2,tx);
1365 fiy2 = _mm_add_pd(fiy2,ty);
1366 fiz2 = _mm_add_pd(fiz2,tz);
1368 fjx1 = _mm_add_pd(fjx1,tx);
1369 fjy1 = _mm_add_pd(fjy1,ty);
1370 fjz1 = _mm_add_pd(fjz1,tz);
1374 /**************************
1375 * CALCULATE INTERACTIONS *
1376 **************************/
1378 if (gmx_mm_any_lt(rsq22,rcutoff2))
1381 r22 = _mm_mul_pd(rsq22,rinv22);
1383 /* EWALD ELECTROSTATICS */
1385 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1386 ewrt = _mm_mul_pd(r22,ewtabscale);
1387 ewitab = _mm_cvttpd_epi32(ewrt);
1388 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1389 ewitab = _mm_slli_epi32(ewitab,2);
1390 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1391 ewtabD = _mm_setzero_pd();
1392 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1393 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1394 ewtabFn = _mm_setzero_pd();
1395 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1396 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1397 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1398 velec = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
1399 felec = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
1401 d = _mm_sub_pd(r22,rswitch);
1402 d = _mm_max_pd(d,_mm_setzero_pd());
1403 d2 = _mm_mul_pd(d,d);
1404 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1406 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1408 /* Evaluate switch function */
1409 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1410 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv22,_mm_mul_pd(velec,dsw)) );
1411 velec = _mm_mul_pd(velec,sw);
1412 cutoff_mask = _mm_cmplt_pd(rsq22,rcutoff2);
1414 /* Update potential sum for this i atom from the interaction with this j atom. */
1415 velec = _mm_and_pd(velec,cutoff_mask);
1416 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1417 velecsum = _mm_add_pd(velecsum,velec);
1421 fscal = _mm_and_pd(fscal,cutoff_mask);
1423 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1425 /* Calculate temporary vectorial force */
1426 tx = _mm_mul_pd(fscal,dx22);
1427 ty = _mm_mul_pd(fscal,dy22);
1428 tz = _mm_mul_pd(fscal,dz22);
1430 /* Update vectorial force */
1431 fix2 = _mm_add_pd(fix2,tx);
1432 fiy2 = _mm_add_pd(fiy2,ty);
1433 fiz2 = _mm_add_pd(fiz2,tz);
1435 fjx2 = _mm_add_pd(fjx2,tx);
1436 fjy2 = _mm_add_pd(fjy2,ty);
1437 fjz2 = _mm_add_pd(fjz2,tz);
1441 /**************************
1442 * CALCULATE INTERACTIONS *
1443 **************************/
1445 if (gmx_mm_any_lt(rsq23,rcutoff2))
1448 r23 = _mm_mul_pd(rsq23,rinv23);
1450 /* EWALD ELECTROSTATICS */
1452 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1453 ewrt = _mm_mul_pd(r23,ewtabscale);
1454 ewitab = _mm_cvttpd_epi32(ewrt);
1455 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1456 ewitab = _mm_slli_epi32(ewitab,2);
1457 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1458 ewtabD = _mm_setzero_pd();
1459 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1460 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1461 ewtabFn = _mm_setzero_pd();
1462 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1463 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1464 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1465 velec = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
1466 felec = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
1468 d = _mm_sub_pd(r23,rswitch);
1469 d = _mm_max_pd(d,_mm_setzero_pd());
1470 d2 = _mm_mul_pd(d,d);
1471 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1473 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1475 /* Evaluate switch function */
1476 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1477 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv23,_mm_mul_pd(velec,dsw)) );
1478 velec = _mm_mul_pd(velec,sw);
1479 cutoff_mask = _mm_cmplt_pd(rsq23,rcutoff2);
1481 /* Update potential sum for this i atom from the interaction with this j atom. */
1482 velec = _mm_and_pd(velec,cutoff_mask);
1483 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1484 velecsum = _mm_add_pd(velecsum,velec);
1488 fscal = _mm_and_pd(fscal,cutoff_mask);
1490 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1492 /* Calculate temporary vectorial force */
1493 tx = _mm_mul_pd(fscal,dx23);
1494 ty = _mm_mul_pd(fscal,dy23);
1495 tz = _mm_mul_pd(fscal,dz23);
1497 /* Update vectorial force */
1498 fix2 = _mm_add_pd(fix2,tx);
1499 fiy2 = _mm_add_pd(fiy2,ty);
1500 fiz2 = _mm_add_pd(fiz2,tz);
1502 fjx3 = _mm_add_pd(fjx3,tx);
1503 fjy3 = _mm_add_pd(fjy3,ty);
1504 fjz3 = _mm_add_pd(fjz3,tz);
1508 /**************************
1509 * CALCULATE INTERACTIONS *
1510 **************************/
1512 if (gmx_mm_any_lt(rsq31,rcutoff2))
1515 r31 = _mm_mul_pd(rsq31,rinv31);
1517 /* EWALD ELECTROSTATICS */
1519 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1520 ewrt = _mm_mul_pd(r31,ewtabscale);
1521 ewitab = _mm_cvttpd_epi32(ewrt);
1522 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1523 ewitab = _mm_slli_epi32(ewitab,2);
1524 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1525 ewtabD = _mm_setzero_pd();
1526 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1527 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1528 ewtabFn = _mm_setzero_pd();
1529 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1530 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1531 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1532 velec = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
1533 felec = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
1535 d = _mm_sub_pd(r31,rswitch);
1536 d = _mm_max_pd(d,_mm_setzero_pd());
1537 d2 = _mm_mul_pd(d,d);
1538 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1540 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1542 /* Evaluate switch function */
1543 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1544 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv31,_mm_mul_pd(velec,dsw)) );
1545 velec = _mm_mul_pd(velec,sw);
1546 cutoff_mask = _mm_cmplt_pd(rsq31,rcutoff2);
1548 /* Update potential sum for this i atom from the interaction with this j atom. */
1549 velec = _mm_and_pd(velec,cutoff_mask);
1550 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1551 velecsum = _mm_add_pd(velecsum,velec);
1555 fscal = _mm_and_pd(fscal,cutoff_mask);
1557 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1559 /* Calculate temporary vectorial force */
1560 tx = _mm_mul_pd(fscal,dx31);
1561 ty = _mm_mul_pd(fscal,dy31);
1562 tz = _mm_mul_pd(fscal,dz31);
1564 /* Update vectorial force */
1565 fix3 = _mm_add_pd(fix3,tx);
1566 fiy3 = _mm_add_pd(fiy3,ty);
1567 fiz3 = _mm_add_pd(fiz3,tz);
1569 fjx1 = _mm_add_pd(fjx1,tx);
1570 fjy1 = _mm_add_pd(fjy1,ty);
1571 fjz1 = _mm_add_pd(fjz1,tz);
1575 /**************************
1576 * CALCULATE INTERACTIONS *
1577 **************************/
1579 if (gmx_mm_any_lt(rsq32,rcutoff2))
1582 r32 = _mm_mul_pd(rsq32,rinv32);
1584 /* EWALD ELECTROSTATICS */
1586 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1587 ewrt = _mm_mul_pd(r32,ewtabscale);
1588 ewitab = _mm_cvttpd_epi32(ewrt);
1589 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1590 ewitab = _mm_slli_epi32(ewitab,2);
1591 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1592 ewtabD = _mm_setzero_pd();
1593 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1594 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1595 ewtabFn = _mm_setzero_pd();
1596 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1597 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1598 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1599 velec = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
1600 felec = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
1602 d = _mm_sub_pd(r32,rswitch);
1603 d = _mm_max_pd(d,_mm_setzero_pd());
1604 d2 = _mm_mul_pd(d,d);
1605 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1607 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1609 /* Evaluate switch function */
1610 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1611 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv32,_mm_mul_pd(velec,dsw)) );
1612 velec = _mm_mul_pd(velec,sw);
1613 cutoff_mask = _mm_cmplt_pd(rsq32,rcutoff2);
1615 /* Update potential sum for this i atom from the interaction with this j atom. */
1616 velec = _mm_and_pd(velec,cutoff_mask);
1617 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1618 velecsum = _mm_add_pd(velecsum,velec);
1622 fscal = _mm_and_pd(fscal,cutoff_mask);
1624 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1626 /* Calculate temporary vectorial force */
1627 tx = _mm_mul_pd(fscal,dx32);
1628 ty = _mm_mul_pd(fscal,dy32);
1629 tz = _mm_mul_pd(fscal,dz32);
1631 /* Update vectorial force */
1632 fix3 = _mm_add_pd(fix3,tx);
1633 fiy3 = _mm_add_pd(fiy3,ty);
1634 fiz3 = _mm_add_pd(fiz3,tz);
1636 fjx2 = _mm_add_pd(fjx2,tx);
1637 fjy2 = _mm_add_pd(fjy2,ty);
1638 fjz2 = _mm_add_pd(fjz2,tz);
1642 /**************************
1643 * CALCULATE INTERACTIONS *
1644 **************************/
1646 if (gmx_mm_any_lt(rsq33,rcutoff2))
1649 r33 = _mm_mul_pd(rsq33,rinv33);
1651 /* EWALD ELECTROSTATICS */
1653 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1654 ewrt = _mm_mul_pd(r33,ewtabscale);
1655 ewitab = _mm_cvttpd_epi32(ewrt);
1656 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
1657 ewitab = _mm_slli_epi32(ewitab,2);
1658 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
1659 ewtabD = _mm_setzero_pd();
1660 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
1661 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
1662 ewtabFn = _mm_setzero_pd();
1663 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
1664 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
1665 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
1666 velec = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
1667 felec = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
1669 d = _mm_sub_pd(r33,rswitch);
1670 d = _mm_max_pd(d,_mm_setzero_pd());
1671 d2 = _mm_mul_pd(d,d);
1672 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
1674 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
1676 /* Evaluate switch function */
1677 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
1678 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv33,_mm_mul_pd(velec,dsw)) );
1679 velec = _mm_mul_pd(velec,sw);
1680 cutoff_mask = _mm_cmplt_pd(rsq33,rcutoff2);
1682 /* Update potential sum for this i atom from the interaction with this j atom. */
1683 velec = _mm_and_pd(velec,cutoff_mask);
1684 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
1685 velecsum = _mm_add_pd(velecsum,velec);
1689 fscal = _mm_and_pd(fscal,cutoff_mask);
1691 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1693 /* Calculate temporary vectorial force */
1694 tx = _mm_mul_pd(fscal,dx33);
1695 ty = _mm_mul_pd(fscal,dy33);
1696 tz = _mm_mul_pd(fscal,dz33);
1698 /* Update vectorial force */
1699 fix3 = _mm_add_pd(fix3,tx);
1700 fiy3 = _mm_add_pd(fiy3,ty);
1701 fiz3 = _mm_add_pd(fiz3,tz);
1703 fjx3 = _mm_add_pd(fjx3,tx);
1704 fjy3 = _mm_add_pd(fjy3,ty);
1705 fjz3 = _mm_add_pd(fjz3,tz);
1709 gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
1711 /* Inner loop uses 647 flops */
1714 /* End of innermost loop */
1716 gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1717 f+i_coord_offset,fshift+i_shift_offset);
1720 /* Update potential energies */
1721 gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
1722 gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
1724 /* Increment number of inner iterations */
1725 inneriter += j_index_end - j_index_start;
1727 /* Outer loop uses 26 flops */
1730 /* Increment number of outer iterations */
1733 /* Update outer/inner flops */
1735 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_VF,outeriter*26 + inneriter*647);
1738 * Gromacs nonbonded kernel: nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_F_sse2_double
1739 * Electrostatics interaction: Ewald
1740 * VdW interaction: LennardJones
1741 * Geometry: Water4-Water4
1742 * Calculate force/pot: Force
1745 nb_kernel_ElecEwSw_VdwLJSw_GeomW4W4_F_sse2_double
1746 (t_nblist * gmx_restrict nlist,
1747 rvec * gmx_restrict xx,
1748 rvec * gmx_restrict ff,
1749 t_forcerec * gmx_restrict fr,
1750 t_mdatoms * gmx_restrict mdatoms,
1751 nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
1752 t_nrnb * gmx_restrict nrnb)
1754 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
1755 * just 0 for non-waters.
1756 * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
1757 * jnr indices corresponding to data put in the four positions in the SIMD register.
1759 int i_shift_offset,i_coord_offset,outeriter,inneriter;
1760 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
1762 int j_coord_offsetA,j_coord_offsetB;
1763 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
1764 real rcutoff_scalar;
1765 real *shiftvec,*fshift,*x,*f;
1766 __m128d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
1768 __m128d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
1770 __m128d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
1772 __m128d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
1774 __m128d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
1775 int vdwjidx0A,vdwjidx0B;
1776 __m128d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
1777 int vdwjidx1A,vdwjidx1B;
1778 __m128d jx1,jy1,jz1,fjx1,fjy1,fjz1,jq1,isaj1;
1779 int vdwjidx2A,vdwjidx2B;
1780 __m128d jx2,jy2,jz2,fjx2,fjy2,fjz2,jq2,isaj2;
1781 int vdwjidx3A,vdwjidx3B;
1782 __m128d jx3,jy3,jz3,fjx3,fjy3,fjz3,jq3,isaj3;
1783 __m128d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
1784 __m128d dx11,dy11,dz11,rsq11,rinv11,rinvsq11,r11,qq11,c6_11,c12_11;
1785 __m128d dx12,dy12,dz12,rsq12,rinv12,rinvsq12,r12,qq12,c6_12,c12_12;
1786 __m128d dx13,dy13,dz13,rsq13,rinv13,rinvsq13,r13,qq13,c6_13,c12_13;
1787 __m128d dx21,dy21,dz21,rsq21,rinv21,rinvsq21,r21,qq21,c6_21,c12_21;
1788 __m128d dx22,dy22,dz22,rsq22,rinv22,rinvsq22,r22,qq22,c6_22,c12_22;
1789 __m128d dx23,dy23,dz23,rsq23,rinv23,rinvsq23,r23,qq23,c6_23,c12_23;
1790 __m128d dx31,dy31,dz31,rsq31,rinv31,rinvsq31,r31,qq31,c6_31,c12_31;
1791 __m128d dx32,dy32,dz32,rsq32,rinv32,rinvsq32,r32,qq32,c6_32,c12_32;
1792 __m128d dx33,dy33,dz33,rsq33,rinv33,rinvsq33,r33,qq33,c6_33,c12_33;
1793 __m128d velec,felec,velecsum,facel,crf,krf,krf2;
1796 __m128d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
1799 __m128d one_sixth = _mm_set1_pd(1.0/6.0);
1800 __m128d one_twelfth = _mm_set1_pd(1.0/12.0);
1802 __m128d ewtabscale,eweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
1804 __m128d rswitch,swV3,swV4,swV5,swF2,swF3,swF4,d,d2,sw,dsw;
1805 real rswitch_scalar,d_scalar;
1806 __m128d dummy_mask,cutoff_mask;
1807 __m128d signbit = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
1808 __m128d one = _mm_set1_pd(1.0);
1809 __m128d two = _mm_set1_pd(2.0);
1815 jindex = nlist->jindex;
1817 shiftidx = nlist->shift;
1819 shiftvec = fr->shift_vec[0];
1820 fshift = fr->fshift[0];
1821 facel = _mm_set1_pd(fr->epsfac);
1822 charge = mdatoms->chargeA;
1823 nvdwtype = fr->ntype;
1824 vdwparam = fr->nbfp;
1825 vdwtype = mdatoms->typeA;
1827 sh_ewald = _mm_set1_pd(fr->ic->sh_ewald);
1828 ewtab = fr->ic->tabq_coul_FDV0;
1829 ewtabscale = _mm_set1_pd(fr->ic->tabq_scale);
1830 ewtabhalfspace = _mm_set1_pd(0.5/fr->ic->tabq_scale);
1832 /* Setup water-specific parameters */
1833 inr = nlist->iinr[0];
1834 iq1 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
1835 iq2 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
1836 iq3 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
1837 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
1839 jq1 = _mm_set1_pd(charge[inr+1]);
1840 jq2 = _mm_set1_pd(charge[inr+2]);
1841 jq3 = _mm_set1_pd(charge[inr+3]);
1842 vdwjidx0A = 2*vdwtype[inr+0];
1843 c6_00 = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A]);
1844 c12_00 = _mm_set1_pd(vdwparam[vdwioffset0+vdwjidx0A+1]);
1845 qq11 = _mm_mul_pd(iq1,jq1);
1846 qq12 = _mm_mul_pd(iq1,jq2);
1847 qq13 = _mm_mul_pd(iq1,jq3);
1848 qq21 = _mm_mul_pd(iq2,jq1);
1849 qq22 = _mm_mul_pd(iq2,jq2);
1850 qq23 = _mm_mul_pd(iq2,jq3);
1851 qq31 = _mm_mul_pd(iq3,jq1);
1852 qq32 = _mm_mul_pd(iq3,jq2);
1853 qq33 = _mm_mul_pd(iq3,jq3);
1855 /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
1856 rcutoff_scalar = fr->rcoulomb;
1857 rcutoff = _mm_set1_pd(rcutoff_scalar);
1858 rcutoff2 = _mm_mul_pd(rcutoff,rcutoff);
1860 rswitch_scalar = fr->rcoulomb_switch;
1861 rswitch = _mm_set1_pd(rswitch_scalar);
1862 /* Setup switch parameters */
1863 d_scalar = rcutoff_scalar-rswitch_scalar;
1864 d = _mm_set1_pd(d_scalar);
1865 swV3 = _mm_set1_pd(-10.0/(d_scalar*d_scalar*d_scalar));
1866 swV4 = _mm_set1_pd( 15.0/(d_scalar*d_scalar*d_scalar*d_scalar));
1867 swV5 = _mm_set1_pd( -6.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
1868 swF2 = _mm_set1_pd(-30.0/(d_scalar*d_scalar*d_scalar));
1869 swF3 = _mm_set1_pd( 60.0/(d_scalar*d_scalar*d_scalar*d_scalar));
1870 swF4 = _mm_set1_pd(-30.0/(d_scalar*d_scalar*d_scalar*d_scalar*d_scalar));
1872 /* Avoid stupid compiler warnings */
1874 j_coord_offsetA = 0;
1875 j_coord_offsetB = 0;
1880 /* Start outer loop over neighborlists */
1881 for(iidx=0; iidx<nri; iidx++)
1883 /* Load shift vector for this list */
1884 i_shift_offset = DIM*shiftidx[iidx];
1886 /* Load limits for loop over neighbors */
1887 j_index_start = jindex[iidx];
1888 j_index_end = jindex[iidx+1];
1890 /* Get outer coordinate index */
1892 i_coord_offset = DIM*inr;
1894 /* Load i particle coords and add shift vector */
1895 gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
1896 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
1898 fix0 = _mm_setzero_pd();
1899 fiy0 = _mm_setzero_pd();
1900 fiz0 = _mm_setzero_pd();
1901 fix1 = _mm_setzero_pd();
1902 fiy1 = _mm_setzero_pd();
1903 fiz1 = _mm_setzero_pd();
1904 fix2 = _mm_setzero_pd();
1905 fiy2 = _mm_setzero_pd();
1906 fiz2 = _mm_setzero_pd();
1907 fix3 = _mm_setzero_pd();
1908 fiy3 = _mm_setzero_pd();
1909 fiz3 = _mm_setzero_pd();
1911 /* Start inner kernel loop */
1912 for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
1915 /* Get j neighbor index, and coordinate index */
1917 jnrB = jjnr[jidx+1];
1918 j_coord_offsetA = DIM*jnrA;
1919 j_coord_offsetB = DIM*jnrB;
1921 /* load j atom coordinates */
1922 gmx_mm_load_4rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
1923 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
1924 &jy2,&jz2,&jx3,&jy3,&jz3);
1926 /* Calculate displacement vector */
1927 dx00 = _mm_sub_pd(ix0,jx0);
1928 dy00 = _mm_sub_pd(iy0,jy0);
1929 dz00 = _mm_sub_pd(iz0,jz0);
1930 dx11 = _mm_sub_pd(ix1,jx1);
1931 dy11 = _mm_sub_pd(iy1,jy1);
1932 dz11 = _mm_sub_pd(iz1,jz1);
1933 dx12 = _mm_sub_pd(ix1,jx2);
1934 dy12 = _mm_sub_pd(iy1,jy2);
1935 dz12 = _mm_sub_pd(iz1,jz2);
1936 dx13 = _mm_sub_pd(ix1,jx3);
1937 dy13 = _mm_sub_pd(iy1,jy3);
1938 dz13 = _mm_sub_pd(iz1,jz3);
1939 dx21 = _mm_sub_pd(ix2,jx1);
1940 dy21 = _mm_sub_pd(iy2,jy1);
1941 dz21 = _mm_sub_pd(iz2,jz1);
1942 dx22 = _mm_sub_pd(ix2,jx2);
1943 dy22 = _mm_sub_pd(iy2,jy2);
1944 dz22 = _mm_sub_pd(iz2,jz2);
1945 dx23 = _mm_sub_pd(ix2,jx3);
1946 dy23 = _mm_sub_pd(iy2,jy3);
1947 dz23 = _mm_sub_pd(iz2,jz3);
1948 dx31 = _mm_sub_pd(ix3,jx1);
1949 dy31 = _mm_sub_pd(iy3,jy1);
1950 dz31 = _mm_sub_pd(iz3,jz1);
1951 dx32 = _mm_sub_pd(ix3,jx2);
1952 dy32 = _mm_sub_pd(iy3,jy2);
1953 dz32 = _mm_sub_pd(iz3,jz2);
1954 dx33 = _mm_sub_pd(ix3,jx3);
1955 dy33 = _mm_sub_pd(iy3,jy3);
1956 dz33 = _mm_sub_pd(iz3,jz3);
1958 /* Calculate squared distance and things based on it */
1959 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1960 rsq11 = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
1961 rsq12 = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
1962 rsq13 = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
1963 rsq21 = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
1964 rsq22 = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
1965 rsq23 = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
1966 rsq31 = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
1967 rsq32 = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
1968 rsq33 = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
1970 rinv00 = gmx_mm_invsqrt_pd(rsq00);
1971 rinv11 = gmx_mm_invsqrt_pd(rsq11);
1972 rinv12 = gmx_mm_invsqrt_pd(rsq12);
1973 rinv13 = gmx_mm_invsqrt_pd(rsq13);
1974 rinv21 = gmx_mm_invsqrt_pd(rsq21);
1975 rinv22 = gmx_mm_invsqrt_pd(rsq22);
1976 rinv23 = gmx_mm_invsqrt_pd(rsq23);
1977 rinv31 = gmx_mm_invsqrt_pd(rsq31);
1978 rinv32 = gmx_mm_invsqrt_pd(rsq32);
1979 rinv33 = gmx_mm_invsqrt_pd(rsq33);
1981 rinvsq00 = _mm_mul_pd(rinv00,rinv00);
1982 rinvsq11 = _mm_mul_pd(rinv11,rinv11);
1983 rinvsq12 = _mm_mul_pd(rinv12,rinv12);
1984 rinvsq13 = _mm_mul_pd(rinv13,rinv13);
1985 rinvsq21 = _mm_mul_pd(rinv21,rinv21);
1986 rinvsq22 = _mm_mul_pd(rinv22,rinv22);
1987 rinvsq23 = _mm_mul_pd(rinv23,rinv23);
1988 rinvsq31 = _mm_mul_pd(rinv31,rinv31);
1989 rinvsq32 = _mm_mul_pd(rinv32,rinv32);
1990 rinvsq33 = _mm_mul_pd(rinv33,rinv33);
1992 fjx0 = _mm_setzero_pd();
1993 fjy0 = _mm_setzero_pd();
1994 fjz0 = _mm_setzero_pd();
1995 fjx1 = _mm_setzero_pd();
1996 fjy1 = _mm_setzero_pd();
1997 fjz1 = _mm_setzero_pd();
1998 fjx2 = _mm_setzero_pd();
1999 fjy2 = _mm_setzero_pd();
2000 fjz2 = _mm_setzero_pd();
2001 fjx3 = _mm_setzero_pd();
2002 fjy3 = _mm_setzero_pd();
2003 fjz3 = _mm_setzero_pd();
2005 /**************************
2006 * CALCULATE INTERACTIONS *
2007 **************************/
2009 if (gmx_mm_any_lt(rsq00,rcutoff2))
2012 r00 = _mm_mul_pd(rsq00,rinv00);
2014 /* LENNARD-JONES DISPERSION/REPULSION */
2016 rinvsix = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
2017 vvdw6 = _mm_mul_pd(c6_00,rinvsix);
2018 vvdw12 = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
2019 vvdw = _mm_sub_pd( _mm_mul_pd(vvdw12,one_twelfth) , _mm_mul_pd(vvdw6,one_sixth) );
2020 fvdw = _mm_mul_pd(_mm_sub_pd(vvdw12,vvdw6),rinvsq00);
2022 d = _mm_sub_pd(r00,rswitch);
2023 d = _mm_max_pd(d,_mm_setzero_pd());
2024 d2 = _mm_mul_pd(d,d);
2025 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2027 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2029 /* Evaluate switch function */
2030 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2031 fvdw = _mm_sub_pd( _mm_mul_pd(fvdw,sw) , _mm_mul_pd(rinv00,_mm_mul_pd(vvdw,dsw)) );
2032 cutoff_mask = _mm_cmplt_pd(rsq00,rcutoff2);
2036 fscal = _mm_and_pd(fscal,cutoff_mask);
2038 /* Calculate temporary vectorial force */
2039 tx = _mm_mul_pd(fscal,dx00);
2040 ty = _mm_mul_pd(fscal,dy00);
2041 tz = _mm_mul_pd(fscal,dz00);
2043 /* Update vectorial force */
2044 fix0 = _mm_add_pd(fix0,tx);
2045 fiy0 = _mm_add_pd(fiy0,ty);
2046 fiz0 = _mm_add_pd(fiz0,tz);
2048 fjx0 = _mm_add_pd(fjx0,tx);
2049 fjy0 = _mm_add_pd(fjy0,ty);
2050 fjz0 = _mm_add_pd(fjz0,tz);
2054 /**************************
2055 * CALCULATE INTERACTIONS *
2056 **************************/
2058 if (gmx_mm_any_lt(rsq11,rcutoff2))
2061 r11 = _mm_mul_pd(rsq11,rinv11);
2063 /* EWALD ELECTROSTATICS */
2065 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2066 ewrt = _mm_mul_pd(r11,ewtabscale);
2067 ewitab = _mm_cvttpd_epi32(ewrt);
2068 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2069 ewitab = _mm_slli_epi32(ewitab,2);
2070 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2071 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2072 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2073 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2074 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2075 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2076 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2077 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2078 velec = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
2079 felec = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
2081 d = _mm_sub_pd(r11,rswitch);
2082 d = _mm_max_pd(d,_mm_setzero_pd());
2083 d2 = _mm_mul_pd(d,d);
2084 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2086 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2088 /* Evaluate switch function */
2089 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2090 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv11,_mm_mul_pd(velec,dsw)) );
2091 cutoff_mask = _mm_cmplt_pd(rsq11,rcutoff2);
2095 fscal = _mm_and_pd(fscal,cutoff_mask);
2097 /* Calculate temporary vectorial force */
2098 tx = _mm_mul_pd(fscal,dx11);
2099 ty = _mm_mul_pd(fscal,dy11);
2100 tz = _mm_mul_pd(fscal,dz11);
2102 /* Update vectorial force */
2103 fix1 = _mm_add_pd(fix1,tx);
2104 fiy1 = _mm_add_pd(fiy1,ty);
2105 fiz1 = _mm_add_pd(fiz1,tz);
2107 fjx1 = _mm_add_pd(fjx1,tx);
2108 fjy1 = _mm_add_pd(fjy1,ty);
2109 fjz1 = _mm_add_pd(fjz1,tz);
2113 /**************************
2114 * CALCULATE INTERACTIONS *
2115 **************************/
2117 if (gmx_mm_any_lt(rsq12,rcutoff2))
2120 r12 = _mm_mul_pd(rsq12,rinv12);
2122 /* EWALD ELECTROSTATICS */
2124 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2125 ewrt = _mm_mul_pd(r12,ewtabscale);
2126 ewitab = _mm_cvttpd_epi32(ewrt);
2127 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2128 ewitab = _mm_slli_epi32(ewitab,2);
2129 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2130 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2131 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2132 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2133 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2134 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2135 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2136 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2137 velec = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
2138 felec = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
2140 d = _mm_sub_pd(r12,rswitch);
2141 d = _mm_max_pd(d,_mm_setzero_pd());
2142 d2 = _mm_mul_pd(d,d);
2143 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2145 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2147 /* Evaluate switch function */
2148 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2149 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv12,_mm_mul_pd(velec,dsw)) );
2150 cutoff_mask = _mm_cmplt_pd(rsq12,rcutoff2);
2154 fscal = _mm_and_pd(fscal,cutoff_mask);
2156 /* Calculate temporary vectorial force */
2157 tx = _mm_mul_pd(fscal,dx12);
2158 ty = _mm_mul_pd(fscal,dy12);
2159 tz = _mm_mul_pd(fscal,dz12);
2161 /* Update vectorial force */
2162 fix1 = _mm_add_pd(fix1,tx);
2163 fiy1 = _mm_add_pd(fiy1,ty);
2164 fiz1 = _mm_add_pd(fiz1,tz);
2166 fjx2 = _mm_add_pd(fjx2,tx);
2167 fjy2 = _mm_add_pd(fjy2,ty);
2168 fjz2 = _mm_add_pd(fjz2,tz);
2172 /**************************
2173 * CALCULATE INTERACTIONS *
2174 **************************/
2176 if (gmx_mm_any_lt(rsq13,rcutoff2))
2179 r13 = _mm_mul_pd(rsq13,rinv13);
2181 /* EWALD ELECTROSTATICS */
2183 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2184 ewrt = _mm_mul_pd(r13,ewtabscale);
2185 ewitab = _mm_cvttpd_epi32(ewrt);
2186 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2187 ewitab = _mm_slli_epi32(ewitab,2);
2188 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2189 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2190 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2191 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2192 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2193 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2194 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2195 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2196 velec = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
2197 felec = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
2199 d = _mm_sub_pd(r13,rswitch);
2200 d = _mm_max_pd(d,_mm_setzero_pd());
2201 d2 = _mm_mul_pd(d,d);
2202 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2204 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2206 /* Evaluate switch function */
2207 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2208 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv13,_mm_mul_pd(velec,dsw)) );
2209 cutoff_mask = _mm_cmplt_pd(rsq13,rcutoff2);
2213 fscal = _mm_and_pd(fscal,cutoff_mask);
2215 /* Calculate temporary vectorial force */
2216 tx = _mm_mul_pd(fscal,dx13);
2217 ty = _mm_mul_pd(fscal,dy13);
2218 tz = _mm_mul_pd(fscal,dz13);
2220 /* Update vectorial force */
2221 fix1 = _mm_add_pd(fix1,tx);
2222 fiy1 = _mm_add_pd(fiy1,ty);
2223 fiz1 = _mm_add_pd(fiz1,tz);
2225 fjx3 = _mm_add_pd(fjx3,tx);
2226 fjy3 = _mm_add_pd(fjy3,ty);
2227 fjz3 = _mm_add_pd(fjz3,tz);
2231 /**************************
2232 * CALCULATE INTERACTIONS *
2233 **************************/
2235 if (gmx_mm_any_lt(rsq21,rcutoff2))
2238 r21 = _mm_mul_pd(rsq21,rinv21);
2240 /* EWALD ELECTROSTATICS */
2242 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2243 ewrt = _mm_mul_pd(r21,ewtabscale);
2244 ewitab = _mm_cvttpd_epi32(ewrt);
2245 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2246 ewitab = _mm_slli_epi32(ewitab,2);
2247 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2248 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2249 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2250 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2251 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2252 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2253 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2254 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2255 velec = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
2256 felec = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
2258 d = _mm_sub_pd(r21,rswitch);
2259 d = _mm_max_pd(d,_mm_setzero_pd());
2260 d2 = _mm_mul_pd(d,d);
2261 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2263 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2265 /* Evaluate switch function */
2266 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2267 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv21,_mm_mul_pd(velec,dsw)) );
2268 cutoff_mask = _mm_cmplt_pd(rsq21,rcutoff2);
2272 fscal = _mm_and_pd(fscal,cutoff_mask);
2274 /* Calculate temporary vectorial force */
2275 tx = _mm_mul_pd(fscal,dx21);
2276 ty = _mm_mul_pd(fscal,dy21);
2277 tz = _mm_mul_pd(fscal,dz21);
2279 /* Update vectorial force */
2280 fix2 = _mm_add_pd(fix2,tx);
2281 fiy2 = _mm_add_pd(fiy2,ty);
2282 fiz2 = _mm_add_pd(fiz2,tz);
2284 fjx1 = _mm_add_pd(fjx1,tx);
2285 fjy1 = _mm_add_pd(fjy1,ty);
2286 fjz1 = _mm_add_pd(fjz1,tz);
2290 /**************************
2291 * CALCULATE INTERACTIONS *
2292 **************************/
2294 if (gmx_mm_any_lt(rsq22,rcutoff2))
2297 r22 = _mm_mul_pd(rsq22,rinv22);
2299 /* EWALD ELECTROSTATICS */
2301 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2302 ewrt = _mm_mul_pd(r22,ewtabscale);
2303 ewitab = _mm_cvttpd_epi32(ewrt);
2304 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2305 ewitab = _mm_slli_epi32(ewitab,2);
2306 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2307 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2308 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2309 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2310 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2311 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2312 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2313 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2314 velec = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
2315 felec = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
2317 d = _mm_sub_pd(r22,rswitch);
2318 d = _mm_max_pd(d,_mm_setzero_pd());
2319 d2 = _mm_mul_pd(d,d);
2320 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2322 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2324 /* Evaluate switch function */
2325 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2326 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv22,_mm_mul_pd(velec,dsw)) );
2327 cutoff_mask = _mm_cmplt_pd(rsq22,rcutoff2);
2331 fscal = _mm_and_pd(fscal,cutoff_mask);
2333 /* Calculate temporary vectorial force */
2334 tx = _mm_mul_pd(fscal,dx22);
2335 ty = _mm_mul_pd(fscal,dy22);
2336 tz = _mm_mul_pd(fscal,dz22);
2338 /* Update vectorial force */
2339 fix2 = _mm_add_pd(fix2,tx);
2340 fiy2 = _mm_add_pd(fiy2,ty);
2341 fiz2 = _mm_add_pd(fiz2,tz);
2343 fjx2 = _mm_add_pd(fjx2,tx);
2344 fjy2 = _mm_add_pd(fjy2,ty);
2345 fjz2 = _mm_add_pd(fjz2,tz);
2349 /**************************
2350 * CALCULATE INTERACTIONS *
2351 **************************/
2353 if (gmx_mm_any_lt(rsq23,rcutoff2))
2356 r23 = _mm_mul_pd(rsq23,rinv23);
2358 /* EWALD ELECTROSTATICS */
2360 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2361 ewrt = _mm_mul_pd(r23,ewtabscale);
2362 ewitab = _mm_cvttpd_epi32(ewrt);
2363 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2364 ewitab = _mm_slli_epi32(ewitab,2);
2365 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2366 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2367 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2368 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2369 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2370 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2371 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2372 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2373 velec = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
2374 felec = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
2376 d = _mm_sub_pd(r23,rswitch);
2377 d = _mm_max_pd(d,_mm_setzero_pd());
2378 d2 = _mm_mul_pd(d,d);
2379 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2381 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2383 /* Evaluate switch function */
2384 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2385 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv23,_mm_mul_pd(velec,dsw)) );
2386 cutoff_mask = _mm_cmplt_pd(rsq23,rcutoff2);
2390 fscal = _mm_and_pd(fscal,cutoff_mask);
2392 /* Calculate temporary vectorial force */
2393 tx = _mm_mul_pd(fscal,dx23);
2394 ty = _mm_mul_pd(fscal,dy23);
2395 tz = _mm_mul_pd(fscal,dz23);
2397 /* Update vectorial force */
2398 fix2 = _mm_add_pd(fix2,tx);
2399 fiy2 = _mm_add_pd(fiy2,ty);
2400 fiz2 = _mm_add_pd(fiz2,tz);
2402 fjx3 = _mm_add_pd(fjx3,tx);
2403 fjy3 = _mm_add_pd(fjy3,ty);
2404 fjz3 = _mm_add_pd(fjz3,tz);
2408 /**************************
2409 * CALCULATE INTERACTIONS *
2410 **************************/
2412 if (gmx_mm_any_lt(rsq31,rcutoff2))
2415 r31 = _mm_mul_pd(rsq31,rinv31);
2417 /* EWALD ELECTROSTATICS */
2419 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2420 ewrt = _mm_mul_pd(r31,ewtabscale);
2421 ewitab = _mm_cvttpd_epi32(ewrt);
2422 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2423 ewitab = _mm_slli_epi32(ewitab,2);
2424 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2425 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2426 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2427 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2428 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2429 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2430 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2431 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2432 velec = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
2433 felec = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
2435 d = _mm_sub_pd(r31,rswitch);
2436 d = _mm_max_pd(d,_mm_setzero_pd());
2437 d2 = _mm_mul_pd(d,d);
2438 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2440 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2442 /* Evaluate switch function */
2443 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2444 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv31,_mm_mul_pd(velec,dsw)) );
2445 cutoff_mask = _mm_cmplt_pd(rsq31,rcutoff2);
2449 fscal = _mm_and_pd(fscal,cutoff_mask);
2451 /* Calculate temporary vectorial force */
2452 tx = _mm_mul_pd(fscal,dx31);
2453 ty = _mm_mul_pd(fscal,dy31);
2454 tz = _mm_mul_pd(fscal,dz31);
2456 /* Update vectorial force */
2457 fix3 = _mm_add_pd(fix3,tx);
2458 fiy3 = _mm_add_pd(fiy3,ty);
2459 fiz3 = _mm_add_pd(fiz3,tz);
2461 fjx1 = _mm_add_pd(fjx1,tx);
2462 fjy1 = _mm_add_pd(fjy1,ty);
2463 fjz1 = _mm_add_pd(fjz1,tz);
2467 /**************************
2468 * CALCULATE INTERACTIONS *
2469 **************************/
2471 if (gmx_mm_any_lt(rsq32,rcutoff2))
2474 r32 = _mm_mul_pd(rsq32,rinv32);
2476 /* EWALD ELECTROSTATICS */
2478 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2479 ewrt = _mm_mul_pd(r32,ewtabscale);
2480 ewitab = _mm_cvttpd_epi32(ewrt);
2481 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2482 ewitab = _mm_slli_epi32(ewitab,2);
2483 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2484 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2485 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2486 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2487 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2488 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2489 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2490 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2491 velec = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
2492 felec = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
2494 d = _mm_sub_pd(r32,rswitch);
2495 d = _mm_max_pd(d,_mm_setzero_pd());
2496 d2 = _mm_mul_pd(d,d);
2497 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2499 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2501 /* Evaluate switch function */
2502 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2503 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv32,_mm_mul_pd(velec,dsw)) );
2504 cutoff_mask = _mm_cmplt_pd(rsq32,rcutoff2);
2508 fscal = _mm_and_pd(fscal,cutoff_mask);
2510 /* Calculate temporary vectorial force */
2511 tx = _mm_mul_pd(fscal,dx32);
2512 ty = _mm_mul_pd(fscal,dy32);
2513 tz = _mm_mul_pd(fscal,dz32);
2515 /* Update vectorial force */
2516 fix3 = _mm_add_pd(fix3,tx);
2517 fiy3 = _mm_add_pd(fiy3,ty);
2518 fiz3 = _mm_add_pd(fiz3,tz);
2520 fjx2 = _mm_add_pd(fjx2,tx);
2521 fjy2 = _mm_add_pd(fjy2,ty);
2522 fjz2 = _mm_add_pd(fjz2,tz);
2526 /**************************
2527 * CALCULATE INTERACTIONS *
2528 **************************/
2530 if (gmx_mm_any_lt(rsq33,rcutoff2))
2533 r33 = _mm_mul_pd(rsq33,rinv33);
2535 /* EWALD ELECTROSTATICS */
2537 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2538 ewrt = _mm_mul_pd(r33,ewtabscale);
2539 ewitab = _mm_cvttpd_epi32(ewrt);
2540 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2541 ewitab = _mm_slli_epi32(ewitab,2);
2542 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2543 ewtabD = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,1) );
2544 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2545 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2546 ewtabFn = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,1) +2);
2547 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2548 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2549 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2550 velec = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
2551 felec = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
2553 d = _mm_sub_pd(r33,rswitch);
2554 d = _mm_max_pd(d,_mm_setzero_pd());
2555 d2 = _mm_mul_pd(d,d);
2556 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2558 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2560 /* Evaluate switch function */
2561 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2562 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv33,_mm_mul_pd(velec,dsw)) );
2563 cutoff_mask = _mm_cmplt_pd(rsq33,rcutoff2);
2567 fscal = _mm_and_pd(fscal,cutoff_mask);
2569 /* Calculate temporary vectorial force */
2570 tx = _mm_mul_pd(fscal,dx33);
2571 ty = _mm_mul_pd(fscal,dy33);
2572 tz = _mm_mul_pd(fscal,dz33);
2574 /* Update vectorial force */
2575 fix3 = _mm_add_pd(fix3,tx);
2576 fiy3 = _mm_add_pd(fiy3,ty);
2577 fiz3 = _mm_add_pd(fiz3,tz);
2579 fjx3 = _mm_add_pd(fjx3,tx);
2580 fjy3 = _mm_add_pd(fjy3,ty);
2581 fjz3 = _mm_add_pd(fjz3,tz);
2585 gmx_mm_decrement_4rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
2587 /* Inner loop uses 617 flops */
2590 if(jidx<j_index_end)
2594 j_coord_offsetA = DIM*jnrA;
2596 /* load j atom coordinates */
2597 gmx_mm_load_4rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
2598 &jx0,&jy0,&jz0,&jx1,&jy1,&jz1,&jx2,
2599 &jy2,&jz2,&jx3,&jy3,&jz3);
2601 /* Calculate displacement vector */
2602 dx00 = _mm_sub_pd(ix0,jx0);
2603 dy00 = _mm_sub_pd(iy0,jy0);
2604 dz00 = _mm_sub_pd(iz0,jz0);
2605 dx11 = _mm_sub_pd(ix1,jx1);
2606 dy11 = _mm_sub_pd(iy1,jy1);
2607 dz11 = _mm_sub_pd(iz1,jz1);
2608 dx12 = _mm_sub_pd(ix1,jx2);
2609 dy12 = _mm_sub_pd(iy1,jy2);
2610 dz12 = _mm_sub_pd(iz1,jz2);
2611 dx13 = _mm_sub_pd(ix1,jx3);
2612 dy13 = _mm_sub_pd(iy1,jy3);
2613 dz13 = _mm_sub_pd(iz1,jz3);
2614 dx21 = _mm_sub_pd(ix2,jx1);
2615 dy21 = _mm_sub_pd(iy2,jy1);
2616 dz21 = _mm_sub_pd(iz2,jz1);
2617 dx22 = _mm_sub_pd(ix2,jx2);
2618 dy22 = _mm_sub_pd(iy2,jy2);
2619 dz22 = _mm_sub_pd(iz2,jz2);
2620 dx23 = _mm_sub_pd(ix2,jx3);
2621 dy23 = _mm_sub_pd(iy2,jy3);
2622 dz23 = _mm_sub_pd(iz2,jz3);
2623 dx31 = _mm_sub_pd(ix3,jx1);
2624 dy31 = _mm_sub_pd(iy3,jy1);
2625 dz31 = _mm_sub_pd(iz3,jz1);
2626 dx32 = _mm_sub_pd(ix3,jx2);
2627 dy32 = _mm_sub_pd(iy3,jy2);
2628 dz32 = _mm_sub_pd(iz3,jz2);
2629 dx33 = _mm_sub_pd(ix3,jx3);
2630 dy33 = _mm_sub_pd(iy3,jy3);
2631 dz33 = _mm_sub_pd(iz3,jz3);
2633 /* Calculate squared distance and things based on it */
2634 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
2635 rsq11 = gmx_mm_calc_rsq_pd(dx11,dy11,dz11);
2636 rsq12 = gmx_mm_calc_rsq_pd(dx12,dy12,dz12);
2637 rsq13 = gmx_mm_calc_rsq_pd(dx13,dy13,dz13);
2638 rsq21 = gmx_mm_calc_rsq_pd(dx21,dy21,dz21);
2639 rsq22 = gmx_mm_calc_rsq_pd(dx22,dy22,dz22);
2640 rsq23 = gmx_mm_calc_rsq_pd(dx23,dy23,dz23);
2641 rsq31 = gmx_mm_calc_rsq_pd(dx31,dy31,dz31);
2642 rsq32 = gmx_mm_calc_rsq_pd(dx32,dy32,dz32);
2643 rsq33 = gmx_mm_calc_rsq_pd(dx33,dy33,dz33);
2645 rinv00 = gmx_mm_invsqrt_pd(rsq00);
2646 rinv11 = gmx_mm_invsqrt_pd(rsq11);
2647 rinv12 = gmx_mm_invsqrt_pd(rsq12);
2648 rinv13 = gmx_mm_invsqrt_pd(rsq13);
2649 rinv21 = gmx_mm_invsqrt_pd(rsq21);
2650 rinv22 = gmx_mm_invsqrt_pd(rsq22);
2651 rinv23 = gmx_mm_invsqrt_pd(rsq23);
2652 rinv31 = gmx_mm_invsqrt_pd(rsq31);
2653 rinv32 = gmx_mm_invsqrt_pd(rsq32);
2654 rinv33 = gmx_mm_invsqrt_pd(rsq33);
2656 rinvsq00 = _mm_mul_pd(rinv00,rinv00);
2657 rinvsq11 = _mm_mul_pd(rinv11,rinv11);
2658 rinvsq12 = _mm_mul_pd(rinv12,rinv12);
2659 rinvsq13 = _mm_mul_pd(rinv13,rinv13);
2660 rinvsq21 = _mm_mul_pd(rinv21,rinv21);
2661 rinvsq22 = _mm_mul_pd(rinv22,rinv22);
2662 rinvsq23 = _mm_mul_pd(rinv23,rinv23);
2663 rinvsq31 = _mm_mul_pd(rinv31,rinv31);
2664 rinvsq32 = _mm_mul_pd(rinv32,rinv32);
2665 rinvsq33 = _mm_mul_pd(rinv33,rinv33);
2667 fjx0 = _mm_setzero_pd();
2668 fjy0 = _mm_setzero_pd();
2669 fjz0 = _mm_setzero_pd();
2670 fjx1 = _mm_setzero_pd();
2671 fjy1 = _mm_setzero_pd();
2672 fjz1 = _mm_setzero_pd();
2673 fjx2 = _mm_setzero_pd();
2674 fjy2 = _mm_setzero_pd();
2675 fjz2 = _mm_setzero_pd();
2676 fjx3 = _mm_setzero_pd();
2677 fjy3 = _mm_setzero_pd();
2678 fjz3 = _mm_setzero_pd();
2680 /**************************
2681 * CALCULATE INTERACTIONS *
2682 **************************/
2684 if (gmx_mm_any_lt(rsq00,rcutoff2))
2687 r00 = _mm_mul_pd(rsq00,rinv00);
2689 /* LENNARD-JONES DISPERSION/REPULSION */
2691 rinvsix = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
2692 vvdw6 = _mm_mul_pd(c6_00,rinvsix);
2693 vvdw12 = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
2694 vvdw = _mm_sub_pd( _mm_mul_pd(vvdw12,one_twelfth) , _mm_mul_pd(vvdw6,one_sixth) );
2695 fvdw = _mm_mul_pd(_mm_sub_pd(vvdw12,vvdw6),rinvsq00);
2697 d = _mm_sub_pd(r00,rswitch);
2698 d = _mm_max_pd(d,_mm_setzero_pd());
2699 d2 = _mm_mul_pd(d,d);
2700 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2702 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2704 /* Evaluate switch function */
2705 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2706 fvdw = _mm_sub_pd( _mm_mul_pd(fvdw,sw) , _mm_mul_pd(rinv00,_mm_mul_pd(vvdw,dsw)) );
2707 cutoff_mask = _mm_cmplt_pd(rsq00,rcutoff2);
2711 fscal = _mm_and_pd(fscal,cutoff_mask);
2713 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2715 /* Calculate temporary vectorial force */
2716 tx = _mm_mul_pd(fscal,dx00);
2717 ty = _mm_mul_pd(fscal,dy00);
2718 tz = _mm_mul_pd(fscal,dz00);
2720 /* Update vectorial force */
2721 fix0 = _mm_add_pd(fix0,tx);
2722 fiy0 = _mm_add_pd(fiy0,ty);
2723 fiz0 = _mm_add_pd(fiz0,tz);
2725 fjx0 = _mm_add_pd(fjx0,tx);
2726 fjy0 = _mm_add_pd(fjy0,ty);
2727 fjz0 = _mm_add_pd(fjz0,tz);
2731 /**************************
2732 * CALCULATE INTERACTIONS *
2733 **************************/
2735 if (gmx_mm_any_lt(rsq11,rcutoff2))
2738 r11 = _mm_mul_pd(rsq11,rinv11);
2740 /* EWALD ELECTROSTATICS */
2742 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2743 ewrt = _mm_mul_pd(r11,ewtabscale);
2744 ewitab = _mm_cvttpd_epi32(ewrt);
2745 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2746 ewitab = _mm_slli_epi32(ewitab,2);
2747 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2748 ewtabD = _mm_setzero_pd();
2749 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2750 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2751 ewtabFn = _mm_setzero_pd();
2752 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2753 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2754 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2755 velec = _mm_mul_pd(qq11,_mm_sub_pd(rinv11,velec));
2756 felec = _mm_mul_pd(_mm_mul_pd(qq11,rinv11),_mm_sub_pd(rinvsq11,felec));
2758 d = _mm_sub_pd(r11,rswitch);
2759 d = _mm_max_pd(d,_mm_setzero_pd());
2760 d2 = _mm_mul_pd(d,d);
2761 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2763 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2765 /* Evaluate switch function */
2766 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2767 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv11,_mm_mul_pd(velec,dsw)) );
2768 cutoff_mask = _mm_cmplt_pd(rsq11,rcutoff2);
2772 fscal = _mm_and_pd(fscal,cutoff_mask);
2774 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2776 /* Calculate temporary vectorial force */
2777 tx = _mm_mul_pd(fscal,dx11);
2778 ty = _mm_mul_pd(fscal,dy11);
2779 tz = _mm_mul_pd(fscal,dz11);
2781 /* Update vectorial force */
2782 fix1 = _mm_add_pd(fix1,tx);
2783 fiy1 = _mm_add_pd(fiy1,ty);
2784 fiz1 = _mm_add_pd(fiz1,tz);
2786 fjx1 = _mm_add_pd(fjx1,tx);
2787 fjy1 = _mm_add_pd(fjy1,ty);
2788 fjz1 = _mm_add_pd(fjz1,tz);
2792 /**************************
2793 * CALCULATE INTERACTIONS *
2794 **************************/
2796 if (gmx_mm_any_lt(rsq12,rcutoff2))
2799 r12 = _mm_mul_pd(rsq12,rinv12);
2801 /* EWALD ELECTROSTATICS */
2803 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2804 ewrt = _mm_mul_pd(r12,ewtabscale);
2805 ewitab = _mm_cvttpd_epi32(ewrt);
2806 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2807 ewitab = _mm_slli_epi32(ewitab,2);
2808 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2809 ewtabD = _mm_setzero_pd();
2810 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2811 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2812 ewtabFn = _mm_setzero_pd();
2813 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2814 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2815 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2816 velec = _mm_mul_pd(qq12,_mm_sub_pd(rinv12,velec));
2817 felec = _mm_mul_pd(_mm_mul_pd(qq12,rinv12),_mm_sub_pd(rinvsq12,felec));
2819 d = _mm_sub_pd(r12,rswitch);
2820 d = _mm_max_pd(d,_mm_setzero_pd());
2821 d2 = _mm_mul_pd(d,d);
2822 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2824 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2826 /* Evaluate switch function */
2827 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2828 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv12,_mm_mul_pd(velec,dsw)) );
2829 cutoff_mask = _mm_cmplt_pd(rsq12,rcutoff2);
2833 fscal = _mm_and_pd(fscal,cutoff_mask);
2835 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2837 /* Calculate temporary vectorial force */
2838 tx = _mm_mul_pd(fscal,dx12);
2839 ty = _mm_mul_pd(fscal,dy12);
2840 tz = _mm_mul_pd(fscal,dz12);
2842 /* Update vectorial force */
2843 fix1 = _mm_add_pd(fix1,tx);
2844 fiy1 = _mm_add_pd(fiy1,ty);
2845 fiz1 = _mm_add_pd(fiz1,tz);
2847 fjx2 = _mm_add_pd(fjx2,tx);
2848 fjy2 = _mm_add_pd(fjy2,ty);
2849 fjz2 = _mm_add_pd(fjz2,tz);
2853 /**************************
2854 * CALCULATE INTERACTIONS *
2855 **************************/
2857 if (gmx_mm_any_lt(rsq13,rcutoff2))
2860 r13 = _mm_mul_pd(rsq13,rinv13);
2862 /* EWALD ELECTROSTATICS */
2864 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2865 ewrt = _mm_mul_pd(r13,ewtabscale);
2866 ewitab = _mm_cvttpd_epi32(ewrt);
2867 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2868 ewitab = _mm_slli_epi32(ewitab,2);
2869 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2870 ewtabD = _mm_setzero_pd();
2871 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2872 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2873 ewtabFn = _mm_setzero_pd();
2874 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2875 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2876 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2877 velec = _mm_mul_pd(qq13,_mm_sub_pd(rinv13,velec));
2878 felec = _mm_mul_pd(_mm_mul_pd(qq13,rinv13),_mm_sub_pd(rinvsq13,felec));
2880 d = _mm_sub_pd(r13,rswitch);
2881 d = _mm_max_pd(d,_mm_setzero_pd());
2882 d2 = _mm_mul_pd(d,d);
2883 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2885 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2887 /* Evaluate switch function */
2888 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2889 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv13,_mm_mul_pd(velec,dsw)) );
2890 cutoff_mask = _mm_cmplt_pd(rsq13,rcutoff2);
2894 fscal = _mm_and_pd(fscal,cutoff_mask);
2896 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2898 /* Calculate temporary vectorial force */
2899 tx = _mm_mul_pd(fscal,dx13);
2900 ty = _mm_mul_pd(fscal,dy13);
2901 tz = _mm_mul_pd(fscal,dz13);
2903 /* Update vectorial force */
2904 fix1 = _mm_add_pd(fix1,tx);
2905 fiy1 = _mm_add_pd(fiy1,ty);
2906 fiz1 = _mm_add_pd(fiz1,tz);
2908 fjx3 = _mm_add_pd(fjx3,tx);
2909 fjy3 = _mm_add_pd(fjy3,ty);
2910 fjz3 = _mm_add_pd(fjz3,tz);
2914 /**************************
2915 * CALCULATE INTERACTIONS *
2916 **************************/
2918 if (gmx_mm_any_lt(rsq21,rcutoff2))
2921 r21 = _mm_mul_pd(rsq21,rinv21);
2923 /* EWALD ELECTROSTATICS */
2925 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2926 ewrt = _mm_mul_pd(r21,ewtabscale);
2927 ewitab = _mm_cvttpd_epi32(ewrt);
2928 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2929 ewitab = _mm_slli_epi32(ewitab,2);
2930 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2931 ewtabD = _mm_setzero_pd();
2932 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2933 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2934 ewtabFn = _mm_setzero_pd();
2935 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2936 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2937 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2938 velec = _mm_mul_pd(qq21,_mm_sub_pd(rinv21,velec));
2939 felec = _mm_mul_pd(_mm_mul_pd(qq21,rinv21),_mm_sub_pd(rinvsq21,felec));
2941 d = _mm_sub_pd(r21,rswitch);
2942 d = _mm_max_pd(d,_mm_setzero_pd());
2943 d2 = _mm_mul_pd(d,d);
2944 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
2946 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
2948 /* Evaluate switch function */
2949 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
2950 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv21,_mm_mul_pd(velec,dsw)) );
2951 cutoff_mask = _mm_cmplt_pd(rsq21,rcutoff2);
2955 fscal = _mm_and_pd(fscal,cutoff_mask);
2957 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
2959 /* Calculate temporary vectorial force */
2960 tx = _mm_mul_pd(fscal,dx21);
2961 ty = _mm_mul_pd(fscal,dy21);
2962 tz = _mm_mul_pd(fscal,dz21);
2964 /* Update vectorial force */
2965 fix2 = _mm_add_pd(fix2,tx);
2966 fiy2 = _mm_add_pd(fiy2,ty);
2967 fiz2 = _mm_add_pd(fiz2,tz);
2969 fjx1 = _mm_add_pd(fjx1,tx);
2970 fjy1 = _mm_add_pd(fjy1,ty);
2971 fjz1 = _mm_add_pd(fjz1,tz);
2975 /**************************
2976 * CALCULATE INTERACTIONS *
2977 **************************/
2979 if (gmx_mm_any_lt(rsq22,rcutoff2))
2982 r22 = _mm_mul_pd(rsq22,rinv22);
2984 /* EWALD ELECTROSTATICS */
2986 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
2987 ewrt = _mm_mul_pd(r22,ewtabscale);
2988 ewitab = _mm_cvttpd_epi32(ewrt);
2989 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
2990 ewitab = _mm_slli_epi32(ewitab,2);
2991 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
2992 ewtabD = _mm_setzero_pd();
2993 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
2994 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
2995 ewtabFn = _mm_setzero_pd();
2996 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
2997 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
2998 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
2999 velec = _mm_mul_pd(qq22,_mm_sub_pd(rinv22,velec));
3000 felec = _mm_mul_pd(_mm_mul_pd(qq22,rinv22),_mm_sub_pd(rinvsq22,felec));
3002 d = _mm_sub_pd(r22,rswitch);
3003 d = _mm_max_pd(d,_mm_setzero_pd());
3004 d2 = _mm_mul_pd(d,d);
3005 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
3007 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
3009 /* Evaluate switch function */
3010 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3011 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv22,_mm_mul_pd(velec,dsw)) );
3012 cutoff_mask = _mm_cmplt_pd(rsq22,rcutoff2);
3016 fscal = _mm_and_pd(fscal,cutoff_mask);
3018 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
3020 /* Calculate temporary vectorial force */
3021 tx = _mm_mul_pd(fscal,dx22);
3022 ty = _mm_mul_pd(fscal,dy22);
3023 tz = _mm_mul_pd(fscal,dz22);
3025 /* Update vectorial force */
3026 fix2 = _mm_add_pd(fix2,tx);
3027 fiy2 = _mm_add_pd(fiy2,ty);
3028 fiz2 = _mm_add_pd(fiz2,tz);
3030 fjx2 = _mm_add_pd(fjx2,tx);
3031 fjy2 = _mm_add_pd(fjy2,ty);
3032 fjz2 = _mm_add_pd(fjz2,tz);
3036 /**************************
3037 * CALCULATE INTERACTIONS *
3038 **************************/
3040 if (gmx_mm_any_lt(rsq23,rcutoff2))
3043 r23 = _mm_mul_pd(rsq23,rinv23);
3045 /* EWALD ELECTROSTATICS */
3047 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
3048 ewrt = _mm_mul_pd(r23,ewtabscale);
3049 ewitab = _mm_cvttpd_epi32(ewrt);
3050 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
3051 ewitab = _mm_slli_epi32(ewitab,2);
3052 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
3053 ewtabD = _mm_setzero_pd();
3054 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
3055 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
3056 ewtabFn = _mm_setzero_pd();
3057 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
3058 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
3059 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
3060 velec = _mm_mul_pd(qq23,_mm_sub_pd(rinv23,velec));
3061 felec = _mm_mul_pd(_mm_mul_pd(qq23,rinv23),_mm_sub_pd(rinvsq23,felec));
3063 d = _mm_sub_pd(r23,rswitch);
3064 d = _mm_max_pd(d,_mm_setzero_pd());
3065 d2 = _mm_mul_pd(d,d);
3066 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
3068 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
3070 /* Evaluate switch function */
3071 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3072 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv23,_mm_mul_pd(velec,dsw)) );
3073 cutoff_mask = _mm_cmplt_pd(rsq23,rcutoff2);
3077 fscal = _mm_and_pd(fscal,cutoff_mask);
3079 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
3081 /* Calculate temporary vectorial force */
3082 tx = _mm_mul_pd(fscal,dx23);
3083 ty = _mm_mul_pd(fscal,dy23);
3084 tz = _mm_mul_pd(fscal,dz23);
3086 /* Update vectorial force */
3087 fix2 = _mm_add_pd(fix2,tx);
3088 fiy2 = _mm_add_pd(fiy2,ty);
3089 fiz2 = _mm_add_pd(fiz2,tz);
3091 fjx3 = _mm_add_pd(fjx3,tx);
3092 fjy3 = _mm_add_pd(fjy3,ty);
3093 fjz3 = _mm_add_pd(fjz3,tz);
3097 /**************************
3098 * CALCULATE INTERACTIONS *
3099 **************************/
3101 if (gmx_mm_any_lt(rsq31,rcutoff2))
3104 r31 = _mm_mul_pd(rsq31,rinv31);
3106 /* EWALD ELECTROSTATICS */
3108 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
3109 ewrt = _mm_mul_pd(r31,ewtabscale);
3110 ewitab = _mm_cvttpd_epi32(ewrt);
3111 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
3112 ewitab = _mm_slli_epi32(ewitab,2);
3113 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
3114 ewtabD = _mm_setzero_pd();
3115 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
3116 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
3117 ewtabFn = _mm_setzero_pd();
3118 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
3119 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
3120 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
3121 velec = _mm_mul_pd(qq31,_mm_sub_pd(rinv31,velec));
3122 felec = _mm_mul_pd(_mm_mul_pd(qq31,rinv31),_mm_sub_pd(rinvsq31,felec));
3124 d = _mm_sub_pd(r31,rswitch);
3125 d = _mm_max_pd(d,_mm_setzero_pd());
3126 d2 = _mm_mul_pd(d,d);
3127 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
3129 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
3131 /* Evaluate switch function */
3132 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3133 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv31,_mm_mul_pd(velec,dsw)) );
3134 cutoff_mask = _mm_cmplt_pd(rsq31,rcutoff2);
3138 fscal = _mm_and_pd(fscal,cutoff_mask);
3140 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
3142 /* Calculate temporary vectorial force */
3143 tx = _mm_mul_pd(fscal,dx31);
3144 ty = _mm_mul_pd(fscal,dy31);
3145 tz = _mm_mul_pd(fscal,dz31);
3147 /* Update vectorial force */
3148 fix3 = _mm_add_pd(fix3,tx);
3149 fiy3 = _mm_add_pd(fiy3,ty);
3150 fiz3 = _mm_add_pd(fiz3,tz);
3152 fjx1 = _mm_add_pd(fjx1,tx);
3153 fjy1 = _mm_add_pd(fjy1,ty);
3154 fjz1 = _mm_add_pd(fjz1,tz);
3158 /**************************
3159 * CALCULATE INTERACTIONS *
3160 **************************/
3162 if (gmx_mm_any_lt(rsq32,rcutoff2))
3165 r32 = _mm_mul_pd(rsq32,rinv32);
3167 /* EWALD ELECTROSTATICS */
3169 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
3170 ewrt = _mm_mul_pd(r32,ewtabscale);
3171 ewitab = _mm_cvttpd_epi32(ewrt);
3172 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
3173 ewitab = _mm_slli_epi32(ewitab,2);
3174 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
3175 ewtabD = _mm_setzero_pd();
3176 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
3177 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
3178 ewtabFn = _mm_setzero_pd();
3179 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
3180 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
3181 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
3182 velec = _mm_mul_pd(qq32,_mm_sub_pd(rinv32,velec));
3183 felec = _mm_mul_pd(_mm_mul_pd(qq32,rinv32),_mm_sub_pd(rinvsq32,felec));
3185 d = _mm_sub_pd(r32,rswitch);
3186 d = _mm_max_pd(d,_mm_setzero_pd());
3187 d2 = _mm_mul_pd(d,d);
3188 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
3190 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
3192 /* Evaluate switch function */
3193 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3194 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv32,_mm_mul_pd(velec,dsw)) );
3195 cutoff_mask = _mm_cmplt_pd(rsq32,rcutoff2);
3199 fscal = _mm_and_pd(fscal,cutoff_mask);
3201 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
3203 /* Calculate temporary vectorial force */
3204 tx = _mm_mul_pd(fscal,dx32);
3205 ty = _mm_mul_pd(fscal,dy32);
3206 tz = _mm_mul_pd(fscal,dz32);
3208 /* Update vectorial force */
3209 fix3 = _mm_add_pd(fix3,tx);
3210 fiy3 = _mm_add_pd(fiy3,ty);
3211 fiz3 = _mm_add_pd(fiz3,tz);
3213 fjx2 = _mm_add_pd(fjx2,tx);
3214 fjy2 = _mm_add_pd(fjy2,ty);
3215 fjz2 = _mm_add_pd(fjz2,tz);
3219 /**************************
3220 * CALCULATE INTERACTIONS *
3221 **************************/
3223 if (gmx_mm_any_lt(rsq33,rcutoff2))
3226 r33 = _mm_mul_pd(rsq33,rinv33);
3228 /* EWALD ELECTROSTATICS */
3230 /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
3231 ewrt = _mm_mul_pd(r33,ewtabscale);
3232 ewitab = _mm_cvttpd_epi32(ewrt);
3233 eweps = _mm_sub_pd(ewrt,_mm_cvtepi32_pd(ewitab));
3234 ewitab = _mm_slli_epi32(ewitab,2);
3235 ewtabF = _mm_load_pd( ewtab + gmx_mm_extract_epi32(ewitab,0) );
3236 ewtabD = _mm_setzero_pd();
3237 GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
3238 ewtabV = _mm_load_sd( ewtab + gmx_mm_extract_epi32(ewitab,0) +2);
3239 ewtabFn = _mm_setzero_pd();
3240 GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
3241 felec = _mm_add_pd(ewtabF,_mm_mul_pd(eweps,ewtabD));
3242 velec = _mm_sub_pd(ewtabV,_mm_mul_pd(_mm_mul_pd(ewtabhalfspace,eweps),_mm_add_pd(ewtabF,felec)));
3243 velec = _mm_mul_pd(qq33,_mm_sub_pd(rinv33,velec));
3244 felec = _mm_mul_pd(_mm_mul_pd(qq33,rinv33),_mm_sub_pd(rinvsq33,felec));
3246 d = _mm_sub_pd(r33,rswitch);
3247 d = _mm_max_pd(d,_mm_setzero_pd());
3248 d2 = _mm_mul_pd(d,d);
3249 sw = _mm_add_pd(one,_mm_mul_pd(d2,_mm_mul_pd(d,_mm_add_pd(swV3,_mm_mul_pd(d,_mm_add_pd(swV4,_mm_mul_pd(d,swV5)))))));
3251 dsw = _mm_mul_pd(d2,_mm_add_pd(swF2,_mm_mul_pd(d,_mm_add_pd(swF3,_mm_mul_pd(d,swF4)))));
3253 /* Evaluate switch function */
3254 /* fscal'=f'/r=-(v*sw)'/r=-(v'*sw+v*dsw)/r=-v'*sw/r-v*dsw/r=fscal*sw-v*dsw/r */
3255 felec = _mm_sub_pd( _mm_mul_pd(felec,sw) , _mm_mul_pd(rinv33,_mm_mul_pd(velec,dsw)) );
3256 cutoff_mask = _mm_cmplt_pd(rsq33,rcutoff2);
3260 fscal = _mm_and_pd(fscal,cutoff_mask);
3262 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
3264 /* Calculate temporary vectorial force */
3265 tx = _mm_mul_pd(fscal,dx33);
3266 ty = _mm_mul_pd(fscal,dy33);
3267 tz = _mm_mul_pd(fscal,dz33);
3269 /* Update vectorial force */
3270 fix3 = _mm_add_pd(fix3,tx);
3271 fiy3 = _mm_add_pd(fiy3,ty);
3272 fiz3 = _mm_add_pd(fiz3,tz);
3274 fjx3 = _mm_add_pd(fjx3,tx);
3275 fjy3 = _mm_add_pd(fjy3,ty);
3276 fjz3 = _mm_add_pd(fjz3,tz);
3280 gmx_mm_decrement_4rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0,fjx1,fjy1,fjz1,fjx2,fjy2,fjz2,fjx3,fjy3,fjz3);
3282 /* Inner loop uses 617 flops */
3285 /* End of innermost loop */
3287 gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
3288 f+i_coord_offset,fshift+i_shift_offset);
3290 /* Increment number of inner iterations */
3291 inneriter += j_index_end - j_index_start;
3293 /* Outer loop uses 24 flops */
3296 /* Increment number of outer iterations */
3299 /* Update outer/inner flops */
3301 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4W4_F,outeriter*24 + inneriter*617);