src/gromacs/gmxlib/nonbonded/nb_kernel_avx_128_fma_double/nb_kernel_ElecEwSh_VdwLJEwSh_GeomW4P1_avx_128_fma_double.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS avx_128_fma_double kernel generator.
  37  */
  38 #include "config.h"
  39
  40 #include <math.h>
  41
  42 #include "../nb_kernel.h"
  43 #include "gromacs/legacyheaders/types/simple.h"
  44 #include "gromacs/math/vec.h"
  45 #include "gromacs/legacyheaders/nrnb.h"
  46
  47 #include "gromacs/simd/math_x86_avx_128_fma_double.h"
  48 #include "kernelutil_x86_avx_128_fma_double.h"
  49
  50 /*
  51  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwLJEwSh_GeomW4P1_VF_avx_128_fma_double
  52  * Electrostatics interaction: Ewald
  53  * VdW interaction:            LJEwald
  54  * Geometry:                   Water4-Particle
  55  * Calculate force/pot:        PotentialAndForce
  56  */
  57 void
  58 nb_kernel_ElecEwSh_VdwLJEwSh_GeomW4P1_VF_avx_128_fma_double
  59                     (t_nblist                    * gmx_restrict       nlist,
  60                      rvec                        * gmx_restrict          xx,
  61                      rvec                        * gmx_restrict          ff,
  62                      t_forcerec                  * gmx_restrict          fr,
  63                      t_mdatoms                   * gmx_restrict     mdatoms,
  64                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  65                      t_nrnb                      * gmx_restrict        nrnb)
  66 {
  67     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
  68      * just 0 for non-waters.
  69      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
  70      * jnr indices corresponding to data put in the four positions in the SIMD register.
  71      */
  72     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
  73     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
  74     int              jnrA,jnrB;
  75     int              j_coord_offsetA,j_coord_offsetB;
  76     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  77     real             rcutoff_scalar;
  78     real             *shiftvec,*fshift,*x,*f;
  79     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
  80     int              vdwioffset0;
  81     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  82     int              vdwioffset1;
  83     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
  84     int              vdwioffset2;
  85     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
  86     int              vdwioffset3;
  87     __m128d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
  88     int              vdwjidx0A,vdwjidx0B;
  89     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  90     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
  91     __m128d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
  92     __m128d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
  93     __m128d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
  94     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
  95     real             *charge;
  96     int              nvdwtype;
  97     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
  98     int              *vdwtype;
  99     real             *vdwparam;
 100     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
 101     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
 102     __m128d           c6grid_00;
 103     __m128d           c6grid_10;
 104     __m128d           c6grid_20;
 105     __m128d           c6grid_30;
 106     real             *vdwgridparam;
 107     __m128d           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 108     __m128d           one_half  = _mm_set1_pd(0.5);
 109     __m128d           minus_one = _mm_set1_pd(-1.0);
 110     __m128i          ewitab;
 111     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 112     real             *ewtab;
 113     __m128d          dummy_mask,cutoff_mask;
 114     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 115     __m128d          one     = _mm_set1_pd(1.0);
 116     __m128d          two     = _mm_set1_pd(2.0);
 117     x                = xx[0];
 118     f                = ff[0];
 119
 120     nri              = nlist->nri;
 121     iinr             = nlist->iinr;
 122     jindex           = nlist->jindex;
 123     jjnr             = nlist->jjnr;
 124     shiftidx         = nlist->shift;
 125     gid              = nlist->gid;
 126     shiftvec         = fr->shift_vec[0];
 127     fshift           = fr->fshift[0];
 128     facel            = _mm_set1_pd(fr->epsfac);
 129     charge           = mdatoms->chargeA;
 130     nvdwtype         = fr->ntype;
 131     vdwparam         = fr->nbfp;
 132     vdwtype          = mdatoms->typeA;
 133     vdwgridparam     = fr->ljpme_c6grid;
 134     sh_lj_ewald      = _mm_set1_pd(fr->ic->sh_lj_ewald);
 135     ewclj            = _mm_set1_pd(fr->ewaldcoeff_lj);
 136     ewclj2           = _mm_mul_pd(minus_one,_mm_mul_pd(ewclj,ewclj));
 137
 138     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
 139     ewtab            = fr->ic->tabq_coul_FDV0;
 140     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
 141     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
 142
 143     /* Setup water-specific parameters */
 144     inr              = nlist->iinr[0];
 145     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
 146     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
 147     iq3              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
 148     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 149
 150     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 151     rcutoff_scalar   = fr->rcoulomb;
 152     rcutoff          = _mm_set1_pd(rcutoff_scalar);
 153     rcutoff2         = _mm_mul_pd(rcutoff,rcutoff);
 154
 155     sh_vdw_invrcut6  = _mm_set1_pd(fr->ic->sh_invrc6);
 156     rvdw             = _mm_set1_pd(fr->rvdw);
 157
 158     /* Avoid stupid compiler warnings */
 159     jnrA = jnrB = 0;
 160     j_coord_offsetA = 0;
 161     j_coord_offsetB = 0;
 162
 163     outeriter        = 0;
 164     inneriter        = 0;
 165
 166     /* Start outer loop over neighborlists */
 167     for(iidx=0; iidx<nri; iidx++)
 168     {
 169         /* Load shift vector for this list */
 170         i_shift_offset   = DIM*shiftidx[iidx];
 171
 172         /* Load limits for loop over neighbors */
 173         j_index_start    = jindex[iidx];
 174         j_index_end      = jindex[iidx+1];
 175
 176         /* Get outer coordinate index */
 177         inr              = iinr[iidx];
 178         i_coord_offset   = DIM*inr;
 179
 180         /* Load i particle coords and add shift vector */
 181         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 182                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 183
 184         fix0             = _mm_setzero_pd();
 185         fiy0             = _mm_setzero_pd();
 186         fiz0             = _mm_setzero_pd();
 187         fix1             = _mm_setzero_pd();
 188         fiy1             = _mm_setzero_pd();
 189         fiz1             = _mm_setzero_pd();
 190         fix2             = _mm_setzero_pd();
 191         fiy2             = _mm_setzero_pd();
 192         fiz2             = _mm_setzero_pd();
 193         fix3             = _mm_setzero_pd();
 194         fiy3             = _mm_setzero_pd();
 195         fiz3             = _mm_setzero_pd();
 196
 197         /* Reset potential sums */
 198         velecsum         = _mm_setzero_pd();
 199         vvdwsum          = _mm_setzero_pd();
 200
 201         /* Start inner kernel loop */
 202         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 203         {
 204
 205             /* Get j neighbor index, and coordinate index */
 206             jnrA             = jjnr[jidx];
 207             jnrB             = jjnr[jidx+1];
 208             j_coord_offsetA  = DIM*jnrA;
 209             j_coord_offsetB  = DIM*jnrB;
 210
 211             /* load j atom coordinates */
 212             gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 213                                               &jx0,&jy0,&jz0);
 214
 215             /* Calculate displacement vector */
 216             dx00             = _mm_sub_pd(ix0,jx0);
 217             dy00             = _mm_sub_pd(iy0,jy0);
 218             dz00             = _mm_sub_pd(iz0,jz0);
 219             dx10             = _mm_sub_pd(ix1,jx0);
 220             dy10             = _mm_sub_pd(iy1,jy0);
 221             dz10             = _mm_sub_pd(iz1,jz0);
 222             dx20             = _mm_sub_pd(ix2,jx0);
 223             dy20             = _mm_sub_pd(iy2,jy0);
 224             dz20             = _mm_sub_pd(iz2,jz0);
 225             dx30             = _mm_sub_pd(ix3,jx0);
 226             dy30             = _mm_sub_pd(iy3,jy0);
 227             dz30             = _mm_sub_pd(iz3,jz0);
 228
 229             /* Calculate squared distance and things based on it */
 230             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 231             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
 232             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
 233             rsq30            = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
 234
 235             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 236             rinv10           = gmx_mm_invsqrt_pd(rsq10);
 237             rinv20           = gmx_mm_invsqrt_pd(rsq20);
 238             rinv30           = gmx_mm_invsqrt_pd(rsq30);
 239
 240             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 241             rinvsq10         = _mm_mul_pd(rinv10,rinv10);
 242             rinvsq20         = _mm_mul_pd(rinv20,rinv20);
 243             rinvsq30         = _mm_mul_pd(rinv30,rinv30);
 244
 245             /* Load parameters for j particles */
 246             jq0              = gmx_mm_load_2real_swizzle_pd(charge+jnrA+0,charge+jnrB+0);
 247             vdwjidx0A        = 2*vdwtype[jnrA+0];
 248             vdwjidx0B        = 2*vdwtype[jnrB+0];
 249
 250             fjx0             = _mm_setzero_pd();
 251             fjy0             = _mm_setzero_pd();
 252             fjz0             = _mm_setzero_pd();
 253
 254             /**************************
 255              * CALCULATE INTERACTIONS *
 256              **************************/
 257
 258             if (gmx_mm_any_lt(rsq00,rcutoff2))
 259             {
 260
 261             r00              = _mm_mul_pd(rsq00,rinv00);
 262
 263             /* Compute parameters for interactions between i and j atoms */
 264             gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,
 265                                          vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
 266             c6grid_00       = gmx_mm_load_2real_swizzle_pd(vdwgridparam+vdwioffset0+vdwjidx0A,
 267                                                                vdwgridparam+vdwioffset0+vdwjidx0B);
 268
 269             /* Analytical LJ-PME */
 270             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 271             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
 272             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
 273             exponent         = gmx_simd_exp_d(ewcljrsq);
 274             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 275             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
 276             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 277             vvdw6            = _mm_mul_pd(_mm_macc_pd(-c6grid_00,_mm_sub_pd(one,poly),c6_00),rinvsix);
 278             vvdw12           = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
 279             vvdw             = _mm_msub_pd(_mm_nmacc_pd(c12_00,_mm_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6),vvdw12),one_twelfth,
 280                                _mm_mul_pd(_mm_sub_pd(vvdw6,_mm_macc_pd(c6grid_00,sh_lj_ewald,_mm_mul_pd(c6_00,sh_vdw_invrcut6))),one_sixth));
 281             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 282             fvdw             = _mm_mul_pd(_mm_add_pd(vvdw12,_mm_msub_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6),vvdw6)),rinvsq00);
 283
 284             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
 285
 286             /* Update potential sum for this i atom from the interaction with this j atom. */
 287             vvdw             = _mm_and_pd(vvdw,cutoff_mask);
 288             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 289
 290             fscal            = fvdw;
 291
 292             fscal            = _mm_and_pd(fscal,cutoff_mask);
 293
 294             /* Update vectorial force */
 295             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 296             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 297             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 298
 299             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 300             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 301             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 302
 303             }
 304
 305             /**************************
 306              * CALCULATE INTERACTIONS *
 307              **************************/
 308
 309             if (gmx_mm_any_lt(rsq10,rcutoff2))
 310             {
 311
 312             r10              = _mm_mul_pd(rsq10,rinv10);
 313
 314             /* Compute parameters for interactions between i and j atoms */
 315             qq10             = _mm_mul_pd(iq1,jq0);
 316
 317             /* EWALD ELECTROSTATICS */
 318
 319             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 320             ewrt             = _mm_mul_pd(r10,ewtabscale);
 321             ewitab           = _mm_cvttpd_epi32(ewrt);
 322 #ifdef __XOP__
 323             eweps            = _mm_frcz_pd(ewrt);
 324 #else
 325             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 326 #endif
 327             twoeweps         = _mm_add_pd(eweps,eweps);
 328             ewitab           = _mm_slli_epi32(ewitab,2);
 329             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 330             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 331             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 332             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 333             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 334             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 335             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 336             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 337             velec            = _mm_mul_pd(qq10,_mm_sub_pd(_mm_sub_pd(rinv10,sh_ewald),velec));
 338             felec            = _mm_mul_pd(_mm_mul_pd(qq10,rinv10),_mm_sub_pd(rinvsq10,felec));
 339
 340             cutoff_mask      = _mm_cmplt_pd(rsq10,rcutoff2);
 341
 342             /* Update potential sum for this i atom from the interaction with this j atom. */
 343             velec            = _mm_and_pd(velec,cutoff_mask);
 344             velecsum         = _mm_add_pd(velecsum,velec);
 345
 346             fscal            = felec;
 347
 348             fscal            = _mm_and_pd(fscal,cutoff_mask);
 349
 350             /* Update vectorial force */
 351             fix1             = _mm_macc_pd(dx10,fscal,fix1);
 352             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
 353             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
 354
 355             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
 356             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
 357             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
 358
 359             }
 360
 361             /**************************
 362              * CALCULATE INTERACTIONS *
 363              **************************/
 364
 365             if (gmx_mm_any_lt(rsq20,rcutoff2))
 366             {
 367
 368             r20              = _mm_mul_pd(rsq20,rinv20);
 369
 370             /* Compute parameters for interactions between i and j atoms */
 371             qq20             = _mm_mul_pd(iq2,jq0);
 372
 373             /* EWALD ELECTROSTATICS */
 374
 375             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 376             ewrt             = _mm_mul_pd(r20,ewtabscale);
 377             ewitab           = _mm_cvttpd_epi32(ewrt);
 378 #ifdef __XOP__
 379             eweps            = _mm_frcz_pd(ewrt);
 380 #else
 381             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 382 #endif
 383             twoeweps         = _mm_add_pd(eweps,eweps);
 384             ewitab           = _mm_slli_epi32(ewitab,2);
 385             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 386             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 387             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 388             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 389             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 390             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 391             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 392             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 393             velec            = _mm_mul_pd(qq20,_mm_sub_pd(_mm_sub_pd(rinv20,sh_ewald),velec));
 394             felec            = _mm_mul_pd(_mm_mul_pd(qq20,rinv20),_mm_sub_pd(rinvsq20,felec));
 395
 396             cutoff_mask      = _mm_cmplt_pd(rsq20,rcutoff2);
 397
 398             /* Update potential sum for this i atom from the interaction with this j atom. */
 399             velec            = _mm_and_pd(velec,cutoff_mask);
 400             velecsum         = _mm_add_pd(velecsum,velec);
 401
 402             fscal            = felec;
 403
 404             fscal            = _mm_and_pd(fscal,cutoff_mask);
 405
 406             /* Update vectorial force */
 407             fix2             = _mm_macc_pd(dx20,fscal,fix2);
 408             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
 409             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
 410
 411             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
 412             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
 413             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
 414
 415             }
 416
 417             /**************************
 418              * CALCULATE INTERACTIONS *
 419              **************************/
 420
 421             if (gmx_mm_any_lt(rsq30,rcutoff2))
 422             {
 423
 424             r30              = _mm_mul_pd(rsq30,rinv30);
 425
 426             /* Compute parameters for interactions between i and j atoms */
 427             qq30             = _mm_mul_pd(iq3,jq0);
 428
 429             /* EWALD ELECTROSTATICS */
 430
 431             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 432             ewrt             = _mm_mul_pd(r30,ewtabscale);
 433             ewitab           = _mm_cvttpd_epi32(ewrt);
 434 #ifdef __XOP__
 435             eweps            = _mm_frcz_pd(ewrt);
 436 #else
 437             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 438 #endif
 439             twoeweps         = _mm_add_pd(eweps,eweps);
 440             ewitab           = _mm_slli_epi32(ewitab,2);
 441             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 442             ewtabD           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,1) );
 443             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 444             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 445             ewtabFn          = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,1) +2);
 446             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 447             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 448             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 449             velec            = _mm_mul_pd(qq30,_mm_sub_pd(_mm_sub_pd(rinv30,sh_ewald),velec));
 450             felec            = _mm_mul_pd(_mm_mul_pd(qq30,rinv30),_mm_sub_pd(rinvsq30,felec));
 451
 452             cutoff_mask      = _mm_cmplt_pd(rsq30,rcutoff2);
 453
 454             /* Update potential sum for this i atom from the interaction with this j atom. */
 455             velec            = _mm_and_pd(velec,cutoff_mask);
 456             velecsum         = _mm_add_pd(velecsum,velec);
 457
 458             fscal            = felec;
 459
 460             fscal            = _mm_and_pd(fscal,cutoff_mask);
 461
 462             /* Update vectorial force */
 463             fix3             = _mm_macc_pd(dx30,fscal,fix3);
 464             fiy3             = _mm_macc_pd(dy30,fscal,fiy3);
 465             fiz3             = _mm_macc_pd(dz30,fscal,fiz3);
 466
 467             fjx0             = _mm_macc_pd(dx30,fscal,fjx0);
 468             fjy0             = _mm_macc_pd(dy30,fscal,fjy0);
 469             fjz0             = _mm_macc_pd(dz30,fscal,fjz0);
 470
 471             }
 472
 473             gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
 474
 475             /* Inner loop uses 208 flops */
 476         }
 477
 478         if(jidx<j_index_end)
 479         {
 480
 481             jnrA             = jjnr[jidx];
 482             j_coord_offsetA  = DIM*jnrA;
 483
 484             /* load j atom coordinates */
 485             gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
 486                                               &jx0,&jy0,&jz0);
 487
 488             /* Calculate displacement vector */
 489             dx00             = _mm_sub_pd(ix0,jx0);
 490             dy00             = _mm_sub_pd(iy0,jy0);
 491             dz00             = _mm_sub_pd(iz0,jz0);
 492             dx10             = _mm_sub_pd(ix1,jx0);
 493             dy10             = _mm_sub_pd(iy1,jy0);
 494             dz10             = _mm_sub_pd(iz1,jz0);
 495             dx20             = _mm_sub_pd(ix2,jx0);
 496             dy20             = _mm_sub_pd(iy2,jy0);
 497             dz20             = _mm_sub_pd(iz2,jz0);
 498             dx30             = _mm_sub_pd(ix3,jx0);
 499             dy30             = _mm_sub_pd(iy3,jy0);
 500             dz30             = _mm_sub_pd(iz3,jz0);
 501
 502             /* Calculate squared distance and things based on it */
 503             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 504             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
 505             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
 506             rsq30            = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
 507
 508             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 509             rinv10           = gmx_mm_invsqrt_pd(rsq10);
 510             rinv20           = gmx_mm_invsqrt_pd(rsq20);
 511             rinv30           = gmx_mm_invsqrt_pd(rsq30);
 512
 513             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 514             rinvsq10         = _mm_mul_pd(rinv10,rinv10);
 515             rinvsq20         = _mm_mul_pd(rinv20,rinv20);
 516             rinvsq30         = _mm_mul_pd(rinv30,rinv30);
 517
 518             /* Load parameters for j particles */
 519             jq0              = _mm_load_sd(charge+jnrA+0);
 520             vdwjidx0A        = 2*vdwtype[jnrA+0];
 521
 522             fjx0             = _mm_setzero_pd();
 523             fjy0             = _mm_setzero_pd();
 524             fjz0             = _mm_setzero_pd();
 525
 526             /**************************
 527              * CALCULATE INTERACTIONS *
 528              **************************/
 529
 530             if (gmx_mm_any_lt(rsq00,rcutoff2))
 531             {
 532
 533             r00              = _mm_mul_pd(rsq00,rinv00);
 534
 535             /* Compute parameters for interactions between i and j atoms */
 536             gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
 537             c6grid_00       = gmx_mm_load_1real_pd(vdwgridparam+vdwioffset0+vdwjidx0A);
 538
 539             /* Analytical LJ-PME */
 540             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
 541             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
 542             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
 543             exponent         = gmx_simd_exp_d(ewcljrsq);
 544             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
 545             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
 546             /* vvdw6 = [C6 - C6grid * (1-poly)]/r6 */
 547             vvdw6            = _mm_mul_pd(_mm_macc_pd(-c6grid_00,_mm_sub_pd(one,poly),c6_00),rinvsix);
 548             vvdw12           = _mm_mul_pd(c12_00,_mm_mul_pd(rinvsix,rinvsix));
 549             vvdw             = _mm_msub_pd(_mm_nmacc_pd(c12_00,_mm_mul_pd(sh_vdw_invrcut6,sh_vdw_invrcut6),vvdw12),one_twelfth,
 550                                _mm_mul_pd(_mm_sub_pd(vvdw6,_mm_macc_pd(c6grid_00,sh_lj_ewald,_mm_mul_pd(c6_00,sh_vdw_invrcut6))),one_sixth));
 551             /* fvdw = vvdw12/r - (vvdw6/r + (C6grid * exponent * beta^6)/r) */
 552             fvdw             = _mm_mul_pd(_mm_add_pd(vvdw12,_mm_msub_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6),vvdw6)),rinvsq00);
 553
 554             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
 555
 556             /* Update potential sum for this i atom from the interaction with this j atom. */
 557             vvdw             = _mm_and_pd(vvdw,cutoff_mask);
 558             vvdw             = _mm_unpacklo_pd(vvdw,_mm_setzero_pd());
 559             vvdwsum          = _mm_add_pd(vvdwsum,vvdw);
 560
 561             fscal            = fvdw;
 562
 563             fscal            = _mm_and_pd(fscal,cutoff_mask);
 564
 565             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 566
 567             /* Update vectorial force */
 568             fix0             = _mm_macc_pd(dx00,fscal,fix0);
 569             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
 570             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
 571
 572             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
 573             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
 574             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
 575
 576             }
 577
 578             /**************************
 579              * CALCULATE INTERACTIONS *
 580              **************************/
 581
 582             if (gmx_mm_any_lt(rsq10,rcutoff2))
 583             {
 584
 585             r10              = _mm_mul_pd(rsq10,rinv10);
 586
 587             /* Compute parameters for interactions between i and j atoms */
 588             qq10             = _mm_mul_pd(iq1,jq0);
 589
 590             /* EWALD ELECTROSTATICS */
 591
 592             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 593             ewrt             = _mm_mul_pd(r10,ewtabscale);
 594             ewitab           = _mm_cvttpd_epi32(ewrt);
 595 #ifdef __XOP__
 596             eweps            = _mm_frcz_pd(ewrt);
 597 #else
 598             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 599 #endif
 600             twoeweps         = _mm_add_pd(eweps,eweps);
 601             ewitab           = _mm_slli_epi32(ewitab,2);
 602             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 603             ewtabD           = _mm_setzero_pd();
 604             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 605             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 606             ewtabFn          = _mm_setzero_pd();
 607             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 608             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 609             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 610             velec            = _mm_mul_pd(qq10,_mm_sub_pd(_mm_sub_pd(rinv10,sh_ewald),velec));
 611             felec            = _mm_mul_pd(_mm_mul_pd(qq10,rinv10),_mm_sub_pd(rinvsq10,felec));
 612
 613             cutoff_mask      = _mm_cmplt_pd(rsq10,rcutoff2);
 614
 615             /* Update potential sum for this i atom from the interaction with this j atom. */
 616             velec            = _mm_and_pd(velec,cutoff_mask);
 617             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 618             velecsum         = _mm_add_pd(velecsum,velec);
 619
 620             fscal            = felec;
 621
 622             fscal            = _mm_and_pd(fscal,cutoff_mask);
 623
 624             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 625
 626             /* Update vectorial force */
 627             fix1             = _mm_macc_pd(dx10,fscal,fix1);
 628             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
 629             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
 630
 631             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
 632             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
 633             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
 634
 635             }
 636
 637             /**************************
 638              * CALCULATE INTERACTIONS *
 639              **************************/
 640
 641             if (gmx_mm_any_lt(rsq20,rcutoff2))
 642             {
 643
 644             r20              = _mm_mul_pd(rsq20,rinv20);
 645
 646             /* Compute parameters for interactions between i and j atoms */
 647             qq20             = _mm_mul_pd(iq2,jq0);
 648
 649             /* EWALD ELECTROSTATICS */
 650
 651             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 652             ewrt             = _mm_mul_pd(r20,ewtabscale);
 653             ewitab           = _mm_cvttpd_epi32(ewrt);
 654 #ifdef __XOP__
 655             eweps            = _mm_frcz_pd(ewrt);
 656 #else
 657             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 658 #endif
 659             twoeweps         = _mm_add_pd(eweps,eweps);
 660             ewitab           = _mm_slli_epi32(ewitab,2);
 661             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 662             ewtabD           = _mm_setzero_pd();
 663             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 664             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 665             ewtabFn          = _mm_setzero_pd();
 666             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 667             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 668             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 669             velec            = _mm_mul_pd(qq20,_mm_sub_pd(_mm_sub_pd(rinv20,sh_ewald),velec));
 670             felec            = _mm_mul_pd(_mm_mul_pd(qq20,rinv20),_mm_sub_pd(rinvsq20,felec));
 671
 672             cutoff_mask      = _mm_cmplt_pd(rsq20,rcutoff2);
 673
 674             /* Update potential sum for this i atom from the interaction with this j atom. */
 675             velec            = _mm_and_pd(velec,cutoff_mask);
 676             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 677             velecsum         = _mm_add_pd(velecsum,velec);
 678
 679             fscal            = felec;
 680
 681             fscal            = _mm_and_pd(fscal,cutoff_mask);
 682
 683             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 684
 685             /* Update vectorial force */
 686             fix2             = _mm_macc_pd(dx20,fscal,fix2);
 687             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
 688             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
 689
 690             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
 691             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
 692             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
 693
 694             }
 695
 696             /**************************
 697              * CALCULATE INTERACTIONS *
 698              **************************/
 699
 700             if (gmx_mm_any_lt(rsq30,rcutoff2))
 701             {
 702
 703             r30              = _mm_mul_pd(rsq30,rinv30);
 704
 705             /* Compute parameters for interactions between i and j atoms */
 706             qq30             = _mm_mul_pd(iq3,jq0);
 707
 708             /* EWALD ELECTROSTATICS */
 709
 710             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
 711             ewrt             = _mm_mul_pd(r30,ewtabscale);
 712             ewitab           = _mm_cvttpd_epi32(ewrt);
 713 #ifdef __XOP__
 714             eweps            = _mm_frcz_pd(ewrt);
 715 #else
 716             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
 717 #endif
 718             twoeweps         = _mm_add_pd(eweps,eweps);
 719             ewitab           = _mm_slli_epi32(ewitab,2);
 720             ewtabF           = _mm_load_pd( ewtab + _mm_extract_epi32(ewitab,0) );
 721             ewtabD           = _mm_setzero_pd();
 722             GMX_MM_TRANSPOSE2_PD(ewtabF,ewtabD);
 723             ewtabV           = _mm_load_sd( ewtab + _mm_extract_epi32(ewitab,0) +2);
 724             ewtabFn          = _mm_setzero_pd();
 725             GMX_MM_TRANSPOSE2_PD(ewtabV,ewtabFn);
 726             felec            = _mm_macc_pd(eweps,ewtabD,ewtabF);
 727             velec            = _mm_nmacc_pd(_mm_mul_pd(ewtabhalfspace,eweps) ,_mm_add_pd(ewtabF,felec), ewtabV);
 728             velec            = _mm_mul_pd(qq30,_mm_sub_pd(_mm_sub_pd(rinv30,sh_ewald),velec));
 729             felec            = _mm_mul_pd(_mm_mul_pd(qq30,rinv30),_mm_sub_pd(rinvsq30,felec));
 730
 731             cutoff_mask      = _mm_cmplt_pd(rsq30,rcutoff2);
 732
 733             /* Update potential sum for this i atom from the interaction with this j atom. */
 734             velec            = _mm_and_pd(velec,cutoff_mask);
 735             velec            = _mm_unpacklo_pd(velec,_mm_setzero_pd());
 736             velecsum         = _mm_add_pd(velecsum,velec);
 737
 738             fscal            = felec;
 739
 740             fscal            = _mm_and_pd(fscal,cutoff_mask);
 741
 742             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
 743
 744             /* Update vectorial force */
 745             fix3             = _mm_macc_pd(dx30,fscal,fix3);
 746             fiy3             = _mm_macc_pd(dy30,fscal,fiy3);
 747             fiz3             = _mm_macc_pd(dz30,fscal,fiz3);
 748
 749             fjx0             = _mm_macc_pd(dx30,fscal,fjx0);
 750             fjy0             = _mm_macc_pd(dy30,fscal,fjy0);
 751             fjz0             = _mm_macc_pd(dz30,fscal,fjz0);
 752
 753             }
 754
 755             gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0);
 756
 757             /* Inner loop uses 208 flops */
 758         }
 759
 760         /* End of innermost loop */
 761
 762         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
 763                                               f+i_coord_offset,fshift+i_shift_offset);
 764
 765         ggid                        = gid[iidx];
 766         /* Update potential energies */
 767         gmx_mm_update_1pot_pd(velecsum,kernel_data->energygrp_elec+ggid);
 768         gmx_mm_update_1pot_pd(vvdwsum,kernel_data->energygrp_vdw+ggid);
 769
 770         /* Increment number of inner iterations */
 771         inneriter                  += j_index_end - j_index_start;
 772
 773         /* Outer loop uses 26 flops */
 774     }
 775
 776     /* Increment number of outer iterations */
 777     outeriter        += nri;
 778
 779     /* Update outer/inner flops */
 780
 781     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_VF,outeriter*26 + inneriter*208);
 782 }
 783 /*
 784  * Gromacs nonbonded kernel:   nb_kernel_ElecEwSh_VdwLJEwSh_GeomW4P1_F_avx_128_fma_double
 785  * Electrostatics interaction: Ewald
 786  * VdW interaction:            LJEwald
 787  * Geometry:                   Water4-Particle
 788  * Calculate force/pot:        Force
 789  */
 790 void
 791 nb_kernel_ElecEwSh_VdwLJEwSh_GeomW4P1_F_avx_128_fma_double
 792                     (t_nblist                    * gmx_restrict       nlist,
 793                      rvec                        * gmx_restrict          xx,
 794                      rvec                        * gmx_restrict          ff,
 795                      t_forcerec                  * gmx_restrict          fr,
 796                      t_mdatoms                   * gmx_restrict     mdatoms,
 797                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 798                      t_nrnb                      * gmx_restrict        nrnb)
 799 {
 800     /* Suffixes 0,1,2,3 refer to particle indices for waters in the inner or outer loop, or
 801      * just 0 for non-waters.
 802      * Suffixes A,B refer to j loop unrolling done with SSE double precision, e.g. for the two different
 803      * jnr indices corresponding to data put in the four positions in the SIMD register.
 804      */
 805     int              i_shift_offset,i_coord_offset,outeriter,inneriter;
 806     int              j_index_start,j_index_end,jidx,nri,inr,ggid,iidx;
 807     int              jnrA,jnrB;
 808     int              j_coord_offsetA,j_coord_offsetB;
 809     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 810     real             rcutoff_scalar;
 811     real             *shiftvec,*fshift,*x,*f;
 812     __m128d          tx,ty,tz,fscal,rcutoff,rcutoff2,jidxall;
 813     int              vdwioffset0;
 814     __m128d          ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 815     int              vdwioffset1;
 816     __m128d          ix1,iy1,iz1,fix1,fiy1,fiz1,iq1,isai1;
 817     int              vdwioffset2;
 818     __m128d          ix2,iy2,iz2,fix2,fiy2,fiz2,iq2,isai2;
 819     int              vdwioffset3;
 820     __m128d          ix3,iy3,iz3,fix3,fiy3,fiz3,iq3,isai3;
 821     int              vdwjidx0A,vdwjidx0B;
 822     __m128d          jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 823     __m128d          dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00;
 824     __m128d          dx10,dy10,dz10,rsq10,rinv10,rinvsq10,r10,qq10,c6_10,c12_10;
 825     __m128d          dx20,dy20,dz20,rsq20,rinv20,rinvsq20,r20,qq20,c6_20,c12_20;
 826     __m128d          dx30,dy30,dz30,rsq30,rinv30,rinvsq30,r30,qq30,c6_30,c12_30;
 827     __m128d          velec,felec,velecsum,facel,crf,krf,krf2;
 828     real             *charge;
 829     int              nvdwtype;
 830     __m128d          rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,sh_vdw_invrcut6;
 831     int              *vdwtype;
 832     real             *vdwparam;
 833     __m128d          one_sixth   = _mm_set1_pd(1.0/6.0);
 834     __m128d          one_twelfth = _mm_set1_pd(1.0/12.0);
 835     __m128d           c6grid_00;
 836     __m128d           c6grid_10;
 837     __m128d           c6grid_20;
 838     __m128d           c6grid_30;
 839     real             *vdwgridparam;
 840     __m128d           ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,f6A,f6B,sh_lj_ewald;
 841     __m128d           one_half  = _mm_set1_pd(0.5);
 842     __m128d           minus_one = _mm_set1_pd(-1.0);
 843     __m128i          ewitab;
 844     __m128d          ewtabscale,eweps,twoeweps,sh_ewald,ewrt,ewtabhalfspace,ewtabF,ewtabFn,ewtabD,ewtabV;
 845     real             *ewtab;
 846     __m128d          dummy_mask,cutoff_mask;
 847     __m128d          signbit   = gmx_mm_castsi128_pd( _mm_set_epi32(0x80000000,0x00000000,0x80000000,0x00000000) );
 848     __m128d          one     = _mm_set1_pd(1.0);
 849     __m128d          two     = _mm_set1_pd(2.0);
 850     x                = xx[0];
 851     f                = ff[0];
 852
 853     nri              = nlist->nri;
 854     iinr             = nlist->iinr;
 855     jindex           = nlist->jindex;
 856     jjnr             = nlist->jjnr;
 857     shiftidx         = nlist->shift;
 858     gid              = nlist->gid;
 859     shiftvec         = fr->shift_vec[0];
 860     fshift           = fr->fshift[0];
 861     facel            = _mm_set1_pd(fr->epsfac);
 862     charge           = mdatoms->chargeA;
 863     nvdwtype         = fr->ntype;
 864     vdwparam         = fr->nbfp;
 865     vdwtype          = mdatoms->typeA;
 866     vdwgridparam     = fr->ljpme_c6grid;
 867     sh_lj_ewald      = _mm_set1_pd(fr->ic->sh_lj_ewald);
 868     ewclj            = _mm_set1_pd(fr->ewaldcoeff_lj);
 869     ewclj2           = _mm_mul_pd(minus_one,_mm_mul_pd(ewclj,ewclj));
 870
 871     sh_ewald         = _mm_set1_pd(fr->ic->sh_ewald);
 872     ewtab            = fr->ic->tabq_coul_F;
 873     ewtabscale       = _mm_set1_pd(fr->ic->tabq_scale);
 874     ewtabhalfspace   = _mm_set1_pd(0.5/fr->ic->tabq_scale);
 875
 876     /* Setup water-specific parameters */
 877     inr              = nlist->iinr[0];
 878     iq1              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+1]));
 879     iq2              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+2]));
 880     iq3              = _mm_mul_pd(facel,_mm_set1_pd(charge[inr+3]));
 881     vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 882
 883     /* When we use explicit cutoffs the value must be identical for elec and VdW, so use elec as an arbitrary choice */
 884     rcutoff_scalar   = fr->rcoulomb;
 885     rcutoff          = _mm_set1_pd(rcutoff_scalar);
 886     rcutoff2         = _mm_mul_pd(rcutoff,rcutoff);
 887
 888     sh_vdw_invrcut6  = _mm_set1_pd(fr->ic->sh_invrc6);
 889     rvdw             = _mm_set1_pd(fr->rvdw);
 890
 891     /* Avoid stupid compiler warnings */
 892     jnrA = jnrB = 0;
 893     j_coord_offsetA = 0;
 894     j_coord_offsetB = 0;
 895
 896     outeriter        = 0;
 897     inneriter        = 0;
 898
 899     /* Start outer loop over neighborlists */
 900     for(iidx=0; iidx<nri; iidx++)
 901     {
 902         /* Load shift vector for this list */
 903         i_shift_offset   = DIM*shiftidx[iidx];
 904
 905         /* Load limits for loop over neighbors */
 906         j_index_start    = jindex[iidx];
 907         j_index_end      = jindex[iidx+1];
 908
 909         /* Get outer coordinate index */
 910         inr              = iinr[iidx];
 911         i_coord_offset   = DIM*inr;
 912
 913         /* Load i particle coords and add shift vector */
 914         gmx_mm_load_shift_and_4rvec_broadcast_pd(shiftvec+i_shift_offset,x+i_coord_offset,
 915                                                  &ix0,&iy0,&iz0,&ix1,&iy1,&iz1,&ix2,&iy2,&iz2,&ix3,&iy3,&iz3);
 916
 917         fix0             = _mm_setzero_pd();
 918         fiy0             = _mm_setzero_pd();
 919         fiz0             = _mm_setzero_pd();
 920         fix1             = _mm_setzero_pd();
 921         fiy1             = _mm_setzero_pd();
 922         fiz1             = _mm_setzero_pd();
 923         fix2             = _mm_setzero_pd();
 924         fiy2             = _mm_setzero_pd();
 925         fiz2             = _mm_setzero_pd();
 926         fix3             = _mm_setzero_pd();
 927         fiy3             = _mm_setzero_pd();
 928         fiz3             = _mm_setzero_pd();
 929
 930         /* Start inner kernel loop */
 931         for(jidx=j_index_start; jidx<j_index_end-1; jidx+=2)
 932         {
 933
 934             /* Get j neighbor index, and coordinate index */
 935             jnrA             = jjnr[jidx];
 936             jnrB             = jjnr[jidx+1];
 937             j_coord_offsetA  = DIM*jnrA;
 938             j_coord_offsetB  = DIM*jnrB;
 939
 940             /* load j atom coordinates */
 941             gmx_mm_load_1rvec_2ptr_swizzle_pd(x+j_coord_offsetA,x+j_coord_offsetB,
 942                                               &jx0,&jy0,&jz0);
 943
 944             /* Calculate displacement vector */
 945             dx00             = _mm_sub_pd(ix0,jx0);
 946             dy00             = _mm_sub_pd(iy0,jy0);
 947             dz00             = _mm_sub_pd(iz0,jz0);
 948             dx10             = _mm_sub_pd(ix1,jx0);
 949             dy10             = _mm_sub_pd(iy1,jy0);
 950             dz10             = _mm_sub_pd(iz1,jz0);
 951             dx20             = _mm_sub_pd(ix2,jx0);
 952             dy20             = _mm_sub_pd(iy2,jy0);
 953             dz20             = _mm_sub_pd(iz2,jz0);
 954             dx30             = _mm_sub_pd(ix3,jx0);
 955             dy30             = _mm_sub_pd(iy3,jy0);
 956             dz30             = _mm_sub_pd(iz3,jz0);
 957
 958             /* Calculate squared distance and things based on it */
 959             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
 960             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
 961             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
 962             rsq30            = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
 963
 964             rinv00           = gmx_mm_invsqrt_pd(rsq00);
 965             rinv10           = gmx_mm_invsqrt_pd(rsq10);
 966             rinv20           = gmx_mm_invsqrt_pd(rsq20);
 967             rinv30           = gmx_mm_invsqrt_pd(rsq30);
 968
 969             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
 970             rinvsq10         = _mm_mul_pd(rinv10,rinv10);
 971             rinvsq20         = _mm_mul_pd(rinv20,rinv20);
 972             rinvsq30         = _mm_mul_pd(rinv30,rinv30);
 973
 974             /* Load parameters for j particles */
 975             jq0              = gmx_mm_load_2real_swizzle_pd(charge+jnrA+0,charge+jnrB+0);
 976             vdwjidx0A        = 2*vdwtype[jnrA+0];
 977             vdwjidx0B        = 2*vdwtype[jnrB+0];
 978
 979             fjx0             = _mm_setzero_pd();
 980             fjy0             = _mm_setzero_pd();
 981             fjz0             = _mm_setzero_pd();
 982
 983             /**************************
 984              * CALCULATE INTERACTIONS *
 985              **************************/
 986
 987             if (gmx_mm_any_lt(rsq00,rcutoff2))
 988             {
 989
 990             r00              = _mm_mul_pd(rsq00,rinv00);
 991
 992             /* Compute parameters for interactions between i and j atoms */
 993             gmx_mm_load_2pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,
 994                                          vdwparam+vdwioffset0+vdwjidx0B,&c6_00,&c12_00);
 995             c6grid_00       = gmx_mm_load_2real_swizzle_pd(vdwgridparam+vdwioffset0+vdwjidx0A,
 996                                                                vdwgridparam+vdwioffset0+vdwjidx0B);
 997
 998             /* Analytical LJ-PME */
 999             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1000             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
1001             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
1002             exponent         = gmx_simd_exp_d(ewcljrsq);
1003             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
1004             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
1005             /* f6A = 6 * C6grid * (1 - poly) */
1006             f6A              = _mm_mul_pd(c6grid_00,_mm_sub_pd(one,poly));
1007             /* f6B = C6grid * exponent * beta^6 */
1008             f6B              = _mm_mul_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6));
1009             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
1010             fvdw              = _mm_mul_pd(_mm_macc_pd(_mm_msub_pd(c12_00,rinvsix,_mm_sub_pd(c6_00,f6A)),rinvsix,f6B),rinvsq00);
1011
1012             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
1013
1014             fscal            = fvdw;
1015
1016             fscal            = _mm_and_pd(fscal,cutoff_mask);
1017
1018             /* Update vectorial force */
1019             fix0             = _mm_macc_pd(dx00,fscal,fix0);
1020             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
1021             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
1022
1023             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
1024             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
1025             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
1026
1027             }
1028
1029             /**************************
1030              * CALCULATE INTERACTIONS *
1031              **************************/
1032
1033             if (gmx_mm_any_lt(rsq10,rcutoff2))
1034             {
1035
1036             r10              = _mm_mul_pd(rsq10,rinv10);
1037
1038             /* Compute parameters for interactions between i and j atoms */
1039             qq10             = _mm_mul_pd(iq1,jq0);
1040
1041             /* EWALD ELECTROSTATICS */
1042
1043             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1044             ewrt             = _mm_mul_pd(r10,ewtabscale);
1045             ewitab           = _mm_cvttpd_epi32(ewrt);
1046 #ifdef __XOP__
1047             eweps            = _mm_frcz_pd(ewrt);
1048 #else
1049             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1050 #endif
1051             twoeweps         = _mm_add_pd(eweps,eweps);
1052             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1053                                          &ewtabF,&ewtabFn);
1054             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1055             felec            = _mm_mul_pd(_mm_mul_pd(qq10,rinv10),_mm_sub_pd(rinvsq10,felec));
1056
1057             cutoff_mask      = _mm_cmplt_pd(rsq10,rcutoff2);
1058
1059             fscal            = felec;
1060
1061             fscal            = _mm_and_pd(fscal,cutoff_mask);
1062
1063             /* Update vectorial force */
1064             fix1             = _mm_macc_pd(dx10,fscal,fix1);
1065             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
1066             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
1067
1068             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
1069             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
1070             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
1071
1072             }
1073
1074             /**************************
1075              * CALCULATE INTERACTIONS *
1076              **************************/
1077
1078             if (gmx_mm_any_lt(rsq20,rcutoff2))
1079             {
1080
1081             r20              = _mm_mul_pd(rsq20,rinv20);
1082
1083             /* Compute parameters for interactions between i and j atoms */
1084             qq20             = _mm_mul_pd(iq2,jq0);
1085
1086             /* EWALD ELECTROSTATICS */
1087
1088             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1089             ewrt             = _mm_mul_pd(r20,ewtabscale);
1090             ewitab           = _mm_cvttpd_epi32(ewrt);
1091 #ifdef __XOP__
1092             eweps            = _mm_frcz_pd(ewrt);
1093 #else
1094             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1095 #endif
1096             twoeweps         = _mm_add_pd(eweps,eweps);
1097             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1098                                          &ewtabF,&ewtabFn);
1099             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1100             felec            = _mm_mul_pd(_mm_mul_pd(qq20,rinv20),_mm_sub_pd(rinvsq20,felec));
1101
1102             cutoff_mask      = _mm_cmplt_pd(rsq20,rcutoff2);
1103
1104             fscal            = felec;
1105
1106             fscal            = _mm_and_pd(fscal,cutoff_mask);
1107
1108             /* Update vectorial force */
1109             fix2             = _mm_macc_pd(dx20,fscal,fix2);
1110             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
1111             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
1112
1113             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
1114             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
1115             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
1116
1117             }
1118
1119             /**************************
1120              * CALCULATE INTERACTIONS *
1121              **************************/
1122
1123             if (gmx_mm_any_lt(rsq30,rcutoff2))
1124             {
1125
1126             r30              = _mm_mul_pd(rsq30,rinv30);
1127
1128             /* Compute parameters for interactions between i and j atoms */
1129             qq30             = _mm_mul_pd(iq3,jq0);
1130
1131             /* EWALD ELECTROSTATICS */
1132
1133             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1134             ewrt             = _mm_mul_pd(r30,ewtabscale);
1135             ewitab           = _mm_cvttpd_epi32(ewrt);
1136 #ifdef __XOP__
1137             eweps            = _mm_frcz_pd(ewrt);
1138 #else
1139             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1140 #endif
1141             twoeweps         = _mm_add_pd(eweps,eweps);
1142             gmx_mm_load_2pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),ewtab+_mm_extract_epi32(ewitab,1),
1143                                          &ewtabF,&ewtabFn);
1144             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1145             felec            = _mm_mul_pd(_mm_mul_pd(qq30,rinv30),_mm_sub_pd(rinvsq30,felec));
1146
1147             cutoff_mask      = _mm_cmplt_pd(rsq30,rcutoff2);
1148
1149             fscal            = felec;
1150
1151             fscal            = _mm_and_pd(fscal,cutoff_mask);
1152
1153             /* Update vectorial force */
1154             fix3             = _mm_macc_pd(dx30,fscal,fix3);
1155             fiy3             = _mm_macc_pd(dy30,fscal,fiy3);
1156             fiz3             = _mm_macc_pd(dz30,fscal,fiz3);
1157
1158             fjx0             = _mm_macc_pd(dx30,fscal,fjx0);
1159             fjy0             = _mm_macc_pd(dy30,fscal,fjy0);
1160             fjz0             = _mm_macc_pd(dz30,fscal,fjz0);
1161
1162             }
1163
1164             gmx_mm_decrement_1rvec_2ptr_swizzle_pd(f+j_coord_offsetA,f+j_coord_offsetB,fjx0,fjy0,fjz0);
1165
1166             /* Inner loop uses 179 flops */
1167         }
1168
1169         if(jidx<j_index_end)
1170         {
1171
1172             jnrA             = jjnr[jidx];
1173             j_coord_offsetA  = DIM*jnrA;
1174
1175             /* load j atom coordinates */
1176             gmx_mm_load_1rvec_1ptr_swizzle_pd(x+j_coord_offsetA,
1177                                               &jx0,&jy0,&jz0);
1178
1179             /* Calculate displacement vector */
1180             dx00             = _mm_sub_pd(ix0,jx0);
1181             dy00             = _mm_sub_pd(iy0,jy0);
1182             dz00             = _mm_sub_pd(iz0,jz0);
1183             dx10             = _mm_sub_pd(ix1,jx0);
1184             dy10             = _mm_sub_pd(iy1,jy0);
1185             dz10             = _mm_sub_pd(iz1,jz0);
1186             dx20             = _mm_sub_pd(ix2,jx0);
1187             dy20             = _mm_sub_pd(iy2,jy0);
1188             dz20             = _mm_sub_pd(iz2,jz0);
1189             dx30             = _mm_sub_pd(ix3,jx0);
1190             dy30             = _mm_sub_pd(iy3,jy0);
1191             dz30             = _mm_sub_pd(iz3,jz0);
1192
1193             /* Calculate squared distance and things based on it */
1194             rsq00            = gmx_mm_calc_rsq_pd(dx00,dy00,dz00);
1195             rsq10            = gmx_mm_calc_rsq_pd(dx10,dy10,dz10);
1196             rsq20            = gmx_mm_calc_rsq_pd(dx20,dy20,dz20);
1197             rsq30            = gmx_mm_calc_rsq_pd(dx30,dy30,dz30);
1198
1199             rinv00           = gmx_mm_invsqrt_pd(rsq00);
1200             rinv10           = gmx_mm_invsqrt_pd(rsq10);
1201             rinv20           = gmx_mm_invsqrt_pd(rsq20);
1202             rinv30           = gmx_mm_invsqrt_pd(rsq30);
1203
1204             rinvsq00         = _mm_mul_pd(rinv00,rinv00);
1205             rinvsq10         = _mm_mul_pd(rinv10,rinv10);
1206             rinvsq20         = _mm_mul_pd(rinv20,rinv20);
1207             rinvsq30         = _mm_mul_pd(rinv30,rinv30);
1208
1209             /* Load parameters for j particles */
1210             jq0              = _mm_load_sd(charge+jnrA+0);
1211             vdwjidx0A        = 2*vdwtype[jnrA+0];
1212
1213             fjx0             = _mm_setzero_pd();
1214             fjy0             = _mm_setzero_pd();
1215             fjz0             = _mm_setzero_pd();
1216
1217             /**************************
1218              * CALCULATE INTERACTIONS *
1219              **************************/
1220
1221             if (gmx_mm_any_lt(rsq00,rcutoff2))
1222             {
1223
1224             r00              = _mm_mul_pd(rsq00,rinv00);
1225
1226             /* Compute parameters for interactions between i and j atoms */
1227             gmx_mm_load_1pair_swizzle_pd(vdwparam+vdwioffset0+vdwjidx0A,&c6_00,&c12_00);
1228             c6grid_00       = gmx_mm_load_1real_pd(vdwgridparam+vdwioffset0+vdwjidx0A);
1229
1230             /* Analytical LJ-PME */
1231             rinvsix          = _mm_mul_pd(_mm_mul_pd(rinvsq00,rinvsq00),rinvsq00);
1232             ewcljrsq         = _mm_mul_pd(ewclj2,rsq00);
1233             ewclj6           = _mm_mul_pd(ewclj2,_mm_mul_pd(ewclj2,ewclj2));
1234             exponent         = gmx_simd_exp_d(ewcljrsq);
1235             /* poly = exp(-(beta*r)^2) * (1 + (beta*r)^2 + (beta*r)^4 /2) */
1236             poly             = _mm_mul_pd(exponent,_mm_macc_pd(_mm_mul_pd(ewcljrsq,ewcljrsq),one_half,_mm_sub_pd(one,ewcljrsq)));
1237             /* f6A = 6 * C6grid * (1 - poly) */
1238             f6A              = _mm_mul_pd(c6grid_00,_mm_sub_pd(one,poly));
1239             /* f6B = C6grid * exponent * beta^6 */
1240             f6B              = _mm_mul_pd(_mm_mul_pd(c6grid_00,one_sixth),_mm_mul_pd(exponent,ewclj6));
1241             /* fvdw = 12*C12/r13 - ((6*C6 - f6A)/r6 + f6B)/r */
1242             fvdw              = _mm_mul_pd(_mm_macc_pd(_mm_msub_pd(c12_00,rinvsix,_mm_sub_pd(c6_00,f6A)),rinvsix,f6B),rinvsq00);
1243
1244             cutoff_mask      = _mm_cmplt_pd(rsq00,rcutoff2);
1245
1246             fscal            = fvdw;
1247
1248             fscal            = _mm_and_pd(fscal,cutoff_mask);
1249
1250             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1251
1252             /* Update vectorial force */
1253             fix0             = _mm_macc_pd(dx00,fscal,fix0);
1254             fiy0             = _mm_macc_pd(dy00,fscal,fiy0);
1255             fiz0             = _mm_macc_pd(dz00,fscal,fiz0);
1256
1257             fjx0             = _mm_macc_pd(dx00,fscal,fjx0);
1258             fjy0             = _mm_macc_pd(dy00,fscal,fjy0);
1259             fjz0             = _mm_macc_pd(dz00,fscal,fjz0);
1260
1261             }
1262
1263             /**************************
1264              * CALCULATE INTERACTIONS *
1265              **************************/
1266
1267             if (gmx_mm_any_lt(rsq10,rcutoff2))
1268             {
1269
1270             r10              = _mm_mul_pd(rsq10,rinv10);
1271
1272             /* Compute parameters for interactions between i and j atoms */
1273             qq10             = _mm_mul_pd(iq1,jq0);
1274
1275             /* EWALD ELECTROSTATICS */
1276
1277             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1278             ewrt             = _mm_mul_pd(r10,ewtabscale);
1279             ewitab           = _mm_cvttpd_epi32(ewrt);
1280 #ifdef __XOP__
1281             eweps            = _mm_frcz_pd(ewrt);
1282 #else
1283             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1284 #endif
1285             twoeweps         = _mm_add_pd(eweps,eweps);
1286             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
1287             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1288             felec            = _mm_mul_pd(_mm_mul_pd(qq10,rinv10),_mm_sub_pd(rinvsq10,felec));
1289
1290             cutoff_mask      = _mm_cmplt_pd(rsq10,rcutoff2);
1291
1292             fscal            = felec;
1293
1294             fscal            = _mm_and_pd(fscal,cutoff_mask);
1295
1296             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1297
1298             /* Update vectorial force */
1299             fix1             = _mm_macc_pd(dx10,fscal,fix1);
1300             fiy1             = _mm_macc_pd(dy10,fscal,fiy1);
1301             fiz1             = _mm_macc_pd(dz10,fscal,fiz1);
1302
1303             fjx0             = _mm_macc_pd(dx10,fscal,fjx0);
1304             fjy0             = _mm_macc_pd(dy10,fscal,fjy0);
1305             fjz0             = _mm_macc_pd(dz10,fscal,fjz0);
1306
1307             }
1308
1309             /**************************
1310              * CALCULATE INTERACTIONS *
1311              **************************/
1312
1313             if (gmx_mm_any_lt(rsq20,rcutoff2))
1314             {
1315
1316             r20              = _mm_mul_pd(rsq20,rinv20);
1317
1318             /* Compute parameters for interactions between i and j atoms */
1319             qq20             = _mm_mul_pd(iq2,jq0);
1320
1321             /* EWALD ELECTROSTATICS */
1322
1323             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1324             ewrt             = _mm_mul_pd(r20,ewtabscale);
1325             ewitab           = _mm_cvttpd_epi32(ewrt);
1326 #ifdef __XOP__
1327             eweps            = _mm_frcz_pd(ewrt);
1328 #else
1329             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1330 #endif
1331             twoeweps         = _mm_add_pd(eweps,eweps);
1332             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
1333             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1334             felec            = _mm_mul_pd(_mm_mul_pd(qq20,rinv20),_mm_sub_pd(rinvsq20,felec));
1335
1336             cutoff_mask      = _mm_cmplt_pd(rsq20,rcutoff2);
1337
1338             fscal            = felec;
1339
1340             fscal            = _mm_and_pd(fscal,cutoff_mask);
1341
1342             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1343
1344             /* Update vectorial force */
1345             fix2             = _mm_macc_pd(dx20,fscal,fix2);
1346             fiy2             = _mm_macc_pd(dy20,fscal,fiy2);
1347             fiz2             = _mm_macc_pd(dz20,fscal,fiz2);
1348
1349             fjx0             = _mm_macc_pd(dx20,fscal,fjx0);
1350             fjy0             = _mm_macc_pd(dy20,fscal,fjy0);
1351             fjz0             = _mm_macc_pd(dz20,fscal,fjz0);
1352
1353             }
1354
1355             /**************************
1356              * CALCULATE INTERACTIONS *
1357              **************************/
1358
1359             if (gmx_mm_any_lt(rsq30,rcutoff2))
1360             {
1361
1362             r30              = _mm_mul_pd(rsq30,rinv30);
1363
1364             /* Compute parameters for interactions between i and j atoms */
1365             qq30             = _mm_mul_pd(iq3,jq0);
1366
1367             /* EWALD ELECTROSTATICS */
1368
1369             /* Calculate Ewald table index by multiplying r with scale and truncate to integer */
1370             ewrt             = _mm_mul_pd(r30,ewtabscale);
1371             ewitab           = _mm_cvttpd_epi32(ewrt);
1372 #ifdef __XOP__
1373             eweps            = _mm_frcz_pd(ewrt);
1374 #else
1375             eweps            = _mm_sub_pd(ewrt,_mm_round_pd(ewrt, _MM_FROUND_FLOOR));
1376 #endif
1377             twoeweps         = _mm_add_pd(eweps,eweps);
1378             gmx_mm_load_1pair_swizzle_pd(ewtab+_mm_extract_epi32(ewitab,0),&ewtabF,&ewtabFn);
1379             felec            = _mm_macc_pd(eweps,ewtabFn,_mm_mul_pd( _mm_sub_pd(one,eweps),ewtabF));
1380             felec            = _mm_mul_pd(_mm_mul_pd(qq30,rinv30),_mm_sub_pd(rinvsq30,felec));
1381
1382             cutoff_mask      = _mm_cmplt_pd(rsq30,rcutoff2);
1383
1384             fscal            = felec;
1385
1386             fscal            = _mm_and_pd(fscal,cutoff_mask);
1387
1388             fscal            = _mm_unpacklo_pd(fscal,_mm_setzero_pd());
1389
1390             /* Update vectorial force */
1391             fix3             = _mm_macc_pd(dx30,fscal,fix3);
1392             fiy3             = _mm_macc_pd(dy30,fscal,fiy3);
1393             fiz3             = _mm_macc_pd(dz30,fscal,fiz3);
1394
1395             fjx0             = _mm_macc_pd(dx30,fscal,fjx0);
1396             fjy0             = _mm_macc_pd(dy30,fscal,fjy0);
1397             fjz0             = _mm_macc_pd(dz30,fscal,fjz0);
1398
1399             }
1400
1401             gmx_mm_decrement_1rvec_1ptr_swizzle_pd(f+j_coord_offsetA,fjx0,fjy0,fjz0);
1402
1403             /* Inner loop uses 179 flops */
1404         }
1405
1406         /* End of innermost loop */
1407
1408         gmx_mm_update_iforce_4atom_swizzle_pd(fix0,fiy0,fiz0,fix1,fiy1,fiz1,fix2,fiy2,fiz2,fix3,fiy3,fiz3,
1409                                               f+i_coord_offset,fshift+i_shift_offset);
1410
1411         /* Increment number of inner iterations */
1412         inneriter                  += j_index_end - j_index_start;
1413
1414         /* Outer loop uses 24 flops */
1415     }
1416
1417     /* Increment number of outer iterations */
1418     outeriter        += nri;
1419
1420     /* Update outer/inner flops */
1421
1422     inc_nrnb(nrnb,eNR_NBKERNEL_ELEC_VDW_W4_F,outeriter*24 + inneriter*179);
1423 }