2 * Note: this file was generated by the Gromacs sse4_1_double kernel generator.
4 * This source code is part of
8 * Copyright (c) 2001-2012, The GROMACS Development Team
10 * Gromacs is a library for molecular simulation and trajectory analysis,
11 * written by Erik Lindahl, David van der Spoel, Berk Hess, and others - for
12 * a full list of developers and information, check out http://www.gromacs.org
14 * This program is free software; you can redistribute it and/or modify it under
15 * the terms of the GNU Lesser General Public License as published by the Free
16 * Software Foundation; either version 2 of the License, or (at your option) any
19 * To help fund GROMACS development, we humbly ask that you cite
20 * the papers people have written on it - you can find them on the website.
28 #include "../nb_kernel.h"
29 #include "types/simple.h"
33 #include "gmx_math_x86_sse4_1_double.h"
34 #include "kernelutil_x86_sse4_1_double.h"
37 * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwCSTab_GeomW4P1_VF_sse4_1_double
38 * Electrostatics interaction: CubicSplineTable
39 * VdW interaction: CubicSplineTable
40 * Geometry: Water4-Particle
41 * Calculate force/pot: PotentialAndForce
44 nb_kernel_ElecCSTab_VdwCSTab_GeomW4P1_VF_sse4_1_double
45 (t_nblist * gmx_restrict nlist,
46 rvec * gmx_restrict xx,
47 rvec * gmx_restrict ff,
48 t_forcerec * gmx_restrict fr,
49 t_mdatoms * gmx_restrict mdatoms,
50 nb_kernel_data_t * gmx_restrict kernel_data,
51 t_nrnb * gmx_restrict nrnb)
53 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
54 * just 0 for non-waters.
55 * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
56 * jnr indices corresponding to data put in the four positions in the SIMD register.
58 int i_shift_offset,i_coord_offset,outeriter,inneriter;
59 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
61 int j_coord_offsetA,j_coord_offsetB;
62 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
64 real *shiftvec,*fshift,*x,*f;
65 __m128d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
67 __m128d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
69 __m128d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
71 __m128d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
73 __m128d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
74 int vdwjidx0A,vdwjidx0B;
75 __m128d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
76 __m128d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
77 __m128d dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
78 __m128d dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
79 __m128d dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
80 __m128d velec,felec,velecsum,facel,crf,krf,krf2;
83 __m128d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
86 __m128d one_sixth = _mm_set1_pd(1.0/6.0);
87 __m128d one_twelfth = _mm_set1_pd(1.0/12.0);
89 __m128i ifour = _mm_set1_epi32(4);
90 __m128d rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
92 __m128d dummy_mask,cutoff_mask;
93 __m128d signbit = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
94 __m128d one = _mm_set1_pd(1.0);
95 __m128d two = _mm_set1_pd(2.0);
101 jindex = nlist->jindex;
103 shiftidx = nlist->shift;
105 shiftvec = fr->shift_vec[0];
106 fshift = fr->fshift[0];
107 facel = _mm_set1_pd(fr->epsfac);
108 charge = mdatoms->chargeA;
109 nvdwtype = fr->ntype;
111 vdwtype = mdatoms->typeA;
113 vftab = kernel_data->table_elec_vdw->data;
114 vftabscale = _mm_set1_pd(kernel_data->table_elec_vdw->scale);
116 /* Setup water-specific parameters */
117 inr = nlist->iinr[0];
118 iq1 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
119 iq2 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
120 iq3 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
121 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
123 /* Avoid stupid compiler warnings */
131 /* Start outer loop over neighborlists */
132 for(iidx=0; iidx<nri; iidx++)
134 /* Load shift vector for this list */
135 i_shift_offset = DIM*shiftidx[iidx];
137 /* Load limits for loop over neighbors */
138 j_index_start = jindex[iidx];
139 j_index_end = jindex[iidx+1];
141 /* Get outer coordinate index */
143 i_coord_offset = DIM*inr;
145 /* Load i particle coords and add shift vector */
146 gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
147 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
149 fix0 = _mm_setzero_pd();
150 fiy0 = _mm_setzero_pd();
151 fiz0 = _mm_setzero_pd();
152 fix1 = _mm_setzero_pd();
153 fiy1 = _mm_setzero_pd();
154 fiz1 = _mm_setzero_pd();
155 fix2 = _mm_setzero_pd();
156 fiy2 = _mm_setzero_pd();
157 fiz2 = _mm_setzero_pd();
158 fix3 = _mm_setzero_pd();
159 fiy3 = _mm_setzero_pd();
160 fiz3 = _mm_setzero_pd();
162 /* Reset potential sums */
163 velecsum = _mm_setzero_pd();
164 vvdwsum = _mm_setzero_pd();
166 /* Start inner kernel loop */
167 for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
170 /* Get j neighbor index, and coordinate index */
173 j_coord_offsetA = DIM*jnrA;
174 j_coord_offsetB = DIM*jnrB;
176 /* load j atom coordinates */
177 gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
180 /* Calculate displacement vector */
181 dx00 = _mm_sub_pd(ix0,jx0);
182 dy00 = _mm_sub_pd(iy0,jy0);
183 dz00 = _mm_sub_pd(iz0,jz0);
184 dx10 = _mm_sub_pd(ix1,jx0);
185 dy10 = _mm_sub_pd(iy1,jy0);
186 dz10 = _mm_sub_pd(iz1,jz0);
187 dx20 = _mm_sub_pd(ix2,jx0);
188 dy20 = _mm_sub_pd(iy2,jy0);
189 dz20 = _mm_sub_pd(iz2,jz0);
190 dx30 = _mm_sub_pd(ix3,jx0);
191 dy30 = _mm_sub_pd(iy3,jy0);
192 dz30 = _mm_sub_pd(iz3,jz0);
194 /* Calculate squared distance and things based on it */
195 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
196 rsq10 = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
197 rsq20 = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
198 rsq30 = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
200 rinv00 = gmx_mm_invsqrt_pd(rsq00);
201 rinv10 = gmx_mm_invsqrt_pd(rsq10);
202 rinv20 = gmx_mm_invsqrt_pd(rsq20);
203 rinv30 = gmx_mm_invsqrt_pd(rsq30);
205 /* Load parameters for j particles */
206 jq0 = gmx_mm_load_2real_swizzle_pd(charge+jnrA+0,charge+jnrB+0);
207 vdwjidx0A = 2*vdwtype[jnrA+0];
208 vdwjidx0B = 2*vdwtype[jnrB+0];
210 fjx0 = _mm_setzero_pd();
211 fjy0 = _mm_setzero_pd();
212 fjz0 = _mm_setzero_pd();
214 /**************************
215 * CALCULATE INTERACTIONS *
216 **************************/
218 r00 = _mm_mul_pd(rsq00,rinv00);
220 /* Compute parameters for interactions between i and j atoms */
221 gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,
222 vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
224 /* Calculate table index by multiplying r with table scale and truncate to integer */
225 rt = _mm_mul_pd(r00,vftabscale);
226 vfitab = _mm_cvttpd_epi32(rt);
227 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
228 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
230 /* CUBIC SPLINE TABLE DISPERSION */
231 vfitab = _mm_add_epi32(vfitab,ifour);
232 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
233 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
234 GMX_MM_TRANSPOSE2_PD(Y,F);
235 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
236 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
237 GMX_MM_TRANSPOSE2_PD(G,H);
238 Heps = _mm_mul_pd(vfeps,H);
239 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
240 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
241 vvdw6 = _mm_mul_pd(c6_00,VV);
242 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
243 fvdw6 = _mm_mul_pd(c6_00,FF);
245 /* CUBIC SPLINE TABLE REPULSION */
246 vfitab = _mm_add_epi32(vfitab,ifour);
247 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
248 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
249 GMX_MM_TRANSPOSE2_PD(Y,F);
250 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
251 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
252 GMX_MM_TRANSPOSE2_PD(G,H);
253 Heps = _mm_mul_pd(vfeps,H);
254 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
255 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
256 vvdw12 = _mm_mul_pd(c12_00,VV);
257 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
258 fvdw12 = _mm_mul_pd(c12_00,FF);
259 vvdw = _mm_add_pd(vvdw12,vvdw6);
260 fvdw = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
262 /* Update potential sum for this i atom from the interaction with this j atom. */
263 vvdwsum = _mm_add_pd(vvdwsum,vvdw);
267 /* Calculate temporary vectorial force */
268 tx = _mm_mul_pd(fscal,dx00);
269 ty = _mm_mul_pd(fscal,dy00);
270 tz = _mm_mul_pd(fscal,dz00);
272 /* Update vectorial force */
273 fix0 = _mm_add_pd(fix0,tx);
274 fiy0 = _mm_add_pd(fiy0,ty);
275 fiz0 = _mm_add_pd(fiz0,tz);
277 fjx0 = _mm_add_pd(fjx0,tx);
278 fjy0 = _mm_add_pd(fjy0,ty);
279 fjz0 = _mm_add_pd(fjz0,tz);
281 /**************************
282 * CALCULATE INTERACTIONS *
283 **************************/
285 r10 = _mm_mul_pd(rsq10,rinv10);
287 /* Compute parameters for interactions between i and j atoms */
288 qq10 = _mm_mul_pd(iq1,jq0);
290 /* Calculate table index by multiplying r with table scale and truncate to integer */
291 rt = _mm_mul_pd(r10,vftabscale);
292 vfitab = _mm_cvttpd_epi32(rt);
293 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
294 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
296 /* CUBIC SPLINE TABLE ELECTROSTATICS */
297 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
298 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
299 GMX_MM_TRANSPOSE2_PD(Y,F);
300 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
301 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
302 GMX_MM_TRANSPOSE2_PD(G,H);
303 Heps = _mm_mul_pd(vfeps,H);
304 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
305 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
306 velec = _mm_mul_pd(qq10,VV);
307 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
308 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
310 /* Update potential sum for this i atom from the interaction with this j atom. */
311 velecsum = _mm_add_pd(velecsum,velec);
315 /* Calculate temporary vectorial force */
316 tx = _mm_mul_pd(fscal,dx10);
317 ty = _mm_mul_pd(fscal,dy10);
318 tz = _mm_mul_pd(fscal,dz10);
320 /* Update vectorial force */
321 fix1 = _mm_add_pd(fix1,tx);
322 fiy1 = _mm_add_pd(fiy1,ty);
323 fiz1 = _mm_add_pd(fiz1,tz);
325 fjx0 = _mm_add_pd(fjx0,tx);
326 fjy0 = _mm_add_pd(fjy0,ty);
327 fjz0 = _mm_add_pd(fjz0,tz);
329 /**************************
330 * CALCULATE INTERACTIONS *
331 **************************/
333 r20 = _mm_mul_pd(rsq20,rinv20);
335 /* Compute parameters for interactions between i and j atoms */
336 qq20 = _mm_mul_pd(iq2,jq0);
338 /* Calculate table index by multiplying r with table scale and truncate to integer */
339 rt = _mm_mul_pd(r20,vftabscale);
340 vfitab = _mm_cvttpd_epi32(rt);
341 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
342 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
344 /* CUBIC SPLINE TABLE ELECTROSTATICS */
345 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
346 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
347 GMX_MM_TRANSPOSE2_PD(Y,F);
348 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
349 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
350 GMX_MM_TRANSPOSE2_PD(G,H);
351 Heps = _mm_mul_pd(vfeps,H);
352 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
353 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
354 velec = _mm_mul_pd(qq20,VV);
355 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
356 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
358 /* Update potential sum for this i atom from the interaction with this j atom. */
359 velecsum = _mm_add_pd(velecsum,velec);
363 /* Calculate temporary vectorial force */
364 tx = _mm_mul_pd(fscal,dx20);
365 ty = _mm_mul_pd(fscal,dy20);
366 tz = _mm_mul_pd(fscal,dz20);
368 /* Update vectorial force */
369 fix2 = _mm_add_pd(fix2,tx);
370 fiy2 = _mm_add_pd(fiy2,ty);
371 fiz2 = _mm_add_pd(fiz2,tz);
373 fjx0 = _mm_add_pd(fjx0,tx);
374 fjy0 = _mm_add_pd(fjy0,ty);
375 fjz0 = _mm_add_pd(fjz0,tz);
377 /**************************
378 * CALCULATE INTERACTIONS *
379 **************************/
381 r30 = _mm_mul_pd(rsq30,rinv30);
383 /* Compute parameters for interactions between i and j atoms */
384 qq30 = _mm_mul_pd(iq3,jq0);
386 /* Calculate table index by multiplying r with table scale and truncate to integer */
387 rt = _mm_mul_pd(r30,vftabscale);
388 vfitab = _mm_cvttpd_epi32(rt);
389 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
390 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
392 /* CUBIC SPLINE TABLE ELECTROSTATICS */
393 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
394 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
395 GMX_MM_TRANSPOSE2_PD(Y,F);
396 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
397 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
398 GMX_MM_TRANSPOSE2_PD(G,H);
399 Heps = _mm_mul_pd(vfeps,H);
400 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
401 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
402 velec = _mm_mul_pd(qq30,VV);
403 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
404 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq30,FF),_mm_mul_pd(vftabscale,rinv30)));
406 /* Update potential sum for this i atom from the interaction with this j atom. */
407 velecsum = _mm_add_pd(velecsum,velec);
411 /* Calculate temporary vectorial force */
412 tx = _mm_mul_pd(fscal,dx30);
413 ty = _mm_mul_pd(fscal,dy30);
414 tz = _mm_mul_pd(fscal,dz30);
416 /* Update vectorial force */
417 fix3 = _mm_add_pd(fix3,tx);
418 fiy3 = _mm_add_pd(fiy3,ty);
419 fiz3 = _mm_add_pd(fiz3,tz);
421 fjx0 = _mm_add_pd(fjx0,tx);
422 fjy0 = _mm_add_pd(fjy0,ty);
423 fjz0 = _mm_add_pd(fjz0,tz);
425 gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
427 /* Inner loop uses 188 flops */
434 j_coord_offsetA = DIM*jnrA;
436 /* load j atom coordinates */
437 gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
440 /* Calculate displacement vector */
441 dx00 = _mm_sub_pd(ix0,jx0);
442 dy00 = _mm_sub_pd(iy0,jy0);
443 dz00 = _mm_sub_pd(iz0,jz0);
444 dx10 = _mm_sub_pd(ix1,jx0);
445 dy10 = _mm_sub_pd(iy1,jy0);
446 dz10 = _mm_sub_pd(iz1,jz0);
447 dx20 = _mm_sub_pd(ix2,jx0);
448 dy20 = _mm_sub_pd(iy2,jy0);
449 dz20 = _mm_sub_pd(iz2,jz0);
450 dx30 = _mm_sub_pd(ix3,jx0);
451 dy30 = _mm_sub_pd(iy3,jy0);
452 dz30 = _mm_sub_pd(iz3,jz0);
454 /* Calculate squared distance and things based on it */
455 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
456 rsq10 = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
457 rsq20 = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
458 rsq30 = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
460 rinv00 = gmx_mm_invsqrt_pd(rsq00);
461 rinv10 = gmx_mm_invsqrt_pd(rsq10);
462 rinv20 = gmx_mm_invsqrt_pd(rsq20);
463 rinv30 = gmx_mm_invsqrt_pd(rsq30);
465 /* Load parameters for j particles */
466 jq0 = _mm_load_sd(charge+jnrA+0);
467 vdwjidx0A = 2*vdwtype[jnrA+0];
469 fjx0 = _mm_setzero_pd();
470 fjy0 = _mm_setzero_pd();
471 fjz0 = _mm_setzero_pd();
473 /**************************
474 * CALCULATE INTERACTIONS *
475 **************************/
477 r00 = _mm_mul_pd(rsq00,rinv00);
479 /* Compute parameters for interactions between i and j atoms */
480 gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
482 /* Calculate table index by multiplying r with table scale and truncate to integer */
483 rt = _mm_mul_pd(r00,vftabscale);
484 vfitab = _mm_cvttpd_epi32(rt);
485 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
486 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
488 /* CUBIC SPLINE TABLE DISPERSION */
489 vfitab = _mm_add_epi32(vfitab,ifour);
490 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
491 F = _mm_setzero_pd();
492 GMX_MM_TRANSPOSE2_PD(Y,F);
493 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
494 H = _mm_setzero_pd();
495 GMX_MM_TRANSPOSE2_PD(G,H);
496 Heps = _mm_mul_pd(vfeps,H);
497 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
498 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
499 vvdw6 = _mm_mul_pd(c6_00,VV);
500 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
501 fvdw6 = _mm_mul_pd(c6_00,FF);
503 /* CUBIC SPLINE TABLE REPULSION */
504 vfitab = _mm_add_epi32(vfitab,ifour);
505 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
506 F = _mm_setzero_pd();
507 GMX_MM_TRANSPOSE2_PD(Y,F);
508 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
509 H = _mm_setzero_pd();
510 GMX_MM_TRANSPOSE2_PD(G,H);
511 Heps = _mm_mul_pd(vfeps,H);
512 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
513 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
514 vvdw12 = _mm_mul_pd(c12_00,VV);
515 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
516 fvdw12 = _mm_mul_pd(c12_00,FF);
517 vvdw = _mm_add_pd(vvdw12,vvdw6);
518 fvdw = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
520 /* Update potential sum for this i atom from the interaction with this j atom. */
521 vvdw = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
522 vvdwsum = _mm_add_pd(vvdwsum,vvdw);
526 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
528 /* Calculate temporary vectorial force */
529 tx = _mm_mul_pd(fscal,dx00);
530 ty = _mm_mul_pd(fscal,dy00);
531 tz = _mm_mul_pd(fscal,dz00);
533 /* Update vectorial force */
534 fix0 = _mm_add_pd(fix0,tx);
535 fiy0 = _mm_add_pd(fiy0,ty);
536 fiz0 = _mm_add_pd(fiz0,tz);
538 fjx0 = _mm_add_pd(fjx0,tx);
539 fjy0 = _mm_add_pd(fjy0,ty);
540 fjz0 = _mm_add_pd(fjz0,tz);
542 /**************************
543 * CALCULATE INTERACTIONS *
544 **************************/
546 r10 = _mm_mul_pd(rsq10,rinv10);
548 /* Compute parameters for interactions between i and j atoms */
549 qq10 = _mm_mul_pd(iq1,jq0);
551 /* Calculate table index by multiplying r with table scale and truncate to integer */
552 rt = _mm_mul_pd(r10,vftabscale);
553 vfitab = _mm_cvttpd_epi32(rt);
554 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
555 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
557 /* CUBIC SPLINE TABLE ELECTROSTATICS */
558 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
559 F = _mm_setzero_pd();
560 GMX_MM_TRANSPOSE2_PD(Y,F);
561 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
562 H = _mm_setzero_pd();
563 GMX_MM_TRANSPOSE2_PD(G,H);
564 Heps = _mm_mul_pd(vfeps,H);
565 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
566 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
567 velec = _mm_mul_pd(qq10,VV);
568 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
569 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
571 /* Update potential sum for this i atom from the interaction with this j atom. */
572 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
573 velecsum = _mm_add_pd(velecsum,velec);
577 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
579 /* Calculate temporary vectorial force */
580 tx = _mm_mul_pd(fscal,dx10);
581 ty = _mm_mul_pd(fscal,dy10);
582 tz = _mm_mul_pd(fscal,dz10);
584 /* Update vectorial force */
585 fix1 = _mm_add_pd(fix1,tx);
586 fiy1 = _mm_add_pd(fiy1,ty);
587 fiz1 = _mm_add_pd(fiz1,tz);
589 fjx0 = _mm_add_pd(fjx0,tx);
590 fjy0 = _mm_add_pd(fjy0,ty);
591 fjz0 = _mm_add_pd(fjz0,tz);
593 /**************************
594 * CALCULATE INTERACTIONS *
595 **************************/
597 r20 = _mm_mul_pd(rsq20,rinv20);
599 /* Compute parameters for interactions between i and j atoms */
600 qq20 = _mm_mul_pd(iq2,jq0);
602 /* Calculate table index by multiplying r with table scale and truncate to integer */
603 rt = _mm_mul_pd(r20,vftabscale);
604 vfitab = _mm_cvttpd_epi32(rt);
605 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
606 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
608 /* CUBIC SPLINE TABLE ELECTROSTATICS */
609 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
610 F = _mm_setzero_pd();
611 GMX_MM_TRANSPOSE2_PD(Y,F);
612 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
613 H = _mm_setzero_pd();
614 GMX_MM_TRANSPOSE2_PD(G,H);
615 Heps = _mm_mul_pd(vfeps,H);
616 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
617 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
618 velec = _mm_mul_pd(qq20,VV);
619 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
620 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
622 /* Update potential sum for this i atom from the interaction with this j atom. */
623 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
624 velecsum = _mm_add_pd(velecsum,velec);
628 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
630 /* Calculate temporary vectorial force */
631 tx = _mm_mul_pd(fscal,dx20);
632 ty = _mm_mul_pd(fscal,dy20);
633 tz = _mm_mul_pd(fscal,dz20);
635 /* Update vectorial force */
636 fix2 = _mm_add_pd(fix2,tx);
637 fiy2 = _mm_add_pd(fiy2,ty);
638 fiz2 = _mm_add_pd(fiz2,tz);
640 fjx0 = _mm_add_pd(fjx0,tx);
641 fjy0 = _mm_add_pd(fjy0,ty);
642 fjz0 = _mm_add_pd(fjz0,tz);
644 /**************************
645 * CALCULATE INTERACTIONS *
646 **************************/
648 r30 = _mm_mul_pd(rsq30,rinv30);
650 /* Compute parameters for interactions between i and j atoms */
651 qq30 = _mm_mul_pd(iq3,jq0);
653 /* Calculate table index by multiplying r with table scale and truncate to integer */
654 rt = _mm_mul_pd(r30,vftabscale);
655 vfitab = _mm_cvttpd_epi32(rt);
656 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
657 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
659 /* CUBIC SPLINE TABLE ELECTROSTATICS */
660 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
661 F = _mm_setzero_pd();
662 GMX_MM_TRANSPOSE2_PD(Y,F);
663 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
664 H = _mm_setzero_pd();
665 GMX_MM_TRANSPOSE2_PD(G,H);
666 Heps = _mm_mul_pd(vfeps,H);
667 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
668 VV = _mm_add_pd(Y,_mm_mul_pd(vfeps,Fp));
669 velec = _mm_mul_pd(qq30,VV);
670 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
671 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq30,FF),_mm_mul_pd(vftabscale,rinv30)));
673 /* Update potential sum for this i atom from the interaction with this j atom. */
674 velec = _mm_unpacklo_pd(velec,_mm_setzero_pd());
675 velecsum = _mm_add_pd(velecsum,velec);
679 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
681 /* Calculate temporary vectorial force */
682 tx = _mm_mul_pd(fscal,dx30);
683 ty = _mm_mul_pd(fscal,dy30);
684 tz = _mm_mul_pd(fscal,dz30);
686 /* Update vectorial force */
687 fix3 = _mm_add_pd(fix3,tx);
688 fiy3 = _mm_add_pd(fiy3,ty);
689 fiz3 = _mm_add_pd(fiz3,tz);
691 fjx0 = _mm_add_pd(fjx0,tx);
692 fjy0 = _mm_add_pd(fjy0,ty);
693 fjz0 = _mm_add_pd(fjz0,tz);
695 gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0);
697 /* Inner loop uses 188 flops */
700 /* End of innermost loop */
702 gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
703 f+i_coord_offset,fshift+i_shift_offset);
706 /* Update potential energies */
707 gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
708 gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
710 /* Increment number of inner iterations */
711 inneriter += j_index_end - j_index_start;
713 /* Outer loop uses 26 flops */
716 /* Increment number of outer iterations */
719 /* Update outer/inner flops */
721 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*26 + inneriter*188);
724 * Gromacs nonbonded kernel: nb_kernel_ElecCSTab_VdwCSTab_GeomW4P1_F_sse4_1_double
725 * Electrostatics interaction: CubicSplineTable
726 * VdW interaction: CubicSplineTable
727 * Geometry: Water4-Particle
728 * Calculate force/pot: Force
731 nb_kernel_ElecCSTab_VdwCSTab_GeomW4P1_F_sse4_1_double
732 (t_nblist * gmx_restrict nlist,
733 rvec * gmx_restrict xx,
734 rvec * gmx_restrict ff,
735 t_forcerec * gmx_restrict fr,
736 t_mdatoms * gmx_restrict mdatoms,
737 nb_kernel_data_t * gmx_restrict kernel_data,
738 t_nrnb * gmx_restrict nrnb)
740 /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
741 * just 0 for non-waters.
742 * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
743 * jnr indices corresponding to data put in the four positions in the SIMD register.
745 int i_shift_offset,i_coord_offset,outeriter,inneriter;
746 int j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
748 int j_coord_offsetA,j_coord_offsetB;
749 int *iinr,*jindex,*jjnr,*shiftidx,*gid;
751 real *shiftvec,*fshift,*x,*f;
752 __m128d tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
754 __m128d ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
756 __m128d ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
758 __m128d ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
760 __m128d ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
761 int vdwjidx0A,vdwjidx0B;
762 __m128d jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
763 __m128d dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
764 __m128d dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
765 __m128d dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
766 __m128d dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
767 __m128d velec,felec,velecsum,facel,crf,krf,krf2;
770 __m128d rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
773 __m128d one_sixth = _mm_set1_pd(1.0/6.0);
774 __m128d one_twelfth = _mm_set1_pd(1.0/12.0);
776 __m128i ifour = _mm_set1_epi32(4);
777 __m128d rt,vfeps,vftabscale,Y,F,G,H,Heps,Fp,VV,FF;
779 __m128d dummy_mask,cutoff_mask;
780 __m128d signbit = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
781 __m128d one = _mm_set1_pd(1.0);
782 __m128d two = _mm_set1_pd(2.0);
788 jindex = nlist->jindex;
790 shiftidx = nlist->shift;
792 shiftvec = fr->shift_vec[0];
793 fshift = fr->fshift[0];
794 facel = _mm_set1_pd(fr->epsfac);
795 charge = mdatoms->chargeA;
796 nvdwtype = fr->ntype;
798 vdwtype = mdatoms->typeA;
800 vftab = kernel_data->table_elec_vdw->data;
801 vftabscale = _mm_set1_pd(kernel_data->table_elec_vdw->scale);
803 /* Setup water-specific parameters */
804 inr = nlist->iinr[0];
805 iq1 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
806 iq2 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
807 iq3 = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
808 vdwioffset0 = 2*nvdwtype*vdwtype[inr+0];
810 /* Avoid stupid compiler warnings */
818 /* Start outer loop over neighborlists */
819 for(iidx=0; iidx<nri; iidx++)
821 /* Load shift vector for this list */
822 i_shift_offset = DIM*shiftidx[iidx];
824 /* Load limits for loop over neighbors */
825 j_index_start = jindex[iidx];
826 j_index_end = jindex[iidx+1];
828 /* Get outer coordinate index */
830 i_coord_offset = DIM*inr;
832 /* Load i particle coords and add shift vector */
833 gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
834 &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
836 fix0 = _mm_setzero_pd();
837 fiy0 = _mm_setzero_pd();
838 fiz0 = _mm_setzero_pd();
839 fix1 = _mm_setzero_pd();
840 fiy1 = _mm_setzero_pd();
841 fiz1 = _mm_setzero_pd();
842 fix2 = _mm_setzero_pd();
843 fiy2 = _mm_setzero_pd();
844 fiz2 = _mm_setzero_pd();
845 fix3 = _mm_setzero_pd();
846 fiy3 = _mm_setzero_pd();
847 fiz3 = _mm_setzero_pd();
849 /* Start inner kernel loop */
850 for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
853 /* Get j neighbor index, and coordinate index */
856 j_coord_offsetA = DIM*jnrA;
857 j_coord_offsetB = DIM*jnrB;
859 /* load j atom coordinates */
860 gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
863 /* Calculate displacement vector */
864 dx00 = _mm_sub_pd(ix0,jx0);
865 dy00 = _mm_sub_pd(iy0,jy0);
866 dz00 = _mm_sub_pd(iz0,jz0);
867 dx10 = _mm_sub_pd(ix1,jx0);
868 dy10 = _mm_sub_pd(iy1,jy0);
869 dz10 = _mm_sub_pd(iz1,jz0);
870 dx20 = _mm_sub_pd(ix2,jx0);
871 dy20 = _mm_sub_pd(iy2,jy0);
872 dz20 = _mm_sub_pd(iz2,jz0);
873 dx30 = _mm_sub_pd(ix3,jx0);
874 dy30 = _mm_sub_pd(iy3,jy0);
875 dz30 = _mm_sub_pd(iz3,jz0);
877 /* Calculate squared distance and things based on it */
878 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
879 rsq10 = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
880 rsq20 = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
881 rsq30 = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
883 rinv00 = gmx_mm_invsqrt_pd(rsq00);
884 rinv10 = gmx_mm_invsqrt_pd(rsq10);
885 rinv20 = gmx_mm_invsqrt_pd(rsq20);
886 rinv30 = gmx_mm_invsqrt_pd(rsq30);
888 /* Load parameters for j particles */
889 jq0 = gmx_mm_load_2real_swizzle_pd(charge+jnrA+0,charge+jnrB+0);
890 vdwjidx0A = 2*vdwtype[jnrA+0];
891 vdwjidx0B = 2*vdwtype[jnrB+0];
893 fjx0 = _mm_setzero_pd();
894 fjy0 = _mm_setzero_pd();
895 fjz0 = _mm_setzero_pd();
897 /**************************
898 * CALCULATE INTERACTIONS *
899 **************************/
901 r00 = _mm_mul_pd(rsq00,rinv00);
903 /* Compute parameters for interactions between i and j atoms */
904 gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,
905 vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
907 /* Calculate table index by multiplying r with table scale and truncate to integer */
908 rt = _mm_mul_pd(r00,vftabscale);
909 vfitab = _mm_cvttpd_epi32(rt);
910 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
911 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
913 /* CUBIC SPLINE TABLE DISPERSION */
914 vfitab = _mm_add_epi32(vfitab,ifour);
915 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
916 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
917 GMX_MM_TRANSPOSE2_PD(Y,F);
918 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
919 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
920 GMX_MM_TRANSPOSE2_PD(G,H);
921 Heps = _mm_mul_pd(vfeps,H);
922 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
923 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
924 fvdw6 = _mm_mul_pd(c6_00,FF);
926 /* CUBIC SPLINE TABLE REPULSION */
927 vfitab = _mm_add_epi32(vfitab,ifour);
928 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
929 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
930 GMX_MM_TRANSPOSE2_PD(Y,F);
931 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
932 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
933 GMX_MM_TRANSPOSE2_PD(G,H);
934 Heps = _mm_mul_pd(vfeps,H);
935 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
936 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
937 fvdw12 = _mm_mul_pd(c12_00,FF);
938 fvdw = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
942 /* Calculate temporary vectorial force */
943 tx = _mm_mul_pd(fscal,dx00);
944 ty = _mm_mul_pd(fscal,dy00);
945 tz = _mm_mul_pd(fscal,dz00);
947 /* Update vectorial force */
948 fix0 = _mm_add_pd(fix0,tx);
949 fiy0 = _mm_add_pd(fiy0,ty);
950 fiz0 = _mm_add_pd(fiz0,tz);
952 fjx0 = _mm_add_pd(fjx0,tx);
953 fjy0 = _mm_add_pd(fjy0,ty);
954 fjz0 = _mm_add_pd(fjz0,tz);
956 /**************************
957 * CALCULATE INTERACTIONS *
958 **************************/
960 r10 = _mm_mul_pd(rsq10,rinv10);
962 /* Compute parameters for interactions between i and j atoms */
963 qq10 = _mm_mul_pd(iq1,jq0);
965 /* Calculate table index by multiplying r with table scale and truncate to integer */
966 rt = _mm_mul_pd(r10,vftabscale);
967 vfitab = _mm_cvttpd_epi32(rt);
968 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
969 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
971 /* CUBIC SPLINE TABLE ELECTROSTATICS */
972 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
973 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
974 GMX_MM_TRANSPOSE2_PD(Y,F);
975 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
976 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
977 GMX_MM_TRANSPOSE2_PD(G,H);
978 Heps = _mm_mul_pd(vfeps,H);
979 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
980 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
981 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
985 /* Calculate temporary vectorial force */
986 tx = _mm_mul_pd(fscal,dx10);
987 ty = _mm_mul_pd(fscal,dy10);
988 tz = _mm_mul_pd(fscal,dz10);
990 /* Update vectorial force */
991 fix1 = _mm_add_pd(fix1,tx);
992 fiy1 = _mm_add_pd(fiy1,ty);
993 fiz1 = _mm_add_pd(fiz1,tz);
995 fjx0 = _mm_add_pd(fjx0,tx);
996 fjy0 = _mm_add_pd(fjy0,ty);
997 fjz0 = _mm_add_pd(fjz0,tz);
999 /**************************
1000 * CALCULATE INTERACTIONS *
1001 **************************/
1003 r20 = _mm_mul_pd(rsq20,rinv20);
1005 /* Compute parameters for interactions between i and j atoms */
1006 qq20 = _mm_mul_pd(iq2,jq0);
1008 /* Calculate table index by multiplying r with table scale and truncate to integer */
1009 rt = _mm_mul_pd(r20,vftabscale);
1010 vfitab = _mm_cvttpd_epi32(rt);
1011 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1012 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1014 /* CUBIC SPLINE TABLE ELECTROSTATICS */
1015 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1016 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
1017 GMX_MM_TRANSPOSE2_PD(Y,F);
1018 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1019 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
1020 GMX_MM_TRANSPOSE2_PD(G,H);
1021 Heps = _mm_mul_pd(vfeps,H);
1022 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1023 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1024 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
1028 /* Calculate temporary vectorial force */
1029 tx = _mm_mul_pd(fscal,dx20);
1030 ty = _mm_mul_pd(fscal,dy20);
1031 tz = _mm_mul_pd(fscal,dz20);
1033 /* Update vectorial force */
1034 fix2 = _mm_add_pd(fix2,tx);
1035 fiy2 = _mm_add_pd(fiy2,ty);
1036 fiz2 = _mm_add_pd(fiz2,tz);
1038 fjx0 = _mm_add_pd(fjx0,tx);
1039 fjy0 = _mm_add_pd(fjy0,ty);
1040 fjz0 = _mm_add_pd(fjz0,tz);
1042 /**************************
1043 * CALCULATE INTERACTIONS *
1044 **************************/
1046 r30 = _mm_mul_pd(rsq30,rinv30);
1048 /* Compute parameters for interactions between i and j atoms */
1049 qq30 = _mm_mul_pd(iq3,jq0);
1051 /* Calculate table index by multiplying r with table scale and truncate to integer */
1052 rt = _mm_mul_pd(r30,vftabscale);
1053 vfitab = _mm_cvttpd_epi32(rt);
1054 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1055 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1057 /* CUBIC SPLINE TABLE ELECTROSTATICS */
1058 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1059 F = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) );
1060 GMX_MM_TRANSPOSE2_PD(Y,F);
1061 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1062 H = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,1) +2);
1063 GMX_MM_TRANSPOSE2_PD(G,H);
1064 Heps = _mm_mul_pd(vfeps,H);
1065 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1066 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1067 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq30,FF),_mm_mul_pd(vftabscale,rinv30)));
1071 /* Calculate temporary vectorial force */
1072 tx = _mm_mul_pd(fscal,dx30);
1073 ty = _mm_mul_pd(fscal,dy30);
1074 tz = _mm_mul_pd(fscal,dz30);
1076 /* Update vectorial force */
1077 fix3 = _mm_add_pd(fix3,tx);
1078 fiy3 = _mm_add_pd(fiy3,ty);
1079 fiz3 = _mm_add_pd(fiz3,tz);
1081 fjx0 = _mm_add_pd(fjx0,tx);
1082 fjy0 = _mm_add_pd(fjy0,ty);
1083 fjz0 = _mm_add_pd(fjz0,tz);
1085 gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
1087 /* Inner loop uses 168 flops */
1090 if(jidx<j_index_end)
1094 j_coord_offsetA = DIM*jnrA;
1096 /* load j atom coordinates */
1097 gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
1100 /* Calculate displacement vector */
1101 dx00 = _mm_sub_pd(ix0,jx0);
1102 dy00 = _mm_sub_pd(iy0,jy0);
1103 dz00 = _mm_sub_pd(iz0,jz0);
1104 dx10 = _mm_sub_pd(ix1,jx0);
1105 dy10 = _mm_sub_pd(iy1,jy0);
1106 dz10 = _mm_sub_pd(iz1,jz0);
1107 dx20 = _mm_sub_pd(ix2,jx0);
1108 dy20 = _mm_sub_pd(iy2,jy0);
1109 dz20 = _mm_sub_pd(iz2,jz0);
1110 dx30 = _mm_sub_pd(ix3,jx0);
1111 dy30 = _mm_sub_pd(iy3,jy0);
1112 dz30 = _mm_sub_pd(iz3,jz0);
1114 /* Calculate squared distance and things based on it */
1115 rsq00 = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1116 rsq10 = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
1117 rsq20 = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
1118 rsq30 = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
1120 rinv00 = gmx_mm_invsqrt_pd(rsq00);
1121 rinv10 = gmx_mm_invsqrt_pd(rsq10);
1122 rinv20 = gmx_mm_invsqrt_pd(rsq20);
1123 rinv30 = gmx_mm_invsqrt_pd(rsq30);
1125 /* Load parameters for j particles */
1126 jq0 = _mm_load_sd(charge+jnrA+0);
1127 vdwjidx0A = 2*vdwtype[jnrA+0];
1129 fjx0 = _mm_setzero_pd();
1130 fjy0 = _mm_setzero_pd();
1131 fjz0 = _mm_setzero_pd();
1133 /**************************
1134 * CALCULATE INTERACTIONS *
1135 **************************/
1137 r00 = _mm_mul_pd(rsq00,rinv00);
1139 /* Compute parameters for interactions between i and j atoms */
1140 gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
1142 /* Calculate table index by multiplying r with table scale and truncate to integer */
1143 rt = _mm_mul_pd(r00,vftabscale);
1144 vfitab = _mm_cvttpd_epi32(rt);
1145 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1146 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1148 /* CUBIC SPLINE TABLE DISPERSION */
1149 vfitab = _mm_add_epi32(vfitab,ifour);
1150 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1151 F = _mm_setzero_pd();
1152 GMX_MM_TRANSPOSE2_PD(Y,F);
1153 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1154 H = _mm_setzero_pd();
1155 GMX_MM_TRANSPOSE2_PD(G,H);
1156 Heps = _mm_mul_pd(vfeps,H);
1157 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1158 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1159 fvdw6 = _mm_mul_pd(c6_00,FF);
1161 /* CUBIC SPLINE TABLE REPULSION */
1162 vfitab = _mm_add_epi32(vfitab,ifour);
1163 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1164 F = _mm_setzero_pd();
1165 GMX_MM_TRANSPOSE2_PD(Y,F);
1166 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1167 H = _mm_setzero_pd();
1168 GMX_MM_TRANSPOSE2_PD(G,H);
1169 Heps = _mm_mul_pd(vfeps,H);
1170 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1171 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1172 fvdw12 = _mm_mul_pd(c12_00,FF);
1173 fvdw = _mm_xor_pd(signbit,_mm_mul_pd(_mm_add_pd(fvdw6,fvdw12),_mm_mul_pd(vftabscale,rinv00)));
1177 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1179 /* Calculate temporary vectorial force */
1180 tx = _mm_mul_pd(fscal,dx00);
1181 ty = _mm_mul_pd(fscal,dy00);
1182 tz = _mm_mul_pd(fscal,dz00);
1184 /* Update vectorial force */
1185 fix0 = _mm_add_pd(fix0,tx);
1186 fiy0 = _mm_add_pd(fiy0,ty);
1187 fiz0 = _mm_add_pd(fiz0,tz);
1189 fjx0 = _mm_add_pd(fjx0,tx);
1190 fjy0 = _mm_add_pd(fjy0,ty);
1191 fjz0 = _mm_add_pd(fjz0,tz);
1193 /**************************
1194 * CALCULATE INTERACTIONS *
1195 **************************/
1197 r10 = _mm_mul_pd(rsq10,rinv10);
1199 /* Compute parameters for interactions between i and j atoms */
1200 qq10 = _mm_mul_pd(iq1,jq0);
1202 /* Calculate table index by multiplying r with table scale and truncate to integer */
1203 rt = _mm_mul_pd(r10,vftabscale);
1204 vfitab = _mm_cvttpd_epi32(rt);
1205 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1206 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1208 /* CUBIC SPLINE TABLE ELECTROSTATICS */
1209 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1210 F = _mm_setzero_pd();
1211 GMX_MM_TRANSPOSE2_PD(Y,F);
1212 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1213 H = _mm_setzero_pd();
1214 GMX_MM_TRANSPOSE2_PD(G,H);
1215 Heps = _mm_mul_pd(vfeps,H);
1216 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1217 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1218 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq10,FF),_mm_mul_pd(vftabscale,rinv10)));
1222 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1224 /* Calculate temporary vectorial force */
1225 tx = _mm_mul_pd(fscal,dx10);
1226 ty = _mm_mul_pd(fscal,dy10);
1227 tz = _mm_mul_pd(fscal,dz10);
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 fjx0 = _mm_add_pd(fjx0,tx);
1235 fjy0 = _mm_add_pd(fjy0,ty);
1236 fjz0 = _mm_add_pd(fjz0,tz);
1238 /**************************
1239 * CALCULATE INTERACTIONS *
1240 **************************/
1242 r20 = _mm_mul_pd(rsq20,rinv20);
1244 /* Compute parameters for interactions between i and j atoms */
1245 qq20 = _mm_mul_pd(iq2,jq0);
1247 /* Calculate table index by multiplying r with table scale and truncate to integer */
1248 rt = _mm_mul_pd(r20,vftabscale);
1249 vfitab = _mm_cvttpd_epi32(rt);
1250 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1251 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1253 /* CUBIC SPLINE TABLE ELECTROSTATICS */
1254 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1255 F = _mm_setzero_pd();
1256 GMX_MM_TRANSPOSE2_PD(Y,F);
1257 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1258 H = _mm_setzero_pd();
1259 GMX_MM_TRANSPOSE2_PD(G,H);
1260 Heps = _mm_mul_pd(vfeps,H);
1261 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1262 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1263 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq20,FF),_mm_mul_pd(vftabscale,rinv20)));
1267 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1269 /* Calculate temporary vectorial force */
1270 tx = _mm_mul_pd(fscal,dx20);
1271 ty = _mm_mul_pd(fscal,dy20);
1272 tz = _mm_mul_pd(fscal,dz20);
1274 /* Update vectorial force */
1275 fix2 = _mm_add_pd(fix2,tx);
1276 fiy2 = _mm_add_pd(fiy2,ty);
1277 fiz2 = _mm_add_pd(fiz2,tz);
1279 fjx0 = _mm_add_pd(fjx0,tx);
1280 fjy0 = _mm_add_pd(fjy0,ty);
1281 fjz0 = _mm_add_pd(fjz0,tz);
1283 /**************************
1284 * CALCULATE INTERACTIONS *
1285 **************************/
1287 r30 = _mm_mul_pd(rsq30,rinv30);
1289 /* Compute parameters for interactions between i and j atoms */
1290 qq30 = _mm_mul_pd(iq3,jq0);
1292 /* Calculate table index by multiplying r with table scale and truncate to integer */
1293 rt = _mm_mul_pd(r30,vftabscale);
1294 vfitab = _mm_cvttpd_epi32(rt);
1295 vfeps = _mm_sub_pd(rt,_mm_round_pd(rt, _MM_FROUND_FLOOR));
1296 vfitab = _mm_slli_epi32(_mm_add_epi32(vfitab,_mm_slli_epi32(vfitab,1)),2);
1298 /* CUBIC SPLINE TABLE ELECTROSTATICS */
1299 Y = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) );
1300 F = _mm_setzero_pd();
1301 GMX_MM_TRANSPOSE2_PD(Y,F);
1302 G = _mm_load_pd( vftab + gmx_mm_extract_epi32(vfitab,0) +2);
1303 H = _mm_setzero_pd();
1304 GMX_MM_TRANSPOSE2_PD(G,H);
1305 Heps = _mm_mul_pd(vfeps,H);
1306 Fp = _mm_add_pd(F,_mm_mul_pd(vfeps,_mm_add_pd(G,Heps)));
1307 FF = _mm_add_pd(Fp,_mm_mul_pd(vfeps,_mm_add_pd(G,_mm_add_pd(Heps,Heps))));
1308 felec = _mm_xor_pd(signbit,_mm_mul_pd(_mm_mul_pd(qq30,FF),_mm_mul_pd(vftabscale,rinv30)));
1312 fscal = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1314 /* Calculate temporary vectorial force */
1315 tx = _mm_mul_pd(fscal,dx30);
1316 ty = _mm_mul_pd(fscal,dy30);
1317 tz = _mm_mul_pd(fscal,dz30);
1319 /* Update vectorial force */
1320 fix3 = _mm_add_pd(fix3,tx);
1321 fiy3 = _mm_add_pd(fiy3,ty);
1322 fiz3 = _mm_add_pd(fiz3,tz);
1324 fjx0 = _mm_add_pd(fjx0,tx);
1325 fjy0 = _mm_add_pd(fjy0,ty);
1326 fjz0 = _mm_add_pd(fjz0,tz);
1328 gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0);
1330 /* Inner loop uses 168 flops */
1333 /* End of innermost loop */
1335 gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1336 f+i_coord_offset,fshift+i_shift_offset);
1338 /* Increment number of inner iterations */
1339 inneriter += j_index_end - j_index_start;
1341 /* Outer loop uses 24 flops */
1344 /* Increment number of outer iterations */
1347 /* Update outer/inner flops */
1349 inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*24 + inneriter*168);